gnucash-docs

[Christian Stimming]

-----BEGIN PGP SIGNED MESSAGE-----

On Mittwoch, 27. November 2002 21:58, Linas Vepstas wrote:
> On Wed, Nov 27, 2002 at 12:15:46PM +0100, Christian Stimming was heard to 
remark:
> > Sorry, I have to object here. Please don't think first about potential
> > translators -- think first about actual users, which make up 99,5% of
> > the downloads. If they click on "Help" and (since an old, crappy
> > transaction happens to exist in their locale) get a really outdated
> > crappy version of docs, they are really pissed off 
>
> Its hard for me to guess how multi-lingual gnucash users really are.
> Can we assume most european gnucash users know some amount of english?

Yes you can.

> I guess this is a requisite for using Linux anyway?

Yes it is.

> I know that if I had to choose between a brand-new russian translation,
> and an old, crappy english text, I'd pick the english text.   But maybe
> this is a bad anology.

It is a good analogy if you make it correct. Let your operating system be 
programmed by russian-speaking programmers, and let all of your day-to-day 
work programs be written originally in Russian. Only recently people have 
started to provide accurate english translations, but you still encounter 
Russian in a lot of places during everyday work (and thus have getten used to 
the language quite a lot). You always have to watch out for places where the 
english translation is badly inaccurate and misleading, in which cases you 
always have to check back with the Russian original.

> > As for translators -- don't underestimate them. Every translator knows
> > what he/she's doing. Every one knows that if there is an exhaustive
> > english documentation, it's gonna be a lot of work to translate it. An
> > existing but crappy documentation doesn't lower the barrier at all. It
> > just turns off users.
>
> I'm surprised.  I'd assume that translators would be happy to do a
> chapter at a time, rather than commit to doing the whole book at once.
> Doing a chapter at a time does not imply laziness, any more than
> accusing a patch submitter of being stupid for not having re-designed
> the whole module.

No, apples and oranges again. First, for programmers there are tools to enable 
partial working (the compiler checks that the code fits together). Which also 
exists for the po files, therefore these are not an issue. Secondly, if a 
patch submitter only submits parts of a feature so that it is still highly 
experimental, crashes often and misleads the user a lot, you might welcome 
the patch, but you would make sure that it isn't in the upcoming stable 
release. For programmers, such stuff would go into experimental/ so that the 
"ordinary" user will not see and touch it. Outdated translations are just 
that. So please don't let the user see and read any outdated (>12 months) and 
unmaintained translation.

Christian
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iQCVAwUBPeVGSGXAi+BfhivFAQH4jwP9GzxTood0I4YzuNrgspAbr4tghskNUUae
U6VpQA0B4UHmSaiGTtQy2bXbQvzmpS9xvhiJAdU2mS6of47qMtEDfixY5cqIV6FD
iPg9lAdN00gKcTRobRxtqjSlH/yZ3YheOgHBhPC7RNTPcqTp9XPiP4ZhxsaY7dl4
aLc21VACW7Q=
=aoLx
-----END PGP SIGNATURE-----