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

Freddy Vulto fvulto at gmail.com
Sat Jan 16 22:10:54 UTC 2010


On 100116 12:29, Crestez Dan Leonard wrote:
> Did anyone else attempt this; perhaps with better results?

I'm running tests nightly, against git head, on Debian (5.0.3).  There's one
unexpected failure concerning rpm, which Ville is working on.  These are the
results of last night:


bash-3.2.39(1)-release
-----------------------------------
# of expected passes            360
# of unexpected failures        1
# of expected failures          1
# of unsupported tests          166


bash-4.0.33(1)-release
-----------------------------------
# of expected passes            359
# of unexpected failures        1
# of expected failures          2
# of unsupported tests          166


bash-4.0.35(2)-release
-----------------------------------
# of expected passes            360
# of unexpected failures        1
# of expected failures          1
# of unsupported tests          166


As said in Alioth #312183, I'm reluctant to run/fix bash-completion-2-tests for
bash < 3.2 if the vast majority of users is using bash >= 3.2 and bash < 3.2
users can rely on bash_completion-1?  There're lots of bugs/improvements for
bash >= 3.2 we'd better pay attention to?


Regards,

Freddy Vulto
http://fvue.nl



More information about the Bash-completion-devel mailing list