[Reproducible-commits] [blog] 01/01: Add some clarifications and fix formatting

Ceridwen ceridwen-guest at moszumanska.debian.org
Tue Jun 14 15:04:00 UTC 2016


This is an automated email from the git hooks/post-receive script.

ceridwen-guest pushed a commit to branch master
in repository blog.

commit f88fecf1408dbfcc9856470de8a1cbd873714454
Author: Ceridwen <ceridwenv at gmail.com>
Date:   Tue Jun 14 11:03:44 2016 -0400

    Add some clarifications and fix formatting
---
 drafts/people/ceridwen/reprotest_week3.mdwn | 23 +++++++++++++----------
 1 file changed, 13 insertions(+), 10 deletions(-)

diff --git a/drafts/people/ceridwen/reprotest_week3.mdwn b/drafts/people/ceridwen/reprotest_week3.mdwn
index 3096caa..0ddc097 100644
--- a/drafts/people/ceridwen/reprotest_week3.mdwn
+++ b/drafts/people/ceridwen/reprotest_week3.mdwn
@@ -4,12 +4,13 @@
 Author: ceridwen
 
 The first, very-alpha release of reprotest is now out at TODO.  While
-it only builds on an existing system, it can now check its own
-reproducibility, which it does in its own tests, both using setuptools
-and debuild.  Unfortunately, setuptools seems to generate
-file-order-dependent binaries, meaning setuptools bdist creates
-unreproducible binaries.  With debuild, reprotest probably would be
-reproducible with the [modified
+it only builds on an existing system (as I'm still working on support
+for virtualization), it can now check its own reproducibility, which
+it does in its own tests, both using setuptools and debuild.
+Unfortunately, setuptools seems to generate file-order-dependent
+binaries, meaning `python setup.py bdist` creates unreproducible
+binaries.  With debuild, reprotest probably would be reproducible with
+the [modified
 packages](https://wiki.debian.org/ReproducibleBuilds/ExperimentalToolchain#Modified_packages)
 from the Reproducible Builds project, though I haven't tested that
 yet.  It tests 'captures_environment', 'fileordering' (renamed from
@@ -21,8 +22,10 @@ system, so they will only be enabled in the containers.
 The next major part of the project is integrating autopkgtests's
 container management system into reprotest.  For the curious,
 autopkgtest is composed of a main program, `adt-run`, which then calls
-other command-line programs, `adt-virt-chroot, `adt-virt-lxd`,
+other command-line programs, `adt-virt-chroot`, `adt-virt-lxd`,
 `adt-virt-schroot`, `adt-virt-null`, `adt-virt-schroot`, and
-`adt-virt-qemu`, that communicate with the containers.  I think this
-is a bit of an odd design but it works well for my purposes since the
-container programs already have existing CLIs that reprotest can use.
+`adt-virt-qemu`, that communicate with the containers.  (The
+autopkgtest maintainer has since renamed the programs, but the
+underlying structure remains the same.)  I think this is a bit of an
+odd design but it works well for my purposes since the container
+programs already have existing CLIs that reprotest can use.

-- 
Alioth's /usr/local/bin/git-commit-notice on /srv/git.debian.org/git/reproducible/blog.git



More information about the Reproducible-commits mailing list