[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: Remove a branch directory completely from cvs



On Tue, 07 Oct 2008 20:31:49 +0200
opensource till name (Till Maas) wrote:

> On Tue October 7 2008, Kevin Fenzi wrote:
> > On Tue, 07 Oct 2008 17:26:38 +0200
> >
> > (Till Maas) wrote:
> 
> > > according to the UsingCvsFaq[1] one can remove a branch directory
> > > from cvs with "cvs remove". I tried this, but it seems not to
> > > work. I removed the "FC-1" directory from pam_mount, but when I
> > > cvs co'ed the pam_mount directory again, an empty "FC-1"
> > > directory appeared. Is there some parameter I can pass to "cvs
> > > remove" to get the directory removed, too?
> >
> > Why are you needing to do this?
> >
> > Granted, not many people should need to look at what was in FC-1
> > much, but it's possible there could be something there someone
> > would want to look at, and now all the files in there are marked
> > dead, so it's more of a chore to get them to look at.
> 
> Using the viewcvs webinterface, it is pretty easy to get the dead
> files:
> http://cvs.fedoraproject.org/viewvc/rpms/pam_mount/FC-1/?hideattic=0
> 
> I guess there is also some commandline option that allows to easily
> access these files.

I suppose. 

> > Perhaps we could look at archiving off all those old branches if
> > they are causing a problem?
> 
> After I got this working, the next step was to suggest to remove all
> old branches (F-7 and older) for every package in the near future.

Why not ask infrastructure to see about moving those all off to a
archive site/location? 

> The reason why I want this done, is that they mess up some packages,
> where there are now 10 or more branch directories. 

Oh? how? 

> Also it takes
> unecessarily long for to checkout packages. Especially packages that
> not only have many branches, but also a lot of patches.

That shouldn't be a function of how many branches there are, but how
many files/revisions are in CVS for the files you are checking out. 

Wouldn't it be better to look at moving to a newer VCS that can handle
this case better? You might look and add that as a desired feature for
the next VCS. 

> Regards,
> Till

kevin

Attachment: signature.asc
Description: PGP signature


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]