September 2002

ISSN 1649-2358

BASQUE COUNTRYL10N and minority language PAGES 6-7

BASQUE COUNTRYL10N and minority language PAGES 6-7

Summer SchoolReport on the 2nd LRC InternationalSummer School PAGES 4-5

Summer SchoolReport on the 2nd LRC InternationalSummer School PAGES 4-5

New TechnologiesThe impact of translation vendor web services PAGES 9-11

New TechnologiesThe impact of translation vendor web services PAGES 9-11

VOL. 1, Issue 2Issue sponsored by:

THE INTERNATIONAL JOURNAL FOR LOCALISATION

LRC 2002 eContent LocalisationConference 12-13 November

PAGES 26-27

Localisation Focus is the publication of theLocalisation Research Centre (LRC). It is dis-tributed free of charge to professionals workingin the localisation industry. Please notify thecentre if you or one of your colleagues wouldlike to receive Localisation Focus regularly.

Editor: Reinhard SchälerAssistant editor: Raphaëlle FraysseProduction editors: Catherine Osborn,

Alan PhelanOrigination: Litho StudiosPublished by: Localisation Research Centre

(LRC), Department of Computer Science andInformation Systems (CSIS), University ofLimerick, Limerick, Ireland.

Tel. +353-61-202881Fax +353-61-202734Email: LRC@ul.ieWeb http://www.localisation.ie

© 2002 Localisation Research Centre

PUBLISHER INFORMATION

Sponsorship and advertising

To advertise in or to sponsor an issue ofLocalisation Focus, contact the LRC for moreinformation at lrc@ul.ie and find out about thebenefits.

Precision Media GlobalizationSystems, part of the PerigordCommunications group, understands the com-plexities involved in the production of any multi-lingual project and offers the experience andexpertise of over 10 years DTP andPrepress localization.

Precision offer fast, efficient, process drivencontrol over the localization of high endpackaging, instructional manuals or promotionalliterature while incorporating IntegratedQuality Control procedures, and a WebManagement Interface that allows pro-ject tracking online.

Precision has progressed beyond the DTParena and has handled multilingual projectsinvolving the localization of multimedia CD-ROMs, courseware and websites.

Precision Media Globalization SystemsKylemore Road, Ballyfermot, Dublin 10

email: peter@precision-mgs.comweb: www.precision-mgs.comphone: 01-6230750

STAR in Taiwan..................................................................

The STAR Group has had offices in PR Chinasince 1995, first in Shanghai and then in Bei-jing. Over the past two years, however, the num-ber of customers requiring translations into bothMainland Chinese and Taiwan Chinese has risendrastically. So STAR decided to set up newoffices in Taiwan. They appointed IrisKleinophorst as Branch Manager. A graduatetranslator in English, Chinese and French fromthe School of Applied Linguistics and Culturalof Mainz University, Germany Studies (FASKGermersheim), she had previously worked atthe STAR offices in Shanghai as project man-ager, so she is conversant with STAR's projectmanagement methods. Iris explains her moveto Taiwan in these terms, "It soon became obvi-ous that it was not a viable long-term solutionto either coordinate translations done in Tai-wan from Shanghai or provide Taiwanese cus-tomers with translations that had merely beencode-converted from Simplified Chinese char-acters (as used in PR China) to the Traditionalcharacters used in Taiwan. Iris adds, "Althoughthe languages of the People's Republic of Chinaand the Republic of Taiwan are very similar, thevocabulary and the linguistic structure differ in

many aspects. We found that companies whowant to address the specific needs of Taiwanesecustomers prefer to have translations done byTaiwanese translators in Taiwan." ContactSTAR Taiwan at taiwan@star-group.net.

T-Remote Memory due for release..................................................................

In October Telelingua Software will release

T-Remote Memory which provides a significant

performance and quality enhancement to trans-

lations being processed through translation mem-

ories (TM). T-Remote is an add-on to TM tools,

allowing multiple translators to work off a sin-

gle corpus in real time via the Internet.According

to Telelingua, the advantages of their product

are the following: PMs don't have to break up

the text for distribution to multiple translators;

all translators are enriching the same db at the

same time; multiple TMs can be used on the same

project; the system works in a pure virtual envi-

ronment, i.e. the TMs can be located in multiple

locations, and the translators in yet other loca-

tions; jobs do not need to be put back together

at the end of the project; all of the alignments of

multiple texts do not need to be carried out at the

end of the project; the customer can easily pos-

sess a single db with all of its TMs thanks to the

central location of the db; the product features a

very high, instantaneous speed. For more infor-

mation, visit http://secure.telelingua.com/soft-

ware/EN/soft_produits.asp

Send your press releases to press-releases-lrc@ul.ie and get your company on LocalisationFocus INDUSTRY.news page.

GET NOTICE FOR YOUR COMPANY

INDUSTRY.news

SEPTEMBER 2002 CONTENTS.editorial 3LOCALISATION FOCUS

Coming back from holidays, many things have changed for manypeople in the localisation industry. Going away for two or eventhree weeks is a risky business in an industry that is always

on the move.Some came back and found they had been relocated to a differ-

ent office or even a different building. Others found themselves work-ing for a different company as a former competitor had acquired theiroriginal employer.

One of the most important events in the localisation business thissummer must be the acquisition of Berlitz GlobalNET by Bowne GlobalSolutions.At the time of writing, this deal was as good as done. BerlitzGlobalNET was, in its original incarnation as "Softrans International",the first large, dedicated localisation service provider. Brian Kelly, Sof-trans International’s MD and more recently Senior VP of Berlitz, becameone of the fathers of the localisation service industry. We hope toassess what this latest acquisition by BGS of the industry veteranmeans for the business as a whole in the next issue of LocalisationFocus and invite you to send us your own, personal assessment of theeffect this acquisition will have on the industry (email lrc@ul.ie).

One of the most dramatic technological developments affectinglocalisation are web services. In the last edition of Localisation Focus,Bill Looby of IBM, provided a technical introduction to web servicestogether with some very useful references for those who wanted tofind out more. Bill also announced the establishment of a workinggroup under the umbrella of the LRC looking into the development ofstandards for translation vendor web services.

In this edition, Jaap van der Meer, another industry veteran, exploresthe potential effect of this new technology on the business modelscurrently used. Jaap supports his analysis with some interesting dataand solid market forecasts.

LRC 2002 – eContent Localisation will take place on 12-13 Novem-ber 2002 in Dublin’s O’Reilly Hall. Europe’s largest and longest run-ning dedicated annual localisation event will provide the platform forthe launch of the Translation Vendor Web Services Group.The confer-ence will feature an industry exhibition, the annual industry dinnerand the announcement of this year’s LRC Best Thesis Award, spon-sored by Symantec Ireland.

Contributors to LRC 2002 will, for the first time, feature localisa-tion colleagues from Canada and India who will join speakers fromUS and European-based companies, including Microsoft and AOL.

Reinhard Schäler

4LRC 2002 International Summer School

6Localisation in the Basque CountryLuistxo Fernandez

8A GALA AffairHans Fenstermacher

Impact of translation web servicesJaap van der Meer

12Documentation localisation costs

14Localisation in the Global VillageDetlev Hoppenrath

16Joining the Localizer communityJuan David Ibáñez Palomar

18Microsoft .NET technology changes the paradigmTony O’Dowd

21The cultural dimension in software localisationReinhard Schäler

Readers’ Forum

30The LRC and its functions

When you change address, remember to update us at LRC@ul.ie! Thisway you will be able to enjoy reading your magazine wherever you go.

From the Editor

25LRC News

28Two Heads are better than OneDamian Scattergood and Paul Quigley

29LocalisationShailendra Musale

SCHOOL.report

COUNTRY.focus

ASSOCIATION.loc

PERSONAL.profile

26LRC 2002 eContent Localisation Conference12-13 November O’Reilly Hall UCD

CONFERENCE.loc

POETRY.loc

LO

CA

LI

SA

TI

ON

C

EN

TR

AL

24

9 LOCALISATION.central

Opinions expressed by indvidual authors do not necessarily reflect those of the editor

SCHOOL.report SEPTEMBER 20024 LOCALISATION FOCUS

LRC 2002 InternationalLocalisation Summer SchoolOver four days, sixty participants from the Americas and a large

number of European countries attended the second LRC

International Localisation Summer School at the University of Lim-

erick 17-20 June 2002.

Two coursesThis year’s summer school offered two courses, each running over

two days: A foundation course and an advanced course in software

localisation. Each course was divided into a number of modules.

The foundation course covered management of the software local-

isation process, localisation document engineering and translation,

localisation engineering, localisation QA and testing, and an intro-

duction to internationalisation issues.

The advanced course covered management of complex, multiple-

language projects, the management of globally operating teams and

the issues arising out of small vendors

servicing large, multinational clients.

Technical discussions focused on the

localisation of .NET applications and

– over a full-day session – localisation

issues around XML.

Challenging tasksWhile the parallel running sessions of

last year’s Summer School had forced

participants to make sometimes dif-

ficult choices in favour of particular

modules and against others, the structure of this year’s course allowed

participants to take part in all modules.

This lead to a very diverse composition of groups ranging from

project managers to engineers and translators to testers. It was a chal-

lenging task for the presenters to present their sometimes complex

material at the right level, so that nobody got lost on the way.

For the participants, it was an opportunity to learn more about

localisation issues that are not part of their day-to-day job. They

learned to appreciate the contribution of their colleagues in other

localisation departments.

Many of the sessions were hands-on and made use of the Univer-

sity’s state-of-the-art computing facilities. Exercises prepared by the

presenters allowed the participants to experiment with the material

and (although in a limited way) learn from experience.

OnlineFor the first time, all presentations have

been recorded on digital video. The LRC

is in the process of making one of the ses-

sions available on its web site (www.local-

isation.ie) on an experimental basis. It is

also planned to make additional sessions

available on CD ROM and online. All of

the sessions will be made available to

researchers in the LRC.

OutlookBased on the positive feedback provided by the participants and on

the experience of this year’s courses, the LRC will again offer a Sum-

mer School in June of 2003.

The LRC will also develop outline curricula for the Summer School

and make these available to potential participants. Together with the

course presenters, we will publish detailed course descriptions to make

it easy for participants to identify the right courses for their require-

ments. Course presenters and the LRC will closely cooperate in the

preparation of the course modules.

Should you have missed this year’s Summer School, visit our web

site where you will find all relevant presentation overheads. Video

files will become available soon. www.localisation.ie

The Summer School was supported by Lotus/IBM Alchemy Software Development , the RWSGroup (translate.com), PASS Engineering, Oracle WPTG,Archetypon and The Institute of

Localisation Professionals (TILP).

All course presenters were professionals with many years of experience in the industry andin the presentation of complex course material. They included Florian Sachse, John Malone,

Lisa Daly, Padraig Bracken, Patrick O'Sullivan, Ray Loughran, Reinhard Schäler, SusanneMaier, Tom Connolly, Tony O'Dowd and Yves Savourel.

The LRC International Localisation Summer School would not have been possible withoutthe generous support of the University of Limerick and the staff at the Department of Com-puter Science and Information Systems (CSIS). Special thanks are due to the Head of CSIS,

Dr Ita Richardson.

The secrets of project management in global teams withRay Loughran (IBM)

Entering the wide worldof XML with Yves Savourel

(RWS Group LLC)

Tony O’Dowd (Alchemy)sharing his enthusiasmfor the localisation of

.NET applications

SEPTEMBER 2002 SCHOOL.report 5LOCALISATION FOCUS

Funtastic! The second international LRC localisation summer school started with the difficulty of overcoming thegreat success that was the first one. I am glad to say that although the expectations were high, the sec-ond edition did not disappoint. I found all the seminars, and particularly the advanced course, extremelyinformative (and useful). I also made some great friends and enjoyed the evenings out (loved the spoontapping concert!).

I sincerely look forward attending next year's summer school.

I attended the Advanced Course in Software Localisation in this year's LRCSummer School.

This year's course was very well organised. It focused on current and upcom-ing technologies in the localisation industry.

The course also highlighted what care should we take to avoid commonmistakes in localisation.

The sight-seeing trip and dinner in traditional Irish environment was a memorable experience.

I enjoyed the Irish music a lot.

Such educational events will greatly help the fellows in the localisationindustry.

First of all I would like to congratulate the entire localisation department atthe University of Limerick for the great organisation of the event and for theirwarm hospitality and attention.

On the academic front, I would like to highlight the conscientious choice ofspeakers, who were all highly experienced in the industry and well preparedboth theoretically and practically. I would also like to congratulate them forthe centre’s impressive facilities, which made the long journey from Barcelonamore than worthwhile.

I would also like to make special mention of the magnificent meal and per-formance on Monday night. It was the first time I had witnessed typical Irishdancing and I found it rather exciting. I even had the opportunity to take partin a short dance thanks to the instructions from the authentic Irish dancers,and although both they and I gave it all we were worth, I think I still have topractice a little more. The night turned out great and ended just as it shouldhave done: with a good, cold Irish beer.

If the night ended well, the morning could not have started off better. I leftthe Kilmurry Lodge Hotel with a Brazilian colleague, ready for the short trip of5 Irish minutes (which, translated/localised to the Spanish metric system,would be no less than 20 minutes) to the University. She was wearing shoesthat were very stylish but little suited to walking, proving that the relationshipbetween women and shoes is a highly complex and, it would appear, ‘global’matter. Her feet hurt a lot and the walk that lay ahead was a tough test for her.It is just as well that we bumped into Geraldine a short way from the Hotel andshe was able to drop us off in the car.Thanks to her we arrived safe and soundfor the second day of the course.

It only remains for me to thank everybody for the experiences they sharedwith me and for all their kindness. I hope to see you all soon.

As a first time participant in LRC activities (and indeed visitor to Limerick!), Iwas looking forward to my 2 days visit in June. And I wasn't disappointed. Istayed in campus accommodation, and promptly wished I had attended UL inmy student days, instead of staying in Cork. University of Limerick grounds mustbe up there with the best of them.When Geraldine Harrahill informed me thata walk around the perimeter whilst following the river, would take me into Co.Clare, I was intrigued.

The Computer Science building where the courses took place was spaciousand modern, and the equipment and facilities top notch. Sitting in the com-puter labs, looking at those coveted flat screen monitors, I recalled my firstvisit to the college computer labs 8 years ago as a first year student in Uni-versity College Cork.We got black and white Macintosh SE machines (yes - thelittle box ones, which provoked much amusement when we were presentedwith an assignment to produce coloured mathematical diagrams in our Pascalprogramming course). It just goes to show what a long way things have comein the past decade. And the setting of all this is a good reflection of the LRCand the work it does.

In several articles in the June issue of Localisation Focus, the writers reflectedon the status of localisation - how it was faring amongst eCulture, globalisa-tion, communication leviathans, etc. I think Limerick is a lot like localisation,I'm amazed about the amount of people who don't know much about either,but are pleasantly surprised when they get the low-down.Also, against the likesof Dublin and Cork, it is a city that has come into its own in the IT industry, andhas more or less left Cork in its wake in that department! It’s good to see local-isation reaching the west of Ireland, withits imagery of misty hills and wild swansof Coole. Maybe LRC can help to staveoff the globalisation effects yet, or atany rate, dispel some of the localisationapathy out there!

Jesús Maroto, Localisation Director, Euro RSCG

Shailendra Musale,F-Secure Corporation

Enric Tortajada, Syntax

FEEDBACK FEEDBACK FEEDBACK FEEDBACK FEEDBACK FEEDBACK FEEDBACK FEEDBACK FEEDBACK FEEDBACK FEEDBACK FEEDBACK

Lisa Daly, Presenter (Oracle)

Participants from around the world enjoying an Irish night

COUNTRY.focus SEPTEMBER 20026 LOCALISATION FOCUS

Probably, most of us first heard

of localisation from Microsoft

around 1995. The Windows 95

and 98 operating systems, as well as

other MS Office packages have since

been localised into Basque. In the

Basque Country several research

groups and companies have con-

tributed to this. But this effort has

been totally subsidised by the Basque

local authorities, and Microsoft has

been the main beneficiary of it.

The localisation process applied the

tools and procedures set up by

Microsoft, and the participation of

Basque experts was largely restricted

to the translation task. Consequently,

the know-how transferred to the local

participants was more related to ter-

minological issues than to real local-

isation procedures. One of the

problems was that the Basque local-

isation effort was undertaken by the

Spanish division of Microsoft, which

had more concern towards general

marketing matters regarding Spain

than with localisation proper. Basque

localisers detected some deficiencies

that they could not mend, such as, for

example, the syllabic partition of the

MS Word processor, which did not

meet with Basque standards.

Microsoft, who is reputed for its well

engineered i18n and l10n projects,

did not give much priority to the

issue. There was no real technical

problem, nor was the Basque locale

too complex. The real matter was one

of organisation, different priorities,

and little concern for real localisation,

besides getting the check from the

Basque administration.

On top of this, the constantly-upgrad-

ing pace of Microsoft software made

the life and profit of the localised

products extremely short (at the

expense of Basque taxpayers, of

course). Apart from that, marketing

and distribution were awful.

Local authorities seem to have

switched channels recently: Now, it

is time to localise free software! There

is already a Mandrake Linux localised

version being distributed, and the

Open Office suite will come next.

This new effort seems fine, at least

Mr. Gates will get zero royalties.

However, the new official push goes

in the same direction, the localisation

Localisation in the Basque Country

The Basque Country, located on the border between Spainand France, is a small industrialised region of around threemillion inhabitants, out of whom one million or less speakBasque. This is the original native language of the region,an isolated pre-Indo-European language, which has the unde-served reputation of being a too difficult language to learn.But this is a wrong perception. It happens to be very differ-ent from neighbouring languages in its syntax, morphology(with its distinctive ergative case mark), and the lexicon.Any Italian speaker reading a Catalan text may be able toroughly grasp its meaning. That would not happen with aBasque text.

Basque is a minority language even in its own territory, interms of number of speakers and social functions. It is an

endangered language that suffered a deep recession dur-ing the 20th century.After the dictatorship of Franco

in Spain, a strong social and cultural movementemerged in support of the language. This movement

was partly promoted by the newly elected localauthorities, with the Basque Government as the

most salient institution, and as a result of itover the past 20 years the language

has made substantial progress inareas such as education, the

media, and the administra-tion. Statistics about the

command and usageby the population are

also improving. Butdespite all these

advances, as we enter the21st century, it is becoming

clear that new technologiesmust be applied to the language

to guarantee its survival. At thefrontline of the technological chal-

lenge is localisation.

Luistxo Fernandez, localisation

consultant at Code & Syntax

writes about localisation in the

Basque country which, in the

main, is subsidised by the Basque

local authorities

SEPTEMBER 2002 COUNTRY.focus 7LOCALISATION FOCUS

of 'big' desktop infrastructure. In my

view, this 'big-desktop' strategy leads

nowhere. I would say that we are

localising just menus. It is okay, but

in the end, what is what we really

need? I, personally, need no Basque

menu. What I really need is Basque

food! A word processor with Basque

menus is fine, but if it cannot cut

Basque syllables correctly, what kind

of Basque text am I going to write

with it? In my opinion, the real

prospects for Basque localisation

should focus in another direction, and

not towards the 'big-desktop' effort.

My hopes for the future are on

research teams specialised in natural

language processing (NLP), and on

some small companies working

around that field. Multidisciplinary

teams of linguists and computer sci-

entists have allied in Basque univer-

sities and have produced interesting

results in areas such as speech recog-

nition and synthesis, morphological

and syntactic analysis, spell check-

ing...

These advances are not what we

would call localisation, properly,

because they pertain to another area

of language engineering. Yet, they pro-

vide the best basis for true Basque

localisation. Translating the huge

amount of message-strings that any

operating system contains is a very

expensive task; however the benefits

are dubious. Adapting a web appli-

cation server to handle dates under

the Basque style, or being able to lem-

matise the root of a term in a search

facility (a very important aspect for

any Basque locale due to the postpo-

sitional nature of its morphology) is

a much more interesting accomplish-

ment. It is nice to be able to develop

web sites or Internet services that can

handle such Basque little quirks. It is

in that manner that language engi-

neering can foster localisation. For-

tunately, some small companies have

joined NLP research teams to develop

this kind of practical localisation soft-

ware for Basque; and it is in the inter-

net where it is becoming most visible

in the form of Basque and multilin-

gual sites.

This trend brings hopeful prospects

to the Basque localisation community.

There is a big concern for multilin-

gualism from the part of the admin-

istration and other public institutions,

such as universities, museums, social

services and the like, since they need

to communicate at least in Basque

and Spanish, both internally and with

the society at large. It is in the realm

of I-net applications that these needs

will have to be solved, in the first

place.

Localising something like Gnome is

a formidable task, with results visible

almost to nobody. Localising a report-

ing tool for an intranet or a public

web site is a much easier undertak-

ing, and besides has immediate and

much more visible results. The desk-

top application must be looked for,

acquired, installed, only to find out

in the end that a new upgraded ver-

sion has just been released. Con-

versely, the localised facility on a web

site can be very easily accessed and

put at work through a simple click on

the Language Change button of the

interface.

There is a clear need for localisation,

and it is up to Basque localisation

companies to join NLP research teams

to meet this demand. This has, in

turn, some important strategic impli-

cations. From the point of view of the

skills, the morphological properties

of Basque and Spanish are different

enough as to make the task of devel-

oping bilingual products an intellec-

tually challenging activity. It is almost

as making a localisation test. Basque

morphology very often creates length-

problems, similar to those faced in

German localisation, and there are

also some tricky syntactic aspects,

such as inverse word-order, post-posi-

tional case marking (similar to that

of Finnish), etc. The expertise we may

have will help very little with encod-

ing issues (as those posed by Asian

languages), but will give us a good

basis for localising on a western-Euro-

pean scale, at least.

In my opinion, developing a Basque

localisation expertise will not just pro-

vide solutions for this marginal lan-

guage that we still speak, but will be

an opportunity for our companies to

open to wider markets, ranging from

local export companies (the local

industry is quite internationalised), to

the participation in European ven-

tures and projects.

This in itself is quite an important

challenge.

Basque alone does not offer a big

enough market to sustain a company.

The 'big-desktop' strategy, the local-

isation of either proprietary or open

source fat systems, is only sustainable

through public subsidy. The ever-

upgrading nature of such systems

shows no prospect of taking us

beyond this subsidy-policy, and offers,

if any, very far-off benefits, which will

anyway fall into the hands of MS roy-

alty collectors, or the free-software

community. It will hardly pay back

to the local users, who will see almost

no change on their computer screens.

It is by no means a sustainable strat-

egy.

A sustainable policy should be based

on real market opportunities, not on

subsidies. The globalised web is that

true market. Multilingual solutions

for that market is where Basque

localisers should focus on. The con-

junction of our local need for multi-

lingualism together with our locally

attained skills will plot the route to

develop solutions for others. And it

is the outside market that will, in turn,

sustain our own local effort.

Luistxo Fernandez can be reached at

LFernandez@codesyntax.com

The famous flower puppy outside theGuggenheim museum in Bilbao

Vitoria-Gasteiz on the occasion of thefeast of the Virgin

ASSOCIATION.loc SEPTEMBER 20028 LOCALISATION FOCUS

T he Globalization, Internationalization,

Localization and Translation (GILT)

industry is in a state of flux. Between

the economic downturn, consolidation, and

the continuing balkanisation of our indus-

try, one of the biggest challenges we face is

understanding and agreeing on what our

industry is. Who are we? What do our cus-

tomers really want? Why are our basic mes-

sages inconsistent or improperly understood?

How can we raise our visibility on the world

business stage?

On April 15 of this year, 15 localisation

companies from 12 countries on four conti-

nents created a new trade association to try

to find answers to these and other questions.

Dubbed the Globalization and Localization

Association (GALA), the association grew

out of a less formal network of small- to

medium-sized localisation service providers

that had been working together for some

time. GALA aims to represent the provider

side of the industry, which consists of thou-

sands of companies all over the world pro-

viding a wide range of products and services.

The GILT industry has several fine organi-

sations already, but most represent a specific

industry segment, cover only a limited geo-

graphic area, or are struggling to reconcile

conflicting constituencies. The founders of

GALA and other like-minded companies have

seen the need for a fully representative, inter-

national trade organisation that works toward

the betterment of the whole industry.

GALA’s philosophy is based on the idea

that industry progress will come from work-

ing more closely together to share informa-

tion, fostering innovative ways to promote

ourselves and our industry as a whole, and

offering our customers unique, collaborative

value. GALA offers a forum for discussing

ideas, developing solutions, and creating

opportunities for joint – and therefore more

visible and powerful – representation within

the industry and to the outside community.

The Association’s mission statement is "to

provide its members with opportunities to

better meet their market needs, while reduc-

ing costs and increasing opportunities, and

to offer their customers better service." This

means taking concrete steps based on simple

ideas:

● cooperating with each other on projects

by means of a mutually agreed-upon Code

of Conduct and a mutual Non-Disclosure

Agreement (sharing a philosophy of qual-

ity and mutual respect);

● providing customers with joint solutions

tailor-made to meet their needs (selling

what customers actually want);

● reducing operational expenses by pool-

ing our purchasing power (investing in

ourselves for growth);

● increasing our members’ knowledge base

by frank information exchange (trusting

and learning from each other); and

● creating business opportunities for our

members so they can win new clients

(making the market bigger).

One of the driving forces behind estab-

lishing GALA is the desire to give companies

in our industry more opportunities to acquire

bandwidth for their products and services.

Many service companies, for example, spe-

cialise in a limited number of language pairs

and/or domains, focusing on producing pre-

mium quality. Some technology companies

have highly specialised tools, but find it hard

to get exposure for their products. Mem-

bership in GALA enables these companies to

expand the scope and volume of their activ-

ities through ongoing or project-based part-

nerships with other companies that share

their philosophy or business model. GALA

gives all companies alike the opportunity to

come to know and — critically — trust part-

ners who can help them deliver what their

clients demand.

GALA is run by an elected 3-person

Board, which manages the affairs of the Asso-

ciation. The founding Board members are:

Hans Fenstermacher (Chairman) of Archi-

Text, Renée Sztabelski of HiText, and Daniel

Carter of International Software Products.

Any company that meets the following eli-

gibility criteria may join the Association:

● Bona fide company of any size provid-

ing translation, localisation, internation-

alisation, or globalisation products or

services, including tools developers, train-

ing suppliers, and consultancies;

● Signature of and adherence to the GALA

Code of Conduct and Non-Disclosure

Agreement; and

● Payment of the annual dues.

GALA will actively help its members work

together, promote themselves, and expand

their reach through activities like: the Asso-

ciation web site (www.gala-global.org); mem-

ber-only events to share information, improve

efficiency and increase industry knowledge;

matchmaking events to bring participating

GALA members together with potential cus-

tomers; joint representation under the GALA

banner at trade events for industry verticals,

so members can extend the scope of their

sales and marketing in an affordable way;

and a hard-hitting on-line newsletter (com-

ing soon) that promotes interactive discus-

sions and solution-building.

The new Association has also begun and

hopes to help lead a cohesive, industry-wide

effort to define and raise the industry’s image.

Other industries have successfully reinvented

themselves by working together, frankly

assessing their strengths and weaknesses, and

pooling their resources to create a better envi-

ronment for their products. So, too, GALA

will work toward a better industry image,

for the benefit of all of us in this space. ■

A GALA Affair

Hans Fenstermacher is president of ArchiTextInc. and Chairman of the Board of GALA. TheGlobalization and Localization Association (GALA)is actively seeking new members from within theworldwide globalisation, internationalisation,localisation and translation community. For moreinformation about GALA, please visit www.gala-global.org or contact info@gala-global.org.

Hans Fenstermacher writes on the

creation of the new Globalization and

Localization Association

SEPTEMBER 2002 LOCALISATION.central 9LOCALISATION FOCUS

IN NOVEMBER of 2001 Bill Looby of

IBM gave a presentation at the LRC con-

ference in Limerick about web services

and how this technology could apply to

the translation market. A lot of talk and

some real developments have happened since

then. Telelingua announced its T-Remote Mem-

ory product, which uses a web services definition

to retrieve translation matches from multiple dis-

tributed translation memory systems. Berlitz uses

a web services definition in its Open Translation

Connector to link content management systems

with BerlitzIT. Lexelnet is about to launch its

translation marketplace using web services defi-

nitions for the providers of the services. Global-

sight announced its .Net edition, which allows

localisation portals to link easily with its global-

isation technology. Those who have not yet imple-

mented web services one way or the other in their

translation technology will most certainly be talk-

ing about it.

Web services create the best opportunity for

the cottage translation market to cross the chasm

JAAP VAN DER MEER’s article is a follow-up to Bill Looby’s

article on web services in the last edition of Localisation Focus.

Bill announced the establishment of a translation vendor web

services standards group under the umbrella of the LRC. This

group has been established and is now in its steering phase. The

group will be officially launched at a special meeting during the

upcoming LRC 2002 Conference in Dublin (12-13 November

2002) and is inviting expressions of interest from interested parties

to join it. In the meantime, for further information please contact:

Reinhard.Schaler@ul.ie

WE would like to invite you to join ourExpert Council on the ELECT project. Theaim of the project is to get a broad spec-trum of eContent development and locali-sation expertise within the council: forexample, eContent publishers,localisation vendors, web designers andarchitects, legal and commercial bodies,standards consortiums, ISPs (global andlocal), TV producers, and representatives of the hotel and tourist industry.

You can subscribe to the Expert Council on our website at http://www.localisation.ie/Experts/Application.htm

We plan to hold annual expert work-shops in 2002 (planned Nov. 14) and in2003 to collect constructive input on thefollowing topics:● Development of global websites (inter-nationalisation);● Localisation of eContent and websites,including cross-cultural aspects.

Proceedings of these workshops willbe localised into four European languagesand published on the ELECT web portal.

Updates and more details will be published on our website at http://www.localisation.ie

Call for expertsand become a professional services industry. Web

services help us to overcome the dilemma that

has suppressed the translation market ever since

we started talking about localisation as an indus-

try (around 1990). The mergers and acquisitions

of recent years and the launch of new market

offerings have only reinforced this dilemma. The

supply chain has become more complex, with

more vendors and more tools — and more incom-

patibility. Any possible gain from the use of trans-

lation memory is lost again because of the increase

in overheads of managing multiple translation

databases. Web services connect tools and sup-

pliers automatically without the need for direct

communication. Articles or sections from web-

sites will find their way to the translator auto-

matically. No phone conversations, no faxes, no

FTP, no emails (except perhaps automatically gen-

erated emails). Translation memory matches will

be retrieved fully automatically from anywhere

in the network. And for real-time translation, the

web services architecture will automatically route

texts to a machine translation engine. No pre-

and post-processing, no translation folders and

no so-called translation engineering. Web services

allow us to build the intelligence into the infra-

structure in standard definitions that communi-

cate openly with any tool and any supplier that

comply with the standard.

Web services will undermine current business

models, especially those of the so-called MLVs.

Ironically speaking, it was in their interest to keep

the dilemma alive. If the infrastructure were to

become intelligent enough to automatically deliver

a quote or final approved translation, the role of

the translation brokerage firm would become less

relevant (unless perhaps they build and rent the

web services infrastructure themselves). In this

article, I will describe the role of web services from

the perspective of the dynamic market we live in.

Market relevanceAs information becomes a corporate asset for

many companies, translation and effective com-

munication in all required business languages also

becomes vitally important. While translation in

many companies is still a relatively isolated activ-

ity, more and more corporate decision-makers

now understand that it must be integrated into

corporate processes. Translation and effective

multilingual communication become part of cor-

porate information strategies. They play a cru-

cial role in managing customer relationships,

human resources and supply chain. Market sur-

veys indicate that people stay four times longer

on sites when the information is provided in their

native language. The extent to which intranets

and extranets are translated has a direct effect on

the success of companies.

However, the current market offering does not

provide for adequate integration in corporate pur-

chasing platforms, in content management or cus-

tomer relationship management. The translation

market is still a cottage industry. Services are

offered by hundreds of thousands of freelancers

working from their homes and thousands of agen-

cies. They all have their specialties, like subject

Impact of translationweb services

Jaap van der Meer Continued on page 10

LOCALISATION.central SEPTEMBER 200210 LOCALISATION FOCUS

knowledge and language pairs. And all have

their own terms and peculiarities when it

comes to delivery and invoicing.

Market sizeAccording to IDC, the size of the locali-

sation, translation and interpretation (LTI)

services market and cross-lingual applica-

tion (CLA) services was $2.981 billion in

2001, predicted to grow to $6 billion by

2006. (IDC, Worldwide Globalization, Inter-

nationalization, and Localization Services

Market, Forecast and Analysis Update,

2001-2006). The diagram at right provides

a breakdown of the major sub-industries,

their 2001 revenues and their projected

growth up to 2006.

Website localisation is a specialised ser-

vice that has emerged in recent years (since

1999). It is basically a packaging of trans-

lation services with technical services that

ensure the proper functioning of the trans-

lated sites. A number of specialised service

companies have entered this marketplace.

Computer-aided translation as a sub-

sector consists of the sales of software and

services that support the automation of

translation processes, such as translation

memory software, machine translation and

dictionary management software. Although

these software products have been around

for a long time (since 1985), the market thus

far has developed slowly. IDC expects this

sub-sector to grow by 35% a year. This

clearly illustrates the corporate need for

process automation and integration.

Software localisation is a specialised ser-

vice that emerged with the proliferation of

desktop computers at the end of the 1980s.

Since that time a new generation of

specialised service companies has emerged

offering packaged translation services and

technical services (like software testing and

quality assurance) that ensure localised soft-

ware will function properly.

Translation is at the forefront of the LTI

services market. It is the basic service pro-

vided by hundreds of thousands of free-

lancers and thousands of agencies. In fact,

most of the software localisation and web-

site localisation companies use the services

of the translation sub-sector on a subcon-

tractual basis rather than using in-house

translators. The market is known for its mul-

tiple layers of subcontracting. The manage-

ment overheads of these multiple transactions

represent a considerable share of the size of

the translation market. While a freelance

translator may in a particular case receive

$0.08 for each word he translates, the ulti-

mate customer may be paying $0.24 — three

times more.

To adequately represent the focus of

translation vendor web services, it is appro-

priate to split the IDC category of the trans-

lation sector into an actual translation service

part and an overheads of transaction man-

agement part. IDC projects an average

annual growth for the translation sub-

sector of 6.5%. With the emergence of web

services standards, it is likely that overhead

expenses will even out over the next couple

of years. Workflow automation, e-procure-

ment, supply chain automation and the

arrival of new marketplaces will help to

bring transaction costs down.

We assume that at our base point rev-

enues of the translation sector are equally

divided between actual translation service

costs and overheads. By 2004 we will al-

ready see the result of transaction automa-

tion and a reduction of overheads. The

proliferation of translation web services stan-

dards will help to bring transaction costs

further down over time.

Cascaded supply chainThe translation market has grown into

a multilayer industry with many interde-

pendencies between the various players.

First, there were the individual profession-

als or freelance translators. They are listed

in the registers of national translator asso-

ciations, but they receive their work usu-

ally through personal contacts and yellow

pages. Since the 1950s and 1960s many

agencies have been established that broker

the services of freelance translators. They

receive their work through yellow page

advertising and now more and more

through the internet. Since the 1980s spe-

cialised software localisation companies

have been established that select suitable

translation resources from the freelance and

agency supply base. Since the 1990s a wave

of consolidation started at the top of the

translation market and a handful of global

multilingual vendors have emerged. They

select and buy the translation services from

agencies and from the smaller specialised

localisation vendors.

A cascaded supply chain has evolved that

is made up of many actors who often play

similar roles in the production of the ser-

vices. At every new link of the supply chain,

project management tasks, quality assur-

ance, procurement, file handling and other

tasks will be performed. But as no industry

standard process definition exists, it is

inevitable that a tremendous amount of

overlap is hidden in these trails of subcon-

tracting and outsourcing.

The table below illustrates a typical out-

sourcing model and the tasks being per-

formed on every level.

Although this supply chain seems to be

an obvious candidate for definition and

automation, very little has been achieved

until now on this level in the globalisation

market. Several of the largest suppliers in

the market have developed workflow

automation products as well as supply chain

and customer portals for online project mon-

itoring and collaboration. However, these

packaged products have not found very

favourable customer response because cus-

tomers in general fear to be locked into a

proprietary service environment.

Customers’ buying criteriaCustomers of multilingual communica-

tion services require a rare combination of

Continued from page 9

Model of outsourcing and tasks performed at each level

ACTOR TASKS PERFORMED

Client Procurement, Vendor Management, Quality Assurance,Translation Memory Management, Project Management

MLV Procurement, Vendor Management, Quality Assurance,Translation Memory Management, Project Management

SLV Procurement, Vendor Management, Quality Assurance,Translation Memory Management, Project Management

Agency Procurement, Vendor Management, Quality Assurance,Translation Memory Management, Project Management

Translator Translation, Translation Memory Management, Quality Assurance, Project Management

Web localisation CLA Software localisation

Translation

SEPTEMBER 2002 LOCALISATION.central 11LOCALISATION FOCUS

the highest level of language quality and

topic specialisation with the lowest level of

overheads and the fastest turnaround. They

want to be able to change their specialist

providers if better quality, better turnaround

or lower cost can be obtained somewhere

else. Therefore, they do not want to be

locked into a one-to-one marketplace or por-

tal from one of their service providers.

The customers’ buying criteria can be

listed as follows:

● Quality of devoted specialist translators;

● Non-proprietary portal for procurement,

vendor management, project management

and monitoring, collaboration, quality assur-

ance;

● Lower cost;

● Faster turnaround;

● The portal or functionality must be inte-

grated into other enterprise applications, like

a procurement platform, intranet, extranet

to allow direct access to all authorised users

within an enterprise;

● The service must offer interoperability for

all translation memory, machine translation,

content management and other enterprise

applications.

To meet all of these customers’ buying

criteria the globalisation market requires a

fresh approach, which comes down to a

separation of the delivery of the high qual-

ity services by the devoted professionals on

the one hand and the facility of a highly

automated and standardised infrastructure

for transaction management on the other

hand.

Market playersThe localisation, translation and inter-

pretation (LTI) and cross-lingual application

(CLA) services market is populated by mul-

tiple layers of service providers and only a

few pure technology providers. The table at

top right provides an overview of the ser-

vice providers in categories of size from large

to small.

A few service providers have entered tech-

nology development, especially at the top of

the supply chain, but most have withdrawn

from it. Customers would not buy their tech-

nology for fear of being locked into one ser-

vice provider’s technology.

The diagram above right illustrates the

rapidly growing importance of cross-lingual

applications services. The percentage of cross-

lingual applications services of the total size

of the localisation, translation and interpre-

tation services market is rising from 1.5% in

2001 to 2% in 2004 and to 3.3% in 2006.

The human translation segment is grow-

ing at only 6.5% annually, thus indicating

the acute need for leveraging software in the

form of transaction management and trans-

lation technology. The industry’s growth is

currently suppressed by the complexity and

cost of managing the translation supplier

base. This reality is stimulating the growth

of translation automation and transaction

and workflow automation.

Hybrid solutionWeb services potentially have a much

greater impact on the translation market

than any other development before, includ-

ing the consolidation and creation of large

MLVs since 1995. The route to scalability

and efficiency does not seem to go through

mergers and acquisitions as it does per-

haps in many other industries. What the

translation market needs is a hybrid solu-

tion, consisting of an open infrastructure

on the one hand and a multitude of highly

specialised human resources on the other

hand. The best quality can be achieved by

those who are nearest to the work being

produced: the translators, the reviewers

and the writers. The most efficient service

management system is an automated open

industry infrastructure.

Web services have the potential to cre-

ate such an intelligent and open infra-

structure that automatically links with

high-quality resources and interoperable

tools and databases.

As some people say: “This industry goes

around in cycles. …” ■

CATEGORY REVENUES NUMBER DIFFERENTIATORS

Global MLVs $30M - $100M < 10 Offer all languages, focus on IT sector, Other services: testing,consulting

MLVs – SMEs $5M - $30M < 50 Offer most languages, specialise in one or two sectors

SLVs $1M - $5M <100 High quality in one language,strong position in “home” market.

Agencies $100K - $1M <5,000 Every major city in the world has between 5 and 20 local agencies.They “own” their networks of freelancers, and have a few strong long-term clients.

Freelancers <$50K <200,000

Overview of service providers from largest to smallest

About the authorIn January Jaap van der Meer and some partners, including Aquarius, ESTeam and Telefonica,formed the plan for COLIN: Collaborative Localisation Industry Network. COLIN was submittedunder the European Commission’s eContent programme with a request for funding. COLINwas one of several built on the deployment of web services. Unfortunately, the proposal didnot meet the Commission’s funding criteria.Until the end of 2001, Jaap van der Meer was President and CEO of ALPNET. Since his departure from ALPNET upon the merger with SDL, he has been advising various technologycompanies.Since his debut in the localisation market in 1980, he has been a great advocate of transla-tion automation. His first company, INK International, launched the first desktop translationmemory software. He also published the Language Technology magazine. In 1990 he launchedthe idea for a localisation industry association, and he funded the establishment of LISA. In1999 he co-founded the SAE TopTec Multilingual Automotive Information conference. At ALPNET he spearheaded the implementation of an end-to-end automated localisationprocess, including machine translation, centralised translation memory and automated workflow.Currently Jaap van der Meer is consulting with SYSTRAN on the promotion of machine transla-tion in the enterprise market. He is also a member of the Translation Vendor Web ServicesSteering Committee.

LOCALISATION.central SEPTEMBER 200212 LOCALISATION FOCUS

Cost control in the current economic climate is

critically important to maintaining an effective

global presence for your products and services while

remaining solvent. In this article, localisation experts

KPS outline 10 steps to effective cost control to be

considered.

WORKING with a number of software publish-

ers, KPS has seen a typical internal/external cost

split of 60/40, and the external 40 is in turn split

by the translation vendor 30/70, where 30 is

the management and engineering cost and 70

is the cost of translating the words. So out of each $1 only 28 cents

is spent on translating the words!

Remember that the margins for straightforward localisation tasks

based on a price per word (of culturally adapting content from one

language into another using the correct terminology and style) is

low (5-10%), so seeking to negotiate lower prices with localisation

companies is not feasible without an unacceptable loss in quality.

Instead, the trend for these companies is to look for revenue from

per hour activities (DTP, engineering) and other services (such as

globalisation assessments and project management). In addition,

your company may be under internal pressure to reduce head count

or to operate under a head count freeze.

You should seek to address lower costs of localisation and con-

trol head count while increasing or maintaining quality and decreas-

ing time to market. An underlying tenet of these solutions is the use

of automation, appropriately deployed with common sense. (Note:

For the sake of simplicity, we assume that the source language is

English. However, these techniques can be adopted for any source-

target language(s) combination.)

One:Educate your colleagues to treat localisation as a busi-

ness process like any other activity critical to the bot-

tom line of your company. Any file can be translated, but not by a

standard process. Start with basic acceptance criteria for localisa-

tion. 1) Is the product internationalised, 2) can the source files or

externalised strings be safely and easily translated by a commercial

and easily available process and 3) can the localised strings and siz-

ing coordinates be recycled and reused as the source files change

between updates. No one-off throwaway localisations please!

Two:Volume — typically documentation projects repre-

sent the biggest localisation expenditure. Whether it

is online help or documentation, the per word cost of localising

these, combined with any related engineering costs can represent

80% of your budget, providing a serious barrier to entry. Start by

promoting the idea of asking two related questions: 1) should there

be documentation written about this? 2) Should it be localised? The

answers to these questions are not the same, and it’s wrong, per-

haps even expensive, to believe that everything written in English

must be localised.

Once you are satisfied that documentation should be written

about a product (for example, when the product’s user interface

cannot be made more intuitive), adopt a selective translation

approach. Structure your documentation’s information into dis-

crete content baskets based on a small number of audience types.

For example, three categories of user could be adopted: installer-

implementer; advanced super user; and self-service user.

● The first type may not need to be translated since it deals with

installation, implementation, consulting and development infor-

mation that is best left in English, dealing with APIs, code samples,

development tools, techniques and so on.

● The second type could deal with infrequently performed tasks

such as changing business parameters. It has a smaller audience.

● The third is typical task information — for example, “how-to”

procedures. It has a wide audience. This needs to be translated if

the product UI is not self-explanatory.

Using XML facilitates this structure and management decision.

Define your data using a DTD (Document Type Definition) that

allows audience attributes to be applied to elements, and then use

XLST to transform the selective elements to a localisable file for-

mat like XLIFF. Since this data definition is not only logical but

flexible, it can be varied according to market requirements and

extended as market conditions vary.

Three:Sorting — alphabetically sorted files are painfully

costly. This is because other languages contain

different characters and numbers of characters from English. Once

translated, the files need to be resorted, extra characters added,

redundant ones deleted and so on.

Have you ever tried sorting HTML alphabetically? It’s a costly

and time-consuming exercise! Instead, externalise sortable content

into a database and use the NLS collating sequence to sort the file

Documentation localisationcosts — who’s managing whom? Ten things to think about as you watch NASDAQ dive

Audience Localised % of total volumeInstaller/implementer Never 30%Advanced super user Sometimes/as market requires 30%Self-service users Yes 50%

Using a simple technique like running the grammar checker in yourword processor may not only catch spelling errors but also detect prob-lematic sentence constructions.

SEPTEMBER 2002 LOCALISATION.central 13LOCALISATION FOCUS

dynamically, or use XML and XLST to do the same.

Four:Reuse of previously translated content is critical.

Develop solutions that not only allow fast reuse but

also secure that reuse by maintaining context. Place identifiers on

data at a level of granularity that is appropriate, for example on

paragraphs.

<ID_ABC123455-->To save the file, press CTL+S<ID_ABC123456-->To run the printing process,

press CTL+P<ID_ABC123457-->Press OK

Use your translation memory tool to analyse both these IDs as

well as the strings’ content so that you can safely update automat-

ically without loss of context: for example, the placeables in

TRADOS Translator’s Workbench. IDs can be either from the data-

base (ideal) or from the authoring environment: for example, IDs

in Frame+SGML. Lever internally — do not externalise the updat-

ing of unchanged content.

Five:Grammar check your work before translation. This

sounds obvious, but it isn’t uncommon that content

is sent for localisation before it has even been spell checked. Using

a simple technique like running the grammar checker in your word

processor may not only catch spelling errors but also detect prob-

lematic sentence constructions: for example, the use of the passive

voice which can only otherwise be detected by using expensive con-

trolled authoring techniques. Even such a simple step means that

some editorial changes can be eliminated before translation starts,

and the English content is better too.

Six:Graphics and images files should be separated from text

to allow the text to be added to the translation memory.

Graphics work is expensive. Treat callouts, captions and explana-

tory text sensibly. Use numbered callouts with corresponding num-

bers on graphics and add the text into the main document. By

separating the text from the graphic, the same graphics file should

be reusable in all language versions without the need to change it.

Only the language of the text will change.

Seven:If you have accessibility requirements, handle this

sensibly. Rewriting HTML documentation is not

required. For tables, you can add summary captions that say “this

table is for formatting only” and for graphics the ALT tag can say,

“this picture is described in the text that follows”. These simple

phrases can be easily localised once and then automatically inserted

into the source code without fear of loss of context or the need to

write and localise expensive descriptions. You can tender product

or service for government and US or European Union contracts

and meet statutory accessibility requirements using these simple

techniques.

<IMG SRC="genericgraphic.gif" ALT="This image isdescribed in the text that follows><TABLE SUMMARY="This table is for formattingonly">

Eight:Use open standards so that your files can be eas-

ily translated by any tool or localisation vendor.

XLIFF is an excellent medium for this purpose. It also allows easy

customisation and the need to develop proprietary filters.

<trans-unit id="bigirishcolumn_145"restype="title" maxwidth="90" size-unit="byte"><source xml:lang="EN">Database manager</source><target xml:lang="GA">Feighlí feasa</target><alt-trans><target xml:lang="GA">Gocamán na ngiotán</target></alt-trans><note>Manager means administration tool - nota person</note></trans-unit>

Nine:Redesign your localisation process so that your

automation is central to reusability, internal check-

ing and validation. Use the power of technology to do as much work

as possible internally by using existing resources or fewer staff. Once

a file leaves your organisation and is sent to localisation for any task,

you will have to pay heavily even for tasks that those companies

have themselves automated (such as updating or leveraging files).

Retain non-language critical tasks such as project management, lever-

aging and quality checking inside your own organisation, and use

solutions designed to perform these tasks safely and automatically,

24 x 7, with minimum manual intervention.

Ten:Terminology — sign off on the terms that will be used

in your translation up front before the actual localisa-

tion progresses too far. Most terminology change requests are based

on a matter of taste, usually by people who don’t have to pay for

the changes. Take control. Extract key terms from the product or

service, localise them first and then have your customer sign off

that they are acceptable. Do a market test with the terms on sam-

ple customers. Most major companies invest heavily in terminol-

ogy. Even if translation isn’t 100%, the documentation is still usable

and the UI perfectly navigable. Get a proper perspective on termi-

nology. Most documentation is read only once. In the days when

there were spelling errors on the front page of the Guardian, no

one complained that they didn’t get the message. String extraction

techniques can improve localisation quality, improving consistency

and usability, giving you that competitive edge. ■

For more information, email info@keyperformancesolutions.comKPS was founded in January 2002 with a vision to be the leadingexpert of software globalisation and process knowledge worldwide.KPS provides consultancy services to the G11N, I18N, L10N andeContent markets.

Visit the LRC at itsnew URL:www.localisation.ie

LRC’s newURL

LOCALISATION.central SEPTEMBER 200214 LOCALISATION FOCUS

WHEN I entered the localisation in-

dustry eight years ago, the first thing

I learned was that localisation is all

about languages (and long hours of

project work). Localisation, I was told,

is translating words and sentences from one language to

another – and “a bit more”. But nobody could actually

tell me precisely what this “a bit more” would be.

After a while we localised Encarta (an electronic ency-

clopaedia) into German, and it instantly became clear.

Pure translation would not have been enough. There

was, for instance, an article about police, showing Amer-

ican police cars and discussing the special situation of

the American police. Now, this would have been use-

less in a German encyclopaedia — the content also had

to be translated and adapted to the culture.

This example became my personal definition of local-

isation — localisation does translate, but not only words.

It translates between cultures. It transports cultural con-

tent from one culture to another. It all starts with trans-

lation, but at the end of the day it’s all about culture.

Language is just the vehicle.

And this is reflected by the people you meet in local-

isation — people from many countries and cultures, liv-

ing and working together, speaking different languages.

Many of the people I have met in localisation constantly

travel between cultures themselves. They live abroad,

are married to people from other countries and work

for companies running offices all over the world. The

localisation community is, at least in my eyes, almost a

miniature model of the global village.

And it faces similar problems. One thing I never fully

understood is that large localisation companies are hav-

ing the same internal communication problems as any

other international company in the world. Localisation

is supposed to know about the subtleties of culture, but

the company I was working for experienced all the cul-

tural issues itself. Miscommunication caused many prob-

lems, delays and excitement. The Germans did well with

the Irish. The Italians liked the Spanish, but nobody

liked some specific other country and everybody smiled

about the Americans, although everybody was doing

his or her best and nobody was really less competent.

It all came down to these basic questions. How can

we communicate with native speakers without having

language disadvantages? How can the meaning of some-

thing be transported from one culture to another in such

a way that it is entirely comprehensible and cannot be

misinterpreted?

That became the core issue of localisation to me: to

talk to each other efficiently in different languages and

across cultural borders. Communication between some-

times very different cultures. It is not a matter of what

is said. It is more a matter of how something is expressed.

It’s a cultural thing.

I had always been looking at localisation as some kind

of a magical creature, even before I knew why. Localisa-

tion can in theory build bridges across these cultural bor-

ders. What I have experienced in a comparatively small

scale in a corporate environment is what happens to the

entire world. We want to bring the world together — into

the “global village” — or at least into something “big-

ger”. This is one of the oldest dreams of mankind (or our

leaders): uniting nations into a larger political entity. We

have many examples, good and bad, in history: Alexan-

der, the Roman Empire, Charlemagne, Genghis Khan,

Napoleon and Hitler. But history also shows that these

large political entities never worked. They only survived

for a longer period where cultural identity could be pre-

served (or achieved) at the same time.

Culture was always the tricky factor in the equations

of politicians. Borders can be deleted on maps. Official

languages can be introduced, but people keep thinking

and living their own way.

Let us assume that mankind has evolved a bit. Let us

assume that political leaders cannot use sheer force any

longer to unite nations and compel a collective culture.

Then the best way to achieve a global village, or at least

a larger social entity, is to build bridges between cul-

tures, to transform and transport content between cul-

tures, to make this content understood by different

cultures, to interconnect them. This is a development in

favour of localisation.

That is where I see one of the important roles of local-

isation in the future — translation and mediation

between cultures, building bridges. Localisation is pro-

viding the vessel in which cultural content can be trans-

formed and carried into the world and to the people,

enabling them to participate and keep their own iden-

tities at the same time.

Despite efforts to build new political structures, like

a new Europe, other social structures are emerging. Cul-

tures are coming together because of improved com-

munication. Already, people can keep their own cultural

identity and participate in a bigger, common cultural

movement at the same time. This must be frightening

for politicians, but it is a big chance for society — and

an even bigger challenge for localisation.

As I write this, the better part of all localisation work

is dedicated to translating software, IT manuals and

similar products. At the same time, the localisation indus-

try is undergoing a major change. The effort to localise

complex software packages “just in time” is constantly

growing from year to year. Staff is more expensive, and

part of the IT industry is in Queer Street (and will not

grow much more).

I am well aware of the fact that some people say local-

isation is just another term for specialised IT industry

translation services. The origin of this opinion is the LISA

definition, which puts a certain emphasis on software and

hardware components. But this is a weak defence for a

lack of growth and adaptation. Of course, translation is

Localisation in the Global VillageMany people say that localisation is facing hard times, not

only because the main market — the IT industry — is in

deep trouble, but because the world is getting smaller.

English is becoming more and more a global language and

translation and localisation will be less crucial to the

software marketplace. But is that true? Or does localisation

only need to change its point of view? DETLEV

HOPPENRATH considers the options.

SEPTEMBER 2002 LOCALISATION.central 15LOCALISATION FOCUS

usually the foundation of localisation. Localisation is mostly

done by translators. But these translators don’t just trans-

late words and sentences but also their idiomatic mean-

ing and the meaning between the lines. Translators who

know about the target market and culture adapt the con-

tent to this target environment. Localisation is a univer-

sal business, and if you look at the many examples, it

becomes obvious that localisation is not limited to the IT

industry. And there are still many new markets for local-

isation on the way to the global village.

But we are far away from that. Localisation has con-

centrated on the IT industry as a (former) area of eco-

nomic growth, so much so that other markets have been

neglected. After working in the localisation industry, I

became CEO of companies in other markets. And I was

surprised how little was known there about localisation

and how proper knowledge about cultural subtleties can

support a business (or politics). Speeches, marketing

material, business plans — all this is usually translated

on the fly, sometimes by semi-professionals, often incon-

sistently and without regard to fine differences. In many

places even the art of translation has deteriorated to a

literal dictionary exercise.

Look at other areas. Who is localising all the EU doc-

uments? Who is transporting the visions of Greenpeace

to other cultures with apt knowledge? One might assume

it must be one of the big players in the localisation world.

Wrong. Not even one of the small players. No player at

all. The EU has built up its own translation and locali-

sation service (although a very small part of the work is

outsourced). I assume Greenpeace did the same.

Large markets like automotive, aerospace, printing

and the public sector are not properly covered by local-

isation. I know that this statement will cause protest

among all the many small localisation companies. In my

country alone are there hundreds, maybe even thousands

of them. They are serving these markets, and their exis-

tence is the proof that the big players have neglected these

markets. Unfortunately, the many small companies have

no lobby and no voice and are not visible as a market

force.

I wonder why the localisation world did not pro-

fessionally cover these markets — markets where local-

isation can be done without huge effort and complex

IT environments, at lower cost. The larger players have

long understood themselves as high-tech outfits, even

going so far as to spend significant percentages of their

revenues on tools and process development, without

getting a return, eventually leading to identity prob-

lems. Would they rather be technology companies or a

service providers? Nobody so far has succeeded in being

both, in the same way that IT conglomerates like

Microsoft and SAP have, and it’s unlikely that anybody

will.

And the main cash-cow, the IT industry, is growing

more slowly and in some cases declining. It is, however,

getting more complex. Localisation for this area becomes

more and more a problem in terms of effort and mar-

gin. Leading localisation companies had to scale down

to be able to continue their services.

The main problem of localisation seems to be that

nobody can say what it really encompasses and where the

markets are. This becomes obvious if you look at the fig-

ures. The estimates for the annual global localisation vol-

ume range between $2bn and $10bn. A variation of 500%

doesn’t suggest everyone shares the same vision.

At the same time there are still so many areas where

cultural issues are important or where cultures are emerg-

ing and growing together. China is opening to the West;

governments, NGOs and large international organisa-

tions like the Red Cross are just a few examples. They

are in need of good value localisation services. Instead

of all the effort and machinery that stores and keeps track

of the million elements in an IT project, they need the

people who know how to use the power of language to

transport an idea or a vision from one culture to another,

very different one.

Many people say this is no longer necessary. English

is becoming the global language, and everyone will speak

English sooner or later. I don’t believe this. First of all,

language is one of the most important elements of cul-

tural and personal identity. We see in history how impor-

tant a distinct language is. Otherwise, we would already

have had one language for hundreds or thousands of

years. But there are still not only many languages, but

many distinct dialects of these languages. Every com-

munity is building its own language to express its iden-

tity. Visit a chat-room on the internet to see how even

on a global communication platform new languages are

evolving.

At the same time there are (and always have been)

auxiliary languages (like English) that are helping to build

bridges. But they will never have the effect and the mean-

ing of the native language — a large area for localisa-

tion. According to Forrester Research, international

companies are losing up to $10m in potential sales a year

if they do not present their products in local languages

(on web pages). Statistics have shown that people pay

more attention to information that is presented in their

native language.

I am very curious what the future of localisation will

be. One thing is certain: localisation will change. It actu-

ally is in the middle of a continual process of transition.

Project management has changed a lot (and become much

more professional). This is an asset. Some localisation

companies have built up professional sales departments.

This is an asset too, because this is their vehicle to enter

new markets.

There are some “active debts” too. There is almost

no marketing, PR and communication — most compa-

nies “in the wild” don’t even know localisation exists.

Margins with existing customers are going up and down,

and everybody knows that some will not survive. I know

only a few other industries where so few players keep

buying and selling one another, trying to adapt to a mar-

ket situation nobody can adapt to.

I can’t look into the future, but I hope localisation will

continue to be the magical creature I love. I hope it will

play a bigger role in communication between cultures.

I hope it will build the bridge to a global village. ■

Detlev Hoppenrath is a management con-sultant and freelance journalist in Munich,Germany. He has several years of firsthand experience in middle management ofa leading localisation company. After that,he was partner in a small localisationcompany and CEO of several other companies.■ His contact address is Detlev.hoppenrath@gmx.net

About the author

LOCALISATION.central SEPTEMBER 200216 LOCALISATION FOCUS

LOCALIZER not only provides

facilities to build multilingual

websites (which we will see

later); it also helps to develop

multilingual Zope products.

An example is Localizer itself, which is a

multilingual product. All its interfaces are

internationalised, and translators have pro-

vided local versions for languages like Span-

ish, German, Basque or Hungarian.

The solution provided by Localizer for

this task is built around the GNU Gettext

tools, which are the de facto standard in the

free software industry. Projects as big as

GNOME and KDE use it.

GettextWith Gettext a special markup is used in

the source code that identifies the translat-

able messages. Adding this markup

represents the part known as international-

isation. The localisation part is done in sev-

eral steps.

First, the source code is parsed using a

command line program, which generates a

text file that contains the translatable mes-

sages. These files are known as PO files.

Second, these PO files (one per language)

are edited by the translator, who adds the

message translations. One of many avail-

able tools can be used to do this task (Emacs,

KBabel and so on).

And finally, the PO files are compiled to

generate the binary files that will be used at

run time. These binary files are known as

MO files.

The markup used in the source code is

actually a function that receives a message,

looks up its translation in the MO files and

returns it.

Using Gettext allows Localizer to profit

from all the expertise and the tools devel-

oped around it. This includes sophisticated

PO files editors for translators, support for

plural forms, fuzzy matching and so on.

WebsitesZope provides a web interface that can

be used to build sites through the web. This

interface is known as the management

screens, in which the different kinds of

objects that can be used within Zope (tem-

plates, folders, images, files etc.) can be man-

aged.

What most Zope products do is to pro-

vide new kinds of objects

instead of the standard ones.

These objects include their own

management screens, which

integrate smoothly with the

rest of the system.

Localizer adds four new

types of objects, which include

their own management screens.

They provide ways to interna-

tionalise the website, to man-

age multilingual content or to

customise the language nego-

tiation policy. This way Localizer allows you

to build multilingual websites via the web.

We will consider these four objects and what

they do in turn.

The Message CatalogThe first one is called the Message Cata-

log and, as its name suggests, it follows a

similar approach to Gettext. It is just a way

of storing messages and their translations,

with a web-based interface to manage them

(See screen shot below). To get the transla-

tions, it provides an API similar to that of

Gettext.

To manage the Message Catalog, it is pos-

sible to export and import the data to and

from PO files, so sophisticated editors can

be used instead of the web interfaces. The

Message Catalog is also accessible through

FTP. In this way it is possible to edit mes-

sages directly with any tool that supports

FTP: for example, the Emacs editor.

Message Catalogs are typically used to

internationalise the user interface: for exam-

ple, the navigation system of the website,

the forms and so on.

ContentContent has different requirements and

thus requires a different solution. For exam-

ple, it needs to be indexed, so it can be

searched. It also needs to use workflow tools

to manage it. To satisfy these and other

requirements, Localizer provides the Local

Content object.

The key difference between the solution

for the interfaces and the solution for the

content is that in the first one all the trans-

lations are stored in a separate database (the

Message Catalog), which is asked for the

message translations.

In local content objects, however, both

the original data and the different language

versions are stored together. This allows

translators to perform translations in con-

text and also satisfies indexing and work-

flow requirements.

Everything we have seen until now deals

with text, either for the user interface or for

the content. But there is more to a website

Localizer is a tool that provides facilities to build multilingual websites. It is based on the Zope web

application server, which is an extensible environment: that is, developers can provide modules that

add new features and services to Zope. These modules are called products. Localizer is one of these

products. There are many Zope products related to internationalisation, and Localizer has become

the most popular — the reference among them for anyone who needs to build a Zope based website

in several languages. As an example, one of the last websites that has used Localizer is

EuroPython.org, which was created to organise the first European congress about the Python

programming language. JUAN DAVID IBÁÑEZ PALOMAR explains how it works.

■ The Message Catalog webinterface.

Joining the Localizer community

■ Juan DavidIbáñez Palomarhas developed away of building multilingualwebsites thatwon’t cost any-one the earth.

SEPTEMBER 2002 LOCALISATION.central 17LOCALISATION FOCUS

than text. What happens, for example, if

you want to localise an image or to show a

date in the local format. In these cases, Local-

izer provides the Locale Folder object, which

is a generic solution to localising anything.

It is a folder that can store different locale

versions of any kind of Zope object and pro-

vides the correct one (based on the language

negotiation rules) when requested. This solu-

tion can be used to localise images, dates

and so on.

In a multilingual website, each time the

client requests a web page the server has to

decide in which language to serve the page.

This is known as language negotiation.

By default, a Localizer-based website will

only consider the browser configuration (the

Accept-Language HTTP header) to choose

the language. But Localizer provides a class

(named Localizer, like the product itself) that

allows you to customise the language nego-

tiation policy.

This class provides several built-in helper

methods for several common tasks. One of

them, for example, stores the language pref-

erence in a cookie, and another displays a

selection box that allows you to change this

cookie. Another built-in method allows the

specification of the language in the URL,

which is the best solution to avoid caching

problems and for the search engines.

Developers can also implement their own

methods to define the language negotiation

policy so that it fits their requirements. These

settings are defined in the Localizer object.

Usually there is only one in each website.

The policy can be changed at any time with-

out affecting the rest of the website, so the

language negotiation is well separated from

the rest of the system.

This also means that Zope product devel-

opers don’t need to worry about it because

language negotiation is something specific

to each website.

Several developers have used this set of

features to create multilingual websites.

The Europython conference website was

created that way. A small company is also

currently developing Basque-Spanish

bilingual sites combining those strikingly

different locales in content-rich websites.

These sites show that the main challenge

for developers is one of organisation. When

entering a term via the search dialog box,

should the user get results listed for all the

locales or just for the locale he or she is cur-

rently using when clicking the button? Local-

izer provides a framework powerful and

flexible enough so that different organisa-

tional models can be implemented. For

instance, websites with multiple-locale organ-

isations are possible (a two-locale website,

with a specific section with four locales).

Full Unicode support is now under

development and will be available in the

next release. (It will probably already

have been released upon publication of

this article.)

It can be seen in the screenshot on the

previous page, which shows a message in

English and its Russian translation. This is

achieved by sending the data to the browser

in UTF-8 encoding.

Internally, the text is stored as Unicode

Python strings, which hide the implementa-

tion details from the developer. When these

Unicode strings are serialised, to be sent to

the browser, or stored in the database for

example, they are encoded in UTF-8.

ExtensibilityWe have quickly seen the different fea-

tures provided by Localizer out of the box,

which can be used by webmasters, content

managers and others. No advanced pro-

gramming skills are needed to build a sim-

ple multilingual website.

But beyond this, developers can create

their own Zope products that extend Local-

izer by using the API: for example, to add

new features or to build custom multilingual

objects.

There are already at least two products,

named CMFLocalizer and Base18, which

add new features to build multilingual web-

sites. The first one, CMFLocalizer, adds some

features specific to other Zope technologies

— the Zope Page Template language and

the Content Management Framework.

Meanwhile, Base18 explores new strategies

to manage multilingual content.

Free softwareLocalizer is free software. It is released

under the terms and conditions of the GNU

General Public Licence. It can be down-

loaded free and it can be used, modified

and redistributed without limits or the need

to pay royalties. The only conditions are

those specified in the licence, the aim of

which is to guarantee these rights to the

user.

I believe this is one of the key reasons for

Localizer’s success — the community sur-

rounding it that has always provided valu-

able feedback and patches, without which

Localizer would not be here.

Zope itself, the framework on which

Localizer is based, is also free, as are most

of the products that extend it.

Past, present and futureI started developing Localizer in the Uni-

versity Jaume I of Castellón in Spain, located

in a bilingual region (Catalan and Spanish

are the two languages there). Localizer was

developed to support the multilingual

requirements of the LLEU project.

Development started at the beginning of

2001. Since then it has been actively

expanded and has become the most popu-

lar tool for building multilingual websites

with Zope. It is being used in countries like

Japan, China, Russia, Kazakhstan, Ukraine,

Canada, France, Spain, Germany and

Belgium.

The community has grown around it.

While I am writing this, 37 people have sub-

scribed to the Localizer mailing list. It has

been presented in several conferences. Many

multilingual websites have been built with

it as people start to get more involved in its

development. As an example, a project is

under way at the University of Deusto that

will add workflow and XLIFF support to

Localizer.

The future looks promising. Future ver-

sions of Localizer will include features cur-

rently available in translation memory

systems. Splitting text into simpler sentences

for translation, the use of external databases

that use fuzzy matching to provide sugges-

tions to the translators, the interconnection

with automatic translation systems, the sup-

port of standards such as XLIFF or TMX,

workflow, creating multiple versions and a

lot more will soon be within the scope of

future releases.

Our goals are to position Localizer as an

industrial strength tool to build multilingual

websites, to manage multilingual content

and to become a standard reference in the

free software community and in the global-

isation market. ■

Juan David Ibáñez Palomar is a softwareengineer, he can be reached at j-david@noos.fr

ReferencesLocalizer’s home — http: //www.j-david.net/localizerZope — http: //www.zope.orgPython — http: //www.python.orgGNU Gettext — http: //www.gnu.org/software/gettext/CMFLocalizer — http: //zope.org/Members/fafhrd/CMFLocalizerBase18 — http: //www.nexedi.org/software/Base18.stxGNU GPL — http: //www.gnu.org/licenses/licenses.html#GPLUniversitat Jaume I — http: //www.uji.esUniversity of Deusto, DELi research group — http: //www.serv-inf.deusto.es/abaitua/deli/ WEBSITES BUILT WITH LOCALIZER: LLEU project — http: //lleu2001.uji.esEuroPython website — http: //www.europython.orgOndarroa.net (Basque/Spanish) — http: //www.ondarroa.net

LOCALISATION.central SEPTEMBER 200218 LOCALISATION FOCUS

The development paradigm has

changed! With the introduction of

Microsoft .NET technology, build-

ing and deploying internet and

mobile applications has become

easier. Using one technology

framework, an easy programming

language, leveraging open

standards like XML and SOAP,

and integrating it all into one

seamless development environ-

ment, Microsoft has released its

most ambitious development

technology yet, says TONY

O’DOWD.

WHILE fundamental

approaches to ap-

p l i c a t i on de s i gn

remain somewhat

consistent with the

approach traditionally chosen by Desktop

Application developers, the LSP (Localisa-

tion Service Provider) community faces a

daunting challenge of upskilling and retool-

ing their localisation teams while embrac-

ing this new Microsoft technology. For

localisation service providers, developing a

localisation strategy for .NET applications

is critical if they are to secure future locali-

sation projects from their customers.

XML technologyAt the core of the Microsoft .NET frame-

work is XML, an open Standard supported

by almost all platforms and all major IT

players such as IBM, LOTUS, Sun Microsys-

tems, Oracle and Novell. It offers the abil-

ity to store data easily and describe the

relationships between this data in a com-

mon, platform-independent vocabulary.

Microsoft has deployed XML as its replace-

ment for the ubiquitous Windows resource

file formats (RC, DLG) for dialogs, menus

and strings. These fundamental UI compo-

nents are now described in a new XML doc-

ument, referred to as .resx documents.The

Alchemy .NET Visual Component provides

plug-ins within Alchemy CATALYST 4.0 that

provide a visual localisation solution for the

Microsoft .NET technology.

WHILE the Microsoft .NET framework

introduces new technologies making the

implementation of globalisation easier to

develop, this technology also redefines local-

isation workflows.

Development overview.NET applications can be developed in

either C# or VB .NET using the Microsoft

Visual Studio .NET Development environ-

ment. Other than procedural and variable

declarative statements, C# and VB .NET are

almost identical in their syntax, and the

process of building .NET application is sym-

metrical.

When an application is created, all the UI

resources (strings, dialogs, menus etc.) are

stored in the applica-

tions’ code files. This

makes localisation

costly and expensive,

as the application has

to be rebuilt for each

language version re-

quired.

To resolve this

problem, Microsoft

has created a solution

whereby the developer

can select the resources

to be translated, and

these will then be

stored in an external resource file, called a

.resx file. These .resx files can then be com-

piled into a resource-only Dynamic Link

Library.

By separating the code files from the UI

resources (.resx files), it is relatively easy to

build multilingual versions of a .NET appli-

cation.

A closer look at .resx filesIn general, for each application resource

that is selected as translatable, VisualStudio

will create one .resx file.

However, the developer also needs to spec-

ify the languages this resource will be trans-

lated into! And for each language a new .resx

file is also created by VisualStudio .NET.

So, suppose a developer selects a Win-

Form (new name for a dialog box) for local-

i sa t ion in to French and German,

VisualStudio will build three .resx files.

These additional .resx files are highly opti-

mised and contain only the differences

between the Base .resx and the individual

language .resx files.

Compilation processEach .resx file is compiled into a .resource

fi le, which in turn is l inked into a

.resource.dll satellite file.

.resx file

.resource file

.resource.dll file

These satellite DLLs contain only the

resources for a single language version of a

.NET application, and they are stored in pre-

determined directories by the Visual Studio

compiler.

When a .NET application is launched on

a client workstation, it will determine the

language of the client operating system and

load the relevant language satellite DLL.

This allows multiple workstations, all run-

ning different language versions of Windows

to launch .NET applications in the same lan-

guage as the workstation OS. You can see

now why the directory structure and nam-

Microsoft .NET technologychanges the paradigm

Section 1: Building .NET applications

■ Using VisualStudio.NET, you can selecteach resource thatneeds to be translated.

Base.resx German French.resx .resx

Form1.resx Form1.de.resx Form1.fr.resx

SEPTEMBER 2002 LOCALISATION.central 19LOCALISATION FOCUS 19

ing convention of the .resx files are so impor-

tant to the development process and corre-

spondingly to the localisation process!

Undoubtedly, Microsoft .NET provides

a powerful way to develop multilingual

applications at the design stage. However,

there are a number of challenges inherent in

this approach.

● Incremental resource files cannot be

translated directly. This is because the

language .resx files and the multiple Satel-

lite Assemblies are not complete copies of

the full resources used by an application.

Incomplete content makes them almost

redundant as a viable localisation strat-

egy.

● Since each WinForm requires an individ-

ual .resx file and each target language

requires a variation of this file, managing

applications with a large number of Win-

forms in many languages becomes impossi-

ble. In addition, consider the development

overheads in maintaining several language

variants. The work involved in keeping

these in-sync with each other would be

unmanageable even in a medium-sized

application.

● Microsoft VisualStudio .NET provides

no version control to help fold back

changes into the main project file of the

translations and new layouts of language

variant .resx files. This has to be done

manually and consequently will be prone

to a high number of errors and repeated

quality assurance problems.

Moreover, if changes are made to the

base application resources, they will not

be duplicated in the translated variant

files automatically. This lack of version

control makes parallel localisation of an

application during its development phase

impractical and unmanageable even in

the smallest of applications.

● Microsoft .NET applications do not

store their resources in standard resource

segments as previously used by Win32

desktop applications. This means that all

existing translation tools and technologies

currently available in the market cannot

handle the translation of .NET applica-

tions, forcing the development and LSP

community to invest in upskilling and

retooling their organisations.

Localisation of .NET applications Alchemy CATALYST provides a com-

prehensive solution for translating, engi-

neering and testing .NET applications. It

provides support for both the .resx XML

based resource files and the .resource binary

equivalent.

Existing client investment in Alchemy

CATALYST 4.0 is maintained, and retrain-

ing of translators and engineers is kept to a

minimum as they’ve integrated a .NET envi-

ronment within their already well known

and market-leading Alchemy CATALYST

4.0 environment.

Localisation process for .resxMicrosoft .resx files can be processed

directly by the Alchemy CATALYST envi-

ronment.

These files will be processed by the

Alchemy XML parser and then displayed in

an XML editor and/or a visual Winform edi-

tor. (See .resx chart on right.)Alchemy CATALYST is aware of the file

name conventions used by the .NET frame-

work so that when a translated .resx file

is extracted from a TTK1, it will be

renamed using the correct .NET language

file name.

For example, if the file Myform.resx is

translated into French and then extracted

from Alchemy CATALYST, the extracted file

name will be Myform.fr.resx. This conforms

to the RFC 1766 file name convention as

used by the Microsoft .NET framework.

Once all the language .resx files have been

extracted from the Alchemy CATALYST

environment, they should be re-introduced

into the .NET application build environ-

ment and the application rebuilt.

Localisation Process for .resourceMicrosoft .resource files can be processed

directly by the Alchemy CATALYST envi-

ronment. As these files are the binary equiv-

alent of the .resx format and contain no

XML data, only the visual Winform editor

is used to translate these file types. (See.resource on right.)

Alchemy CATALYST is aware of the file

name conventions used by the .NET frame-

work so that when a translated .resource file

is extracted from a TTK , it will be renamed

using the correct .NET language file name.

For example, if the file Myform.resourceis translated into French and then extracted

from Alchemy CATALYST, the extracted file

name will be Myform.fr.resource. This con-

forms to the RFC 1766 file name conven-

tion as used by the Microsoft .NET

framework.

The major advantage of localising

.resource files is that it is simple to build

satellite assemblies using the AL2 utility pro-

vided by the Microsoft .NET SDK3 .

AL /embed:myform.fr.resources/culture:fr /out:file.resources.dll

In the example just cited, a French satel-

lite assembly is being created.

ADVANTAGES■ You have com-plete control overchanges to each.resx file

■ Translations caneasily be put backinto an applica-tion build system.

■ For each lan-guage a set of .resxfiles is created.

■ You can quicklybuild satelliteDLLs by changingstatements in theVisualStudio pro-ject file.

DISADVANTAGES■ A large numberof .resx files needto be processed foreven a medium-sized application.

■ Each .resx fileneeds to be compiled into a.resource filebefore a satelliteDLL can be created.

■ Translators canhave access tothe XML .resx fileduring the transla-tion process.

ADVANTAGES■ No source codeneeds to be sent totranslation agen-cies.

■ Faster build system is requiredas the .resourcefiles can be con-verted into a satel-lite DLL using theAssembly Linker.

■ Since no XMLfiles are involved,translation is lesscomplex.

DISADVANTAGES■ A large number of.resource files needto be processed foreven a medium-sized application.

■ No access is avail-able to the under-lying XML sourcefile.

■ No language .resxis available for eachtranslation.

1 Translation Toolkit (TTK) is the object store that AlchemyCATALYST uses internally.2 Assembly Linker is the tool provided by Microsoft to buildSatellite DLLs.3 Software Development Kit. (SDK)

.resx chart

.resource chart

Section 2 continues on page 20

Figure above: Directory structure used byVisualStudio when building multilingual .NETapplications.

Root .Net application

de

de-AT

de-DE

de-LU

de-CH

en

LOCALISATION.central SEPTEMBER 200220 LOCALISATION FOCUS

ALCHEMY CATALYST is equally at home

processing . resx or .resource files, and the

process for translating, engineering and test-

ing these file types is identical.

The .NET localisation componentWhile Alchemy CATALYST 4.0 can

process .resx files using its internal XML

editor, it is not visual and cannot render

WinForms4 in a WYSIWYG manner. The

Alchemy .NET localisation component needs

to be activated so that these files can be

localised in a WYSIWYG environment.

When Alchemy CATALYST 4.0 is

launched, it will automatically detect plug-

in components. If the component is not acti-

vated for that l icence of Alchemy

CATALYST, a dialog will appear requesting

an activation code.

These activation codes can be requested

by sending an emai l to support@

AlchemySoftware.ie and specifying the

number of your Alchemy CATALYST

dongle.

Localising .resx filesMicrosoft .resx files are resource docu-

ments for applications developed using the

.NET framework. Using XML as a foun-

dation technology for .NET applications,

strings, dialogue forms, menus and other

user interface components is now repre-

sented within these .resx XML documents.

Selecting target languagesThe .NET Framework expects file names

to adhere to the RFC 1766 file name con-

vention. The .NET Framework uses this

standard to manage satellite assemblies and

language deployment directory structures.

When you insert files into Alchemy

CATALYST, it is advisable to select a target

language for the TTK so that these RFC

1766 compliant file names can be created

automatically on file extraction.

You can select the target language for a

TTK simply by clicking on the Translator

Toolbar and selecting a target language from

the list provided.

Extracting filesWhen a file is selected for extraction,

Alchemy CATALYST will create a file name

compliant with the RFC 1766 standard.

Localising .NET WinformsWinforms are the new term Microsoft is

using to refer to .NET Dialog boxes. To view

a Winform in Alchemy CATALYST, simply

point to the Winform name in the Naviga-

tor panel and it will be displayed in the

WorkSpace window on the right.

Double click on text in the Translated

field to edit it. The standard CATALYST fea-

tures for localising dialogues are available

— for example, resizing and aligning con-

trols, inserting hotkeys, validation features,

features for protecting untranslatable text

and so on.

Localising .NET menusWhen building a .NET application, the

hierarchical structure of a menu resource

is not stored within the .resx file but within

the C# or VB.NET source code. This

means that while the .resx file contains the

text for each menu item, it contains no

information as to the position of the item

within the menu resource.

As a direct consequence, .NET menu

resources can only be displayed as a list of

strings.

Localising .NET stringsThe most fundamental of .NET resources

are strings and these are translated, as in

.NET menus within a string editor.

XML synchronisation As .NET resources are translated within

the Alchemy CATALYST environment, the

XML source files are simultaneously updated

within the TTK. In addition, if engineering

changes are made within the XML source

files, the visual editors are also updated. This

mechanism makes it possible for engineers

to work both directly on the .resx files and

through the visual editors. ■

Section 2: Localising Microsoft .NET applications

Translators and engineers can now work in a WYSIWYG environment within Alchemy CATALYST 4.0. 4 WinForms are the equivalent of dialog boxes.

ConclusionsWith the introduction of .NET, Microsofthas redefined the computing paradigmfor developing and deploying multilingualapplications across the web.Alchemy Software Development has created a robust and elegant solution tohelp localisation service providers andpublishers localise, engineer and test.NET applications using the already wellestablished market leader Alchemy CATALYST.Using Alchemy CATALYST simplifies thetranslation process for .NET applicationsand will ensure that LSPs can developlocalisation strategies for their clientsquickly and cost effectively.

About theauthorTony O’Dowd, BSc Computer Science, previouslyGeneral Manager and Executive Vice Presidentof Corel Corporation, has over 14 years’ experi-ence as a senior manager in the localisationindustry. Before working with Corel Corporation,he was Technology Manager for SymantecCorporation and Lotus Development Corporationwhere he worked on the development of Lotus’sinternal localisation technologies and strategies. Tony established Alchemy SoftwareDevelopment in November 2000 and thenacquired CATALYST from Corel Corporation.Today Corel has a 25% equity stake in AlchemySoftware Development, based in Dublin, Ireland.Tony O’Dowd can be reached at Tonyod@alchemysoftware.ie

SEPTEMBER 2002 LOCALISATION.central 21LOCALISATION FOCUS

1. L10N is top downMANY software developers in Silicon Valley still wish the rest of

the world would be just like them: share the same values, speak the

same language, have the same taste and preferences, use the same

character set. Over the past 15 years, they could only be convinced

to internationalise their code when managers — whose salary was

closely linked to the sales figures for their international products

— wrote this into their objectives. In fairness, they have been

extremely successful. The overwhelming majority of publishers in

the digital world now make more money from the sales of their

localised products than they make from the sales of the original

product. More than 60% of the overall revenues of Microsoft now

originate from international sales. In 2001, and in Ireland alone,

Microsoft had revenues of US$1.9 billion from its international

sales.

Localisation provides access to the digital world “made in the

USA” to non-English speakers. This is good news — but obviously

a one-way street. The flow of products, of content, is in one direc-

tion only. The contribution to maintaining linguistic and cultural

diversity is — if you can still call it a contribution — extremely

limited.

2. L10N is ‘make-believe’SOME localisers define localisation as the effort to “adapt prod-

ucts so that they look and feel as if they were developed in the coun-

try, the language and the culture that they are localised for”.

A trendy website in France will have a black background, while

bright colours and a geometrical layout give a site a German feel.

Dutch surfers are keen on video downloads, and Scandinavians

seem to have a soft spot for images of nature, according Ben Vick-

ers in the Wall Street Journal (26 March 2001), reporting on the

EU-funded seed project Multilingual Digital Culture, or MUDICU

(http://www.mudicu.org). Interestingly but hardly surprising —

MUDICU is coordinated by Helene Abrand, an internet consultant

working for Real Media France, the French subsidiary of the US

company Real Media Inc.

Was it really Columbus who discovered America? Was the tele-

phone invented by Alexander Graham Bell or by rival pioneer and

inventor Marconi? What is the height of Mont Blanc — 4,808

meters as the French believe, 4,807 meters as the Dutch believe, or

4,810 meters as the Italian believe? What about such issues as:

● Date, time and number formats

● Colour schemes

● Pictures and images

● Hand signals

● Symbols

● Sounds

● Historical data

● Product names and acronyms

The result of the cultural adaptation effort is that a programme

or a web page sends out all the right signals to you, the user —

something like a chameleon that tricks its enemies. But you know,

and most of all you feel, that something is not quite right — because

no matter how much you change the colour of a website, the hand

signals, the symbols and the sounds used, the content will remain

the same — and people in Finland will still not laugh at a joke from

Texas.

3. L10N is a US driven industryWITH very few exceptions, the most notable one probably being

the case of the German software developer SAP, most large-scale

localisation projects are those where US material is localised for

other, non-US markets. Around 90% of the overall global local-

isation effort is — and this is a conservative estimate — invested

in the localisation of US-developed digital material. In other words,

localisation is currently used almost exclusively by large US cor-

porations as a vehicle to increase their profits. This is only logi-

cal and easy to understand. Localisation is relatively cheap

compared to the development costs of the original product, yet it

provides access to markets much larger than the original US domes-

tic market.

Clearly, market expansion and higher revenue are the primary

drivers for the localisation effort of the large IT publishers, not lan-

guage, not culture, and certainly not diversity.

4. L10N widens gap between rich and poorTHE argument has been made that access to the power of the inter

Continued on page 22

The cultural dimension insoftware localisation

Over the past 16 years the concept of locali-sation has changed. Around 1985, whensoftware localisation as we know itemerged, it was defined as “the adaptationof software to the requirements of a foreignmarket”. Today it has to be redefined toreflect the changing circumstances as “theprovision of services and technologies forthe management of multilinguality acrossthe global information flow”. (Schäler 1999)

Localisation is often seen as an activity thatwill help to preserve the linguistic and cultural diversity of Europe and, indeed, therest of the world. This is why the EuropeanUnion has recently launched a multi-millioneuro programme, the eContent Programme,which supports, under its Action Line 2,initiatives in the area of eLocalisation. Theaim of this programme is to encouragelocalisation initiatives targeting specifically

electronic content that represents Europe’srich cultural heritage.Without a doubt, however, the idea that current localisation efforts help to preservelinguistic and cultural diversity is a mis-conception. Localisation certainly has thepotential to preserve linguistic and culturaldiversity. However, Reinhard Schäler believesthat localisation – in its present form – andcultural diversity are mutually exclusive.

Background

Ten reasons why localisation and cultural diversity are mutually exclusive

LOCALISATION.central SEPTEMBER 200222 LOCALISATION FOCUS

Continued from page 21net and the web will narrow the gap between the haves and the

have-nots. It will lead to a better world, with more equality and

prosperity for everybody. If that were true, localisation would sup-

port this development because it facilitates access and use of the

web for non-English speakers, who otherwise could not participate

in the Information Society. Sadly, the opposite is true.

In a world where half of the population has never made a tele-

phone call — a world where only 3% out of 80% of the total pop-

ulation has a telephone connection — access to the internet is

restricted to those who have the money to pay for it, to those who

are already among the better off. While localisation could help to

overcome the language and cultural barriers that can prevent access

to the web, it cannot solve the fundamental financial and develop-

ment barriers that prevent access for the vast majority of people.

The internet and the web are actually helping to widen the gap

between rich and poor in our world. Localisation as we know it

does not counteract this development; it supports it.

5. L10N favours “northern” languagesSOFTWARE and web pages are not localised into languages that

are spoken by the majority of people on our globe. They are mostly

localised into what could be called the northern languages, those

spoken by the affluent people of the northern hemisphere.

Today approximately 80% of web pages are in English, accord-

ing to Julian Perkin in the Financial Times (7 February 2001). Most

of the remaining web pages are available only in major northern

languages like Japanese, German, French, Spanish or Portuguese.

This contrasts sharply with the 5.4% — little more than 1 in 20

— of the world’s population who speak English as their mother

tongue (according to IDC, the Boston IT market researchers). Not

surprisingly, the majority of these are Americans. A further 7% of

the world’s population are proficient English speakers, so the lan-

guage is a good means of communication for just around one eighth

of the world’s inhabitants.

There are fractionally more native Spanish speakers in the world,

and the Chinese languages dwarf English as a native language, spo-

ken by 20.7 per cent of the world population. Three-quarters of

these, or 15 per cent of the world population, speak Mandarin alone

— nearly three times as many people as have English as a first lan-

guage.

6. L10N contributes to creation of the McWorldIN 1995, the political scientist Benjamin Barber first used the term

McWorld to describe “the cosmopolitan, international, consumerist,

multinationalised, advertising-based culture of cable TV, popular

magazines, and Hollywood films”. This culture, according to Bar-

ber, aims at universal accessibility, in which billions watch the same

World Cup finals. It is a culture in which MTV (translated), drama-

tisations of the lives of imaginary American millionaires, CNN and

films like Titanic dominate and homogenise local cultures, pro-

ducing a thin but powerful layer of consumerist, advertiser-driven,

entertainment-based, and perhaps in the last analysis, American-

influenced culture with great popular appeal, backed by enormous

financial and technological resources.

He did not mention localisation explicitly, but, without a doubt,

localisation facilitates access to this culture and facilitates the pen-

etration of this culture into other local cultures, changing and adapt-

ing them in the process. In a “simple twist of fate”, almost magically,

localisation is no longer just a process of linguistic and cultural

adaptation but becomes itself an active player that adapts and

changes (localises) other languages and cultures.

There is no doubt that the plethora of Americanisms, especially

in the terminology dealing with IT in many “northern” languages

has been introduced to these languages through localised products.

Similar patterns can be observed in consumer behaviour, busi-

ness attitudes and social change. The American concept of “stock

holder value” runs directly against the traditional attitude of social

responsibility that has characterised the way large German corpo-

rations run their businesses. Yet it took less than a decade during

the eBoom emerging from California’s Silicon Valley towards the

end of the 1990s to force German businesses to dramatically change

the way they run their businesses.

7. Lowest common denominator and global symbolismWHEREAS culture needs diversity and thrives on difference, local-

isation, internationalisation, and globalisation often aim for the

lowest common denominator and a global symbolism. While diver-

sity should be interesting and stimulating, one of the golden rules

of localisation is that of reusability. Reuse as much material and

change as little material as possible. Designers of global products,

of websites aimed at the global customer, use globally acceptable

standards, symbols, and conventions. They refer to what they believe

are globally applicable appeals to needs and desires, and they cre-

ate global icons, references and stereotypes.

These stereotypes strongly influence the design, development and

presentation of web content. By contrast, culture is about the spice

of life, culture is about difference, culture is about contrasts. A pre-

Three CocaCola web-sites: (topleft) Dutch;(left)Germanand (right)French.

Designers of global products, of websites aimed at the global customer, … refer to whatthey believe are globally applicable appeals to needs and desires, and they create glob-al icons, references and stereotypes.

SEPTEMBER 2002 LOCALISATION.central 23LOCALISATION FOCUS

vailing lowest common denominator of tastes, preferences and opin-

ions, nurtured by the global web, marks the end of diversity and

contrast and, ultimately, the end of culture.

8. Global icons suppress local contents and create the‘accidental web’BECAUSE of the power, persuasiveness and the resources behind

them, global icons, global symbols and global content are pushing

aside and suppressing local content. When you travel to Spain, do

you want to find out from a web-based, localised US travel guide

where to eat in Santiago, Madrid or Seville? Or when you travel to

the Middle East, do you want to read up on the history of the region

on a localised US web page? Regrettably, this is what you will most

likely be offered when searching the web for this kind of informa-

tion.

If one searches for background information on any region of the

world, the likelihood is that the information presented will not orig-

inate from the region but from US, localised websites that are reg-

istered with the major search engines and shown at the top of the

list of search results.

Like travel writer Macon Leary in Anne Tyler’s book The Acci-dental Tourist, who hates both travel and anything out of the ordi-

nary, many global eContent publishers dislike diversity and

divergence from standards — for the simple reason that it makes

their lives more difficult and (even more crucially) their projects

more expensive. They create perfect “accidental” websites that are

acceptable to every global citizen’s tastes, beliefs and customs. There

are no surprises, no deviation from the norm. There is an almost

clinical feel of global political over-correctness to them.

Macon Leary needed Muriel, a wacky dog-obedience trainer, to

end his insular world and thrust him headlong into a remarkable

engagement with life. Local content producers, local cultures need

the technical experts to bring their content to the world so that the

world can enjoy the different perspectives and approaches offered

by them.

9. Contents must be local — L10N must be bottom upTO PRESERVE diversity, to preserve culture, digital content must

be local and must be presented and accessible in as many languages

as possible. In Europe, for example, there is no shortage of digital

content. In fact, many market researchers have highlighted Europe’s

advantage over many other regions in the world in this area. How-

ever, this content must be made easily available to the rest of the

digital world. It must be made available to people living in differ-

ent countries, speaking different languages.

Imagine how interesting and entertaining it would be to be able

to listen to the Spanish and the English commentaries of a soccer

match between Barcelona and Manchester United. And there is no

reason why the sources of information on which we base our inter-

pretation of contemporary events should exclusively be supplied

by and filtered through the large news networks like Sky and CNN.

A recent example of the important role, and indeed the success, of

a relatively unknown digital content provider (in this case a local

satellite news station) has been the Al Jazeera satellite news chan-

nel, which is supplying up-to-date and often exclusive reports and

pictures from the Arabic world (see www.aljazeera.net).

There are reasons why major European television stations reg-

ularly report on torrential rains and severe flooding in the Missis-

sippi Delta causing great damage to local houses and sweeping away

half a dozen cars, while minimum coverage is given to human cat-

astrophes, death, hunger and disease in Africa and Asia.

These reasons are directly linked to availability and cost.

Diverse, locally produced, eContents can only be offered if this

content finds its way into the digital world and is then localised to

make it accessible to speakers of other languages. It is “bottom-up

localisation” (and not the top-down localisation that we have become

so used to) that will contribute to preserve the diversity of languages

and cultures.

10. Why L10N efforts must fundamentally changeMOST of the reasons I’ve covered so far why localisation efforts

must fundamentally change and be redirected concerned mainly

political and cultural areas. As I said earlier, in the world of big IT

business, a world dominated by young, hard-working, no-nonsense

professionals whose decisions are mostly determined by stock mar-

ket analysts’ reports, it is difficult — if not impossible — to inter-

est anybody in diversity, never mind culture. So let’s talk about

revenues, new markets and cost models.

It has been established that most large digital publishers make

most of their revenue from their international sales. These pub-

lishers would never have got where they are now, they would never

have achieved their current revenue figures, they would never have

been able to achieve that level of market penetration — without

their highly successful localisation strategies.

The success stories of companies like Microsoft, Lotus/IBM,

Oracle and Symantec, which by now have long moved from being

simple software developers to becoming electronic publishing houses,

should encourage publishers in other regions in the world.

Europe has the knowledge, the experience and the expertise to

localise its vast amounts of electronic content and make it available

to the rest of the world, creating new business opportunities and

markets.

What is needed is the political foresight and the industrial ini-

tiative to make this happen. ■

ReferencesMultilingual Digital Culture, EU funded seed project,

http://www.mudicu.orgSchäler, R. (1999). ‘New Media Localisation – a LINGLINK

Report for the European Commission DGXIII’, http://www.hlt-central.org/usr_docs/studies/Localisation.pdf (last consulted 19 April2002)

Vickers, Ben (2001). ‘Internet becomes Battleground for Europeto defend Culture’, The Wall Street Journal, 26 March 2001.

ConclusionsI hope to have shown that localisation certainly has the potentialto preserve linguistic and cultural diversity. To fulfil this potential,however, its efforts have to be redirected. Initiatives and actionssuch as those undertaken within the European Union eContentProgramme (http://www.hltcentral.org and http://www.electonline.org) as well as the sharing of the lessons learnt during the course of successful and unsuccessful localisation projects (especially those for languages and cultures ignored bymainstream localisation efforts) could, in my opinion, help to usethe large pool of expertise and experience available within thelocalisation community to preserve cultural diversity and to prevent a slow but certain demise of cultures currently not present in the digital world.Earlier on, I quoted David Brooks, who said any language that isnot captured in this electronic world will soon become obsolete. Ibelieve that any culture not captured in this electronic world willeventually become obsolete.In addition, dominance and exclusivity cannot ultimately benefitanybody, not even to the big players, because no culture can live ifit attempts to be exclusive (Ghandi).I believe we should not remain at the level of mere recipients ofthe localised content of others. We should become actors, localiseour own content and make it accessible to everybody in the digitalworld. We should all become localisers.The author can be reached at Reinhard.Schaler@ul.ie

Readers’ ForumLOCALISATION.central SEPTEMBER 200224 LOCALISATION FOCUS

TRANSLATION is critical to the process of local-isation (or “localization” in the US, or morecommonly L10n). Of that there can be nodoubt. It is also perfectly clear that in manyways translators have been getting the shortend of the stick of this emerging and growingarea of real-life translations.I have been providing support to translatorsonline, helping them to understand the biggerpictures around L10n since about 1995 when Ifirst joined the translator’s mailing list, LANTRA.I believe few non-translators in the L10n indus-try have more sensitivity to, or a longer historyof, supporting the needs of translators’ workingconditions than I do.Even so, I feel uncomfortable with the topic“Localisation IS translation”. People who workfor software companies who make internationalversions don’t see it that way. Ask any of themwhat localisation is and you will hear somethinglike: “L10n is the process of inserting locale-specific instances of data into a generic, locale-neutral software product (i.e “internation-alised”). This definition is accepted by softwaredevelopment companies and L10n vendors alike.Because they are not aware of the bigger pic-ture and the value that others provide in theL10n processes, it is often true that translatorsare marginalised in L10n projects. Ironicallyenough, L10n managers often complain thatthey are marginalised from the overall softwaredevelopment process. That happens for similarreasons: failure to adapt to the bigger process.

There are some things I would love to see fromthe translation industry that would be extreme-ly valuable to L10n folks and would make iteasier to become stronger participants in theoverall process.● Quality metrics: I know that translation isnot an exact science. But software managers,like managers in any industry, need to measurethe quality of the input to their processes sothat the rest of the project can be scheduledand budgeted accordingly. Delivery of a stan-dard estimated quality metric that can beapplied to any translation is the number onething translators could do to “reclaim” at leasta part of the L10n process.● Intellectual property rights: I often see ontranslator and agency mailing lists the con-tention that translators “own” the intellectualproperty rights of their translations and trans-lation memories. Without discussing legalissues, it is important to realise that L10n pro-jects use multiple translators for each piece oforiginal text, if not on the first version, then onthe next. The software company needs to beunburdened in its ability to proceed with itsbusiness.I recognise that this may be different fromtranslating other types of material for otherindustries. It often surprises me that transla-tors and agencies do not ask for credit for thetranslation, which is common for other types oftranslations. A software company should behappy to provide that recognition.

● Cultural advice: translators make manyimportant cultural decisions for software com-panies. Most of these essential decisions areinvisible. I believe there would be tremendousvalue if some of those decisions could also bequantified. Translators might be in the bestposition to get this information to UI designersor software product managers. Doing so in noway affects the amount of translation thatneeds to be done.● Pricing models: the translation needs to finda new model for pricing for L10n projects.Pricing by the word, or other small unit, helpsno one. It does not reflect the value translatorsactually add to the project, but it gives everyincentive to the buyer of the translation to drivethe price as low as possible. Prices tied toquality metrics will make for a more productivealliance of L10n managers and translators.Most importantly, I would like to see translatorswho enjoy working in the software industryreach out to their colleagues who performother tasks and roles. Understanding the over-all process and the roles of others in theprocess is far more productive than claiming“Localisation is translation”. ■

— Barry Caplan

Barry publishes www.i18n.com, a website thatenables interaction with others in the overallinternational software process, so that all participants may learn from one another.

CAN adverts be translated? As I was looking for an answer to this question, I quickly realisedthat in international advertising the fashionable motto was “Think global, act local”.Therefore — not much room for translation. The principle put forward is to create an advertadapted to each market. Localising and translating can appear as opposing methods, thefirst one being costly and the second left out or reduced to a bare minimum to saving a fewpennies.I then “discovered” what translating means in international advertising and marketing:transcoding, changing words from one language to the other. Obviously, with such a concep-tion, one can only refuse translation. This helps to understand the success of localisation: itallows us to respect the local culture.Word by word translation is far from being the only translation method. The InterpretiveTheory of Translation (ESIT) and the German Skopostheorie are two good examples. Forthose who would contest the theory, there is the practice. I have recently been hired as anediting assistant in Portuguese language to localise a method of learning English as a for-eign language. What did I do? I have purely and simply translated, sorry localised! (to usethe politically correct wording). ■

— Isabel Pereira Da CostaPhD Student (ESIT-Paris III)

Next issueProposition: Translation vendor web serviceswill deliver what localisation management platforms promised. They will ultimately makemultilingual localisation vendors redundant.● Localisation Focus invites its readers to comment on this statement and send theircontributions to lrc@ul.ie by Monday, 7 October2002.

Understanding the overall process

Is localisation justtranslation? Tworeaders begin alively debate.

Why translators should, finally,reclaim translation as their own

SEPTEMBER 2002 LRC.news 25LOCALISATION FOCUS

localisation.ieInformation on the LRC’s activities is now

available on the LRC’s new web address:

www.localisation.ie

TILPThe Institute of Localisation Professionals

is now up and running and open for mem-

bership applications. More information on

www.tilponline.org

LRC Advisory BoardThe LRC is very pleased to welcome

Dirk Metzger, Translation Manager of

the SAP Language Services in Walldorf

(Germany), as a new member of its

Advisory Board. Dirk will add substan-

tial weight and an international dimen-

sion to the expertise of the Board. As

the Translation Manager of Europe's

largest software publisher, he will also

be an important voice for the transla-

tors in the industry."

LRC staff newsThe LRC welcomes two new members of

staff who are joining the ELECT project as

assistant researchers responsible for content

management and web development.

Joanne Cheung, originally from Hong

Kong, has a BSc in Applied Mathematics

and Computing (honours) and a Graduate

Diploma in Software Localisation. She has

previously worked in the Aviation Services

Industry.

Adrienne McCarthy has an honours

degree in Business Studies and a full CIPD

professional qualification. Adrienne has pre-

viously worked in human resources pro-

viding a full range of front-line support to

assigned client groups in EMEA and

Asia/Pacific with direct responsibility for all

French and non-professional Irish staff.

More room for the LRCThe office space occupied by the LRC will

double over the summer. This will enable

the LRC to take on more project staff and

to provide better facilities for visitors. In

addition to the office space, the LRC will

also open the new Localisation Technology

Laboratory and Showcase with more than

ten dedicated computers, running on a vari-

ety of platforms and in different languages.

The LRC Annual Worldwide Localisation Salary SurveySurvey forms are currently being circulated

to companies interested in participating in

this annual LRC survey. The survey results

will only be made available to participants

in the survey. Should your company be inter-

ested in participating in this survey and not

have received a survey form, please contact:

LRC@ul.ieJoanne Cheung Adrienne McCarthy

Workshops and meetings (pre- and post-conference)

This year’s conference will offer ameeting place for people and

organisations involved in different aspectsof the industry. Amongst these will be the

ELECT Expert Group, The Institute ofLocalisation Professionals (TILP), the

Localisation Teaching and Training Network (LttN) and the Translation Vendor

Web Services Group.

Annual Localisation DinnerThe 3rd Annual Localisation Dinner

will take place on Monday, 11 November2002. This is a unique opportunity tomeet old friends and make new ones,

exchange the latest news and just having a good time.As in previous years,there will be a draw for a special prize

amongst the participants.

Exhibition (12-13 November 2002)

Localisation Industry Exhibition and Showcase

Localisation service providers and publishers, tools and technology developers, teaching and training

organisations as well as publishing companieswill exhibit in the LRC 2002 Localisation

Exhibition and Showcase.European eContent Village

Researchers and European eContent projects will present their work in the dedicated European eContent Village

Product demonstrations (12-13 November 2002)Product demonstrations will be

organised covering:● eContent Localisation tools

● Multilingual eContent technologyapplications

● Translation technology● Web Services

Best Thesis AwardThe winner of the 6th Annual

Best Thesis Award, sponsored bySymantec Ireland, will beannounced at the Annual Localisation Dinner on

Monday, 11 November 2002.

Organised by the LRC and the European ELECT project in cooperation with the Department of Computer Science, University College Dublin (UCD).

Supported by GALA, TILP, PAL, LttN, and Multilingual Computing

The 7th Annual International Localisation Conferenceand Exhibition organised by the LRC is Europe’s largest

and best-established annual localisation event. This year’sconference is organised in co-operation with the European Localisation Exchange Centre (ELECT),

supported by the European Union eContent Programme.

12-13 November 2002O’Reilly HallUniversity College Dublin

Information for participantsRegistrations forms and information on accommodationavailable in Dublin for conference participants isavailable on the LRC’s web site(www.localisation.ie). For further information please contact LRC@ul.ie.

Information for exhibitorsAn information pack forexhibitors is available on theLRC’s web site (www.localisation.ie)

LRC 2002 eCont

Pre-conference Workshop

LRC 2002 Conference

Monday, 11 November 2002Localisation Teaching and Training

Tuesday, 12 November 2002

Wednesday, 13 November 2002

Thursday, 14 November 2002

ELECT Expert Council Meeting

CONFERENCE PROGRAMME

Organised by the Localisation Teaching and Training Networkunder the umbrella of the LRC

08:30 Registration09:00 Opening09:15 Session 1

The Localisation Teaching and Training Network (LttN)10:30 Break

11:00 Session 2Localisation Teaching and Training in Europe

12:30 Lunch14:00 Session 3

Localisation Teaching and Training in America and Asia-Pacific15:30 Break16:00 Discussion17:00 Close

08:30 Registration09:30 Official Opening of Conference, Exhibition and European

eContent Village09:50 Keynote Address: The Global Market for eContent10:30 Break, Exhibition and Product Demonstrations11:00 Session 1: Case Studies

Developers and localisers present detailed reports on the challenges of eContent localisation

12:30 Lunch, Exhibition and Product Demonstrations

14:00 Session 2: Localisation issues in AsiaIndian localisation experts report, together with Prof Pat Hall ofthe Open University (U.K.), on the issues encountered when localising into and out of Indian languages.

15:30 Break, Exhibition and Product Demonstrations

16:00 Session 3: eContent Localisation – Projects and InfrastructureProfiles of European and Canadian Localisation Initiatives

17:30 Close

16:00 Special Session (by invitation only): The Localisation ForumLocalisation organisations explore opportunities for co-operation.

09:00 Keynote Address: Developing eContent for the Global Market10:30 Break, Exhibition and Product Demonstrations11:00 Session 1: eContent Localisation Technologies

Developers and researchers of eContent localisation technologiespresent new approaches to the automation of localisation tasks,including .NET technologies and XLIFF.

12:30 Lunch, Exhibition and Product Demonstrations

14:00 Session 2: Translation Vendor Web ServicesDuring this session, the Translation Vendor Web Services Stan-dards Group operating under the umbrella of the LRC will belaunched. Speakers will address strategic business questions aswell as detailed technical questions in the context of the devel-opment of a standard for translation vendor web services.

15:30 Close

Post-conference Workshop15:45-17:00 Translation Vendor Web Services Standards organised by the TVWS Group under the umbrella of the LRC

09:30 Welcome09:45 ELECT Overview10:15 Review of work and material produced

10:45 Discussion12:30 Outlook13:00 Lunch

ent Localisation

PERSONAL.profile SEPTEMBER 200228 LOCALISATION FOCUS

Both paul and Damian have beendominant figures in the localisationtools arena in Ireland for the past 10

years, having developed many of the lead-ing industry translation tools.

Whilst both are instantly recognised fortheir individual styles, characters and contributions to the industry their visionof the future is remarkably similar.Operating as competitors in the past theyearlier this year combined forces to formKey Performance Solutions. KPS specialisein the analysis, design and implementationof Automated Localisation Solutions(ALS). It is the world’s first AutomatedLocalisation Solutions Provider.

“We empower our customers withprocesses and technologies to enable themdevelop and deliver competitively and efficiently to a global marketplace.”Paul Quigley

Both began life as Assembler programmersmany years ago and followed various soft-ware development paths. In their earlycareers they jointly caught the entrepreneurial bug.

1986 saw Paul become lead developer andone of four founding members of the start-

up VISION computing. Damian has run anumber of his own software development companies in the past. Hisearly ventures saw him focus on ComputerGraphics and CAD development.

The early 90s saw both join major multinational corporations in develop-ment roles. Paul joined Lotus developinglocalisation tools and later OracleCorporation as Director of LocalisationTools. Symantec became the home forDamian as localisation tools development manager.

“Sometimes it is amazing to look back onyour career. I was recently asked howmany localised products I shipped in mycareer? Over my 20 year career I’ve beenresponsible for over 3000 localised masters! That’s a lot of software” Damian Scattergood

In the last few years Damian did theDot.Com roller-coaster as Director ofOperations with Worldport communications. “Data centres can be great fun and excellent technology learning places.When you run lots of sites at the sametime, you get to understand exactly howthe Internet works.” Damian Scattergood

Remarkably both have kept a keen eyeon each other’s careers over the last 10years. “Damian and I have always had aninteresting relationship. Managing theLocalisation Tools developmentprocess for our companies places us ina position of both competitors andpartners”. Together we worked on anumber of projects evaluating newtechnology, but we also developedcompeting tools. It’s almost a love/haterelationship.”Paul Quigley

Today they share a combined vision to“Automate The Localisation Industry”.It is time for the localisation tools industry to mature and take the nextleap forward. The major translationcompanies have grown to optimumproportions and need to be optimised to meet the very demanding

marketplace. With KPS we believethat automated localisation engi-

neering delivers “More Translation ForLess Cost”. That’s what themarket wants today.

Their spare time is whereboth individuals show

their different charac-ters. Paul exudes hisindependence commut-ing around the countryon his motorbike mainlybetween Dublin and

Wexford. Damian onthe other hand is a StarWars fan who is alwaysimproving his collectionof memorabilia.

The personal traits areinteresting. Damian hasrecently been refereed toas a maven of companyinformation. He is a

data collector who has an unusual skillfor memorising vast amounts of informa-tion.

“I believe information is the key to futurevalue. It makes me sound nerdy, but I’mnot really – No really I’m not!”Damian Scattergood

Paul has the same level of cognitive skillwith technology.

“I find that Paul can learn any new technology amazingly quickly, but moreimportantly has the ability to instantlyrecognise where it fits in the bigger picture.” Damian Scattergood

“About nine months ago Damian and Ihad a discussion over coffee about theindustry in the Jury’s Hotel at DublinAirport after a long flight home. Mainlytrying to work out our next strategicmoves against each other. We could beboth great chess players. We came to theconclusion that two heads were betterthan one. Our combined forces would bevery interesting. KPS was born, and hopefully the rest will be history.”Paul Quigley ■

Two Heads are Better than OneIn this issue we take an unusual view of personal profile by looking at twoindividuals simultaneously:Damian Scattergood CEO and Paul Quigley CTO of Key Performance Solutions.

SEPTEMBER 2002 POETRY.loc 29LOCALISATION FOCUS

LocalisationBy Shailendra Musale

W hen many hear localisation, they think it s translation but there is more in localisation, & it needs everybody’s participation

W ith an aim of globalisation, a company decides market expansion manager does a project initiation, without any plans & preparation

The project starts with existing version, which has no internationalisation developers opt for instant solutions Ending in frustration, company gives up localisation

This is a common situation, but we can change this condition

With everybody’s contribution, We can do better localisation

Have a proper internationalisation, respect user’s culture & tradition

perform string isolation, avoid any concatenation allow more room for text-expansion

With everybody’s contribution, We can do better localisation

Contact experts in localisation, plan & have a discussion Give all members a proper education, boost their confidence & motivation

With everybody’s contribution, We can do better localisation

Before starting translation, have a glossary collection Putting tools into action, for every future version you will save time with more repetition

With everybody’s contribution, We can do better localisation

Check for any file corruption don’t forget an important function to conduct in-country validation & have proper verification

With everybody’s contribution, We can do better localisation

There is an awareness & realisation we now have courses for education Universities are giving certification, So many are joining industry of localisation

With everybody’s contribution, We can do better localisation

Tools are now available as free-distribution, there is more to explore about machine translation We will seek better solutions, to achieve goal of globalisation

W ith everybody’s contribution, We can do better localisation

Incorporating these notations, company can achieve cost-reduction & enjoy profit maximisation Thats the road to better localisation

With everybody’s contribution, We can do better localisation

SEPTEMBER 200230 LOCALISATION FOCUS

The LRC Industrial Advisory Board meets at least twice ayear to review the work of the LRC, advise on potentialprojects and strategies, and provide support for itsactions. Members of the board recognise the importanceof the LRC’s activities for the localisation industry andsupport its aims and objectives.The board’s chairperson is Alan Barrett (IBM). He waselected at the board’s first meeting in December 1999

LRC Industrial Advisory Board

THE LRC at UL is the focal point and the research and educational

centre for localisation. It is one of the world’s leading intelligence,

technology and educational localisation centres. The LRC was estab-

lished in 1995 at University College Dublin under the Irish

Government and European Union funded Technology Centres

Programme as the Localisation Resources Centre. When the centre

moved to the University of Limerick (UL) in 1999, it merged with

UL’s Centre for Language Engineering and was renamed the

Localisation Research Centre (LRC).

The LRC is owned by UL. It has a director, faculty members and

project staff. Its Industrial Advisory Board represents a large section

of the localisation industry. The LRC is supported by UL, its

Industrial Advisory Board, subscribers to its services and Enterprise

Ireland. It’s main areas of research are:

Industry intelligenceLocalisation Exchange Point

Localisation Directory

Contact Database

Surveys and industry studies

Education and trainingGraduate Diploma / MSc in Software Localisation

Professional Certification

Professional Development Courses

Localisation Teaching and Training Network (LttN)

International Localisation Summer School

TechnologyLocalisation Technology Laboratory and Showcase

Tools and technology evaluation and certification

Translation and test automation

Annual LRC Best Thesis Award sponsored by Symantec Ireland

The LRC and its functions

Alan Barrett Director of Technology,International Product Development, IBM

Gerry Carty General Manager, Vivendi Universal Publishing Ireland

Ian Dunlop VP & GM, Provisioning Services & Solutions Group, Novell Inc.

Mervyn Dyke Managing Director, VistaTecSeamus Gallen National Software DirectorateJames Grealis Director EMEA Localisation,

SymantecWendy Hamilton Vice President Business

Development, Bowne Global Solutions

Martin Hynes Senior Analyst, FORFÁSBrian Kelly Vice President, Berlitz John Malone Director International Sales and

Marketing, ArchetyponPaul McBride Director European Operations,

VeriTest (Division of Lionbridge)Patti McCann Director of Localisation, Business

Tools Division, MicrosoftDave MacDonald ConsultantDirk Metzger Translation Manager, SAPDavid Murphy Director Localisation, Siebel

SystemsMichael O'Callaghan Vice President, OracleAnthony O'Dowd President, Alchemy SoftwareKevin Ryan VP Academic Affairs, ULReinhard Schäler Director, LRCJim Seward Head of Professional Services, ETP

Europe’s biggest annual localisation event12-13 November 2002

Dublin – Irelandinfo: www.localisation.ie

www.tilponline.org

See you at

LRC 2002