Should selftest success depend on verbosity?

[Andrew Bennetts]

Maritza Mendez wrote:
[...]
>    Observation #3
>    What's even more remarkable is that the test_wrong_pass which stalled the
>    non-verbose run actually passes in verbose mode:
> 
>    ���
>    test_http.TestAuth.test_wrong_pass(pycurl,HTTP/1.1,basicdigest)�����������
>    OK����������������� 49ms
> 
>    Each of these observations was 100% reproducible in ten consecutive
>    trials.
> 
>    I suppose this could be explained by test order (shouldn't order by the
>    same for the commands quoted above?) or by sensitivity to timing
>    differences between verbose and non-verbose modes.� But I have no evidence
>    to support this pure speculation.

My wild guess would be timing sensitivity, but the complete determinism (100%
reproducible in 10 attempts) suggests not.

Hmm, perhaps it's a garbage collection issue?  Turning on -v will have the
effect of producing slightly different object allocation/deallocation counts,
which will cause the periodic cyclic garbage collector to kick in at slightly
different points... it's not the first time I've seen an issue like that...

I don't have time to do this right now, but if I were investigating this I'd be
experimenting with adding "import gc; gc.disable()" early in the 'bzr' script
and see how that affects results.

(It might be useful to add a --no-python-gc global option to do this?)

-Andrew.