https://sourceware.org/bugzilla/show_bug.cgi?id=34566

--- Comment #7 from Mark Wielaard <mark at klomp dot org> ---
(In reply to Frank Ch. Eigler from comment #5)
> (In reply to Mark Wielaard from comment #4)
> > Might this be the same as bug #34158
> 
> Nope.  In this case, libcurl can make https connection to their server, and
> then sends the request away.  But there's no response.

aha, sorry. Then this patch:

(In reply to Frank Ch. Eigler from comment #6)
> Created attachment 16974 [details]
> proposed patch
> 
> Hand-tested with dead server as well as mixture of live/dead via gdb, seems
> tricky to automate.

Does make sense, just always call the progressfn in the big do-while loop so
the user can interrupt. Looks good to me.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

Reply via email to