Re: Idea: each FAQ as its own xml file

Michèle Garoche <[email protected]>
Newsgroups gmane.os.apple.fink.i18n
Message-ID <[email protected]>
Le 19 août 2004, à 4:17, Daniel Macks a écrit :

> (Sorry I haven't had a chance to respond to the past day's messages)
>
> First, it is not my goal to force this (or any) new system on the doc
> team. However, there's no denying that some translations appear to
> have stagnated. Whether that's because nobody is volunteering to do
> the work or nobody is coordinating the effort, I don't know.
Both certainly. And here we can make a change.

>  But I do
> know that the current system seems (to me) complicated and overly
> administrative. If the problem is lack of coordination, then changing
> to a system that requires less coordination seems smart. If the
> problem is lack of volunteer translator time, then switching to a
> system that keeps the ongoing work as easily-identified bite-sized
> chunks seems smart. That's why I proposed what I did. I'm happy to
> work on this, but really, it's up to the doc team to figure out what
> (if any) changes to make.
Problem is where is the doc team? We are alone (baba, you and me) to 
discuss this here. That's where things are not satisfactory.

> So, onward. Baba's thought of having each language default to the
> English if the native translation isn't available is interesting.
I think this is the better thing to do, because finally that the result 
which is wanted. I simply say I don't see how you can do that in such a 
way that you are sure you don't put additional burden on the translator 
or the documentarian.

>  I
> wasn't even thinking of something that drastic, though...just put each
> FAQ in its own file and then have Makefile string them all together
> with a simple skeleton to get back to what we have now.
That for me is additional burden on translator or documentarian.

>
> Consider: each FAQ is a (mostly) self-contained entity.
That's the mostly which is the problem here.

>  In a given
> non-.en language, it could {not exist, be outdated, be up-to-date,
> exist but be deleted in .en}, and this is fairly independent of the
> state of any other FAQ entry.
Not always independent, especially in the case of deletion.

>  By tagging (CVS commit message) each
> with what .en version it corresponds to, it removes the need to keep
> any sort of master record of what changes to "the FAQ" still need to
> be translated: just compare the .ja commit msg to the .en HEAD
> revision number.
But that you can do it right now. No need to split the files.

Observe what happens if a change is made in English file because of 
misspelling. Will you tag the other language files with outdated? This 
is fully wrong here.

>  It also makes it easier for a team to translate "a
> part at a time" and require less coordination (both manpower and CVS):
> there's no need to have to re-integrate a translator's changes for one
> FAQ into a huge .zh that may haev changed in other places in the mean
> time.
I don't agree here. It's fare more easier to translate a part of a huge 
file and integrate new changes than to keep traces of a huge numbers of 
files and know at any time which one needs translation or which one is 
up to date.

I mean the less files you have, the easier is the job.

> Our build system is already quite complicated and a bit fragile, and
> there are certainly modifications that could be made, and even whole
> other ways of doing things (with both pros and cons, and attendant
> migration issues and learning curve). But if people are happy with the
> current system, that's fine too.
No, I don't say I'm happy with the current system, though it is not 
that bad. But I really would like to hear some other translators here. 
I feel like the bad girl who always contradicts.

Michèle
<http://micmacfr.homeunix.org>
PGP.sig (application/pgp-signature, 186 B) - not displayed
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.