Message94105
Antoine asked me to give my opinion on the non-blocking SSL API. I can
say that the current behavior of SSLSocket.recv is certainly not very
good and probably makes SSLSocket useless for any non-blocking
application. Letting the underlying want-read/want-write error bubble
up to the caller definitely makes more sense than blocking.
I haven't had a chance to look into the new SSL code in depth, so don't
construe this as a thorough review of the change, though. |
|
Date |
User |
Action |
Args |
2009-10-15 18:49:25 | exarkun | set | recipients:
+ exarkun, barry, rhettinger, josiahcarlson, amaury.forgeotdarc, janssen, giampaolo.rodola, josiah.carlson, ddvoinikov, srid, qwavel |
2009-10-15 18:49:25 | exarkun | set | messageid: <1255632565.31.0.660937057079.issue3890@psf.upfronthosting.co.za> |
2009-10-15 18:49:24 | exarkun | link | issue3890 messages |
2009-10-15 18:49:23 | exarkun | create | |
|