Trialing KDE on bzr

Thiago Macieira thiago at kde.org
Sun Jul 20 20:10:34 BST 2008


Let's please move this discussion to the mailing list we set up to discuss 
this very subject.

Andreas Hartmetz wrote:
>My current opinion is that no DVCS is worth it as long as it requires us
> to split some things up instead of using subdirectories like SVN.
> Subdirectories are already set up, easy to use, don't force unnecessary
> decisions, handle file moves better and probably have some other
> advantages over any kind of split projects.
>Not using the file system's capabilities and semantics [they are
> well-known and work] is Wrong IMHO.
>Unfortunately subdirectory support also seems to be really hard to
> implement in DVCSes - maybe one or more will pull it off anyway, we'll
> see.

The only reason we use one huge SVN repository today is because it was 
necessary in order to support copy-with-history across modules. That's 
the only showstopper. (BTW, when Linus criticised us for having the big 
repository, he forgot/didn't know about this factor)

I know Git can support it more or less easily and we have developer buy-in 
to that feature. I know that, given the nature of the thing, it's also 
possible in other DVCS, though I have no idea whether it's easy to 
implement or if it's a crude hack.

The one thing that the filesystem buys us is that we don't have to decide 
whether we split at kdelibs or the library; extragear, extragear/category 
or extragear/category/appname.

But we *can* decide on something. I made a proposal, Oswald made a 
different proposal. I think mine is workable: the split is at the tarball 
level (each main KDE module, plus koffice; kdesupport, extragear, 
playground and review are all "a collection of applications/libraries" 
each packaged individually and must build alone).

I see no technical reason to not switch to a DVCS. There's a lot of people 
who want to have a DVCS -- many are already using one. What's stopping us 
now is inertia and having people to do the actual conversion work.

-- 
  Thiago Macieira  -  thiago (AT) macieira.info - thiago (AT) kde.org
    PGP/GPG: 0x6EF45358; fingerprint:
    E067 918B B660 DBD1 105C  966C 33F5 F005 6EF4 5358
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part.
URL: <http://mail.kde.org/pipermail/kde-core-devel/attachments/20080720/50f2174c/attachment.sig>


More information about the kde-core-devel mailing list