[Zip] Initial Mail after ZIP revival at EuroZope Conference

[Juan David Ibáñez Palomar]

> 
> Zope can be localized using the current tools. There might be some things 
> to make this nicer, such as automatic language detection (which people at 
> DC work on), but in general we got everything.
> 

I don't know about ZBabel, but Localizer is far to get everything.


> I think we should push to get one tool, build a Demo localized Zope version 
> and push DC to accept this Localization product in the core.
> 

"push DC to accept" sounds like a too much aggresive expression to me.
DC is not stupid, if we do things right and in the right way we'll get
it, without "pushing to accept".


> >ZBabel and Localizer are much more advanced solutions than what we had 
> >before (zzLocale and the others) but it's still a patching on Zope. I 
> >think that I18N support should come in the Zope core and DC will agree if 
> >we get to explain it enough clear.
> 
> Yes, via a product. But we need a product that covers 95% of all the cases 
> first, before it gets into the core. For example SQL Methods are part of 
> the core, but they are still in the Products Directory.
> 

Yes and no. Some things like the Unicode support or the internationalization
and localization of Zope itself must be in the core. While other facilities
to build multilingual web sites should be in a product.


> >- Much discussion about what does Zope need to be fully Localisable. But 
> >don't forget that we should profit the work and experience of other 
> >software developements ( We are not discovering anything. The software 
> >industry and the Free Software community is dealing with those problems 
> >for many many years. )
> 
> That's the reason Localizer as well as ZBabel both use very similar 
> approaches to the other projects. In fact, now that I think about it again, 
> I think the internationalization should be optional in the Zope core as the 
> ZCatalog is.
> 
> >- I don't think that we need convergence of solutions NOW. I think each of 
> >the approaches should follow his way and see the results after a bit of 
> >testing. We should try, test and think about each solution before starting 
> >to discuss. As far as I know Localizer 0.5 with Localizable Content and 
> >Properties are available only for a couple of hours...
> 
> WHY? We have far too many solutions already. Instead of me and a couple 
> other guys working on ZBabel and some guys working on Localizer, why should 
> we not join together and have a programming team of 15 people or so? The 
> results will be much better!!!
> 

It would be very nice to merge ZBabel and Localizer into one product, but
it's not so much essential. There's nothing wrong to have several products
to experiment with.


> >So... let's discuss about what does Zope need to be a fully multilingual 
> >application.
> 
> It needs the functionality of the translation tools already done! It's that 
> easy! All we need, is to get one into the core!
> 
> Let's take KDE for example. The Qt and KDE libs are not localized, but 
> there is an additional lib to localize your programs. That is exactly how 
> Localizer and ZBabel work. The Zoep libs are not localized, but with an 
> additional library (Localizer or ZBabel) you can do it!
> 

It's not that easy, it's difficult, very difficult.

If tomorrow DC says me: "we want to put Localizer in the core" I'd
reply: "are you crazy?". As Gari said Localizer is, in some way, a
patch to Zope.

I much prefer the following:

 - Continue discussion about ZBabel/Localizer/.. and try to merge.

 - If we merge: let's celebrate! If not, nothing bad happens,
   there's nothing wrong to have several tools to experiment with.

 - In any case, start a more general discussion, in this mailing
   list and in some ZWiki (at eurozope.org or in the Fishbowl for
   example), where we'll try to address *everything* related to
   i18n/l10n.

 - From this discussion we'll make different proposals in a non
   random order to the Fishbowl and develop them further (for
   this it will be important to have the CVS opened).

 - Perhaps some time in the future we'll have a product good enough
   to be distributed in the standard Zope distribution.

Summary:

  Several things need to be implemented in the core, others would
  be better addressed in a separate product. About the things that
  must be in the core, we must follow the Fishbowl process and we
  need the DC help, specially in the implementation details.


Regards,
jdavid



> Regards,
> Stephan
> 
> --
> Stephan Richter
> CBU - Physics and Chemistry Student
> Web2k - Web Design/Development & Technical Project Management
> 
>