Message226237
> We can leave these undocumented at the Python level if you prefer.
I'd rather that indeed. If there's a specific need, we can expose them as a separate issue.
> Maybe just "SSLInstance", would that be better than "SSLObject"?
That doesn't sound much better :-) Ok, let's keep SSLObject then.
> I believe that the overall _ssl/ssl code could be simplified by: [snip]
That would be nice. Would that also handle e.g. socket timeouts?
> To use SSLObject as a mixin it would have to be aware of these two uses of its subclasses. It could be done but I don't think it's 100% clean either.
Fair enough. We just have to make sure to implement and test new APIs twice (e.g the version() method in issue20421). |
|
Date |
User |
Action |
Args |
2014-09-01 17:43:41 | pitrou | set | recipients:
+ pitrou, gvanrossum, geertj, janssen, vstinner, giampaolo.rodola, christian.heimes, ezio.melotti, alex, sbt, Ben.Darnell, yselivanov, dstufft |
2014-09-01 17:43:41 | pitrou | set | messageid: <1409593421.14.0.103928574374.issue21965@psf.upfronthosting.co.za> |
2014-09-01 17:43:41 | pitrou | link | issue21965 messages |
2014-09-01 17:43:40 | pitrou | create | |
|