2013/4/15 Dominique Pellé <[email protected]>
>
> I see. Even without the github limit, cloning such
> a huge git repository would be a pain. Whereas
> with svn, we only checkout one version (smaller).
>
> In the LanguageTool source tree, we checkin
> several binary files which do not compress well in
> delta. I suppose that's why git's repository is so big.
> All the POS tag dictionaries and FSA spelling dictionaries
> must be quite big, with multiple versions and multiple
> languages. I always hesitate before checking
> a new version of a binary dictionary and try to
> delay it if possible to make one checkin instead
> of multiple small checkins. It would make sense
> to build them from scripts as part of the build,
> rather than checking-in binary dictionaries. But
> that would also make the build very slow. Source
> code and binary files could be in different
> repositories but that's not convenient either.
> Well, I have no ideal solution :-(
>
>
In another project, we already faced this problem, and we decoupled
binaries and soruce in order fix that size-increase our repository had (
https://github.com/Softcatala/Catalanitzador and
https://github.com/Softcatala/Catalanitzador-binaries)
Maybe that can't be done here, as th binaries might be spread across
several directories, so using --ignore-paths when svn cloning the
repositories might make no sense, but could you confirm it?
--
< Xavi Ivars >
< http://xavi.ivars.me >
------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
_______________________________________________
Languagetool-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/languagetool-devel
