> -----Original Message-----
> From: fcoe-devel [mailto:[email protected]] On Behalf Of
> Dev, Vasu
> Sent: Friday, January 30, 2015 9:14 AM
> To: Chris Leech
> Cc: [email protected]
> Subject: Re: [Open-FCoE] [PATCH] fcoe-utils: automake fixes for
> dist/distcheck
> 
> > -----Original Message-----
> > From: Chris Leech [mailto:[email protected]]
> > Sent: Thursday, January 29, 2015 9:35 PM
> > To: Dev, Vasu
> > Cc: [email protected]
> > Subject: Re: [Open-FCoE] [PATCH] fcoe-utils: automake fixes for
> > dist/distcheck
> >
> > On Fri, Jan 30, 2015 at 01:54:42AM +0000, Dev, Vasu wrote:
> > > >
> > > > I thought I'd give make distcheck a try on fcoe-utils and see if
> > > > it produced a working release.  It didn't, so here are some fixes for 
> > > > that.
> > > >
> > > > Missing header files added to git are included in the release archive.
> > > >
> > > > The systemd unit files are marked dist_DATA so they get included,
> > > > otherwise data files are assumed to be generated at build time and
> > > > left out of the release.
> > > >
> > > > A few other missing doc files are added to dist_noinst_DATA.
> > > >
> > > > The manual rules for handling bash completion files are replaced
> > > > with dist_DATA definitions letting automake handle generating rules.
> > > >
> > > > Any change of applying build system fixes, and getting a new tag
> > > > and release tarball anytime in the near future?
> > > >
> > >
> > > I announced last year of doing away  with upstream tar ball release
> > > and in fact when we had it, then also it was simply tar of three
> > > open-fcoe user tools git trees instead of using distcheck, so I
> > > won't be surprised if this doesn't work for other two user tools git
> > > trees but any
> > case good to fix all three git repos.
> >
> > Ah, I knew the kernel code wasn't using a separate tree anymore but I
> > guess I missed the bit about not doing user-space releases.
> >
> > libhbaapi/libhbalinux look to have some minor autotools issues, I can
> > send a patch if you want.
> >
> 
> Always, that would be good.
> 
> > > I'm not sure if we need tags but sure user tools version bump make
> > > sense and I'm thinking of using same for tag instead of we used to
> > > have tag every kernel cycle even when no change to user tools.
> >
> > Yeah, the kernel version tags aren't all that useful - I keep deleting
> > them locally so they don't mess with git-describe.  I get that Rob was
> > trying to keep the same cadence as the kernel releases when there were
> > active changes going on.
> >
> > fcoe-utils is now 33 patches ahead of the last tag, with the bulk of
> > those being in git for over a year.  I just thought it would be nice
> > to bump the version to 1.0.30 and tag it, maybe put out a tarball.  I
> > don't expect this to be a frequent thing.
> >
> 
> Yes looks like won't be frequent. All good suggestions, I agree with all and
> doing tarball release along any major version bump makes sense given that it
> should be infrequent.
> 
I mean any version bump not major version, I was considering later but will
defer that for now.

//Vasu
_______________________________________________
fcoe-devel mailing list
[email protected]
http://lists.open-fcoe.org/mailman/listinfo/fcoe-devel

Reply via email to