Re: [pcp] pmrep: extend QA

[Nathan Scott <nathans@xxxxxxxxxx>]

----- Original Message -----
> [...]
> The second one is that, as you perhaps noticed,

Yep, these test output differences were what triggered looking into -z.

> your 1069.out update
> changed the timestamps for the test case, you see 20:36:45 but I see
> 13:36:45 here so the test is now failing for me.

Yeah thats definitely not right then - the idea was to have determinism
from -z use -- I think the python tzset from your patch is probably the
subtle piece I missed, keeping the PCP and python date tz in sync.

> I remember I used surprisingly lot of time with write_ext_header() when
> I originally wrote it but seems that it's still not working optimally.
> 
> What do you think of the patch below?

Looks good and passing for me also now - thanks Marko.

cheers.

--
Nathan