libqalculate soname change

2020-03-07 Thread Mukundan Ragavan
libqalculate soname bump is happening with v3.8.0.

The following packages are affected -

- qalculate-gtk
- plasma-workspace
- step
- cantor

I will rebuild all these packages on rawhide.

qalculate-kde will also affected but this is already FTBFS. So, I won't
touch this.

Mukundan.



signature.asc
Description: OpenPGP digital signature
___
devel mailing list -- devel@lists.fedoraproject.org
To unsubscribe send an email to devel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org


libqalculate soname change

2019-11-29 Thread Mukundan Ragavan
libqalculate soname bump is happening with v3.6.0.

The following packages are affected -

plasma-workspace
step
cantor
qalculate-kde

I will rebuild all the packages on rawhide.

Mukundan.



signature.asc
Description: OpenPGP digital signature
___
devel mailing list -- devel@lists.fedoraproject.org
To unsubscribe send an email to devel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org


Re: libqalculate soname change

2019-08-26 Thread Adam Williamson
On Mon, 2019-08-26 at 17:47 -0400, Mukundan Ragavan wrote:
> libqalculate soname bump is happening with v3.3 (libqalculate.so.21.2.0).
> 
> The following packages are affected -
> 
> plasma-workspace
> step
> cantor
> qalculate-kde
> 
> I will rebuild the packages on both rawhide and F31 tomorrow.
> Maintainers, if you do not want me to touch F31, please let me know.

It'd be great if you try and time this so all the builds are done
between composes - if the soname bump gets into a compose attempt but
the plasma-workspace rebuild does not, it will likely fail. thanks!
-- 
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . net
http://www.happyassassin.net
___
devel mailing list -- devel@lists.fedoraproject.org
To unsubscribe send an email to devel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org


libqalculate soname change

2019-08-26 Thread Mukundan Ragavan
libqalculate soname bump is happening with v3.3 (libqalculate.so.21.2.0).

The following packages are affected -

plasma-workspace
step
cantor
qalculate-kde

I will rebuild the packages on both rawhide and F31 tomorrow.
Maintainers, if you do not want me to touch F31, please let me know.

Mukundan.



signature.asc
Description: OpenPGP digital signature
___
devel mailing list -- devel@lists.fedoraproject.org
To unsubscribe send an email to devel-le...@lists.fedoraproject.org
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org


libqalculate soname change

2019-04-17 Thread Mukundan Ragavan
libqalculate soname bump is happening with update to v3.1. The following
packages are affected -

plasma-workspace
step
cantor
qalculate-kde

I will rebuild these in the coming days.Note that qalculate-kde is FTBFS
right now.



signature.asc
Description: OpenPGP digital signature
___
devel mailing list -- devel@lists.fedoraproject.org
To unsubscribe send an email to devel-le...@lists.fedoraproject.org
Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org


libqalculate - soname change

2019-02-23 Thread Mukundan Ragavan
libqalculate soname bump is happening with v2.9. The following
packages are affected -

plasma-workspace
step
cantor
qalculate-kde

I will rebuild these packages and file bug reports if needed.

Mukundan.



signature.asc
Description: OpenPGP digital signature
___
devel mailing list -- devel@lists.fedoraproject.org
To unsubscribe send an email to devel-le...@lists.fedoraproject.org
Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org


Re: libqalculate soname change

2018-04-11 Thread Mukundan Ragavan
Since 2.4.0 is out, soname change will be to libqalculate.so.16. I will submit 
pull requests for the affected packages. I might miss F28.
___
devel mailing list -- devel@lists.fedoraproject.org
To unsubscribe send an email to devel-le...@lists.fedoraproject.org


Re: libqalculate soname change

2018-04-05 Thread Mukundan Ragavan
I will be updating libqalculate to the latest upstream release (v2.3.0 instead 
of 2.2.1 as per my previous email). This involves a soname change. 

The following packages are affected - 

plasma-workspace
step
cantor
qalculate-kde

I will build qalculate and the affected packages in rawhide and Fedora 28 in 
the coming days (most likely during the weekend well before the final freeze).


Mukundan.
___
devel mailing list -- devel@lists.fedoraproject.org
To unsubscribe send an email to devel-le...@lists.fedoraproject.org


Re: libqalculate soname change

2018-03-08 Thread Kevin Kofler
Adam Williamson wrote:
> This cycle was actually kind of an interesting case, because on the one
> hand, there's a pretty good argument that we were in a very bad state
> at freeze time this cycle. On the *other* hand, one of the main reasons
> we were in a bad state is because people kept frickin' landing de-
> stabilizing changes while we were still dealing with the fallout from
> the previous de-stabilizing change: we've had GCC 8, various
> unannounced soname bumps, basically all of nu-Modularity showing up in
> the middle there somewhere, anaconda modularization, new versions of
> pungi and pykickstart and systemd, and that's just the things I
> *remember*. It's been something of a wild ride.

The reason why there are so many major changes in F28 is pretty clear: the 
F27 cycle was too short! Hardly any major change made it in there. So we now 
get double the changes.

Kevin Kofler
___
devel mailing list -- devel@lists.fedoraproject.org
To unsubscribe send an email to devel-le...@lists.fedoraproject.org


Re: libqalculate soname change

2018-03-08 Thread Mukundan Ragavan
> On Wed, 2018-03-07 at 20:46 -0500, Mukundan Ragavan wrote:
> 
> Thanks for the notification.
> 
> Note F28 is currently in freeze for Beta; this is a long freeze, and
> honestly, might get longer, given the current state of things. So I'm
> not sure whether it'd be optimal to have this committed and built for
> F28, but not able to go stable due to the freeze. Might be better to
> wait till after Beta, *or* get a freeze exception. Is there any
> significant benefit for us in the new libqalculate, which would justify
> a freeze exception?

I was worried about pushing this during freeze as well. While I do want to get 
this pushed out to F28 (this and previous two releases that never made to 
Fedora have bug fixes and new features), I can certainly do so after beta is 
out and freeze is lifted. 

I will build only for rawhide on Friday and leave F28 updates for after the 
freeze is lifted.

Thanks for replying.
___
devel mailing list -- devel@lists.fedoraproject.org
To unsubscribe send an email to devel-le...@lists.fedoraproject.org


Re: libqalculate soname change

2018-03-08 Thread Adam Williamson
On Thu, 2018-03-08 at 10:15 -0500, Przemek Klosowski wrote:
> 
> Having said that, I could see there might be a case for delaying the 
> freeze before it starts, if the overall schedule is delayed for 
> well-understood reasons.

This cycle was actually kind of an interesting case, because on the one
hand, there's a pretty good argument that we were in a very bad state
at freeze time this cycle. On the *other* hand, one of the main reasons
we were in a bad state is because people kept frickin' landing de-
stabilizing changes while we were still dealing with the fallout from
the previous de-stabilizing change: we've had GCC 8, various
unannounced soname bumps, basically all of nu-Modularity showing up in
the middle there somewhere, anaconda modularization, new versions of
pungi and pykickstart and systemd, and that's just the things I
*remember*. It's been something of a wild ride.

Freezing at least stops any *more* wacky new crap landing while we're
still trying to sort out the remains of this last batch.
-- 
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . net
http://www.happyassassin.net
___
devel mailing list -- devel@lists.fedoraproject.org
To unsubscribe send an email to devel-le...@lists.fedoraproject.org


Re: libqalculate soname change

2018-03-08 Thread Adam Williamson
On Thu, 2018-03-08 at 15:51 +0100, Kevin Kofler wrote:
> Adam Williamson wrote:
> > Note F28 is currently in freeze for Beta; this is a long freeze, and
> > honestly, might get longer, given the current state of things.
> 
> IMHO, if we know that we are not in a good enough state to release, we 
> should delay the start of the freeze (in the current case where it has 
> already started: cancel the freeze and try again later) until we are in a 
> releasable state. It just does not make sense to extend the freezes forever. 
> We should only freeze if we are actually ready to prepare a release (e.g. 
> F28 Beta).

This is actually being kicked around informally ATM, but I don't know
where we're going to end up yet.

> And there really needs to be a maximum freeze period. If we are still not 
> ready to go gold by then, we need to take all the pending packages and 
> restart the process from there.

In most cases this is a *really bad* idea as it will almost inevitably
mean the freeze (once restored) gets longer again, as it's almost
certainly going to be the case that some of the changes blocked by the
freeze cause new issues. That's *why we have the freeze*, after all.
-- 
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . net
http://www.happyassassin.net
___
devel mailing list -- devel@lists.fedoraproject.org
To unsubscribe send an email to devel-le...@lists.fedoraproject.org


Re: libqalculate soname change

2018-03-08 Thread Randy Barlow
On 03/08/2018 09:51 AM, Kevin Kofler wrote:
> IMHO, if we know that we are not in a good enough state to release, we 
> should delay the start of the freeze (in the current case where it has 
> already started: cancel the freeze and try again later) until we are in a 
> releasable state. It just does not make sense to extend the freezes forever. 
> We should only freeze if we are actually ready to prepare a release (e.g. 
> F28 Beta).

This would be counter to the goal of the freeze - we freeze so we can
stabilize things and get the release out the door. If we unfreeze, we
only decrease the stability.



signature.asc
Description: OpenPGP digital signature
___
devel mailing list -- devel@lists.fedoraproject.org
To unsubscribe send an email to devel-le...@lists.fedoraproject.org


Re: libqalculate soname change

2018-03-08 Thread Przemek Klosowski

On 03/08/2018 09:51 AM, Kevin Kofler wrote:

Adam Williamson wrote:

Note F28 is currently in freeze for Beta; this is a long freeze, and
honestly, might get longer, given the current state of things.

IMHO, if we know that we are not in a good enough state to release, we
should delay the start of the freeze (in the current case where it has
already started: cancel the freeze and try again later) until we are in a
releasable state. It just does not make sense to extend the freezes forever.
We should only freeze if we are actually ready to prepare a release (e.g.
F28 Beta).


You are speaking from the point of view of scheduling, but the reason 
for the freeze is to stabilize things during the release process.


Clearly it is not acceptable to extend the freezes forever if the 
release process is not going well, but the solution is to fix the 
problems which delay the release, not compound them by unfreezing in the 
middle of the process.


Having said that, I could see there might be a case for delaying the 
freeze before it starts, if the overall schedule is delayed for 
well-understood reasons.


___
devel mailing list -- devel@lists.fedoraproject.org
To unsubscribe send an email to devel-le...@lists.fedoraproject.org


Re: libqalculate soname change

2018-03-08 Thread Kevin Kofler
Adam Williamson wrote:
> Note F28 is currently in freeze for Beta; this is a long freeze, and
> honestly, might get longer, given the current state of things.

IMHO, if we know that we are not in a good enough state to release, we 
should delay the start of the freeze (in the current case where it has 
already started: cancel the freeze and try again later) until we are in a 
releasable state. It just does not make sense to extend the freezes forever. 
We should only freeze if we are actually ready to prepare a release (e.g. 
F28 Beta).

And there really needs to be a maximum freeze period. If we are still not 
ready to go gold by then, we need to take all the pending packages and 
restart the process from there.

It is just not workable to have everything shut down (including updates to 
stable releases if we care about upgrade paths!) for 3+ weeks at a time.

Kevin Kofler
___
devel mailing list -- devel@lists.fedoraproject.org
To unsubscribe send an email to devel-le...@lists.fedoraproject.org


Re: libqalculate soname change

2018-03-07 Thread Adam Williamson
On Wed, 2018-03-07 at 20:46 -0500, Mukundan Ragavan wrote:
> I will be updating libqalculate to the latest upstream release (v2.2.1).
> This involves a soname change. The following packages are affected.
> 
> plasma-workspace
> step
> cantor
> qalculate-kde
> 
> I have built all these packages for rawhide and F28 in a COPR [0]. Since
> we have the same versions on both rawhide and F28, I hope to push
> libqalculate to both and rebuild all these packages.
> 
> If building for F28 is not preferred, please let me know. I anticipate
> building everything by Friday (2018-03-08) and will proceed unless I
> hear otherwise.

Thanks for the notification.

Note F28 is currently in freeze for Beta; this is a long freeze, and
honestly, might get longer, given the current state of things. So I'm
not sure whether it'd be optimal to have this committed and built for
F28, but not able to go stable due to the freeze. Might be better to
wait till after Beta, *or* get a freeze exception. Is there any
significant benefit for us in the new libqalculate, which would justify
a freeze exception?
-- 
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | identi.ca: adamwfedora
http://www.happyassassin.net
___
devel mailing list -- devel@lists.fedoraproject.org
To unsubscribe send an email to devel-le...@lists.fedoraproject.org


libqalculate soname change

2018-03-07 Thread Mukundan Ragavan

I will be updating libqalculate to the latest upstream release (v2.2.1).
This involves a soname change. The following packages are affected.

plasma-workspace
step
cantor
qalculate-kde

I have built all these packages for rawhide and F28 in a COPR [0]. Since
we have the same versions on both rawhide and F28, I hope to push
libqalculate to both and rebuild all these packages.

If building for F28 is not preferred, please let me know. I anticipate
building everything by Friday (2018-03-08) and will proceed unless I
hear otherwise.

Mukundan.


[0] https://copr.fedorainfracloud.org/coprs/nonamedotc/libqalculate/builds/



signature.asc
Description: OpenPGP digital signature
___
devel mailing list -- devel@lists.fedoraproject.org
To unsubscribe send an email to devel-le...@lists.fedoraproject.org