Tue, 18 Mar 2014 12:59:49 -0700 evolve: fix evolve with mercurial default
Matt Mackall <mpm@selenic.com> [Tue, 18 Mar 2014 12:59:49 -0700] rev 886
evolve: fix evolve with mercurial default MODERNCAPS have been renamed to moderncaps upstream.
Wed, 12 Mar 2014 13:07:37 -0700 simple4server: update tests
Pierre-Yves David <pierre-yves.david@fb.com> [Wed, 12 Mar 2014 13:07:37 -0700] rev 885
simple4server: update tests Forgot to amend before pushing.
Wed, 12 Mar 2014 13:03:35 -0700 sinmple4server: advertise new cap only when advertising obsolescence
Pierre-Yves David <pierre-yves.david@fb.com> [Wed, 12 Mar 2014 13:03:35 -0700] rev 884
sinmple4server: advertise new cap only when advertising obsolescence This is more consistent.
Wed, 12 Mar 2014 13:01:32 -0700 evolve: do not exchange obsolescence markers when server refuse it
Pierre-Yves David <pierre-yves.david@fb.com> [Wed, 12 Mar 2014 13:01:32 -0700] rev 883
evolve: do not exchange obsolescence markers when server refuse it The server can be configured to refuse obsolescence markers exchange. See changeset 48804a8d7472 for details about it. We respect server will in the client.
Wed, 12 Mar 2014 12:44:33 -0700 simple4server: option to advertise obsolescence support
Pierre-Yves David <pierre-yves.david@fb.com> [Wed, 12 Mar 2014 12:44:33 -0700] rev 882
simple4server: option to advertise obsolescence support We had an option to control the advertising of the obsolescence markers support from the server. The goal of this option is to allow large scale repo hosting service to control the use of obsolescence markers exchange.
Mon, 10 Mar 2014 19:47:24 -0700 export: add test for simple4server
Pierre-Yves David <pierre-yves.david@fb.com> [Mon, 10 Mar 2014 19:47:24 -0700] rev 881
export: add test for simple4server We now officially test it over http
Mon, 10 Mar 2014 21:57:41 -0700 exchange: fix push markers over wire protocol for http peer
Pierre-Yves David <pierre-yves.david@fb.com> [Mon, 10 Mar 2014 21:57:41 -0700] rev 880
exchange: fix push markers over wire protocol for http peer Http actually interpret the data passed to _callpush expecting them to a bundle. Interpretation of a markers stream as a bundle obviously led to crash.
Mon, 10 Mar 2014 21:55:41 -0700 exchange: expose length of pushed markers
Pierre-Yves David <pierre-yves.david@fb.com> [Mon, 10 Mar 2014 21:55:41 -0700] rev 879
exchange: expose length of pushed markers http wireprotocol will need it.
Mon, 10 Mar 2014 22:09:26 -0700 simple4server: update "from <changeset>" statement
Pierre-Yves David <pierre-yves.david@fb.com> [Mon, 10 Mar 2014 22:09:26 -0700] rev 878
simple4server: update "from <changeset>" statement Could obviously not do it in the same changeset than the original changes.
Mon, 10 Mar 2014 21:35:54 -0700 exchange: fix changeset pulling over http
Pierre-Yves David <pierre-yves.david@fb.com> [Mon, 10 Mar 2014 21:35:54 -0700] rev 877
exchange: fix changeset pulling over http http compress its stream. The were building a bogus stream where the first 10 bytes where uncompressed. Confusing the client.
(0) -300 -100 -10 +10 +100 +300 +1000 +3000 tip