Re: qt_gui_core

47 views
Skip to first unread message

Isaac Saito

unread,
Apr 2, 2013, 1:20:36 PM4/2/13
to Dirk Thomas, ros-s...@googlegroups.com
Dirk,

# I'm moving the discussion to rqt ML.

Coming right to the point first, I have to admit that I didn't test
enough the latest patch [0] that was commited 3 days before my latest
release, and understand the issue caused by that [1] is critical for
certain users. While we try to fix this, should we rollback to the
status before [0] and release? What's the best solution here?

I apologize first, and I'll be more cautious for the future release.

[0] https://github.com/ros-visualization/qt_gui_core/commit/295a040845dd2290de9c22e80bd41d7c53f413db
[1] https://github.com/yujinrobot/kobuki_desktop/issues/10

Inline from here,

On Tue, Apr 2, 2013 at 8:00 AM, Dirk Thomas <dth...@osrfoundation.org> wrote:
>
> Why did you released the packages twice - between both versions was not a single change?

I did release twice almost by accident. I don't exactly remember why I
had to do so, but it was because I had a problem in releasing due to
my familliarity with bloom at that time.


> Also was the release really necessary?
> In this case the latest changes are responsible for a regression.
> It definitely needs more testing after patches are being applied.
> Due to the very short time between the changes from Dorian and the release
> the issue was not found before.
> But now the problem is released and also synced to Groovy and Hydro.

Yes the release was necessary to apply a patch that was commited
almost 2 months ago. But again I should have tested the all commits
applied since last release.

--
Isaac

Isaac Saito

unread,
Apr 2, 2013, 2:55:24 PM4/2/13
to Dirk Thomas, ros-s...@googlegroups.com
(Update) William & I decided as a quick workaround that we rollback
qt_gui_core public debian to the 0.2.12 where all commits that are not
well tested aren't included (i.e. all commits after March 21
http://goo.gl/eRCWa). Thank you William & possibly Tully who are now
handling this.

Meanwhile we need for sure a permanent solution to the issue raised.

Isaac

William Woodall

unread,
Apr 2, 2013, 3:05:29 PM4/2/13
to ros-s...@googlegroups.com, Dirk Thomas
To be clear, because it is already in public I reverted the changes in question and release it as 0.2.15. (this is required as apt-get will not auto downgrade). Once this issue is solved those commits can be added back, fixed, and then release as 0.2.16.

--

--
--
To unsubscribe from this group, send email to ros-sig-rqt...@googlegroups.com
Google group at http://groups.google.com/group/ros-sig-rqt?hl=en
rqt: http://ros.org/wiki/rqt
---
You received this message because you are subscribed to the Google Groups "SIG for rqt (aka ROS GUI) developers and users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ros-sig-rqt...@googlegroups.com.
Visit this group at http://groups.google.com/group/ros-sig-rqt?hl=en.





--
William Woodall
ROS Development Team
Reply all
Reply to author
Forward
0 new messages