1:46 a.m.
>>>> "CA" == Christopher Aillon writes:
CA> Author: caillon Update of /cvs/extras/rpms/blam/F-9 In directory
CA> cvs1.fedora.phx.redhat.com:/tmp/cvs-serv30824
CA> Modified Files: blam.spec Log Message: * Wed Sep 24 2008
CA> Christopher Aillon <caillon(a)redhat.com> - 1.8.5-2 - Rebuild
CA> against newer gecko
Are you going to bump and rebuild the rawhide/F-10 branch? rawhide is
on:
blam-1.8.5-1.fc10
which will cause a broken upgrade path as
blam-1.8.5-2.fc9 > blam-1.8.5-1.fc10
Can you consider bumping the release tag *after* the disttag, i.e.:
blam-1.8.5-1.fc9.1
this would avoid the need to rebuild the rawhide branch.
Alex
2:32 a.m.
New subject: rpms/blam/F-9 blam.spec,1.24,1.25
On 24.09.2008 08:46, Alex Lancaster wrote:
>>>>> "CA" == Christopher Aillon writes:
CA> Author: caillon Update of /cvs/extras/rpms/blam/F-9 In directory
CA> cvs1.fedora.phx.redhat.com:/tmp/cvs-serv30824
CA> Modified Files: blam.spec Log Message: * Wed Sep 24 2008
CA> Christopher Aillon <caillon(a)redhat.com> - 1.8.5-2 - Rebuild
CA> against newer gecko
Are you going to bump and rebuild the rawhide/F-10 branch? rawhide is
on:
blam-1.8.5-1.fc10
which will cause a broken upgrade path as
blam-1.8.5-2.fc9 > blam-1.8.5-1.fc10
Can you consider bumping the release tag *after* the disttag, i.e.:
blam-1.8.5-1.fc9.1
this would avoid the need to rebuild the rawhide branch.
In case anybody is interested: a rough and not much tested patch(¹) that
makes rpmdev-bumpspec from rpmdevtools optionally bump the rightmost
part of the release field can be found in:
https://fedorahosted.org/rpmdevtools/ticket/1
CU
knurd
(¹) done by someone with only basic python and programming skills ;-)
3:43 p.m.
New subject: rpms/blam/F-9 blam.spec,1.24,1.25
Alex Lancaster wrote:
>>>>> "CA" == Christopher Aillon writes:
CA> Author: caillon Update of /cvs/extras/rpms/blam/F-9 In directory
CA> cvs1.fedora.phx.redhat.com:/tmp/cvs-serv30824
CA> Modified Files: blam.spec Log Message: * Wed Sep 24 2008
CA> Christopher Aillon <caillon(a)redhat.com> - 1.8.5-2 - Rebuild
CA> against newer gecko
Are you going to bump and rebuild the rawhide/F-10 branch? rawhide is
on:
Yes, I intend on doing this for all packages that need it but my first
priority is for F8 and F9.
blam-1.8.5-1.fc10
which will cause a broken upgrade path as
blam-1.8.5-2.fc9 > blam-1.8.5-1.fc10
Can you consider bumping the release tag *after* the disttag, i.e.:
blam-1.8.5-1.fc9.1
this would avoid the need to rebuild the rawhide branch.
I'm using bumpspecfile.py[1] to do this automatically because manually
doing it for 30ish packages just isn't my idea of fun...
[1]
http://fedoraproject.org/wiki/PackageMaintainers/UsefulScripts#bumpspecfile
3:51 p.m.
New subject: rpms/blam/F-9 blam.spec,1.24,1.25
Thorsten Leemhuis wrote:
On 24.09.2008 08:46, Alex Lancaster wrote:
>>>>>> "CA" == Christopher Aillon writes:
>
> CA> Author: caillon Update of /cvs/extras/rpms/blam/F-9 In directory
> CA> cvs1.fedora.phx.redhat.com:/tmp/cvs-serv30824
>
> CA> Modified Files: blam.spec Log Message: * Wed Sep 24 2008
> CA> Christopher Aillon <caillon(a)redhat.com> - 1.8.5-2 - Rebuild
> CA> against newer gecko
>
> Are you going to bump and rebuild the rawhide/F-10 branch? rawhide is
> on:
>
> blam-1.8.5-1.fc10
>
> which will cause a broken upgrade path as
> blam-1.8.5-2.fc9 > blam-1.8.5-1.fc10
>
> Can you consider bumping the release tag *after* the disttag, i.e.:
>
> blam-1.8.5-1.fc9.1
>
> this would avoid the need to rebuild the rawhide branch.
In case anybody is interested: a rough and not much tested patch(¹) that
makes rpmdev-bumpspec from rpmdevtools optionally bump the rightmost
part of the release field can be found in:
https://fedorahosted.org/rpmdevtools/ticket/1
I'd also like to see it bump 1.1{?dist} -> 1.2{?dist} instead of
2.1{?dist} without needing to pass an option.
4:56 p.m.
New subject: rpms/blam/F-9 blam.spec,1.24,1.25
On Wed, 24 Sep 2008 16:51:39 -0400, Christopher Aillon wrote:
[bumpspecfile]
I'd also like to see it bump 1.1{?dist} -> 1.2{?dist} instead
of
2.1{?dist} without needing to pass an option.
It was made to cover the most common versioning schemes used in the Fedora
package collection without passing an option.
How does it know whether Y in X.Y is not a cvs/svn/post-release/date
identifier or number that it must not touch?
Such as:
1.0-1.a%{?dist} for version 1.0a (post-1.0 update)
1.0-1.20080901%{?dist} for a scm checkout
Side-note: 1.1%{?dist} to begin with may be fine, but only if you
use exactly the same versioning scheme for all %dists. Else you
break the upgrade path too easily, e.g. if an older dist bumps
1%{?dist} to 1.1%{?dist}, which then compares .1 to %{?dist},
and 1 is greater than all fedora dist tags.
5:18 p.m.
New subject: rpms/blam/F-9 blam.spec,1.24,1.25
On Wed, 24 Sep 2008 09:32:31 +0200, Thorsten Leemhuis wrote:
In case anybody is interested: a rough and not much tested patch(¹)
that
makes rpmdev-bumpspec from rpmdevtools optionally bump the rightmost
part of the release field can be found in:
https://fedorahosted.org/rpmdevtools/ticket/1
CU
knurd
(¹) done by someone with only basic python and programming skills ;-)
The patch looks harmless, but I wonder whether the script's users
will remember to use that option when bumping old branches only.
I wonder even more why rpmdev-bumpspec says
Copyright (c) 2005-2008 Red Hat Inc.
which is not correct. 2005 may be true for its origin, the pre-Fedora
Extras era. For the first mass-rebuilds and beyond that it was developed
further by me, and just because I don't want to claim full copyright,
it should not be given to Red Hat for 2006-2008. That sounds very wrong
to me, because nobody at Red Hat has been an author of changes to the
script in those years.
12:55 a.m.
New subject: rpms/blam/F-9 blam.spec,1.24,1.25
On Thursday 25 September 2008, Michael Schwendt wrote:
I wonder even more why rpmdev-bumpspec says
Copyright (c) 2005-2008 Red Hat Inc.
which is not correct. 2005 may be true for its origin, the pre-Fedora
Extras era. For the first mass-rebuilds and beyond that it was developed
further by me, and just because I don't want to claim full copyright,
it should not be given to Red Hat for 2006-2008. That sounds very wrong
to me, because nobody at Red Hat has been an author of changes to the
script in those years.
This was discussed in several mails between you, me and notting in March when
I was about to add the script in rpmdevtools. I asked you to add copyright
notices to the script, and you added it but without mentioning a copyright
holder.
When discussing it further, you concluded "(c) Fedora Project ... sounds
reasonable. Who is/are the original author(s)? [...]" and did not mention
that you would like to have your name there. To your question (which
mentioned someone at Red Hat giving it out for the pre-Fedora Extras 3
period), Bill answered "If it was given out by someone @RH, you can just put
(c) Red Hat on it." and there were no further replies. I wanted to have
something as the copyright holder in it, and since your version didn't have
one (it still doesn't, it seems), I put what's currently in it there based on
Bill's suggestion. Maybe adding the years was not a good thing to do, but I
don't think there's any irreparable harm done.
If you or anyone who knows the history of the script can suggest a more
truthful set of copyright holder lines to have in it, patches welcome. I'm
not going to invent anything based on informal discussions for the 2nd time,
so explicit patches or specified lines are needed to change it. Or if you
like, the offer for commit access to rpmdevtools upstream (also discussed in
March) still stands so you could fix it yourself, just request it in FAS and
I'll approve it.
12:58 a.m.
New subject: rpms/blam/F-9 blam.spec,1.24,1.25
On Wednesday 24 September 2008, Thorsten Leemhuis wrote:
Ah, I had no idea that a ticket was waiting there, I was not notified by email
of its existence - the fedorahosted trac setup does not seem to notify anyone
useful in new out of the box projects. It should be configured to mail me
now, and I'll look into the ticket soon.
2:52 a.m.
New subject: rpms/blam/F-9 blam.spec,1.24,1.25
On Thu, 25 Sep 2008 08:55:38 +0300, Ville Skyttä wrote:
On Thursday 25 September 2008, Michael Schwendt wrote:
> I wonder even more why rpmdev-bumpspec says
>
> Copyright (c) 2005-2008 Red Hat Inc.
>
> which is not correct. 2005 may be true for its origin, the pre-Fedora
> Extras era. For the first mass-rebuilds and beyond that it was developed
> further by me, and just because I don't want to claim full copyright,
> it should not be given to Red Hat for 2006-2008. That sounds very wrong
> to me, because nobody at Red Hat has been an author of changes to the
> script in those years.
This was discussed in several mails between you, me and notting in March when
I was about to add the script in rpmdevtools. I asked you to add copyright
notices to the script, and you added it but without mentioning a copyright
holder.
I've found those mails, but they don't cover such a very specific
copyright line that differs from the one I called "reasonable".
It's simply not accurate.
When discussing it further, you concluded "(c) Fedora Project
... sounds
reasonable. Who is/are the original author(s)? [...]"
Because I'm not a lawyer, and the script is derived work based on something
that didn't have any legal stuff attached to it at all. I wrote the
following and added Bill:
: (c) Fedora Project ... sounds reasonable.
:
: Who is/are the original author(s)? The skeleton of the original code is
: still present. If memory serves correctly, the first version of the script
: was given out by someone at Red Hat for the pre-Fedora Extras 3 period and
: then improved for the first mass-rebuilds. But who exactly was the original
: author, I don't know. notting or sopwith or mkj or gafton? Neither one?
You see? With an existing list of authors, I could have added myself.
But claiming full copyright and credits [for any original bits that
may be left] is an entirely different thing.
and did not mention
that you would like to have your name there.
I don't care about my name in there. I don't get my name attached to
lots of patches for F/LOSS either. And this is just a script that can be
rewritten from scratch if someone insists on doing that.
To your question (which
mentioned someone at Red Hat giving it out for the pre-Fedora Extras 3
period), Bill answered "If it was given out by someone @RH, you can just put
(c) Red Hat on it." and there were no further replies.
True, but Bill apparently was the wrong one to ask. And a different
proposal was made earlier, too. However, the added line changes history.
12:33 p.m.
New subject: rpms/blam/F-9 blam.spec,1.24,1.25
On Thu, 25 Sep 2008 20:11:55 +0300, Ville Skyttä wrote:
On Thursday 25 September 2008, Michael Schwendt wrote:
> However, the added line changes history.
Would you mind just telling me how to fix it so we can move on?
Replacing it with (c) 2005-2008 Fedora Project would be reasonable
as pointed out before. It doesn't fill in the blanks, but is closer
to the truth.
Oldest public mention I could find of the original script is this message
where Seth Vidal attached it after he had added one sys.arg:
http://www.redhat.com/archives/fedora-maintainers/2005-February/msg00042....
No author information, no licence, no copyright.
12:51 p.m.
New subject: rpms/blam/F-9 blam.spec,1.24,1.25
On Thursday 25 September 2008, Michael Schwendt wrote:
On Thu, 25 Sep 2008 20:11:55 +0300, Ville Skyttä wrote:
> On Thursday 25 September 2008, Michael Schwendt wrote:
> > However, the added line changes history.
>
> Would you mind just telling me how to fix it so we can move on?
Replacing it with (c) 2005-2008 Fedora Project would be reasonable
as pointed out before. It doesn't fill in the blanks, but is closer
to the truth.
Ok, done in git, thanks.
5712
days inactive
5713
days old
11 comments
5 participants
participants (5)
-
Alex Lancaster -
Christopher Aillon -
Michael Schwendt -
Thorsten Leemhuis -
Ville Skyttä