xmlrpc improvements

Hello there. I've been trying to, in the last weeks, to pass on to the trunk the improvements I've made to XMLRPC. I've created several issues in order to do these changes incrementally but have got no comments. Perhaps it is best to show the whole thing in context, then. I´ve gathered all the changes in a single patch, here: http://codereview.appspot.com/73041/show I've also created a corresponding issue in the issue tracker: http://bugs.python.org/issue6267 I'd really like to get this stuff in. The performance gains allowing http1.1 and gzip for xmlrpc are quite significant. Also, there are bugfixes in there. Cheers, Kristján

I‘d really like to get this stuff in. The performance gains allowing http1.1 and gzip for xmlrpc are quite significant.
I think you really need to get Fredrik Lundh to comment on it. If he can't predict when he'll be able to review the changes, maybe he can accept releasing control of xmlrpclib. Regards, Martin

Fredrik Lundh wrote:
http://bugs.python.org/issue6267 While I have your attention, please also review http://bugs.python.org/issue6233 Regards, Martin

On Sat, Jun 20, 2009 at 6:57 PM, "Martin v. Löwis"<martin@v.loewis.de> wrote:
While I have your attention, please also review
I'm pretty sure that fix is the wrong fix - afaik, _encode is used to encode tag/attribute names, and charrefs don't work in that context. </F>

I‘d really like to get this stuff in. The performance gains allowing http1.1 and gzip for xmlrpc are quite significant.
I think you really need to get Fredrik Lundh to comment on it. If he can't predict when he'll be able to review the changes, maybe he can accept releasing control of xmlrpclib. Regards, Martin

Fredrik Lundh wrote:
http://bugs.python.org/issue6267 While I have your attention, please also review http://bugs.python.org/issue6233 Regards, Martin

On Sat, Jun 20, 2009 at 6:57 PM, "Martin v. Löwis"<martin@v.loewis.de> wrote:
While I have your attention, please also review
I'm pretty sure that fix is the wrong fix - afaik, _encode is used to encode tag/attribute names, and charrefs don't work in that context. </F>
participants (4)
-
"Martin v. Löwis"
-
Fredrik Lundh
-
Kristján Valur Jónsson
-
Terry Reedy