This site is now 100% read-only, and retired.

XML logo

Challenge: one reproducible package a week
Posted by dkg on Thu 4 Jun 2015 at 02:31
I encourage anyone interested in debian development to get involved with the Reproducible Builds project. My own project is to try to diagnose (and hopefully provide patches for) two unreproducible packages a week. Maybe you can do one package a week?

Reproducible Builds is another example of the kind of audacious project that I celebrated in my last blog post.

It's a fun way to learn a little bit about different parts of the archive, and to help on an incrementally-improving process. My workflow below is meant to encourage folks to join in. The documentation on the wiki is certainly more authoritative (and will be more up-to-date in the future).

My usual workflow is this:

  • Visit the list of unreproducible packages "with no notes" -- these are packages that are known to not build reproducibly, but no one else has diagnosed.
  • Click on a package basically at random (but if you notice one that you are familiar with in that list, go ahead and pick it!)
  • That will take you to a page showing the debbindiff (difference between binary packages) of the two builds. Sometimes, this shows an obvious difference, like the difference for cloc, which shows that the man page embeds the build time.
    • If you have a guess about what's wrong, but don't feel sure, pop over to the #debian-reproducible IRC channel on -- usually there are friendly people there who will discuss it with you.
    • If you find a debbindiff that is completely indecipherable, go back and pick another package
  • Fetch the package source. I like to use debcheckout from the devscripts package, so that i can work with the maintainer's revision control system of choice. With cloc above, i'd do:
    debcheckout cloc
    If that failed for some reason, I would use:
    apt-get source cloc
  • Change into the source directory you just unpacked and use your preferred tools (i like good old grep and find) to figure out where the affected file is created, and how it derives its variance. You might want to look through the list of known causes of unreproducibility to see if there are any similar suggestions there.
  • If you can fix the variance, make a patch!
  • File a bug report with your diagnosis (and with your patch if you have one). I write my bug reports as a simple e-mail, with the subject line describing what i found, and with a header referring it to the r-b project, like this:
    Source: packagename
    Version: packageversion
    Tags: patch
    Severity: wishlist
    Usertags: closest-usertag
    When choosing the closest usertag, i look at the "Usertagged bugs" block on the R-B continuous integration dashboard or the wiki documentation to see what makes sense.
  • If you find that this is an example of one of the known issues, you can also note it in the notes.git repository, or just mailing a description of what you found to the reproducible-builds mailing list.

The information at the R-B wiki page is quite detailed if you want more info.

Finally, many many thanks to all of the people involved in the project, and particularly to h01ger and Lunar^, who have both contributed a ton of work, and have also made it easy to plug in and help as a less-involved contributor. The nice automatically-updated statistics provided by the team's continuous integration work makes it a fun game to help out.