[Bash-completion-devel] Running tests against multiple bash versions.

[Crestez Dan Leonard]

I tried to run bash-completion tests against multiple versions of bash.
I did this for every version I found a tarball for on the official ftp:
3.0, 3.0.16, 3.1, 3.2, 3.2.48, 4.0 and 4.1. Bash-completion version is
latest HEAD.

Here are the results I got:

bash-4.1:
# of expected passes            491
# of unexpected failures        1
# of expected failures          1
# of unresolved testcases       1
# of unsupported tests          101

Bash 4.0 segfaults very quickly; for example on "ls <TAB>". This is
probably the cause of the bug:
http://mirrors.aixtools.net/gnu/bash/bash-4.0-patches/bash40-002 . My
guess is that all distros that ship bash 4.0 also ship that patch. It
would be worthwhile looking for a 4.0.* release.

bash-3.2.48:
# of expected passes            490
# of unexpected failures        1
# of expected failures          1
# of unresolved testcases       2
# of unsupported tests          101

bash-3.2:
# of expected passes            489
# of unexpected failures        7
# of expected failures          1
# of unresolved testcases       2
# of unsupported tests          101

bash-3.1:
# of expected passes            436
# of unexpected failures        62
# of expected failures          1
# of unresolved testcases       4
# of unsupported tests          101

Bash 3.0 and 3.0.16 time out on all completion tests. This is probably
because of the printf -v bug:
https://alioth.debian.org/tracker/?func=detail&atid=413095&aid=312183&group_id=100114

Did anyone else attempt this; perhaps with better results?