bug#3953: 23.0.94; GPG interaction fails

reopen 2412
merge 2412 3953
thanks

>>>>> In <20090727194515.F1958E42E9@xxxxxxxxxxxxxxxxxxx> 
>>>>>   Torsten Bronger <bronger@xxxxxxxxxxxxxxxxxxxxx> wrote:
> >>> -    (accept-process-output (epg-context-process context) 1)))
> >>> +    (accept-process-output (epg-context-process context) 1))
> >>> +  (sleep-for 0.1))
> >>
> >> That seems to do the trick.  At least, I had no problem in ~15 attempts
> >> to preview/send signed messages, whereas without your patch the "Sign
> >> failed" error would happen about every second time.
> >
> > Thanks for testing.  I've checked in the patch.

> I experience the same problem but the above patch doesn't really
> help me.  Without the sleeping time, I get failures with signature
> checking of incoming emails almost always; with the sleeping time,
> it works at least in 20% of the cases.  I increased the time to even
> 10 seconds but the success rate didn't increase further.  The
> behaviour remains probabilistic.

Thanks for further testing.  Well, in that case an essential part of
process-filter might not be executed in 20% of the cases.

Does replacing (sleep-for 0.1) with:

(epg--process-filter (epg-context-process context) "")

change the situation?

Regards,
-- 
Daiki Ueno