On Sat, Jun 25, 2011 at 11:18 AM, Dick Groskamp <[email protected]>wrote:
> Op 25-6-2011 17:23, Rob Weir schreef: > >> Thanks for the list. I looked around. Some lists are very active. >> Some have not seen activity for a year or more. Some seem to never >> have been active. And some are just full of spam :-( >> >> I can see three ways to decide what to do (but maybe someone has other >> ideas?): >> >> 3) Create lists only when there is a sufficient number of project >> members on the Apache list asking for new list. >> > +1 > >> Right now we have just a single discussion pubic list, ooo-dev. I can >> easily imagine, that once we have some code checked in and start >> actively working on making our first release, that the traffic in that >> one list will be larger enough that we'll want to split into >> specialized functional lists, maybe: >> [snip] >> >> >> I don't think we're there yet, but I can certainly see that happening >> in the next few weeks/months. >> > I can surely agree with that > > 1) Do we want Apache to host a Pootle server? If so, we need to put >> together that request and make it happen. >> > O yes please...... (but I was the main cinbtributor for that on the "old" > project for the Dutch language.) > > But there was another reason. SUN/Oracle made the builds with about 8 > langiages EMBEDDED in them. > Yes, indeed one of them was Dutch. :-) > > Furthermore the proces was more and more set to be automated. Source > strings were fed to POOTLE. > The translators for each language could then do their job in POOTLE and > engineers from SUN/Oracle > pulled, when they could use them in htier proces, the strings from POOTLE > and proces them towards the build. > Last minute translations could mostly be delivered on time for the final > pull for releases > > Since localization seemed to be always at the end of the proces translators > had to work under duress. > This way the pressure was less at the end of the proces since strings were > already for the biggest part already > translated. > At the same time, there was efforts to do the continuos l10n which was meant to have localization at any time of the process. http://wiki.services.openoffice.org/wiki/ContinuousL10n server for the continuos l10n will give out reports on false strings. ftp://qa-upload.services.openoffice.org/l10n/download/ There was also special KeID builds to do QA on the locale which ease the context of the locale: http://wiki.services.openoffice.org/wiki/KeyID_Build > If not using POOTLE you will have to think of a way to collect the > tranlations to put them into the build. > Will APACHE build the same EMBEDDED builds as SUN/Oracle did or willl there > only be languagepacks. > If the latter who will make them ? > >> 2) Did the Oracle SGA include all of the language translation sources? >> If not, we need to identify what is missing >> [snip] >> Apache is different. >> > No problem with that at all. Most of us just want to produce a nice working > officesuite > > > -- > DiGro > > Windows 7 and OpenOffice.org 3.3 > Scanned with Ziggo uitgebreide Internetbeveiliging (F-Secure) > > -- *Alexandro Colorado* *OpenOffice.org* Español http://es.openoffice.org
