On Sun 2019-05-05 19:44:01 +0300, Tomi Ollila wrote:
>> I actually think that CI and autobuilders *should* exercise the parallel
>> tests, as annoying as that might be initially, because it seems likely
>> to catch any other potential entanglements.
>
> ... after it has been proven a bit more to work...

i tend to think the other way around -- the way to prove it is to push
it into the CI pipelines on the development branch, rather than trying
to prove it on individual developer systems.

> One thing more, the "perverse" pre-caching done in add_email_corpus
> can be retained -- just do that step *before* going to parallelism...

You could do that, but i'm not sure how much you'd gain from it -- its
current implementation is all mixed up with the parallelized tests,
anyway.  If you wanted to propose a concrete re-optimization like this,
i'd be happy to review it, test it and report back on the speedup
gained.

thanks for thinking this through, Tomi!

     --dkg