Lessons Learned from GSoC 2017

• Umesh and Zero presented their projects and what went well and not so well during GSoC 2017.
• We should use some sort of issue tracking to make progress more visible and measurable. A good idea might be Trello or GitHub projects
  • Some projects do a lot of planning beforehand and use a burndown chart to track student progress. We don't think this will work very well for us, since our projects often change after they start.
• Mojca and Clemens attended the Mentor Summit
  • We asked for a period after the initial submission deadline where students can improve their proposals but no new proposals can be added. There used to be a button to enable editing in the application review period that could be used to achieve this. Let's see whether this will be the case this year, and if it is, use it.
  • Clemens sent a mail to the list with a report: ​https://lists.macports.org/pipermail/macports-dev/2017-October/036720.html

Hacking session ideas

• Umesh had problems with selfupdate not working on college network behind proxy and it doesn't even tell you that
  • selfupdate needs to be able to fetch from tarball, not just rsync (Rainer started working on that a while ago, but it wasn't finished yet) #38265
• Intercept the hook from Travis telling you about the failure and write notification to the pull request
• figure out what goes wrong with maintainer timeout label
• portindex should return a non-zero value when it fails and the code in Travis CI should be simplified (not to use the workaround)
• Is there any way to get rid of dependency on getopt?