SeaMonkey 1.0 pl-PL for Mac

classic Classic list List threaded Threaded
24 messages Options
12
Reply | Threaded
Open this post in threaded view
|

SeaMonkey 1.0 pl-PL for Mac

Adrian (Adrianer) Kalla
Hi,

polish SeaMonkey 1.0 for Mac is avaiable:
http://prdownloads.sourceforge.net/mozillapl/seamonkey-1.0.pl-PL.mac.dmg?download

Please, upload them to ftp.mozilla.org.

Regards,
Adrian Kalla
_______________________________________________
dev-l10n mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-l10n
Reply | Threaded
Open this post in threaded view
|

Re: SeaMonkey 1.0 pl-PL for Mac

Mad Maks
Adrian (Adrianer) Kalla schreef:
> Hi,
>
> polish SeaMonkey 1.0 for Mac is avaiable:
> http://prdownloads.sourceforge.net/mozillapl/seamonkey-1.0.pl-PL.mac.dmg?download
>
> Please, upload them to ftp.mozilla.org.
>
> Regards,
> Adrian Kalla
uploaded
_______________________________________________
dev-l10n mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-l10n
Reply | Threaded
Open this post in threaded view
|

How to get a voucher?

Stephen Holt

Hello all,

I've filed a bug (https://bugzilla.mozilla.org/show_bug.cgi?id=330026)
to try to get CVS access for the Kinyarwanda Mozilla l10n project.

I'm told, however, that I need a voucher who's already inside the
project. How can I go about doing that, given that I live in Rwanda and
don't know anybody connected with the project? Is it possible for
someone from this list to QA the work and vouch? Not sure on procedure
here.

Thanks,
Steve.


               
___________________________________________________________
NEW Yahoo! Cars - sell your car and browse thousands of new and used cars online! http://uk.cars.yahoo.com/
_______________________________________________
dev-l10n mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-l10n
Reply | Threaded
Open this post in threaded view
|

Re: How to get a voucher?

Gervase Markham
Stephen Holt wrote:

> Hello all,
>
> I've filed a bug (https://bugzilla.mozilla.org/show_bug.cgi?id=330026)
> to try to get CVS access for the Kinyarwanda Mozilla l10n project.
>
> I'm told, however, that I need a voucher who's already inside the
> project. How can I go about doing that, given that I live in Rwanda and
> don't know anybody connected with the project? Is it possible for
> someone from this list to QA the work and vouch? Not sure on procedure
> here.

Well, the real answer is that we need an l10n mechanism where people can
update their localisations without having full access to the CVS tree.
Something like Pootle, in fact :-)

But the short answer is that I'll vouch for Stephen, if that's
permitted; we were at University together.

Gerv
_______________________________________________
dev-l10n mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-l10n
Reply | Threaded
Open this post in threaded view
|

Re: How to get a voucher?

Axel Hecht
Gervase Markham wrote:

> Stephen Holt wrote:
>> Hello all,
>>
>> I've filed a bug (https://bugzilla.mozilla.org/show_bug.cgi?id=330026)
>> to try to get CVS access for the Kinyarwanda Mozilla l10n project.
>>
>> I'm told, however, that I need a voucher who's already inside the
>> project. How can I go about doing that, given that I live in Rwanda and
>> don't know anybody connected with the project? Is it possible for
>> someone from this list to QA the work and vouch? Not sure on procedure
>> here.
>
> Well, the real answer is that we need an l10n mechanism where people can
> update their localisations without having full access to the CVS tree.
> Something like Pootle, in fact :-)

Well, they don't, they only have access to the l10n repository. I would
like to give a sneak preview that I don't think that the Mozilla
development process should be changed to match some webtool, too. We
have an established process to localize via DTD and properties files
(and some defines at odd ends, plus some custom files for bootstrapping
profiles), hosted on CVS. Changing that would give rise to more friction
between l10n and the development of the core product, which I don't
think is worth it. I don't see that much good coming out of some
aside-or-on-top process. It may be able to boost the shear amount of
locales available, but I don't see it helping the real problems, which
are mostly non-code paperwork processes and l10n-reviews for new code
etc. I.e., our code won't get simple to localize just because we put
another layer to it, we need to make the existing layers easier and
maybe even documented. Though, being moody as I am these days, there
doesn't seem to be any interest in docs.

> But the short answer is that I'll vouch for Stephen, if that's
> permitted; we were at University together.

Vouching for l10n includes a technical review of the localization in
question, so it's a tad different.

Axel
_______________________________________________
dev-l10n mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-l10n
Reply | Threaded
Open this post in threaded view
|

Re: How to get a voucher?

Pavel Franc
> Vouching for l10n includes a technical review of the localization in
> question, so it's a tad different.

And some must wait more 9 months to get one -
https://bugzilla.mozilla.org/show_bug.cgi?id=291275

Pavel
_______________________________________________
dev-l10n mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-l10n
Reply | Threaded
Open this post in threaded view
|

Re: How to get a voucher?

Stephen Holt
> > Vouching for l10n includes a technical review of the localization
> in
> > question, so it's a tad different.
>
> And some must wait more 9 months to get one -
> https://bugzilla.mozilla.org/show_bug.cgi?id=291275

Mmm... that's a really long time. I appreciate the need to review
carefully all inputs, but if it's going to take that long then it's
basically an unworkable system as it stands and you'll end up with
localised versions of the code sitting all over the place on local
teams' websites with all the consequent versioning issues rather than
in the central repository.

Is there nothing that can be done to speed up the process?

Steve.


               
___________________________________________________________
To help you stay safe and secure online, we've developed the all new Yahoo! Security Centre. http://uk.security.yahoo.com
_______________________________________________
dev-l10n mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-l10n
Reply | Threaded
Open this post in threaded view
|

Re: How to get a voucher?

Dwayne Bailey
In reply to this post by Axel Hecht
On Fri, 2006-03-17 at 17:49 +0100, Axel Hecht wrote:
> Gervase Markham wrote:

[snip]

> > Well, the real answer is that we need an l10n mechanism where people can
> > update their localisations without having full access to the CVS tree.
> > Something like Pootle, in fact :-)
>
> Well, they don't, they only have access to the l10n repository. I would
> like to give a sneak preview that I don't think that the Mozilla
> development process should be changed to match some webtool, too. We
> have an established process to localize via DTD and properties files
> (and some defines at odd ends, plus some custom files for bootstrapping
> profiles), hosted on CVS.

But of course you'd be open to changing the process to meet the needs of
localisers?  Especially if it meant we could have more languages of
higher quality for less work from programmers and localisers?

> Changing that would give rise to more friction
> between l10n and the development of the core product, which I don't
> think is worth it.

By the way Pootle and the Translate Toolkit don't change any of that
underlying file types as it still outputs .dtd and .properties files so
I'm not 100% sure I understand how that creates friction?  Since to a
coder they wouldn't even be aware that those translations come from PO
files.

Have I missed something?

>  I don't see that much good coming out of some
> aside-or-on-top process. It may be able to boost the shear amount of
> locales available, but I don't see it helping the real problems, which
> are mostly non-code paperwork processes and l10n-reviews for new code
> etc.

The Real Problems (TM) that you mention are mostly the l10n problems as
seen from your perspective.  Localisers feel those but also have their
own set of problems:

- How to increasing quality
- Managing delegating of work
- Working to deadlines
- Reusing translations
- etc

> I.e., our code won't get simple to localize just because we put
> another layer to it, we need to make the existing layers easier and
> maybe even documented.

Programming is not easy and neither is good localisation.  But I've
never seen a programmer not take any opportunity to make things easier
for themselves.

--
Dwayne Bailey
Translate.org.za

+27-12-460-1095 (w)
+27-83-443-7114 (cell)

_______________________________________________
dev-l10n mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-l10n
Reply | Threaded
Open this post in threaded view
|

Re: How to get a voucher?

Axel Hecht
In reply to this post by Stephen Holt
Stephen Holt wrote:

>>> Vouching for l10n includes a technical review of the localization
>> in
>>> question, so it's a tad different.
>> And some must wait more 9 months to get one -
>> https://bugzilla.mozilla.org/show_bug.cgi?id=291275
>
> Mmm... that's a really long time. I appreciate the need to review
> carefully all inputs, but if it's going to take that long then it's
> basically an unworkable system as it stands and you'll end up with
> localised versions of the code sitting all over the place on local
> teams' websites with all the consequent versioning issues rather than
> in the central repository.
>
> Is there nothing that can be done to speed up the process?
>
> Steve.
>

Oh Pavel, that was such a bad comment. Totally off. Georgian is not
waiting on technical review, but has two competing teams. That doesn't
mean that this problem is outstanding too long, but it is a completely
different one.

I'm working on making reviews more reliable, but the only way to get to
that is to have documents that describe the process. I have about three
posts recently to this group talking about docs, and I didn't get a
single comment, neither on the pages nor in the group.

There is hardly any way to speed up the process, what can be done is to
request approval-l10n on the zip, that way, it doesn't get lost. That
looks like speed up, still.

Axel
_______________________________________________
dev-l10n mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-l10n
Reply | Threaded
Open this post in threaded view
|

Re: How to get a voucher?

Axel Hecht
In reply to this post by Axel Hecht
Dwayne Bailey wrote:

> On Fri, 2006-03-17 at 17:49 +0100, Axel Hecht wrote:
>> Gervase Markham wrote:
>
> [snip]
>
>>> Well, the real answer is that we need an l10n mechanism where people can
>>> update their localisations without having full access to the CVS tree.
>>> Something like Pootle, in fact :-)
>> Well, they don't, they only have access to the l10n repository. I would
>> like to give a sneak preview that I don't think that the Mozilla
>> development process should be changed to match some webtool, too. We
>> have an established process to localize via DTD and properties files
>> (and some defines at odd ends, plus some custom files for bootstrapping
>> profiles), hosted on CVS.
>
> But of course you'd be open to changing the process to meet the needs of
> localisers?  Especially if it meant we could have more languages of
> higher quality for less work from programmers and localisers?

If that happened, sure. From my experience, though, we don't have a
problem with the processes itself as much. What we lack at is to
actually define those processes (yes, stuff like that floats with
contradicting expectations from all sides), and write them down. Insert
doc-rant-mantra.

As a comment on the basic guideline, number of languages is about as a
good argument as download numbers.

>> Changing that would give rise to more friction
>> between l10n and the development of the core product, which I don't
>> think is worth it.
>
> By the way Pootle and the Translate Toolkit don't change any of that
> underlying file types as it still outputs .dtd and .properties files so
> I'm not 100% sure I understand how that creates friction?  Since to a
> coder they wouldn't even be aware that those translations come from PO
> files.
>
> Have I missed something?

What you're doing is adding a layer on top of the existing complexity.
In my math head, that can only add to the complexity of the whole system.

Technically, there are other file formats in l10n, like bookmarks,
search plugins (format changing, and not final yet), include files,
binary images (screen shots, search, for now), help files (html and
rdf). Maybe we end up with yet another format for places. If I ever find
out what needs to be in there [2].

>>  I don't see that much good coming out of some
>> aside-or-on-top process. It may be able to boost the shear amount of
>> locales available, but I don't see it helping the real problems, which
>> are mostly non-code paperwork processes and l10n-reviews for new code
>> etc.
>
> The Real Problems (TM) that you mention are mostly the l10n problems as
> seen from your perspective.  Localisers feel those but also have their
> own set of problems:
>
> - How to increasing quality

Interesting and tricky point. Not so much from the actual work, again
from my side. But when to land those changes. There is some ongoing
discussion inside those driving the security releases on whether
improving the quality of a localization is actually something that
should ship with a security release. The answer is currently "not
really". We had one or two fixes of real problems (broken bookmarks, for
example), though. Any change comes with risk, and for our release
management, localization changes are mostly just risk, with limited benefit.
The challenge I see is to get the l10n quality for a release to a point
where we don't have to bother about this. This includes things like
reliable nightlies from the right trees, good l10n freezes, properly
laid out trademarks policies (to free resources for the real work),
probably others.

> - Managing delegating of work
> - Working to deadlines

Oh, right, deadlines.

> - Reusing translations
> - etc
>
>> I.e., our code won't get simple to localize just because we put
>> another layer to it, we need to make the existing layers easier and
>> maybe even documented.
>
> Programming is not easy and neither is good localisation.  But I've
> never seen a programmer not take any opportunity to make things easier
> for themselves.
>

Cute story about "easier for themselves". default_places.html [1]. It is
up to now totally unspecified what should be in that file [2]. It
doesn't have a trademarks policy, either. It reuses some parts of the
bookmarks.html syntax, but adds more attributes. It ends up in en-US.jar
at chrome://browser/locale/places/default_places.html [3]. Nice for
multiple locale installs, but does the CCK know?

"Easier" is a very local measure when it comes down to getting people to
buy in for change.

This is of course not a general veto to change, I just want to remind
you that changing people is something that needs really good incentives.
And that you won't make me buy change by convincing me, you have to
convince me to fight with some 60 people for about a quarter, and nag
them for about a year. OTH, it would probably be another business trip
to the US ;-).

Axel

[1]
http://lxr.mozilla.org/mozilla1.8/source/browser/locales/en-US/chrome/browser/places/default_places.html
[2] https://bugzilla.mozilla.org/show_bug.cgi?id=329657
[3] http://lxr.mozilla.org/mozilla1.8/source/browser/locales/jar.mn#29
_______________________________________________
dev-l10n mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-l10n
Reply | Threaded
Open this post in threaded view
|

Re: How to get a voucher?

Gervase Markham
Axel Hecht wrote:
 > What you're doing is adding a layer on top of the existing complexity.
> In my math head, that can only add to the complexity of the whole system.

That's certainly not always true - or rather, it depends on from whose
viewpoint you see "complexity".

LyX is a layer on top of LaTeX. Does that make the system more complex?
Yes, in one sense - there's more software. But definitely not in another
sense - LyX is far easier to use than typing raw LaTeX into a text editor.

> This is of course not a general veto to change, I just want to remind
> you that changing people is something that needs really good incentives.

I'm not certain that Dwayne is necessarily suggesting that people have
to change. As I understand it, translating via Pootle could be provided
as an option for those groups which wish to use it.

Gerv
_______________________________________________
dev-l10n mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-l10n
Reply | Threaded
Open this post in threaded view
|

Re: How to get a voucher?

Axel Hecht
Gervase Markham wrote:

> Axel Hecht wrote:
>  > What you're doing is adding a layer on top of the existing complexity.
>> In my math head, that can only add to the complexity of the whole system.
>
> That's certainly not always true - or rather, it depends on from whose
> viewpoint you see "complexity".
>
> LyX is a layer on top of LaTeX. Does that make the system more complex?
> Yes, in one sense - there's more software. But definitely not in another
> sense - LyX is far easier to use than typing raw LaTeX into a text editor.

The difference is, LyX can safely ignore those parts of LaTeX it doesn't
want to care about. And neither does LyX have any impact on developments
in LaTeX.

Note that we already ran into problems with core developers like darin
or Annie not worrying about the l10n toolchain and just assuming their
tools to be good enough.
(Introducing preprocessor defines in l10n, that gave rise to
http://developer.mozilla.org/en/docs/Writing_localizable_code)

>> This is of course not a general veto to change, I just want to remind
>> you that changing people is something that needs really good incentives.
>
> I'm not certain that Dwayne is necessarily suggesting that people have
> to change. As I understand it, translating via Pootle could be provided
> as an option for those groups which wish to use it.

Given that we don't give write access to our repositories to pserver, do
we want to grant write access to a web app with a different stack of
apps to watch for vulnerabilities? Just one aspect of many of that kind.

I'm not saying that there isn't a good use for something like pootle for
some communities, but I don't see mozilla hosting that, let alone give
it direct write access. At least not in the short-to-mid term. I think
that the value is bigger for cross-application l10n teams which use the
same scheme for other applications. Those teams can benefit from such a
tool much more than others, for which the break of toolchain for
help/search and images etc (see other post) may be more of an impact.

Axel
_______________________________________________
dev-l10n mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-l10n
Reply | Threaded
Open this post in threaded view
|

Re: How to get a voucher?

Damjan Georgievski
> Given that we don't give write access to our repositories to pserver,
                                        ^^^^^^

Axel, are you the voice of the community in the MoFo (like you said when you
were appointed the l10n officer), or the other way around?


--
damjan
_______________________________________________
dev-l10n mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-l10n
Reply | Threaded
Open this post in threaded view
|

Re: How to get a voucher?

Axel Hecht
Damjan Georgievski wrote:
>> Given that we don't give write access to our repositories to pserver,
>                                         ^^^^^^
>
> Axel, are you the voice of the community in the MoFo (like you said when you
> were appointed the l10n officer), or the other way around?

Both ways. The "sheppard" picture fits in some aspects, though I won't
stick to that image when you come up with an aspect that I see not fitting.

In addition, I also do some abstracting, that is, I try to limit the
load of l10n on both QA and build. I do this in a pretty opaque way so
that those only get heat from me, which may not be the best solution,
but I haven't convinced myself that doing that more transparent would
yield an overall improvement. To some extent that is because I know that
both departments are understaffed, and that l10n doesn't have a "budget"
in either of them. That is, I can't go there and say "you owe l10n N
hours this week still, jump!" So instead I assume they're flooded and I
may end up over-managing the workload. Which ends up in an odd mix of me
intentionally not forwarding stuff and dropping some.

Axel
_______________________________________________
dev-l10n mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-l10n
Reply | Threaded
Open this post in threaded view
|

L10n docs (was Re: How to get a voucher?)

Ricardo Palomares Martí­nez
In reply to this post by Axel Hecht
Axel Hecht escribió:
> I'm working on making reviews more reliable, but the only way to get to
> that is to have documents that describe the process. I have about three
> posts recently to this group talking about docs, and I didn't get a
> single comment, neither on the pages nor in the group.


I can only agree, we are not being very helpful with you in this
topic. Are we (the translation teams) really so autonomus that we
don't need to share our experiences? I'm getting help requests by
e-mail from a guy interested in translating Nvu 1.0 to portuguese, and
I'm realizing that not only the doc at
http://glazman.org/nvu/NVU-L10N.html is not enough for some people,
but even my own helper scripts for Nvu translation are
under-documented to simply hand them to anyone else without further
explanations.

At FOSDEM, IIRC, it turned out that about four people are still using
Mozilla Translator as its main L10n tool, two or three use Pootle and
most teams use simply UTF-8 editors (and a lot of helper scripts, I
suppose). Given that Mozilla Translator has some serious limitations
(and it's not trivial at all to fix them) and that using plain text
editors must involve a lot of repetitive work and "tricks" to reuse
translations, there *must* be things that we can share.

Speaking for myself, I use MT and have Bash scripts for these things:

- to merge all the locales in a (configurable) number of JARs (like,
en-US.jar, chatzilla.jar, reporter.jar, etc.) in repackaged en-US.jar

- to fix a MT generated JAR in order to add inside it "verbatim" en-US
copies of the infamous brand.dtd that MT is unable to parse (I have
used this also for Sunbird).

- to automatically unzip the MT generated langpack, run the previous
script, and repack the langpack with the fixed JAR

A document with a proposed directory layout for using MT to translate
SeaMonkey and other products is something I could also prepare.

The point is, would these things be useful for anyone? I'm already
short of time and I wouldn't like to spend time in something no one
really needs.

Overall, I must say, there is still a really sheer learning curve for
a newcomer that wants to help but doesn't know much about computers
(and I say not only in reference to english documents, but also
talking about our own docs in es-ES translation team website). IMHO,
translating Mozilla (I don't know if it is the case for every L10n
project) is more a task for a computer-savvy guy than for a good
translator.

I think that, if we want to take the l10n documentation seriously, we
need to start by writing down a list of main subcategories for the
entry page at
http://developer.mozilla.org/en/docs/Category:Localization [1]. Then,
each team should browse it and make an attempt to write something
useful (or translate documents they already have in their native
language) in an area in which they feel that fit in. Some areas (like
organization, registering as a L10n team, etc.) may have to be left to
you, Axel, or other people close or inside MoFo.

[1] A minimal draft list for subcategories:

- Products and projects covered (grouping them by L10n procedure)
- Registering as a translator (for official MoFo products and for
  projects and extensions)
- Different L10n Processes
- Tools, Scripts and Hacks
- Other L10n not related to products/projects/extensions (MDC, Mozilla
  Europe/Japan/etc...)
- L10n references and guides for Mozilla developers
- Non-Mozilla L10n References

This probably could be shortened further. IMHO, ideally shouldn't be
any document directly hanging from the main Localization category.

Two more things to do:

- physically delete old documentation, it may confuse people. For
instance, in
http://developer.mozilla.org/en/docs/Category:Localization:Tools there
is a link to http://www.mozilla.org/projects/l10n/mlp_tools.html that
lists tools that I don't really think that still work (like MozExpTool
or L10nZilla). It also doesn't mention the problems that may arise by
using some of the tools, like MT.

- start with gathering the docs related to the source l10n process
(Fx/Tb), then complete them with specific instructions for every tool
in use as of today, and then expand to other processes, like the one
used for SeaMonkey and Sunbird, or the different processes for extensions.

Thank you for reading so far. :-)

--
If it's true that we are here to help others,
then what exactly are the OTHERS here for?
_______________________________________________
dev-l10n mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-l10n
Reply | Threaded
Open this post in threaded view
|

Re: L10n docs (was Re: How to get a voucher?)

Simos Xenitellis
O/H Ricardo Palomares Martinez έγραψε:

> Axel Hecht escribió:
>  
>> I'm working on making reviews more reliable, but the only way to get to
>> that is to have documents that describe the process. I have about three
>> posts recently to this group talking about docs, and I didn't get a
>> single comment, neither on the pages nor in the group.
>>    
>
>
> I can only agree, we are not being very helpful with you in this
> topic. Are we (the translation teams) really so autonomus that we
> don't need to share our experiences? I'm getting help requests by
> e-mail from a guy interested in translating Nvu 1.0 to portuguese, and
> I'm realizing that not only the doc at
> http://glazman.org/nvu/NVU-L10N.html is not enough for some people,
> but even my own helper scripts for Nvu translation are
> under-documented to simply hand them to anyone else without further
> explanations.
>
> At FOSDEM, IIRC, it turned out that about four people are still using
> Mozilla Translator as its main L10n tool, two or three use Pootle and
> most teams use simply UTF-8 editors (and a lot of helper scripts, I
> suppose). Given that Mozilla Translator has some serious limitations
> (and it's not trivial at all to fix them) and that using plain text
> editors must involve a lot of repetitive work and "tricks" to reuse
> translations, there *must* be things that we can share.
>
> Speaking for myself, I use MT and have Bash scripts for these things:
>
> - to merge all the locales in a (configurable) number of JARs (like,
> en-US.jar, chatzilla.jar, reporter.jar, etc.) in repackaged en-US.jar
>
> - to fix a MT generated JAR in order to add inside it "verbatim" en-US
> copies of the infamous brand.dtd that MT is unable to parse (I have
> used this also for Sunbird).
>
> - to automatically unzip the MT generated langpack, run the previous
> script, and repack the langpack with the fixed JAR
>
> A document with a proposed directory layout for using MT to translate
> SeaMonkey and other products is something I could also prepare.
>
> The point is, would these things be useful for anyone? I'm already
> short of time and I wouldn't like to spend time in something no one
> really needs.
>
> Overall, I must say, there is still a really sheer learning curve for
> a newcomer that wants to help but doesn't know much about computers
> (and I say not only in reference to english documents, but also
> talking about our own docs in es-ES translation team website). IMHO,
> translating Mozilla (I don't know if it is the case for every L10n
> project) is more a task for a computer-savvy guy than for a good
> translator.
>
> I think that, if we want to take the l10n documentation seriously, we
> need to start by writing down a list of main subcategories for the
> entry page at
> http://developer.mozilla.org/en/docs/Category:Localization [1]. Then,
> each team should browse it and make an attempt to write something
> useful (or translate documents they already have in their native
> language) in an area in which they feel that fit in. Some areas (like
> organization, registering as a L10n team, etc.) may have to be left to
> you, Axel, or other people close or inside MoFo.
>
> [1] A minimal draft list for subcategories:
>
> - Products and projects covered (grouping them by L10n procedure)
> - Registering as a translator (for official MoFo products and for
>   projects and extensions)
> - Different L10n Processes
> - Tools, Scripts and Hacks
> - Other L10n not related to products/projects/extensions (MDC, Mozilla
>   Europe/Japan/etc...)
> - L10n references and guides for Mozilla developers
> - Non-Mozilla L10n References
>
> This probably could be shortened further. IMHO, ideally shouldn't be
> any document directly hanging from the main Localization category.
>
> Two more things to do:
>
> - physically delete old documentation, it may confuse people. For
> instance, in
> http://developer.mozilla.org/en/docs/Category:Localization:Tools there
> is a link to http://www.mozilla.org/projects/l10n/mlp_tools.html that
> lists tools that I don't really think that still work (like MozExpTool
> or L10nZilla). It also doesn't mention the problems that may arise by
> using some of the tools, like MT.
>
> - start with gathering the docs related to the source l10n process
> (Fx/Tb), then complete them with specific instructions for every tool
> in use as of today, and then expand to other processes, like the one
> used for SeaMonkey and Sunbird, or the different processes for extensions.
>
> Thank you for reading so far. :-)
>  
Thanks for writing this.
Such a guide would be really useful.
If you find it difficult to write a guide in the form of documentation,
you can provide a complete checklist
for the process (with scripts and so on) having in mind a rather
advanced user.
This way they will be able to help out writing the proper documentation.

I am not sure what the state is regarding the conversion to PO and back,
using the tools from translate.sf.net.
Dwayne, David, are you around?
Is MT not developed any more, so you have to do some of the work with
scripts?
There are existing tools and experience from other l10n projects with
the PO format, shall we stress there?


Simos
_______________________________________________
dev-l10n mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-l10n
Reply | Threaded
Open this post in threaded view
|

Re: L10n docs (was Re: How to get a voucher?)

Axel Hecht
In reply to this post by Ricardo Palomares Martí­nez
Simos Xenitellis wrote:
> O/H Ricardo Palomares Martinez έγραψε:
>> Axel Hecht escribió:

<...>

>>  
> Thanks for writing this.
> Such a guide would be really useful.
> If you find it difficult to write a guide in the form of documentation,
> you can provide a complete checklist
> for the process (with scripts and so on) having in mind a rather
> advanced user.
> This way they will be able to help out writing the proper documentation.

This is what I'm doing ;-)

> I am not sure what the state is regarding the conversion to PO and back,
> using the tools from translate.sf.net.
> Dwayne, David, are you around?
> Is MT not developed any more, so you have to do some of the work with
> scripts?
> There are existing tools and experience from other l10n projects with
> the PO format, shall we stress there?

Our documentation should be tool-agnostic. That is, we should describe
as many schemes as possible, introducing them, describing them, but not
recommending them. NPOV in wikipedia-speak,
http://en.wikipedia.org/wiki/Wikipedia:NPOV.

Axel
_______________________________________________
dev-l10n mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-l10n
Reply | Threaded
Open this post in threaded view
|

Re: L10n docs (was Re: How to get a voucher?)

Oscar Manuel Gómez Senovilla
In reply to this post by Ricardo Palomares Martí­nez
Simos Xenitellis escribió:

> Thanks for writing this.
> Such a guide would be really useful.
> If you find it difficult to write a guide in the form of documentation,
> you can provide a complete checklist
> for the process (with scripts and so on) having in mind a rather
> advanced user.
> This way they will be able to help out writing the proper documentation.
>
> I am not sure what the state is regarding the conversion to PO and back,
> using the tools from translate.sf.net.
> Dwayne, David, are you around?
> Is MT not developed any more, so you have to do some of the work with
> scripts?
> There are existing tools and experience from other l10n projects with
> the PO format, shall we stress there?


Well, meanwhile David or Dwayne arise, I can give you some cookies :)


I described the process I did for thunderbird 1.5 with potools in

http://sourceforge.net/mailarchive/message.php?msg_id=14675170

That needs some fix, mostly because the current cvs fixes some of the
issues, and the scripts have some (minor) improvement. I have also
pending to organize that with existing wiki docs.


Also, I wouldn't like to "break" David's role, but next Thursday there
will be a bug-day. You can read the full text here:

http://sourceforge.net/mailarchive/forum.php?thread_id=9994757&forum_id=36744

Personally, I'd like feedback about the tools you know and/or intend to
use for po files (not sure is this ng is the best place).


Regards.
_______________________________________________
dev-l10n mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-l10n
Reply | Threaded
Open this post in threaded view
|

Re: L10n docs (was Re: How to get a voucher?)

Axel Hecht
Oscar Manuel Gómez Senovilla wrote:

> Simos Xenitellis escribió:
>> Thanks for writing this.
>> Such a guide would be really useful.
>> If you find it difficult to write a guide in the form of documentation,
>> you can provide a complete checklist
>> for the process (with scripts and so on) having in mind a rather
>> advanced user.
>> This way they will be able to help out writing the proper documentation.
>>
>> I am not sure what the state is regarding the conversion to PO and back,
>> using the tools from translate.sf.net.
>> Dwayne, David, are you around?
>> Is MT not developed any more, so you have to do some of the work with
>> scripts?
>> There are existing tools and experience from other l10n projects with
>> the PO format, shall we stress there?
>
>
> Well, meanwhile David or Dwayne arise, I can give you some cookies :)
>
>
> I described the process I did for thunderbird 1.5 with potools in
>
> http://sourceforge.net/mailarchive/message.php?msg_id=14675170
>
> That needs some fix, mostly because the current cvs fixes some of the
> issues, and the scripts have some (minor) improvement. I have also
> pending to organize that with existing wiki docs.
>

That sounds like a good start, could you join that together with some of
the tricks in
http://wiki.mozilla.org/Draft:L10n:Creating_a_new_Localization (I should
probably move that over to devmo) for check-out etc on devmo?

That'd be great.

Axel

>
> Also, I wouldn't like to "break" David's role, but next Thursday there
> will be a bug-day. You can read the full text here:
>
> http://sourceforge.net/mailarchive/forum.php?thread_id=9994757&forum_id=36744
>
> Personally, I'd like feedback about the tools you know and/or intend to
> use for po files (not sure is this ng is the best place).
>
>
> Regards.
_______________________________________________
dev-l10n mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-l10n
Reply | Threaded
Open this post in threaded view
|

Re: L10n docs (was Re: How to get a voucher?)

Axel Hecht
In reply to this post by Ricardo Palomares Martí­nez
Crossposting to .l10n and .mdc to join forces here.

I'd be thankful for the mdc folks to share some experience on
- creating a new block of docs
- migrating existing docs to devmo.

I added a XXX devmo start to my full copy for those that want to easily
skip the more l10n specific part.

Some of my own comments are below.

Ricardo Palomares Martinez wrote:

> Axel Hecht escribió:
>> I'm working on making reviews more reliable, but the only way to get to
>> that is to have documents that describe the process. I have about three
>> posts recently to this group talking about docs, and I didn't get a
>> single comment, neither on the pages nor in the group.
>
>
> I can only agree, we are not being very helpful with you in this
> topic. Are we (the translation teams) really so autonomus that we
> don't need to share our experiences? I'm getting help requests by
> e-mail from a guy interested in translating Nvu 1.0 to portuguese, and
> I'm realizing that not only the doc at
> http://glazman.org/nvu/NVU-L10N.html is not enough for some people,
> but even my own helper scripts for Nvu translation are
> under-documented to simply hand them to anyone else without further
> explanations.
>
> At FOSDEM, IIRC, it turned out that about four people are still using
> Mozilla Translator as its main L10n tool, two or three use Pootle and
> most teams use simply UTF-8 editors (and a lot of helper scripts, I
> suppose). Given that Mozilla Translator has some serious limitations
> (and it's not trivial at all to fix them) and that using plain text
> editors must involve a lot of repetitive work and "tricks" to reuse
> translations, there *must* be things that we can share.
>
> Speaking for myself, I use MT and have Bash scripts for these things:
>
> - to merge all the locales in a (configurable) number of JARs (like,
> en-US.jar, chatzilla.jar, reporter.jar, etc.) in repackaged en-US.jar
>
> - to fix a MT generated JAR in order to add inside it "verbatim" en-US
> copies of the infamous brand.dtd that MT is unable to parse (I have
> used this also for Sunbird).
>
> - to automatically unzip the MT generated langpack, run the previous
> script, and repack the langpack with the fixed JAR
>
> A document with a proposed directory layout for using MT to translate
> SeaMonkey and other products is something I could also prepare.
>
> The point is, would these things be useful for anyone? I'm already
> short of time and I wouldn't like to spend time in something no one
> really needs.
>
> Overall, I must say, there is still a really sheer learning curve for
> a newcomer that wants to help but doesn't know much about computers
> (and I say not only in reference to english documents, but also
> talking about our own docs in es-ES translation team website). IMHO,
> translating Mozilla (I don't know if it is the case for every L10n
> project) is more a task for a computer-savvy guy than for a good
> translator.


XXX devmo start


> I think that, if we want to take the l10n documentation seriously, we
> need to start by writing down a list of main subcategories for the
> entry page at
> http://developer.mozilla.org/en/docs/Category:Localization [1]. Then,
> each team should browse it and make an attempt to write something
> useful (or translate documents they already have in their native
> language) in an area in which they feel that fit in. Some areas (like
> organization, registering as a L10n team, etc.) may have to be left to
> you, Axel, or other people close or inside MoFo.
>
> [1] A minimal draft list for subcategories:
>
> - Products and projects covered (grouping them by L10n procedure)
> - Registering as a translator (for official MoFo products and for
>   projects and extensions)
> - Different L10n Processes
> - Tools, Scripts and Hacks
> - Other L10n not related to products/projects/extensions (MDC, Mozilla
>   Europe/Japan/etc...)
> - L10n references and guides for Mozilla developers
> - Non-Mozilla L10n References
>
> This probably could be shortened further. IMHO, ideally shouldn't be
> any document directly hanging from the main Localization category.

The way I understand it, we should just create content and then
iteratively tie it to a web of integrated docs. Your plan just looks
like too much planning and too little documentation to me.

The scheme I follow is, if I get a question to be answered by mail, or
news, I write the answer (somewhat generalized) to the wiki and reply
with a link. At least that's what I try.

Which brings up a question to devmo's, do we have a clear distinction of
what should go on wiki.m.o and what should go on devmo?
http://wiki.mozilla.org/Draft:L10n:Creating_a_new_Localization for
example looks to me like something I should have created over at devmo,
is that right?

> Two more things to do:
>
> - physically delete old documentation, it may confuse people. For
> instance, in
> http://developer.mozilla.org/en/docs/Category:Localization:Tools there
> is a link to http://www.mozilla.org/projects/l10n/mlp_tools.html that
> lists tools that I don't really think that still work (like MozExpTool
> or L10nZilla). It also doesn't mention the problems that may arise by
> using some of the tools, like MT.

This is the "migrate existing pages" part. How did we did that for other
devmo parts? I'm especially interested on how to split up the work, when
to fix links on the imported content etc.

> - start with gathering the docs related to the source l10n process
> (Fx/Tb), then complete them with specific instructions for every tool
> in use as of today, and then expand to other processes, like the one
> used for SeaMonkey and Sunbird, or the different processes for extensions.
>
> Thank you for reading so far. :-)
>

Thanks for your reply, it was a pleasure reading.

Axel
_______________________________________________
dev-l10n mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-l10n
12