We’re trying to add a feature to snapcraft for a customer. Part of that was a catkin bugfix that has since landed. However, we can’t finish the feature until it’s actually contained in a catkin release and sync’d. @dirk-thomas would you mind spinning a new release of catkin so we can get it in kinetic and melodic?
I will do a new release once the pending PRs around
gmock have been reviewed and merged.
Thanks @dirk-thomas. Respectfully, those PRs have been up for a while. The one that has been reviewed has been blocked on you for a month, and the other hasn’t moved for longer than that. Any ETA for when you think you’ll be able to get to them? I understand you’re busy, of course, I just want to get an idea for timeline so we can organize our own work.
The point is not my level of busyness since making a release takes only a few minutes. A
catkin release has two aspects which make it fairly expensive:
- all downstream packages have to be rebuild (Jenkins will take care of us and the resources aren’t what I am concerned about even though they are significant)
- every user will have to download and install the new Debian packages on the next
apt dist-upgradewhich is a lot of bandwidth for a lot of users.
Therefore we try to minimize releases of low level packages like this as much as we can.
The pending PRs are relevant to support Debian Buster which was recently released as well as Ubuntu 20.04 which is targeted by ROS 1 Noetic. Since we are already working on Noetic / Python 3 support this should land in the near future (maybe a couple of weeks). Making two
catkin releases shortly after one another is simply not desired.
Especially the changes for
gmock can be pretty difficult to review since there are so many ways it can be used / fail that it is almost impossible to test them all.
That makes perfect sense. I’ll put this on our backlog for a few weeks out, then. Thanks for the information!