Multi-language to be or not to be

I'm working on a rewrite and major cleanup of parts of the code that
runs the main website, and came across a thought.

We currently have support for translation of news, events, surveys,
quotes and pages. So far, we have zero news, zero events, zero surveys,
zero quotes and zero pages translated.

The only translated pages are the press information ones, and I don't
count them because we don't actually use the translation infrastructure
for them - we link specifically to each and every page.

All the other translations live on completely separate sites - I think
because they only translate parts of the information and also provide a
lot of native information in that language.

I know adding translation capabilities was one of the big requirements
for the last rewrite. But I'd still like to bring up the question -
should we bother keeping it, or just get rid of it? We thought we had a
need then, but do we still have it?

Getting rid of it will significantly simplify some code, and it will
make the static mirrors about half the size they are now (we currently
mirror every page both in English and in language-neutral (which means
English again)).

Thoughts?

//Magnus

On Sunday 11 February 2007 08:22, Magnus Hagander wrote:
> I'm working on a rewrite and major cleanup of parts of the code that
> runs the main website, and came across a thought.
>
> We currently have support for translation of news, events, surveys,
> quotes and pages. So far, we have zero news, zero events, zero surveys,
> zero quotes and zero pages translated.
>
> The only translated pages are the press information ones, and I don't
> count them because we don't actually use the translation infrastructure
> for them - we link specifically to each and every page.
>
> All the other translations live on completely separate sites - I think
> because they only translate parts of the information and also provide a
> lot of native information in that language.
>
> I know adding translation capabilities was one of the big requirements
> for the last rewrite. But I'd still like to bring up the question -
> should we bother keeping it, or just get rid of it? We thought we had a
> need then, but do we still have it?
>
> Getting rid of it will significantly simplify some code, and it will
> make the static mirrors about half the size they are now (we currently
> mirror every page both in English and in language-neutral (which means
> English again)).
>

See recent post on advocacy from italian community asking about providing
translation capabilities. We need to talk to them and see if we have
something that they want to use; if we don't then I have no objections to
removing it (though fwiw I don't find it a terrible burden as it currently
stands)

--
Robert Treat
Build A Brighter LAMP :: Linux Apache {middleware} PostgreSQL

Magnus Hagander wrote:
> We currently have support for translation of news, events, surveys,
> quotes and pages. So far, we have zero news, zero events, zero
> surveys, zero quotes and zero pages translated.

Where do we find information on how to translate the website? Searching
the web site for "website translation" doesn't show anything obvious.

--
Peter Eisentraut
http://developer.postgresql.org/~petere/

> I know adding translation capabilities was one of the big requirements
> for the last rewrite. But I'd still like to bring up the question -
> should we bother keeping it, or just get rid of it? We thought we had a
> need then, but do we still have it?

No we don't need them. Most of the non-english sites (I say most because
I am sure there is one, on some island in the Pacific that requires this
ability but only has 1 user) have their own language specific websites
already.

The reality is this, the majority (likely > 98%) is English speaking.
All development is done, in English, the code is in English.

The only substantial community out there that *may* not have english
speakers is JPUG and they have an entire infrastructure on their own.

The sites out there such as PostgreSQLFR that are in different languages
don't use the main Web code base anyway. They use their own.

Just my 10 cents.

Sincerely,

Joshua D. Drake

--

=== The PostgreSQL Company: Command Prompt, Inc. ===
Sales/Support: +1.503.667.4564 || 24x7/Emergency: +1.800.492.2240
Providing the most comprehensive PostgreSQL solutions since 1997
http://www.commandprompt.com/

Donate to the PostgreSQL Project: http://www.postgresql.org/about/donate
PostgreSQL Replication: http://www.commandprompt.com/products/

Joshua D. Drake wrote:
> The reality is this, the majority (likely > 98%) is English speaking.

of what?

--
Peter Eisentraut
http://developer.postgresql.org/~petere/

Peter Eisentraut wrote:
> Joshua D. Drake wrote:
>> The reality is this, the majority (likely > 98%) is English speaking.
>
> of what?
>

The postgresql community.

Sincerely,

Joshua D. Drake

--

=== The PostgreSQL Company: Command Prompt, Inc. ===
Sales/Support: +1.503.667.4564 || 24x7/Emergency: +1.800.492.2240
Providing the most comprehensive PostgreSQL solutions since 1997
http://www.commandprompt.com/

Donate to the PostgreSQL Project: http://www.postgresql.org/about/donate
PostgreSQL Replication: http://www.commandprompt.com/products/

On 2/11/07, Magnus Hagander <magnus(at)hagander(dot)net> wrote:
> I'm working on a rewrite and major cleanup of parts of the code that
> runs the main website, and came across a thought.
>
> We currently have support for translation of news, events, surveys,
> quotes and pages. So far, we have zero news, zero events, zero surveys,
> zero quotes and zero pages translated.

Hello,

In my opinion the translation infrastructure is not complete and does not allow
translations to be done easily. It's lacking some way of tracking
which English pages
were modified, so that the translator would know what pages should be updated.

About 2 years ago, I have started to translate the website to Romanian. At that
time I had almost all the pages done, except for the weekly news.

I had trouble with setting up a local copy of the website for testing
purposes ,
then the time passed by without any progress.

The big question is : how is someone supposed to synchronise his translation
with the original English pages (which are updated frequently) ?

A way to accomplish this might be to write inside the translated page the CVS
revision of the English one, and write some tool which would report the pages
were modified in the meantime.
But this is ugly , and the feature should be provided by the
translation infrastructure.

Cheers,
Adrian Maier

Joshua D. Drake wrote:
> Peter Eisentraut wrote:
> > Joshua D. Drake wrote:
> >> The reality is this, the majority (likely > 98%) is English
> >> speaking.
> >
> > of what?
>
> The postgresql community.

And that statistic is derived from what sources?

--
Peter Eisentraut
http://developer.postgresql.org/~petere/

Peter Eisentraut wrote:
> Joshua D. Drake wrote:
>> Peter Eisentraut wrote:
>>> Joshua D. Drake wrote:
>>>> The reality is this, the majority (likely > 98%) is English
>>>> speaking.
>>> of what?
>> The postgresql community.
>
> And that statistic is derived from what sources?

*sigh*, I am not going to argue the obvious with you. If you don't
agree, don't agree.

Joshua D. Drake

>

--

=== The PostgreSQL Company: Command Prompt, Inc. ===
Sales/Support: +1.503.667.4564 || 24x7/Emergency: +1.800.492.2240
Providing the most comprehensive PostgreSQL solutions since 1997
http://www.commandprompt.com/

Donate to the PostgreSQL Project: http://www.postgresql.org/about/donate
PostgreSQL Replication: http://www.commandprompt.com/products/

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

- --On Sunday, February 11, 2007 16:33:21 -0800 "Joshua D. Drake"
<jd(at)commandprompt(dot)com> wrote:

> Peter Eisentraut wrote:
>> Joshua D. Drake wrote:
>>> Peter Eisentraut wrote:
>>>> Joshua D. Drake wrote:
>>>>> The reality is this, the majority (likely > 98%) is English
>>>>> speaking.
>>>> of what?
>>> The postgresql community.
>>
>> And that statistic is derived from what sources?
>
> *sigh*, I am not going to argue the obvious with you. If you don't
> agree, don't agree.

I personally don't agree either ... I would guess that the majority of
PostgreSQL users are non-English as *first* language ... just because they can
communicate in English doesn't mean that their are either fluent in, or
comfortable with, the language ...

I'd love to see our demographics some day, but would guess that the Japanese
themselves are a fairly high percentage of users, based on how they've
organized over there ... Spanish is probably another high percentage language
... French, German, Chinese (Mandarin?), Arabic ... if you look at 'first
languages', I suspect that "English" is alot smaller then 98% ... whether its
less then 50%, that I wouldn't hazard a guess ...

Personally, I think it almost does us a dis-service statistically not to have
more translations available on our main web site ...

- ----
Marc G. Fournier Hub.Org Networking Services (http://www.hub.org)
Email . scrappy(at)hub(dot)org MSN . scrappy(at)hub(dot)org
Yahoo . yscrappy Skype: hub.org ICQ . 7615664
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (FreeBSD)

iD8DBQFFz9+x4QvfyHIvDvMRAuyXAJ9t5Y6bcwaOuw+HCFVstOLSsuL4VACgi9iP
azNDNscxt7jNIpmJID1a8g8=
=P6cB
-----END PGP SIGNATURE-----

On Sunday 11 February 2007 16:22, Adrian Maier wrote:
> On 2/11/07, Magnus Hagander <magnus(at)hagander(dot)net> wrote:
> > I'm working on a rewrite and major cleanup of parts of the code that
> > runs the main website, and came across a thought.
> >
> > We currently have support for translation of news, events, surveys,
> > quotes and pages. So far, we have zero news, zero events, zero surveys,
> > zero quotes and zero pages translated.
>
> Hello,
>
> In my opinion the translation infrastructure is not complete and does not
> allow translations to be done easily. It's lacking some way of tracking
> which English pages
> were modified, so that the translator would know what pages should be
> updated.
>

I've not heard of anyone come up with a better idea than following the website
commit mailing list... it might be cumbersome, but wouldn't that work?

> About 2 years ago, I have started to translate the website to Romanian. At
> that time I had almost all the pages done, except for the weekly news.
>
> I had trouble with setting up a local copy of the website for testing
> purposes ,
> then the time passed by without any progress.
>

Yes, I think we all have that problem. My thinking is that anyone serious
about doing this would be given an install on one of our development servers,
but istr there was an issue with you working on a remote host?

> The big question is : how is someone supposed to synchronise his
> translation with the original English pages (which are updated frequently)
> ?
>
> A way to accomplish this might be to write inside the translated page the
> CVS revision of the English one, and write some tool which would report the
> pages were modified in the meantime.
> But this is ugly , and the feature should be provided by the
> translation infrastructure.
>

Yeah, seems troublesome, though not insurmountable.

--
Robert Treat
Build A Brighter LAMP :: Linux Apache {middleware} PostgreSQL

>>>>> The postgresql community.
>>>> And that statistic is derived from what sources?
>>> *sigh*, I am not going to argue the obvious with you. If you don't
>>> agree, don't agree.
>
> I personally don't agree either ... I would guess that the majority of
> PostgreSQL users are non-English as *first* language ... just because they can
> communicate in English doesn't mean that their are either fluent in, or
> comfortable with, the language ...

Then you maintain the code :). Which I believe is one of the arguments.

>
> I'd love to see our demographics some day, but would guess that the Japanese
> themselves are a fairly high percentage of users, based on how they've
> organized over there ...

I would say outside of the US the Japanese are probably the largest of
our users, and as I said -- they have their own infrastructure.

> Spanish is probably another high percentage language
> ... French, German, Chinese (Mandarin?), Arabic ... if you look at 'first
> languages', I suspect that "English" is alot smaller then 98% ... whether its
> less then 50%, that I wouldn't hazard a guess ...

England, Canada, US, Australia ...

French... have their own infrastructure
German? I have not idea what they have.. Peter?

>
> Personally, I think it almost does us a dis-service statistically not to have
> more translations available on our main web site ...

I would not disagree, *if* more of the international community was
actually helping keep up with that infrastructure but that isn't why it
is happening. They are creating their own infrastructure.

I know that PostgreSQLFR for example is considering (as we are) creating
a new postgresql website dedicated to being a international docs portal.

The long and short is that without the internationalization the website
becomes quite a bit less complex, that is valid enough argument.

Nobody is stopping people from creating alternative websites, no one is
stopping the main .Org community from insuring that those sub projects
for France, Germany, Austria (where ever) are appropriately linked off
the main site.

Sincerely,

Joshua D. Drake

>
> ----
> Marc G. Fournier Hub.Org Networking Services (http://www.hub.org)
> Email . scrappy(at)hub(dot)org MSN . scrappy(at)hub(dot)org
> Yahoo . yscrappy Skype: hub.org ICQ . 7615664

--

=== The PostgreSQL Company: Command Prompt, Inc. ===
Sales/Support: +1.503.667.4564 || 24x7/Emergency: +1.800.492.2240
Providing the most comprehensive PostgreSQL solutions since 1997
http://www.commandprompt.com/

Donate to the PostgreSQL Project: http://www.postgresql.org/about/donate
PostgreSQL Replication: http://www.commandprompt.com/products/

On 2/11/07, Joshua D. Drake <jd(at)commandprompt(dot)com> wrote:
> > Spanish is probably another high percentage language
>
> I would not disagree, *if* more of the international community was
> actually helping keep up with that infrastructure but that isn't why it
> is happening. They are creating their own infrastructure.
>

well actually it was a discussion about that in the spanish list few
month ago... some were talking about creating a new page and the
concensus was to update what historically has been the spanish
postgres page (www.postgresql.cl) but maybe having a translate of the
*oficial* page is better idea...

--
regards,
Jaime Casanova

"Programming today is a race between software engineers striving to
build bigger and better idiot-proof programs and the universe trying
to produce bigger and better idiots.
So far, the universe is winning."
Richard Cook

On 2/12/07, Robert Treat <xzilla(at)users(dot)sourceforge(dot)net> wrote:
> On Sunday 11 February 2007 16:22, Adrian Maier wrote:
> > On 2/11/07, Magnus Hagander <magnus(at)hagander(dot)net> wrote:
> > > I'm working on a rewrite and major cleanup of parts of the code that
> > > runs the main website, and came across a thought.
> > >
> > > We currently have support for translation of news, events, surveys,
> > > quotes and pages. So far, we have zero news, zero events, zero surveys,
> > > zero quotes and zero pages translated.
> >
> > Hello,
> >
> > In my opinion the translation infrastructure is not complete and does not
> > allow translations to be done easily. It's lacking some way of tracking
> > which English pages
> > were modified, so that the translator would know what pages should be
> > updated.
> >
>
> I've not heard of anyone come up with a better idea than following the website
> commit mailing list... it might be cumbersome, but wouldn't that work?

It is possible to do that , but it is not "easy".

What I am personally missing is just one thing : a nice coloured table showing
the translation status for every page (similar to the NLS status tables, except
that it would contain a revision number instead of a percentage).

To achieve this I have the following idea : every translated page
would contain a
comment like : <!-- REVISION_NUMBER=xxxxxx --> , where xxxxx is the
CVS revision of the corresponding English page. It would be the
translator's job
to manually write the comment and modify it when he updates the translation .
A script would later scan the files to get those revisions, then get
the latest revision
numbers from CVS and present them all as a table.

> > About 2 years ago, I have started to translate the website to Romanian. At
> > that time I had almost all the pages done, except for the weekly news.
> >
> > I had trouble with setting up a local copy of the website for testing
> > purposes ,
> > then the time passed by without any progress.
> >
>
> Yes, I think we all have that problem. My thinking is that anyone serious
> about doing this would be given an install on one of our development servers,
> but istr there was an issue with you working on a remote host?
>
> > The big question is : how is someone supposed to synchronise his
> > translation with the original English pages (which are updated frequently)
> > ?
> >
> > A way to accomplish this might be to write inside the translated page the
> > CVS revision of the English one, and write some tool which would report the
> > pages were modified in the meantime.
> > But this is ugly , and the feature should be provided by the
> > translation infrastructure.

> Yeah, seems troublesome, though not insurmountable.

My point is that the core of the translation infrastructure is there,
but it lacks
an user interface and is vaguely promoted. It's no surprise that the large
communities that already have their own websites haven't rushed to translate
the main site for the past 2(?) years .

Cheers,
Adrian Maier

On Mon, Feb 12, 2007 at 12:11:02AM -0500, Jaime Casanova wrote:
> On 2/11/07, Joshua D. Drake <jd(at)commandprompt(dot)com> wrote:
> >> Spanish is probably another high percentage language
> >
> >I would not disagree, *if* more of the international community was
> >actually helping keep up with that infrastructure but that isn't why it
> >is happening. They are creating their own infrastructure.
> >
>
> well actually it was a discussion about that in the spanish list few
> month ago... some were talking about creating a new page and the
> concensus was to update what historically has been the spanish
> postgres page (www.postgresql.cl) but maybe having a translate of the
> *oficial* page is better idea...

Just for the record, if the translators want it, I'm *not* suggesting we
remove the translation infrastructure. But most people who've previously
suggested it's important we keep it are people who would *not* be doing
any of the translations themselves.

And I think you've found the real question here - is it better to have a
language/region-specific page, or is it better to translate the main
one. I personally do not think it's much point to having *both*, but as
for which one of the two is best I think that's best left to those
actually doing it.

//Magnus

On Mon, Feb 12, 2007 at 10:16:31AM +0200, Adrian Maier wrote:
> On 2/12/07, Robert Treat <xzilla(at)users(dot)sourceforge(dot)net> wrote:
> >On Sunday 11 February 2007 16:22, Adrian Maier wrote:
> >> On 2/11/07, Magnus Hagander <magnus(at)hagander(dot)net> wrote:
> >> > I'm working on a rewrite and major cleanup of parts of the code that
> >> > runs the main website, and came across a thought.
> >> >
> >> > We currently have support for translation of news, events, surveys,
> >> > quotes and pages. So far, we have zero news, zero events, zero surveys,
> >> > zero quotes and zero pages translated.
> >>
> >> Hello,
> >>
> >> In my opinion the translation infrastructure is not complete and does not
> >> allow translations to be done easily. It's lacking some way of tracking
> >> which English pages
> >> were modified, so that the translator would know what pages should be
> >> updated.
> >>
> >
> >I've not heard of anyone come up with a better idea than following the
> >website
> >commit mailing list... it might be cumbersome, but wouldn't that work?
>
> It is possible to do that , but it is not "easy".
>
> What I am personally missing is just one thing : a nice coloured table
> showing
> the translation status for every page (similar to the NLS status tables,
> except
> that it would contain a revision number instead of a percentage).
>
> To achieve this I have the following idea : every translated page
> would contain a
> comment like : <!-- REVISION_NUMBER=xxxxxx --> , where xxxxx is the
> CVS revision of the corresponding English page. It would be the
> translator's job
> to manually write the comment and modify it when he updates the translation
> .
> A script would later scan the files to get those revisions, then get
> the latest revision
> numbers from CVS and present them all as a table.

Would it not be much easier to just check the *dates* on the files? Once
a file is translated, you commit it. Then it can be considered
up-to-date up until the point that the base file is newer than the
translated one (when you change the base file).

You can do this today, but it might certainly help if someone put
togheter a small script to show it in a nice way. There's not much point
to that until someone has actually done some translations, though ;-)

> >> A way to accomplish this might be to write inside the translated page the
> >> CVS revision of the English one, and write some tool which would report
> >the
> >> pages were modified in the meantime.
> >> But this is ugly , and the feature should be provided by the
> >> translation infrastructure.
>
> >Yeah, seems troublesome, though not insurmountable.
>
> My point is that the core of the translation infrastructure is there,
> but it lacks
> an user interface and is vaguely promoted. It's no surprise that the large
> communities that already have their own websites haven't rushed to translate
> the main site for the past 2(?) years .

Right. But will they rush to do that even if we have easier tools to do
it?

There's also the question of wether it's a good thing to have a say 15%
translated site, vs a 0% translated one. If we have a 15% translation,
that will give a very strange impression for people going there with a
browser set for that language - some pages come up in their language,
the majority comes up in a completely different language.

//Magnus

Am Montag, 12. Februar 2007 09:50 schrieb Magnus Hagander:
> Right. But will they rush to do that even if we have easier tools to do
> it?

I would love to translate some pages, but frankly anything short of a PO file
or some equivalent technology isn't going to excite any translators for long.

> There's also the question of wether it's a good thing to have a say 15%
> translated site, vs a 0% translated one. If we have a 15% translation,
> that will give a very strange impression for people going there with a
> browser set for that language - some pages come up in their language,
> the majority comes up in a completely different language.

By that theory, no translation of open-source software would ever take place.

--
Peter Eisentraut
http://developer.postgresql.org/~petere/

Am Montag, 12. Februar 2007 09:45 schrieb Magnus Hagander:
> And I think you've found the real question here - is it better to have a
> language/region-specific page, or is it better to translate the main
> one.

I think it's perfectly fine to have both.

--
Peter Eisentraut
http://developer.postgresql.org/~petere/

On Mon, Feb 12, 2007 at 10:16:10AM +0100, Peter Eisentraut wrote:
> Am Montag, 12. Februar 2007 09:50 schrieb Magnus Hagander:
> > Right. But will they rush to do that even if we have easier tools to do
> > it?
>
> I would love to translate some pages, but frankly anything short of a PO file
> or some equivalent technology isn't going to excite any translators for long.

A .po file is easier than a plaintext file? Well, I'm don't do much
translating myself, but I can't see how translating a webpage can be
easier than translating the actual text of the webpage in the file...

(We do have .po files for the strings that come out of the PHP code
directly)

As for news and such, there's a web interface. Are you saying that's
also too complex? Frankly, I don't see how a .po would make that better,
but again I'm not used to working with these things.

> > There's also the question of wether it's a good thing to have a say 15%
> > translated site, vs a 0% translated one. If we have a 15% translation,
> > that will give a very strange impression for people going there with a
> > browser set for that language - some pages come up in their language,
> > the majority comes up in a completely different language.
>
> By that theory, no translation of open-source software would ever take place.

Well, I stand by that opinion. I know for examlpe pgAdmin only ships
translations that are n percent or better (iirc, it's 85% or so), which
makes it reasonable. Shipping something that only has 15% translation
rate does the user a disservice, imho.

//Magnus

On 2/12/07, Magnus Hagander <magnus(at)hagander(dot)net> wrote:
> On Mon, Feb 12, 2007 at 10:16:31AM +0200, Adrian Maier wrote:
> > To achieve this I have the following idea : every translated page
> > would contain a
> > comment like : <!-- REVISION_NUMBER=xxxxxx --> , where xxxxx is the
> > CVS revision of the corresponding English page. It would be the
> > translator's job
> > to manually write the comment and modify it when he updates the translation
> > .
> > A script would later scan the files to get those revisions, then get
> > the latest revision
> > numbers from CVS and present them all as a table.
>
> Would it not be much easier to just check the *dates* on the files? Once
> a file is translated, you commit it. Then it can be considered
> up-to-date up until the point that the base file is newer than the
> translated one (when you change the base file).

Revision number or commit date is almost the same thing. When I am
looking at a certain translated file i want to know which is the date or
revision of the corresponding English file. This piece of information makes
it possible to use cvs diff for seeing the modifications that took place in
the meantime on the original file.

> You can do this today, but it might certainly help if someone put
> togheter a small script to show it in a nice way. There's not much point
> to that until someone has actually done some translations, though ;-)

Yep , you are right.

> > >> A way to accomplish this might be to write inside the translated page the
> > >> CVS revision of the English one, and write some tool which would report
> > >the
> > >> pages were modified in the meantime.
> > >> But this is ugly , and the feature should be provided by the
> > >> translation infrastructure.
> >
> > >Yeah, seems troublesome, though not insurmountable.
> >
> > My point is that the core of the translation infrastructure is there,
> > but it lacks
> > an user interface and is vaguely promoted. It's no surprise that the large
> > communities that already have their own websites haven't rushed to translate
> > the main site for the past 2(?) years .
>
> Right. But will they rush to do that even if we have easier tools to do
> it?

Maybe, who knows... At the present it's quite scary to begin translating.

> There's also the question of wether it's a good thing to have a say 15%
> translated site, vs a 0% translated one. If we have a 15% translation,
> that will give a very strange impression for people going there with a
> browser set for that language - some pages come up in their language,
> the majority comes up in a completely different language.

I'm seeing this quite differently: at first any translation is supposed to
be almost 100% complete, but in time the contents will become old if
the english pages get updated but the translated pages aren't modified
accordingly.

--
Adrian Maier

On Mon, Feb 12, 2007 at 11:37:16AM +0200, Adrian Maier wrote:
> On 2/12/07, Magnus Hagander <magnus(at)hagander(dot)net> wrote:
> >On Mon, Feb 12, 2007 at 10:16:31AM +0200, Adrian Maier wrote:
> >> To achieve this I have the following idea : every translated page
> >> would contain a
> >> comment like : <!-- REVISION_NUMBER=xxxxxx --> , where xxxxx is the
> >> CVS revision of the corresponding English page. It would be the
> >> translator's job
> >> to manually write the comment and modify it when he updates the
> >translation
> >> .
> >> A script would later scan the files to get those revisions, then get
> >> the latest revision
> >> numbers from CVS and present them all as a table.
> >
> >Would it not be much easier to just check the *dates* on the files? Once
> >a file is translated, you commit it. Then it can be considered
> >up-to-date up until the point that the base file is newer than the
> >translated one (when you change the base file).
>
> Revision number or commit date is almost the same thing. When I am
> looking at a certain translated file i want to know which is the date or
> revision of the corresponding English file. This piece of information makes
> it possible to use cvs diff for seeing the modifications that took place in
> the meantime on the original file.

Not in cvs. Each file has it's own revision number, whereas the date is
ever-increasing. If you had repository revision numbers, this would be
easier.
And yes, we've been talking abuot moving the pgweb stuff into svn
somewhere, but it's not done. And yes, if repo version numbers would
help a lot here, it might be a tipping factor.

What I'm against is having to store the revision number manually in the
file, it seems like a very ugly and unmaintainable solution to me.

> >There's also the question of wether it's a good thing to have a say 15%
> >translated site, vs a 0% translated one. If we have a 15% translation,
> >that will give a very strange impression for people going there with a
> >browser set for that language - some pages come up in their language,
> >the majority comes up in a completely different language.
>
> I'm seeing this quite differently: at first any translation is supposed
> to
> be almost 100% complete, but in time the contents will become old if
> the english pages get updated but the translated pages aren't modified
> accordingly.

Well, once a page is out of date, I'd consider it "not translated". That
said, there needs to be a very clear policy and a way to deal with
it. Certain updates wouldn't need to invalidate the translatino (say a
spelling or grammar fix), whereas certain others will need it to be
completely invalidated right away (say information about a security
issue, where you really don't want out-of-date information in different
languages).

One way I've seen other sites do it is have a banner on the page that
says "this page is out of date compared to the english version. Click
here to see the original one", but that just annoys the hell out of me
everytime. It's ok if it sits there for a couple of hours, but unelss
you have a really responsive translation team it can be stuck like that
for months.

So you really need a policy for how to deal with that. I think that's
another one of the reasons why most translation teams so far have been
rolling their own instead.

//Magnus

hat theory, no translation of open-source software would ever take place.
> Magnus Hagander wrote:
> Well, I stand by that opinion. I know for examlpe pgAdmin only ships
> translations that are n percent or better (iirc, it's 85% or so), which
> makes it reasonable. Shipping something that only has 15% translation
> rate does the user a disservice, imho.

Thats entirely correct - we don't ship translations that are less than
85% - it looks unprofessional and doesn't help the user.

Regards, Dave.

Am Montag, 12. Februar 2007 10:36 schrieb Magnus Hagander:
> A .po file is easier than a plaintext file?

Definitely.

> Well, I'm don't do much
> translating myself, but I can't see how translating a webpage can be
> easier than translating the actual text of the webpage in the file...

One advantage is that it is disassociated from the particulars of the source
format and the details of how to get it and how to send it back. Another
advantage is that there is an established and powerful toolset for editing,
merging, and managing translations.

> Well, I stand by that opinion. I know for examlpe pgAdmin only ships
> translations that are n percent or better (iirc, it's 85% or so), which
> makes it reasonable. Shipping something that only has 15% translation
> rate does the user a disservice, imho.

It's one thing to reject web *pages* that are less than 85% or even 100%
translated. But no one can require a 85% or 100% translated web *site*. Not
even Microsoft or IBM can manage that.

--
Peter Eisentraut
http://developer.postgresql.org/~petere/

On Mon, Feb 12, 2007 at 11:51:34AM +0100, Peter Eisentraut wrote:
> Am Montag, 12. Februar 2007 10:36 schrieb Magnus Hagander:
> > A .po file is easier than a plaintext file?
>
> Definitely.
>
> > Well, I'm don't do much
> > translating myself, but I can't see how translating a webpage can be
> > easier than translating the actual text of the webpage in the file...
>
> One advantage is that it is disassociated from the particulars of the source
> format and the details of how to get it and how to send it back. Another
> advantage is that there is an established and powerful toolset for editing,
> merging, and managing translations.

That's actually an argument *for* ripping out the current translation
infrastructure. Because if we want to use .po and related tools, we need
a complete rewrite of it.
(Today, you edit the templated HTML which comes down to a plaintext
file, but the entire system depends on being able to parse it as a
template)

> > Well, I stand by that opinion. I know for examlpe pgAdmin only ships
> > translations that are n percent or better (iirc, it's 85% or so), which
> > makes it reasonable. Shipping something that only has 15% translation
> > rate does the user a disservice, imho.
>
> It's one thing to reject web *pages* that are less than 85% or even 100%
> translated. But no one can require a 85% or 100% translated web *site*. Not
> even Microsoft or IBM can manage that.

I'm not saying reject <100%. I'm saying reject 15%.

Oh, and the fact that MS for example can't do it is one of the reasons
that *everybody* I know in Sweden goes to the US site and not the
Swedish one.
And I also notice that they are *separate* sites, and *not* just
translations of the same site...

//Magnus

On 2/12/07, Magnus Hagander <magnus(at)hagander(dot)net> wrote:
> On Mon, Feb 12, 2007 at 11:37:16AM +0200, Adrian Maier wrote:
> > On 2/12/07, Magnus Hagander <magnus(at)hagander(dot)net> wrote:
> > >On Mon, Feb 12, 2007 at 10:16:31AM +0200, Adrian Maier wrote:
> > >> To achieve this I have the following idea : every translated page
> > >> would contain a
> > >> comment like : <!-- REVISION_NUMBER=xxxxxx --> , where xxxxx is the
> > >> CVS revision of the corresponding English page. It would be the
> > >> translator's job
> > >> to manually write the comment and modify it when he updates the
> > >translation
> > >> .
> > >> A script would later scan the files to get those revisions, then get
> > >> the latest revision
> > >> numbers from CVS and present them all as a table.
> > >
> > >Would it not be much easier to just check the *dates* on the files? Once
> > >a file is translated, you commit it. Then it can be considered
> > >up-to-date up until the point that the base file is newer than the
> > >translated one (when you change the base file).
> >
> > Revision number or commit date is almost the same thing. When I am
> > looking at a certain translated file i want to know which is the date or
> > revision of the corresponding English file. This piece of information makes
> > it possible to use cvs diff for seeing the modifications that took place in
> > the meantime on the original file.
>
> Not in cvs. Each file has it's own revision number, whereas the date is
> ever-increasing. If you had repository revision numbers, this would be
> easier.

I'm not sure why per-file revision numbers are unsatisfactory.

The problem is that a "cvs diff" on the english pages doesn't answer the
question "i have this particular translated page. Which is the version of the
corresponding english text ? "

Giving an answer based on the commit dates is only an approximation,
because it takes some time until the translation is committed . It will
get painful if the english page is modified again before having the translation
comitted.

In particular, translating for the first time takes some time; the original
files are updated in the meantime. It is getting painful when you realise that
you are translating a moving target and you don't have any real tools for
associating your translated file to the corresponding version of the original
file. "cvs diff" is helpful but it's not enough

> And yes, we've been talking abuot moving the pgweb stuff into svn
> somewhere, but it's not done. And yes, if repo version numbers would
> help a lot here, it might be a tipping factor.
>
> What I'm against is having to store the revision number manually in the
> file, it seems like a very ugly and unmaintainable solution to me.

It is ugly indeed, but i'm not sure what else can be done without requiring
some work . A real solution would be having a table in the
database with the status of the translations and a web interface for uploading
the translations (so that the database would be updated ).
But I can't expect anyone to work on this, particularly since noone
ever provided
any translation so far.

> > >There's also the question of wether it's a good thing to have a say 15%
> > >translated site, vs a 0% translated one. If we have a 15% translation,
> > >that will give a very strange impression for people going there with a
> > >browser set for that language - some pages come up in their language,
> > >the majority comes up in a completely different language.
> >
> > I'm seeing this quite differently: at first any translation is supposed
> > to
> > be almost 100% complete, but in time the contents will become old if
> > the english pages get updated but the translated pages aren't modified
> > accordingly.

> Well, once a page is out of date, I'd consider it "not translated". That
> said, there needs to be a very clear policy and a way to deal with
> it. Certain updates wouldn't need to invalidate the translatino (say a
> spelling or grammar fix), whereas certain others will need it to be
> completely invalidated right away (say information about a security
> issue, where you really don't want out-of-date information in different
> languages).

Yes, a policy will be needed. But probably some translations need to
show up before discussing about the policy.

In 2-3 weeks from now i'll have some time to look at bringing up to date
the files. But before spending time on them it would be nice to know
whether the translation facility will be removed or not .. .

--
Adrian Maier

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

- --On Monday, February 12, 2007 11:37:16 +0200 Adrian Maier
<adrian(dot)maier(at)gmail(dot)com> wrote:

> I'm seeing this quite differently: at first any translation is supposed to
> be almost 100% complete, but in time the contents will become old if
> the english pages get updated but the translated pages aren't modified
> accordingly.

So, why not a committers list just for translators, so that whenever the web
pages *are* updated, a messages goes out letting ppl know that? then, it would
be a simple matter of doing a cvs diff to see what changed ..

Marc G. Fournier Hub.Org Networking Services (http://www.hub.org)
Email . scrappy(at)hub(dot)org MSN . scrappy(at)hub(dot)org
Yahoo . yscrappy Skype: hub.org ICQ . 7615664
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (FreeBSD)

iD8DBQFF0F1W4QvfyHIvDvMRAuv6AKCaohZ3wdjXaywV4j+F/J3Dkz8zJwCfXf9J
9gSp4bkV962d1vJ0V1tKVf4=
=/p/f
-----END PGP SIGNATURE-----

On Mon, Feb 12, 2007 at 08:28:05AM -0400, Marc G. Fournier wrote:

> > I'm seeing this quite differently: at first any translation is supposed to
> > be almost 100% complete, but in time the contents will become old if
> > the english pages get updated but the translated pages aren't modified
> > accordingly.
>
> So, why not a committers list just for translators, so that whenever the web
> pages *are* updated, a messages goes out letting ppl know that? then, it would
> be a simple matter of doing a cvs diff to see what changed ..

That would be the current pgweb-commits list. The amount of changes that
go into the code that would not affect the translators is close to zero.

//Magnus

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

- --On Monday, February 12, 2007 09:50:06 +0100 Magnus Hagander
<magnus(at)hagander(dot)net> wrote:

> Would it not be much easier to just check the *dates* on the files? Once
> a file is translated, you commit it. Then it can be considered
> up-to-date up until the point that the base file is newer than the
> translated one (when you change the base file).
>
> You can do this today, but it might certainly help if someone put
> togheter a small script to show it in a nice way. There's not much point
> to that until someone has actually done some translations, though ;-)

There should be some mechanism whereby translators are informed that there have
been changes (ie. pgsql-committers kind of list) ...

> There's also the question of wether it's a good thing to have a say 15%
> translated site, vs a 0% translated one. If we have a 15% translation,
> that will give a very strange impression for people going there with a
> browser set for that language - some pages come up in their language,
> the majority comes up in a completely different language.

You could argue that ppl are doing that now ... they go to a 'language /
regional site' and then have to cut over to the main site in english to read,
say, the docs ...

At least having 15% visible would hopefully encourage work to be done on the
other 85% over time ... whereas right, just getting started might seem a
daunting task ...

- ----
Marc G. Fournier Hub.Org Networking Services (http://www.hub.org)
Email . scrappy(at)hub(dot)org MSN . scrappy(at)hub(dot)org
Yahoo . yscrappy Skype: hub.org ICQ . 7615664
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (FreeBSD)

iD8DBQFF0Gcx4QvfyHIvDvMRApyUAKDIfYP30jsUHbbzRZWoA/owkOTEPACfWv8t
jJqJWei356toSljaGsbgAgo=
=rDe5
-----END PGP SIGNATURE-----

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

- --On Monday, February 12, 2007 11:51:34 +0100 Peter Eisentraut
<peter_e(at)gmx(dot)net> wrote:

> It's one thing to reject web *pages* that are less than 85% or even 100%
> translated. But no one can require a 85% or 100% translated web *site*. Not
> even Microsoft or IBM can manage that.

I have to agree with PeterE here ... even 15% shows that we are *attempting* to
accommodate other languages ... which would hopefully act as encouragement for
others to submit translations ...

If you expect someone to sit down and translate 85% of the web site into a
language before we consider it useful, it will never get done ... if we accept
"one page at a time" so that ppl see their work / effort is being made use of,
then its less daunting of a project, and more ppl will hopefully get involved
...

- ----
Marc G. Fournier Hub.Org Networking Services (http://www.hub.org)
Email . scrappy(at)hub(dot)org MSN . scrappy(at)hub(dot)org
Yahoo . yscrappy Skype: hub.org ICQ . 7615664
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (FreeBSD)

iD8DBQFF0GkQ4QvfyHIvDvMRAl/gAJ40yvI8shS2Qs1v98EDGl1+U5CHZwCgtTqw
z6A+Ve+HKS8Fzq7CShO8/pc=
=a5OQ
-----END PGP SIGNATURE-----

Magnus Hagander a ecrit le 12/02/2007 10:36:
> On Mon, Feb 12, 2007 at 10:16:10AM +0100, Peter Eisentraut wrote:
>> Am Montag, 12. Februar 2007 09:50 schrieb Magnus Hagander:
>>> Right. But will they rush to do that even if we have easier tools to do
>>> it?
>> I would love to translate some pages, but frankly anything short of a PO file
>> or some equivalent technology isn't going to excite any translators for long.
>
> A .po file is easier than a plaintext file? Well, I'm don't do much
> translating myself, but I can't see how translating a webpage can be
> easier than translating the actual text of the webpage in the file...
>
> (We do have .po files for the strings that come out of the PHP code
> directly)
>
> As for news and such, there's a web interface. Are you saying that's
> also too complex? Frankly, I don't see how a .po would make that better,
> but again I'm not used to working with these things.
>

Honestly, I didn't know there was a way to translate the main website.
Renaud Fortier and I did the translation on pgAdmin's website because
Dave told us it was possible. I think keeping the translation part of
the code is really necessary. I would be glad to provide a french
translation of the www.postgresql.org website. But I don't know how. And
a few days back, I didn't know we could do it. Is there a document
describing how to take care of this ?

Using po files is really easier when you don't have big pages or
documents. They are really useful for apps' translations and websites.
But they could be nightmare for things like manuals, books.

>>> There's also the question of wether it's a good thing to have a say 15%
>>> translated site, vs a 0% translated one. If we have a 15% translation,
>>> that will give a very strange impression for people going there with a
>>> browser set for that language - some pages come up in their language,
>>> the majority comes up in a completely different language.
>> By that theory, no translation of open-source software would ever take place.
>
> Well, I stand by that opinion. I know for examlpe pgAdmin only ships
> translations that are n percent or better (iirc, it's 85% or so), which
> makes it reasonable. Shipping something that only has 15% translation
> rate does the user a disservice, imho.
>

I agree on this. And it's really easy to get the percentage with po files.

Regards.

--
Guillaume.

> At least having 15% visible would hopefully encourage work to be done on the
> other 85% over time ... whereas right, just getting started might seem a
> daunting task ...

No, it wouldn't. It makes us look like people who don't care enough to
do the work on the translation to get it right.

I would rather us just say, "We don't have the resources to do
translations" than expect the sub communities at some point to pick up
the slack and hope it gets done.

Sincerely,

Joshua D. Drake

>
> ----
> Marc G. Fournier Hub.Org Networking Services (http://www.hub.org)
> Email . scrappy(at)hub(dot)org MSN . scrappy(at)hub(dot)org
> Yahoo . yscrappy Skype: hub.org ICQ . 7615664

---------------------------(end of broadcast)---------------------------
TIP 3: Have you checked our extensive FAQ?

http://www.postgresql.org/docs/faq

--

=== The PostgreSQL Company: Command Prompt, Inc. ===
Sales/Support: +1.503.667.4564 || 24x7/Emergency: +1.800.492.2240
Providing the most comprehensive PostgreSQL solutions since 1997
http://www.commandprompt.com/

Donate to the PostgreSQL Project: http://www.postgresql.org/about/donate
PostgreSQL Replication: http://www.commandprompt.com/products/

Guillaume Lelarge wrote:
> Magnus Hagander a ecrit le 12/02/2007 10:36:
>> On Mon, Feb 12, 2007 at 10:16:10AM +0100, Peter Eisentraut wrote:
>>> Am Montag, 12. Februar 2007 09:50 schrieb Magnus Hagander:
>>>> Right. But will they rush to do that even if we have easier tools to do
>>>> it?
>>> I would love to translate some pages, but frankly anything short of a
>>> PO file or some equivalent technology isn't going to excite any
>>> translators for long.
>>
>> A .po file is easier than a plaintext file? Well, I'm don't do much
>> translating myself, but I can't see how translating a webpage can be
>> easier than translating the actual text of the webpage in the file...
>>
>> (We do have .po files for the strings that come out of the PHP code
>> directly)
>>
>> As for news and such, there's a web interface. Are you saying that's
>> also too complex? Frankly, I don't see how a .po would make that better,
>> but again I'm not used to working with these things.
>>
>
> Honestly, I didn't know there was a way to translate the main website.

Really? You mustn't have followed -www very closely :-) I specifically
recall having told several people about that - though it's been a while
since someone last asked.

> Renaud Fortier and I did the translation on pgAdmin's website because
> Dave told us it was possible. I think keeping the translation part of
> the code is really necessary. I would be glad to provide a french
> translation of the www.postgresql.org website. But I don't know how. And
> a few days back, I didn't know we could do it. Is there a document
> describing how to take care of this ?

There's some stuff in the README file:
http://gborg.postgresql.org/cgi-bin/cvsweb.cgi/portal/README?rev=1.8;cvsroot=pgweb

> Using po files is really easier when you don't have big pages or
> documents. They are really useful for apps' translations and websites.
> But they could be nightmare for things like manuals, books.

Right, that's what I thought. BTW, I don't think the website currently
has the infrastructure to translate the docs part ;-)

>> Well, I stand by that opinion. I know for examlpe pgAdmin only ships
>> translations that are n percent or better (iirc, it's 85% or so), which
>> makes it reasonable. Shipping something that only has 15% translation
>> rate does the user a disservice, imho.
>>
>
> I agree on this. And it's really easy to get the percentage with po files.

Well, you're one of the maintainers of postgresqlfr.org, right? What are
your thoughts on that vs a translated www.postgresql.org? I'd really
like to hear more from the people who're actually maintaining the
non-english websites..

//Magnus

Hi Magnus :)

Magnus Hagander a écrit :
> Guillaume Lelarge wrote:
>> Magnus Hagander a ecrit le 12/02/2007 10:36:
>>> On Mon, Feb 12, 2007 at 10:16:10AM +0100, Peter Eisentraut wrote:
>>>> Am Montag, 12. Februar 2007 09:50 schrieb Magnus Hagander:
>>>>> Right. But will they rush to do that even if we have easier tools to do
>>>>> it?
>>>> I would love to translate some pages, but frankly anything short of a
>>>> PO file or some equivalent technology isn't going to excite any
>>>> translators for long.
>>> A .po file is easier than a plaintext file? Well, I'm don't do much
>>> translating myself, but I can't see how translating a webpage can be
>>> easier than translating the actual text of the webpage in the file...
>>>
>>> (We do have .po files for the strings that come out of the PHP code
>>> directly)
>>>
>>> As for news and such, there's a web interface. Are you saying that's
>>> also too complex? Frankly, I don't see how a .po would make that better,
>>> but again I'm not used to working with these things.
>>>
>> Honestly, I didn't know there was a way to translate the main website.
>
> Really? You mustn't have followed -www very closely :-) I specifically
> recall having told several people about that - though it's been a while
> since someone last asked.
>

I think it would have triggered something in me :) Perhaps, I wasn't on
pgsql-www at this time. And perhaps, I didn't read these mails really
carefuly :)

>> Renaud Fortier and I did the translation on pgAdmin's website because
>> Dave told us it was possible. I think keeping the translation part of
>> the code is really necessary. I would be glad to provide a french
>> translation of the www.postgresql.org website. But I don't know how. And
>> a few days back, I didn't know we could do it. Is there a document
>> describing how to take care of this ?
>
> There's some stuff in the README file:
> http://gborg.postgresql.org/cgi-bin/cvsweb.cgi/portal/README?rev=1.8;cvsroot=pgweb
>

OK, I'll read this.

>> Using po files is really easier when you don't have big pages or
>> documents. They are really useful for apps' translations and websites.
>> But they could be nightmare for things like manuals, books.
>
> Right, that's what I thought. BTW, I don't think the website currently
> has the infrastructure to translate the docs part ;-)
>

You're right. And I won't translate once again just to use .po files :)

>>> Well, I stand by that opinion. I know for examlpe pgAdmin only ships
>>> translations that are n percent or better (iirc, it's 85% or so), which
>>> makes it reasonable. Shipping something that only has 15% translation
>>> rate does the user a disservice, imho.
>>>
>> I agree on this. And it's really easy to get the percentage with po files.
>
> Well, you're one of the maintainers of postgresqlfr.org, right? What are
> your thoughts on that vs a translated www.postgresql.org? I'd really
> like to hear more from the people who're actually maintaining the
> non-english websites..
>

At the beginning of this thread, I thought we could rip the translation
part of postgresql.org. Really, why would we need two websites ? But, as
the thread goes on and on, I now think we need both of them. I think
many french people first go to the main website, ie www.postgresql.org,
for example to download the tarball. That's where they discover the
french mailing list and the french website. The french website is the
association's website. We put in it a translated PGWN, release news and
french success stories... and I think that's all. Hmmm, no, we also add
news on meetings (Solutions Linux, RMLL, etc.) And there's a cookbook
and a forum. That's all. Nothing like what's available on postgresql.org.

So I think we need both of them.

And I think we need a documentation's website but that's another story.

Regards.

--
Guillaume.
<!-- http://abs.traduc.org/
http://lfs.traduc.org/
http://docs.postgresqlfr.org/ -->

>> There's some stuff in the README file:
>> http://gborg.postgresql.org/cgi-bin/cvsweb.cgi/portal/README?rev=1.8;cvsroot=pgweb
>>
>
> OK, I'll read this.

Please do and let us know what you think. If the way it's done today is
broken, we need to fix that rather than just live on with what we have.
But that requires an idea of what to fix it into :-)

>> Well, you're one of the maintainers of postgresqlfr.org, right? What are
>> your thoughts on that vs a translated www.postgresql.org? I'd really
>> like to hear more from the people who're actually maintaining the
>> non-english websites..
>>
>
> At the beginning of this thread, I thought we could rip the translation
> part of postgresql.org. Really, why would we need two websites ? But, as
> the thread goes on and on, I now think we need both of them. I think
> many french people first go to the main website, ie www.postgresql.org,
> for example to download the tarball. That's where they discover the
> french mailing list and the french website. The french website is the
> association's website. We put in it a translated PGWN, release news and
> french success stories... and I think that's all. Hmmm, no, we also add
> news on meetings (Solutions Linux, RMLL, etc.) And there's a cookbook
> and a forum. That's all. Nothing like what's available on postgresql.org.
>
> So I think we need both of them.
>
> And I think we need a documentation's website but that's another story.

Okay. Let me then ask the other thing that we really haven't discussed:
Do we want the language site to be served up automatically. That's
really what the code is all about today. If your browser is set to
prefer French, we'll serve up the French version of the page (if
available). While convenient, I personally don't like sites that do this
for me (I've reconfigured my browser to put English as the top language
for that reason).
Now if we *don't* want this, the comment about incomplete translations
fall - then you will only see that which is translated.

//Magnus

> In particular, translating for the first time takes some time; the original
> files are updated in the meantime. It is getting painful when you
> realise that
> you are translating a moving target and you don't have any real tools for
> associating your translated file to the corresponding version of the
> original
> file. "cvs diff" is helpful but it's not enough

Not really that big a problem I think, unless you take a long time. Most
of our pages don't change often at all. The only stuff with a high
throughput is news and events, and they're just additions and not
modifications.

>> And yes, we've been talking abuot moving the pgweb stuff into svn
>> somewhere, but it's not done. And yes, if repo version numbers would
>> help a lot here, it might be a tipping factor.
>>
>> What I'm against is having to store the revision number manually in the
>> file, it seems like a very ugly and unmaintainable solution to me.
>
> It is ugly indeed, but i'm not sure what else can be done without requiring
> some work . A real solution would be having a table in the
> database with the status of the translations and a web interface for
> uploading
> the translations (so that the database would be updated ).
> But I can't expect anyone to work on this, particularly since noone
> ever provided
> any translation so far.

Right. Well, one option is to rip out what we have now in favor of
putting in something else - provided we do that in the end :-)

>> Well, once a page is out of date, I'd consider it "not translated". That
>> said, there needs to be a very clear policy and a way to deal with
>> it. Certain updates wouldn't need to invalidate the translatino (say a
>> spelling or grammar fix), whereas certain others will need it to be
>> completely invalidated right away (say information about a security
>> issue, where you really don't want out-of-date information in different
>> languages).
>
> Yes, a policy will be needed. But probably some translations need to
> show up before discussing about the policy.
>
> In 2-3 weeks from now i'll have some time to look at bringing up to date
> the files. But before spending time on them it would be nice to know
> whether the translation facility will be removed or not .. .

Well, it won't be removed if people like it and intend to use it. It
will be removed if nobody wants to use it. If will be removed if it's
bad and needs to be replaced with something better.

//Magnus

Magnus Hagander a écrit :
>>> There's some stuff in the README file:
>>> http://gborg.postgresql.org/cgi-bin/cvsweb.cgi/portal/README?rev=1.8;cvsroot=pgweb
>>>
>> OK, I'll read this.
>
> Please do and let us know what you think. If the way it's done today is
> broken, we need to fix that rather than just live on with what we have.
> But that requires an idea of what to fix it into :-)
>

OK, I've read it. It seems good to me. And I would like to translate the
website in french.

>>> Well, you're one of the maintainers of postgresqlfr.org, right? What are
>>> your thoughts on that vs a translated www.postgresql.org? I'd really
>>> like to hear more from the people who're actually maintaining the
>>> non-english websites..
>>>
>> At the beginning of this thread, I thought we could rip the translation
>> part of postgresql.org. Really, why would we need two websites ? But, as
>> the thread goes on and on, I now think we need both of them. I think
>> many french people first go to the main website, ie www.postgresql.org,
>> for example to download the tarball. That's where they discover the
>> french mailing list and the french website. The french website is the
>> association's website. We put in it a translated PGWN, release news and
>> french success stories... and I think that's all. Hmmm, no, we also add
>> news on meetings (Solutions Linux, RMLL, etc.) And there's a cookbook
>> and a forum. That's all. Nothing like what's available on postgresql.org.
>>
>> So I think we need both of them.
>>
>> And I think we need a documentation's website but that's another story.
>
> Okay. Let me then ask the other thing that we really haven't discussed:
> Do we want the language site to be served up automatically. That's
> really what the code is all about today. If your browser is set to
> prefer French, we'll serve up the French version of the page (if
> available). While convenient, I personally don't like sites that do this
> for me (I've reconfigured my browser to put English as the top language
> for that reason).

I kind of agreed with you on this one. I particularly hate when google
put me on their french search website when I really asked to be in the
english one. I hate this because, with the french search website, french
pages are ranked higher than others. But I don't think this will be the
point here. If you have a browser configured to put french as top
language, that's just because you prefer french pages. And if you don't
want that, you will have a combo to choose another language.

> Now if we *don't* want this, the comment about incomplete translations
> fall - then you will only see that which is translated.
>

I want this. But others should tell us what they prefer. It's also
possible to first go to the browser's language page if this language's
po file is at least 85% translated... and go to the english one if it isn't.

--
Guillaume.
<!-- http://abs.traduc.org/
http://lfs.traduc.org/
http://docs.postgresqlfr.org/ -->

Magnus Hagander wrote:
> Okay. Let me then ask the other thing that we really haven't
> discussed: Do we want the language site to be served up
> automatically. That's really what the code is all about today. If
> your browser is set to prefer French, we'll serve up the French
> version of the page (if available). While convenient, I personally
> don't like sites that do this for me (I've reconfigured my browser to
> put English as the top language for that reason).

That's just the point. If you don't want it, you configure your browser
to make that preference known.

--
Peter Eisentraut
http://developer.postgresql.org/~petere/

On 2/12/07, Magnus Hagander <magnus(at)hagander(dot)net> wrote:
> Guillaume Lelarge wrote:
> > Using po files is really easier when you don't have big pages or
> > documents. They are really useful for apps' translations and websites.
> > But they could be nightmare for things like manuals, books.
>
> Right, that's what I thought. BTW, I don't think the website currently
> has the infrastructure to translate the docs part ;-)
>
>

but Mario Gonzales and Alvaro Herrera has:
http://pgfoundry.org/projects/webtranslator

you can find the project actually in use in
http://l10n.postgresql.cl/es it is still in development but the
infraestructure is there

--
regards,
Jaime Casanova

"Programming today is a race between software engineers striving to
build bigger and better idiot-proof programs and the universe trying
to produce bigger and better idiots.
So far, the universe is winning."
Richard Cook

Magnus Hagander schrieb:
...
> Okay. Let me then ask the other thing that we really haven't discussed:
> Do we want the language site to be served up automatically. That's
> really what the code is all about today. If your browser is set to
> prefer French, we'll serve up the French version of the page (if
> available). While convenient, I personally don't like sites that do this
> for me (I've reconfigured my browser to put English as the top language
> for that reason).
> Now if we *don't* want this, the comment about incomplete translations
> fall - then you will only see that which is translated.

Well, you see this happen on Debian.org for example and I find it
ok. (While I usually set my browser to english as first language
anyway). This also makes it easy to see where translation might
be needed so anybody can jump in.

Speaking of jump in - so how do we do translation? I might take
some of the manual pages to translate to german at least. Maybe
some russian too but I'm not a native speaker, so probably not ;-)

Regards
Tino

Magnus,

I have to say "I told you so". When the existing translation scheme was
built two+ years ago, I pointed out that it was cumbersome, confusing and
inaccessable and predicted that none of our non-English communities would use
it.

So, my vote is that whether or not we have *an* translation infrastructure,
the current incomplete and non-standard infrastructure be junked. It's never
going to be used in its current form.

Further, we're going to have to expect that some language communities will
never translate the main site, since that puts them in a position of having
all new content generated in English and just having their site mirror that
without the ability to add new content originating in their group. And I
don't think that most of our language groups are large enough to sustain both
organizing content for their site and keeping translations of the English
site updated. Also, a couple of the local site groups told me that they want
an easy-to-use CMS like Drupal, so they're not merging to our site for that
reason.

P.S. Josh D, you are absolutely wrong about our language composition. The
majority of our community speaks a first language other than English, and at
least half of the non-English speakers aren't fluent in English. There are
large communities in Brazil, Spanish-speaking South America, Italy and
Germany -- as well as Japan -- which you aren't aware of because they don't
join the English-speaking MLs for obvious reasons. When I used to answer
webmaster@, for example, I got *more* questions in Portuguese than in
English.

--
Josh Berkus
PostgreSQL @ Sun
San Francisco

Josh Berkus wrote:
> Magnus,
>
> I have to say "I told you so". When the existing translation scheme was
> built two+ years ago, I pointed out that it was cumbersome, confusing and
> inaccessable and predicted that none of our non-English communities would use
> it.

Nah, you can't say "i told you so", because I wasn't around the webteam
when that was built :-P

> So, my vote is that whether or not we have *an* translation infrastructure,
> the current incomplete and non-standard infrastructure be junked. It's never
> going to be used in its current form.

ok.
If we're doing this, we should definitely decide so *before* the French
and German guys who just said they might do it get started.

So perhaps the deal is junk what we have now, and then think hard about
how it should *really* work? With proper input from the people who are
actually going to use it?

> Further, we're going to have to expect that some language communities will
> never translate the main site, since that puts them in a position of having
> all new content generated in English and just having their site mirror that
> without the ability to add new content originating in their group. And I
> don't think that most of our language groups are large enough to sustain both
> organizing content for their site and keeping translations of the English
> site updated.

That was what I thought originally in this thread. Seems at least the
French community disagrees with us.

> P.S. Josh D, you are absolutely wrong about our language composition. The
> majority of our community speaks a first language other than English, and at
> least half of the non-English speakers aren't fluent in English. There are
> large communities in Brazil, Spanish-speaking South America, Italy and
> Germany -- as well as Japan -- which you aren't aware of because they don't
> join the English-speaking MLs for obvious reasons. When I used to answer
> webmaster@, for example, I got *more* questions in Portuguese than in
> English.
>

Right. In case you didn't noticed, we added some google analytics stuff
to the website (because it was the easiest way) and will be collecting
some visitor statistics over the next couple of days - wrt where in the
world people are coming from, and what languages their browsers are
configured for.

//Magnus

Hi,

Josh Berkus wrote:
> I have to say "I told you so". When the existing translation scheme was
> built two+ years ago, I pointed out that it was cumbersome, confusing and
> inaccessable and predicted that none of our non-English communities would use
> it.

The main problem as "I told you back then" is that a person willing to
contribute to the website has to jump through a lot of hoops. As you probably
noticed, potential translators who participated in this thread had no clue about
the possibility of website translation. To know about that requires either
searching the archives of pgsql-www or looking at pgweb module on gborg (which
is itself "deprecated" for quite a bit of time).

So to learn about translation infrastructure one essentially has to already know
about translation infrastructure.

> So, my vote is that whether or not we have *an* translation infrastructure,
> the current incomplete and non-standard infrastructure be junked. It's never
> going to be used in its current form.

Well, even if you create a complete and standard infrastructure you'll still
need to translate at least
1) Script messages and words from common templates. This can be done now through
complete and standard gettext.
2) Content stored in database (news and such). There is an interface for this
now, though it may require polishing (no one can say for sure, 'cause no one
actually *used* that).

The only real problem IMO is "static" pages containing a lot of text and stored
in CVS currently. It wasn't the brightest idea back then and they probably
belong in the database, alongside all other stuff.

Magnus Hagander a écrit :
> Josh Berkus wrote:
>> Magnus,
>>
>> I have to say "I told you so". When the existing translation scheme was
>> built two+ years ago, I pointed out that it was cumbersome, confusing and
>> inaccessable and predicted that none of our non-English communities would use
>> it.
>
> Nah, you can't say "i told you so", because I wasn't around the webteam
> when that was built :-P
>

And I didn't say it was confusing. It seems pretty simple to me.

>> So, my vote is that whether or not we have *an* translation infrastructure,
>> the current incomplete and non-standard infrastructure be junked. It's never
>> going to be used in its current form.
>
> ok.
> If we're doing this, we should definitely decide so *before* the French
> and German guys who just said they might do it get started.
>

I can wait :) I have enough things to do actually : a paper on
PostgreSQL 8.2, some slides (a neat idea from Devrim and Magnus :) ),
some more documentation-oriented stuff.

> So perhaps the deal is junk what we have now, and then think hard about
> how it should *really* work? With proper input from the people who are
> actually going to use it?
>
>
>> Further, we're going to have to expect that some language communities will
>> never translate the main site, since that puts them in a position of having
>> all new content generated in English and just having their site mirror that
>> without the ability to add new content originating in their group. And I
>> don't think that most of our language groups are large enough to sustain both
>> organizing content for their site and keeping translations of the English
>> site updated.
>
> That was what I thought originally in this thread. Seems at least the
> French community disagrees with us.
>

You're right. As far as I can say, we can do it. But we are one
community beyond many others. I would really like to know what others
thought about this. Do someone know how many local PostgreSQL
communities exist ?

And I don't think we should keep code for only one community.

>> P.S. Josh D, you are absolutely wrong about our language composition. The
>> majority of our community speaks a first language other than English, and at
>> least half of the non-English speakers aren't fluent in English. There are
>> large communities in Brazil, Spanish-speaking South America, Italy and
>> Germany -- as well as Japan -- which you aren't aware of because they don't
>> join the English-speaking MLs for obvious reasons. When I used to answer
>> webmaster@, for example, I got *more* questions in Portuguese than in
>> English.
>>
>
> Right. In case you didn't noticed, we added some google analytics stuff
> to the website (because it was the easiest way) and will be collecting
> some visitor statistics over the next couple of days - wrt where in the
> world people are coming from, and what languages their browsers are
> configured for.
>

Having a booth at Solutions Linux is really a great thing. We can get in
touch with people who use PostgreSQL but don't use mailing lists, people
who are not members of the association. They really use the french
translation, mostly because they aren't fluent in english. Really a lot
more people than what I thought. They are technical people but they
really need a translation. So, yes, I still think we would make use of a
translated www.postgresql.org website.

--
Guillaume.
<!-- http://abs.traduc.org/
http://lfs.traduc.org/
http://docs.postgresqlfr.org/ -->

Alexey Borzov a écrit :
> Josh Berkus wrote:
>> I have to say "I told you so". When the existing translation scheme
>> was built two+ years ago, I pointed out that it was cumbersome,
>> confusing and inaccessable and predicted that none of our non-English
>> communities would use it.
>
> The main problem as "I told you back then" is that a person willing to
> contribute to the website has to jump through a lot of hoops. As you
> probably noticed, potential translators who participated in this thread
> had no clue about the possibility of website translation. To know about
> that requires either searching the archives of pgsql-www or looking at
> pgweb module on gborg (which is itself "deprecated" for quite a bit of
> time).
>
> So to learn about translation infrastructure one essentially has to
> already know about translation infrastructure.
>

+1

>> So, my vote is that whether or not we have *an* translation
>> infrastructure, the current incomplete and non-standard infrastructure
>> be junked. It's never going to be used in its current form.
>
> Well, even if you create a complete and standard infrastructure you'll
> still need to translate at least
> 1) Script messages and words from common templates. This can be done now
> through complete and standard gettext.
> 2) Content stored in database (news and such). There is an interface for
> this now, though it may require polishing (no one can say for sure,
> 'cause no one actually *used* that).
>
> The only real problem IMO is "static" pages containing a lot of text and
> stored in CVS currently. It wasn't the brightest idea back then and they
> probably belong in the database, alongside all other stuff.
>

+1

--
Guillaume.
<!-- http://abs.traduc.org/
http://lfs.traduc.org/
http://docs.postgresqlfr.org/ -->

>
> That was what I thought originally in this thread. Seems at least the
> French community disagrees with us.
>
>
>> P.S. Josh D, you are absolutely wrong about our language composition. The
>> majority of our community speaks a first language other than English,

I never said first language. I said they speak English. There is a big
difference. If you are going to correct me, please at least read the
post that I made completely.

Sincerely,

Joshua D. Drake

Joshua D. Drake

--

=== The PostgreSQL Company: Command Prompt, Inc. ===
Sales/Support: +1.503.667.4564 || 24x7/Emergency: +1.800.492.2240
Providing the most comprehensive PostgreSQL solutions since 1997
http://www.commandprompt.com/

Donate to the PostgreSQL Project: http://www.postgresql.org/about/donate
PostgreSQL Replication: http://www.commandprompt.com/products/

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

- --On Tuesday, February 13, 2007 20:18:27 +0100 Magnus Hagander
<magnus(at)hagander(dot)net> wrote:

> ok.
> If we're doing this, we should definitely decide so *before* the French
> and German guys who just said they might do it get started.

Ya, the feel I got from those 'non-English' that have participated in this
thread wasn't that the current way was difficult, but that they just couldnt'
find the documentation for it ... again, my impression was that once they were
able to find that, they were eager to get moving ...

- ----
Marc G. Fournier Hub.Org Networking Services (http://www.hub.org)
Email . scrappy(at)hub(dot)org MSN . scrappy(at)hub(dot)org
Yahoo . yscrappy Skype: hub.org ICQ . 7615664
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (FreeBSD)

iD8DBQFF0pUl4QvfyHIvDvMRAigIAJ9007tHZcX4cb/u2Gql1XD3FqfELACg7g+D
juddHID3ILZF+bWA3cQML8Y=
=nX7f
-----END PGP SIGNATURE-----

Josh Berkus wrote:

> P.S. Josh D, you are absolutely wrong about our language composition. The
> majority of our community speaks a first language other than English, and at
> least half of the non-English speakers aren't fluent in English. There are
> large communities in Brazil, Spanish-speaking South America, Italy and
> Germany -- as well as Japan -- which you aren't aware of because they don't
> join the English-speaking MLs for obvious reasons. When I used to answer
> webmaster@, for example, I got *more* questions in Portuguese than in
> English.
>

That certainly isn't the case now - the vast majority of webmaster@
emails are in English - perhaps that indicates that the regional sites
are more widely used these days.

I do agree with what you are saying though.

Regards, Dave.

Alexey Borzov wrote:
> Hi,
>
> Josh Berkus wrote:
>> I have to say "I told you so". When the existing translation scheme
>> was built two+ years ago, I pointed out that it was cumbersome,
>> confusing and inaccessable and predicted that none of our non-English
>> communities would use it.
>
> The main problem as "I told you back then" is that a person willing to
> contribute to the website has to jump through a lot of hoops. As you
> probably noticed, potential translators who participated in this thread
> had no clue about the possibility of website translation. To know about
> that requires either searching the archives of pgsql-www or looking at
> pgweb module on gborg (which is itself "deprecated" for quite a bit of
> time).

Err, no it's not, though we don't tend to use the task manager any more.
Or do you mean GBorg itself?

> So to learn about translation infrastructure one essentially has to
> already know about translation infrastructure.

There are 'translation people' who know about the infrastructure who
still have chosen not to work on translating the website - I suspect
that part of the issue is simply the size of the task rather than
difficultly in doing the job - the po files are there for the dynamic
stuff, the admin site for the stuff that comes and goes on a regular
basis, and the vast majority of the static pages never change (which
means there is not necessarily any need for gettext type tools to
monitor the changes).

>> So, my vote is that whether or not we have *an* translation
>> infrastructure, the current incomplete and non-standard infrastructure
>> be junked. It's never going to be used in its current form.
>
> Well, even if you create a complete and standard infrastructure you'll
> still need to translate at least
> 1) Script messages and words from common templates. This can be done now
> through complete and standard gettext.
> 2) Content stored in database (news and such). There is an interface for
> this now, though it may require polishing (no one can say for sure,
> 'cause no one actually *used* that).
>
> The only real problem IMO is "static" pages containing a lot of text and
> stored in CVS currently. It wasn't the brightest idea back then and they
> probably belong in the database, alongside all other stuff.

That would make management easier, but I don't think it will make a huge
difference to translatability of the site - whilst you could check a
page on the admin site periodically to check for changes since the last
translation update, it would probably be easier to just monitor the
pgweb-commit list and update translations reactively.

Regards, Dave.

On 2/14/07, Marc G. Fournier <scrappy(at)hub(dot)org> wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
>
>
> - --On Tuesday, February 13, 2007 20:18:27 +0100 Magnus Hagander
> <magnus(at)hagander(dot)net> wrote:
>
> > ok.
> > If we're doing this, we should definitely decide so *before* the French
> > and German guys who just said they might do it get started.
>
> Ya, the feel I got from those 'non-English' that have participated in this
> thread wasn't that the current way was difficult, but that they just couldnt'
> find the documentation for it ... again, my impression was that once they were
> able to find that, they were eager to get moving ...

The translation interface is not a confusing mess of unusable junk. It has some
problems but this doesn't mean that it must be rewritten from scratch :

- it's poorly advertised : the fact that the site can be translated
needs to be
clearly mentioned on the website, not only sporadically on pgsql-www .

- setting up a local copy of the website is tricky ( the installation
readme is probably
missing some details ).

- keeping the content up to date is difficult at the moment because
tracking the
changes needs to be done manually by watching the CVS commits.

The first two points have nothing to do with the exact implementation: the new
one can also be widely unknown ...

So: we can start translating using the current infrastructure. The translated
content will not get lost even if later it becomes absolutely clear that the
infrastructure needs to be rewritten : importing the text content to the new
one will be possible one way or another .

Cheers,
Adrian Maier

Hi,

Dave Page wrote:
>> The main problem as "I told you back then" is that a person willing to
>> contribute to the website has to jump through a lot of hoops. As you
>> probably noticed, potential translators who participated in this thread
>> had no clue about the possibility of website translation. To know about
>> that requires either searching the archives of pgsql-www or looking at
>> pgweb module on gborg (which is itself "deprecated" for quite a bit of
>> time).
>
> Err, no it's not, though we don't tend to use the task manager any more.
> Or do you mean GBorg itself?

Yes, I meant GBorg itself, sorry for lack of clarity.

>> So to learn about translation infrastructure one essentially has to
>> already know about translation infrastructure.
>
> There are 'translation people' who know about the infrastructure who
> still have chosen not to work on translating the website - I suspect
> that part of the issue is simply the size of the task rather than
> difficultly in doing the job - the po files are there for the dynamic
> stuff, the admin site for the stuff that comes and goes on a regular
> basis, and the vast majority of the static pages never change (which
> means there is not necessarily any need for gettext type tools to
> monitor the changes).

I do understand the task is quite big, but attracting more people to it will
allow it to be split in more manageable chunks.

>> Well, even if you create a complete and standard infrastructure you'll
>> still need to translate at least
>> 1) Script messages and words from common templates. This can be done now
>> through complete and standard gettext.
>> 2) Content stored in database (news and such). There is an interface for
>> this now, though it may require polishing (no one can say for sure,
>> 'cause no one actually *used* that).
>>
>> The only real problem IMO is "static" pages containing a lot of text and
>> stored in CVS currently. It wasn't the brightest idea back then and they
>> probably belong in the database, alongside all other stuff.
>
> That would make management easier, but I don't think it will make a huge
> difference to translatability of the site - whilst you could check a
> page on the admin site periodically to check for changes since the last
> translation update, it would probably be easier to just monitor the
> pgweb-commit list and update translations reactively.

That will make a difference, as we gain 2 types of content to translate rather
than current 3. Also we can add customized notifications rather than make people
monitor the whole commits list.

Of course that leads us to implementing a means to add new pages and sections to
the website without having to manually add them to the navigation templates.
Then we need to have a flexible access rights infrastructure that will allow
translators to work on their stuff without the possibility that one day the
front page of postgresql.org will be presented in a mix of Japanese, Russian and
Turkish.

Also please note, I'm not trying to defend the current website architecture, it
definitely isn't the best one: my goal back then was to get the site out ASAP
and reuse as much previously written code as possible. But claiming that
translation "doesn't work" when no one ever tried to use it is laughable.

>> Okay. Let me then ask the other thing that we really haven't discussed:
>> Do we want the language site to be served up automatically. That's
>> really what the code is all about today. If your browser is set to
>> prefer French, we'll serve up the French version of the page (if
>> available). While convenient, I personally don't like sites that do this
>> for me (I've reconfigured my browser to put English as the top language
>> for that reason).
>
> I kind of agreed with you on this one. I particularly hate when google
> put me on their french search website when I really asked to be in the
> english one. I hate this because, with the french search website, french
> pages are ranked higher than others. But I don't think this will be the
> point here. If you have a browser configured to put french as top
> language, that's just because you prefer french pages. And if you don't
> want that, you will have a combo to choose another language.

Well, google is much worse than that. They put you in the language
corresponding to *where you are*, and completely ignore your browsers
setting. So when I'm in France, I get google in French. Imagine the joy
when you're in Japan...

But nobody is suggesting we do that, I hope :-)

//Magnus

Magnus Hagander a écrit :
>>> Okay. Let me then ask the other thing that we really haven't discussed:
>>> Do we want the language site to be served up automatically. That's
>>> really what the code is all about today. If your browser is set to
>>> prefer French, we'll serve up the French version of the page (if
>>> available). While convenient, I personally don't like sites that do this
>>> for me (I've reconfigured my browser to put English as the top language
>>> for that reason).
>> I kind of agreed with you on this one. I particularly hate when google
>> put me on their french search website when I really asked to be in the
>> english one. I hate this because, with the french search website, french
>> pages are ranked higher than others. But I don't think this will be the
>> point here. If you have a browser configured to put french as top
>> language, that's just because you prefer french pages. And if you don't
>> want that, you will have a combo to choose another language.
>
> Well, google is much worse than that. They put you in the language
> corresponding to *where you are*, and completely ignore your browsers
> setting. So when I'm in France, I get google in French. Imagine the joy
> when you're in Japan...
>
> But nobody is suggesting we do that, I hope :-)
>

Well, I know I'm not suggesting this :-D

--
Guillaume.
<!-- http://abs.traduc.org/
http://lfs.traduc.org/
http://docs.postgresqlfr.org/ -->