🚀 go-pugleaf

Here's the consolidated responses to the original statements and
questions from Simon North (north@hgl.signaal.nl):

     Over the past few weeks the discussion seens to have turned on the
     idea of localization as if were an _optional_ extra ...

     I am a member of a Dutch standards commission which has been
     addressing the subject of documentation for Consumer Products. Faced
     with a rapidly increasing influx of 'foreign' products we were
     concerned about the standard and nature of the documentation
     accompanying them. After two years of hard work we have now started
     to circulate the final draft of a new standard covering 'instructions
     for use'. Like a lot of other European countries (we have sourced
     most of the international and Common Market standards), we are
     stating it as a mandatory requirement that any product appearing on
     the Dutch market MUST have its documentation written in Dutch.
     [We will shortly be starting work on a new standard to cover software
      in which a similar demand will appear.]

     I would therefore like to address a (semi-hypothetical) question to
     all the representatives of software companies subscribing to this
     list: just how much longer do you think you can continue to avoid
     localizing software ?

     Disclaimer: The opinions expressed herein do not represent those of
                 either my employer or the NNI.


Jeff
INSOFT-L List Manager

*========================================================================*
 Jeff Bowyer                         EMail:
 Computing Center		     jbowyer@cis.vutbr.cz
 Technical University of Brno
 Udolni 19, 602 00 BRNO
 Czech Republic
*========================================================================*


Sender: Stanislav Fritz <stanf@microsoft.com>
        International Program Manager

A set of two interesting points in this message, which I feel compelled
to respond to:

The point is well taken that software companies, to expand their
markets, need to localize their software.  If they do not they will
meet competition and/or as the writer mentions legal requirements to do so.

>From a policy and economics view point I have to say that the
"mandatory requirement that any product
appearing on the Dutch market MUST have its documentation written in
Dutch" is--I won't mince words--stupid.   The U.S. and the EEC have for
years complained of Japan's non-tariff barriers to trade.  This is what
this is.  If the consumer **WANTS** a product and it is not localized,
what are they supposed to do?   The law better be VERY carefully
written.   Does this mean a French diplomat wanting to buy a copy of
French Windows in Holland can't because it is not "localized"?

Does this mean a consumer has to buy the *translated* works of Shakespear?

I could go on, and on, with a long tirade on the bad economics and loss
of consumer freedom in the name of "protecting" the consumer.  But, I
think you get my feelings on the matter.


Sender: Anders Thulin (ath@linkoping.trab.se)

I think the idea is a good one - documentation should be adapted to the
users rather than vice versa.  (Note: the users! I mean precisely that!)

But ...

Standards are as far as I know, still optional, except possibly in
cases where it is *vitally* important to follow one. In order to make
this standard widely adopted, someting extra is probably necessary.
Either a whip or a carrot.

Will it be a civil offence to market software without documentation in
Dutch? Or lower VAT (or whatever) for the final product?

There's also a quality problem: QA is usually applied on the base
product (call it the US version).  The localized product should be
submitted to the same procedure, but (I'm sometimes inclined to
believe) seldom or never is.

Any ideas on these aspects?


Sender: Bob Peterson <"gemgrp::peterson"@tle.enet.dec.com>

> Just how
> much longer do you think you can continue to avoid localizing software ?

One year, I hope.  That is how long it might take to do a massive I18N of our
products.  I suppose the real answer is "we can't wait", it'll just take time.


Sender: John Clifford <johnclif@microsoft.com>

Just a quick question:

Why should it be mandatory for all products sold in a country to
include documentation for that country's official language?  I'm
thinking of products like HP calculators, which, by the amount of email
traffic on various mailing lists, are in great demand in Europe
(Europeans are always asking how they can buy one from overseas, or
what procedure to buy one while they're here visiting).  Should HP be
forbidden to sell their calculators in the Netherlands until they
localize the documents?  What about the English-speaking engineering
firm who wants to buy HP-48s for their staff in order to be able to
compete with US, UK, et al, firms?

I think to codify this as a law is bad.  Let the consumer decide.  If
HP's target market is made up of engineers and businessmen who are
fluent in English and can understand the US documentation, your law
would prevent them from buying something that may be of great utility.
Thus, your law, grounded as it is in the desire to help the Dutch
consumer, actually ends up hurting him -- because US engineers and
businessmen (and those in the UK, and any other countries where HP _is_
allowed to sell their product) can make use of HP's calculators to
further their ability to compete while the Dutch businessman, despite
his English fluency, can't.  The law you propose forces the Dutch
consumer to either stick with pencil and paper (and reference books
galore) and be less competitive, thus depriving your country of jobs
(the firm can't grow if it can't get business because it can't compete
against other firms) and revenue (the taxes paid by those so employed),
or to buy the equipment he needs overseas thus depriving your country
of jobs (local representatives/distributors of foreign companies) and
revenues (taxes paid by those so employed, and taxes paid on the items bought).

If the nature of a product and its intended market is such that the
manuals should be localized before the product can be of use, then lack
of sales will either compel the seller to localize the manual, or take
the product off the market.  Either way, the right thing will be done
for both the consumer and the company -- and your country.

Re 'just how much longer do you think you can continue to avoid
localizing software,' let me just say two things:

Thing #1:  I do not represent Microsoft's official viewpoint.  See
BillG, et al, if you wish that.

Thing #2:  Having made the disclaimer in Thing #1 above, I'd like to
offer my own personal $0.02 (two cents)...

Microsoft does indeed localize its software and has made a considerable
investment in order to do so.  This should be self-evident by perusing
the sources of software in European, Asian, Middle East, South
American, etc., countries -- you'll see localized versions of our
various products. Also, if you look at our annual company reports, you
will see that the international marketplace is a vital part of our
business.  As our bottom line shows, localization of our products is
just good business sense.

Nevertheless, localization of nontrivial software is itself nontrivial.
I get the impression that you (S. North) think software companies
don't localize software because it is somehow 'undignified.'  Again, in
my opinion, I don't think this is true at all.  Having worked at more
than one company that produced software sold internationally, I can
tell you that if a company determines it's worth the considerable
investment to localize software (based on an expected increase in sales
to justify the expense)  _AND_ they have the capital to make the
investment, software _WILL_ be localized.  But that second part, having
the money to make the investment, is what stops a lot of smaller
companies from localizing their products.

In summary, my personal opinion is that companies that don't localize
software sold in different countries fail to do so because it is, for
whatever reason, economically infeasable.  If you prohibit the sale of
unlocalized software, or products with unlocalized documentation, you
are only hurting the consumer.


Sender: Arne Larsson (LARSSON@ntcclu.ntc.nokia.com)

> I am a member of a Dutch standards commission which has been
> addressing the subject of documentation for Consumer Products. Faced with
> a rapidly increasing influx of 'foreign' products we were concerned about
> the standard and nature of the documentation accompanying them. After two
> years of hard work we have now started to circulate the final draft of a
> new standard covering 'instructions for use'.

	This is interesting news. Would copies be available, and
	if so, on what terms?

	Nokia Telecommunications produces embedded and distributed
	software for telecommunications equipment in Finnish, French,
	German, Russian, Spanish, and Swedish with accompanying
	manuals in these same languages. Documentation in Chinese is
	developed in situ. Technical writing is done in more than
	one country.

	With experiences from these efforts in mind, I would like
	to suggest that the answer (or at least one answer) to
	Simon's question concerning doing vs. avoiding localization
	of software, lies in the adequate use of scarce resources.

	In addition to AVOIDING localization, the real options
	include, but are not limited to,

	internationalization	i.e. making the same piece of
				software 'switchable' between
				several languages

	localization		i.e. making several releases of
				basically the 'same' software
				available with menues, manuals
				help systems, etc, in the users'
				native languages

	targetization		i.e. releasing the (more or less)
				'same' software product adapted to
				different groups and subgroups of
				users. A spreadsheet, for example,
				targeted to, say, Swedish business
				consultants would be different from
				the 'same' speadsheet sold to Swedish
				schools for educational purposes, which
				in turn would be different from this
				'same' educational spreadsheet sold
				to schools in Ukraina, or Poland, or ...;
				and also different from versions sold
				business consultants in these same
				countries

	ease of use		i.e. providing the user not only with
				facilities for getting started, but
				also with tutorials showing how
				increasingly more complicated and
				sofisticated processing can be
				accomplished using the specific
				software product

	domain knowledge	i.e. providing the user with databases
				of a more general, educational, or
				professional nature. A mathematical
				program might be accompanied with
				databases pertaining to higher math,
				finite elements, or a wordprocessor
				could be delivered with a copy of
				the national encyclopedia of the
				user's choice; just to take a few,
				not too ingenious examples.

	Any of these, and many other routes can be choosen, and will,
	eventually, make localized software less ugly and more
	faithful to the real needs of our users.

	The fact that there are currently between 4000 and 6000 different
	languages in the world to day lets us understand what kind of
	resources are needed. Formidable, to say the least.

	Additional discussion on the list would be very welcome.


Sender: Lyle Seaman (Lyle_Seaman@transarc.com)

Oh, great.  I hope you're not recommending penalties for smuggling
"unapproved" software in from foreign countries.

You realise, of course, that this will drive up the cost of products
which appear on the Dutch market.  Or do you think that you can
mandate that vendors sell their products at the same prices in all
markets?

Imagine a small company, with three engineers and a part-time
documentation person working on a software project.  Say, a high-end
Fortran compiler for supercomputers.  Their product is sold in small
volumes in 15 or 20 countries.  There *might* be 50 potential
customers in NL.

The size of that market in no way justifies the cost of producing
entirely new documentation in a foreign language, unless the cost is
entirely borne by a customer.

Given that they probably will never produce Dutch documentation, what
do you propose as recourse for the few Dutch sites who wish to use their
software?  I doubt the customers will be too proud to purchase it in a
foreign country.  Of course, nobody reads the documentation anyway.
:-/

Localisation to improve the quality of a software product is good.  If
it helps users to complete their tasks more quickly, or more easily,
or more accurately, it's great.  To the extent that I can acheive
those goals subject to my budget constraints, I will do so.
Localisation to satisfy jingoistic bureaucratic requirements is
pointless.


Sender: Martin "J." Duerst <mduerst@ifi.unizh.ch>

I remember having seen an interesting answer to that question, perhaps it
was in the "Guide to Macintosh Software Localization":

You can avoid localization as long as your competitors do.

This especially applies to the Japanese market, but probably elsewhere,
too. It means that the fact that a product is localized is strongly
dominating purchase decisions. For a country like the Netherlands, where
people are much more at ease with English, the above might not be true.

Generally, I think that imposing localization by standards and legislation
is a two-sided sword. You gain something for the everyday user of mass
products (if you can assure that localization of those products is done
with high quality). You might also help the local industry, as localization
might be done by local firms, and as local software gets some advantage
(it is already localized, the others are shut out until they localize).

However, what about all those nice shareware products, and what about highly
specialized scientific and technical software? It is too expensive to
localize it, but even if a user is willing and completely able to use
the foreign version, he is not allowed to do it. Hopefully your standard
has some exceptions for such cases. Another problem is delays of updates
and new versions, and a restriction of the market size on the side of
supply. With a strict standard, there are only two alternatives: Localize
or stay away. Without standard, a company has one alternative more:
Test the foreign version. The test might reveal that the local users
particularly like one program or the other, whereupon localization is
started. So you get those programs localized that suit the local user.
Otherwise, you just get programs of those big companies that can afford
localization by the dozen.

All in all, I think that forcing localization has a positive effect esp.
in big countries, where it might help the local software industry without
adversly affecting the users. But in big markets, the market forces towards
localization should be strong enough, so that standards and legislation are
not really necessary. On the other hand, in smaller countries, where
market forces towards localization are small, standardization has the
unwanted effect of reducing the choice of the user.


Sender: Robin Gordinier (esri!hobbes!robin@uunet.UU.NET)

I do not believe that it is a question of whether software and
documentation should be localized or not.  I believe it is truly a
marketing decision.  Localization is quite expensive and if the
market demand for a product can not justify the resources and funds
to localize a product - well....And then on the other side of the coin,
is the issue of how can a software company expect to compete or even
enter a market if the local issues are not addressed.

One problem I have been struggling with is the fact that standards continue
to be a rather big issue.  Perhaps if those who represent country standards
organizations could provide software companies guidelines of how to
approach such a problem.

The types of information which would be helpful
from my point of view (keeping in mind I am a localization neophyte), is
not only a collection of country standards (with examples of industry
standards) but the following:

- codeset definitions across all operating systems (this would save a
  considerable amount of development/research time)

- if multiple codesets are used (I understand this is common), examples
  of how industry leaders code software to adapt/convert between them

- availability of quality fonts that can published for end users and those
  which can be obtained at modest costs (for distribution).  I understand
  Japanese distributors can pay up to $25USD/character (x8,000+) for a
  font to distribute to their users.

- a forum of experiences (good and bad) of how a company evolves to be
  truly international.  This would provide guidelines with which to
  gently nudge a company to make well informed decisions based on success
  stories.

I am a firm believer in creating truly localized products for all, however,
how do you convince the higher beings that be.