Re: cvs and XFree86 (Was Re: PATCH: -multiwindow with root window hidden...)

[Igor Pechtchanski <pechtcha at cs dot nyu dot edu>]

Alan,

That's good to know.  So then developers could simply use the xoncygwin
CVS instead of the source tarballs.  Except...  I went to
http://xoncygwin.sourceforge.net/ and all I get is an empty directory
listing.  What gives?
	Igor

On Thu, 1 May 2003, Alan Hourihane wrote:

> Actually Igor,
>
> We do. Goto http://xoncygwin.sourceforge.net.
>
> The CVS is there, and I've used it to write the NativeGDI code.
>
> So, to encourage more development there, I suggest adding Kensuke
> to the CVS committers at the very least.
>
> Alan.
>
> On Thu, May 01, 2003 at 11:19:21AM -0400, Igor Pechtchanski wrote:
> > Harold,
> >
> > I know it's kinda off-topic, but have you considered setting up a local
> > cvs repository and doing "cvs import"s from the main CVS (your comment
> > below seems to indicate that you don't do that)?  It works quite nicely
> > (for me, on another project), and you get the benefit of revision control
> > for your local changes, among other things.
> >
> > I know that you having a local CVS repository doesn't really help Early or
> > any other developers, although nothing really stops them from doing the
> > same, i.e.,
> > - unpack the source, cd to the source directory
> > - "cvs -d .repository init",
> > - "find . -name .repository -prune -o -name CVS -prune -o -type d -exec cvs -d .repository add {} \;",
> > - "find . -name .repository -prune -o -name CVS -prune -o -type f -print | xargs cvs add",
> > - "cvs commit" from the top level,
> > and then any changes they make after that can be committed to the local
> > repository).
> >
> > FWIW, there were also offers of setting up separate branches (not in the
> > XFree86 CVS tree, though, AFAIR), so that developers can work on their
> > code and then someone in charge would merge their changes into the
> > trunk...  Maybe the powers that be would consent to that instead?
> >
> > Just a few thoughts...
> >       Igor
> >
> > On Thu, 1 May 2003, Harold L Hunt II wrote:
> >
> > > Ruth,
> > >
> > > No, cvs diff would have been useless.
> > >
> > > The XFree86 CVS tree limits commits to only a few "core" members, which
> > > I am not one of which.  This means that I cannot just "cvs commit" my
> > > changes, I have to periodically create a set of patches (which I use
> > > "cvs diff" for) and a change log and send them to "patches at
> > > xfree86.org".   These patches are eventually committed.
> > >
> > > The problem here is that I haven't done that in awhile, so someone
> > > sending me a "cvs diff" wouldn't do me a whole lot of good, since the
> > > code in cvs is quite old at the moment.  What they would have to do
> > > instead is pick a recent Test release, stick it in hw/xwin-TestXX then:
> > >
> > > cd xc/programs/Xserver/hw
> > > diff -U3 -N xwin-TestXX xwin > xwin-TestXX-PlusMyChanges.diff
> > >
> > > That works quite nicely.
> > >
> > > Harold
> > >
> > > Ruth Ivimey-Cook wrote:
> > > > At 22:26 30/04/2003, you wrote:
> > > >
> > > >> Heh... the only problem with that would be that, in this case, CVS is
> > > >> way out of date :)  I will probably synch up with CVS again soon.
> > > >
> > > > If the cvs source you're using is the one you compiled from, then cvs
> > > > diff will do the right thing: compare your modified sources against the
> > > > cvs source you started from, not the current version (whatever that
> > > > happens to be).
> > > >
> > > > So cvs diff is (probably) your friend, in this case.
> > > >
> > > > Ruth

-- 
				http://cs.nyu.edu/~pechtcha/
      |\      _,,,---,,_		pechtcha@cs.nyu.edu
ZZZzz /,`.-'`'    -.  ;-;;,_		igor@watson.ibm.com
     |,4-  ) )-,_. ,\ (  `'-'		Igor Pechtchanski
    '---''(_/--'  `-'\_) fL	a.k.a JaguaR-R-R-r-r-r-.-.-.  Meow!

Knowledge is an unending adventure at the edge of uncertainty.
  -- Leto II