On Sunday, 4 July 2010 at 16:00, Guangcong Luo wrote: > This commit incremented the netcode version for 2.3-branch. From > conversations with Safety0ff, it appears this was unnecessary - > netcode compatibility was not broken in that commit. > > Furthermore, I note that netcode version was incremented without > informing ML, tracker, forums, or mentioning it in the commit log, and > apparently even Safety0ff didn't know it was done. I only noticed it > myself because I was debugging svn/2.3 with some 2.3.1 players.
Did you notice that the netcode version was already increased after 2.3.1? Ok, that doesn't really address your point, but... > First: I believe that incrementing netcode version in a stable branch > is a big thing. I think that incrementing netcode version should > always be mentioned in the commit log (no matter which branch), and if > done to a stable branch, should be brought up in at least the ML. In > addition, patches that break netcode compatibility should note this in > the tracker. Can we agree to make this policy? ... personally I don't think it's a big deal (sync is crap anyway, but no need to make it even worse by allowing versions that differ much to play together), if you really need to connect to players with an earlier version you can change it back locally. > Second: Buggy, is incrementing netcode version necessary? Safety0ff > says that none of the changes break netcode compatibility, and that None of our recent changes break net_code_ compatibility. But anything that changes how the engine behaves will make sync worse. > netcode version wasn't incremented in any of the other branches. If > not, I'd like to change it back. If so, I have a few other > netcode-compatibility-breaking patches I'd like to commit. :P Fine by me, if you post them first. :P _______________________________________________ Warzone-dev mailing list Warzone-dev@gna.org https://mail.gna.org/listinfo/warzone-dev