Just the 'Rowan' title and the choices, but not what Rowan actually says.Įdit: Also, the src/chocolate-*-setup binaries are not cleaned up in the distclean rule - as are the manpages, the desktop files and some others.Įdit 2: Sorry, the missing dialog was my fault. If(nickname != player_names)Ģ) When talking to Rowan, his text isn't displayed at all. I just gave chocolate-strife beta 3 a try in Linux and I believe it is not ready yet.ġ) The following fix is required to keep it from instantly segfaulting whenever a (new) game is started or loaded: In DOSBox, for me the addresses were always the same: Use choco to break on the tick and mobj update and step through and see what is happening (I added a mobj counter so I could break at the mobj). The tick and mobj responsible for the desync can then be determined. I made a similar mod to choco-heretic so then the data for both tics could be compared. I made a program to load the file and walk the thinkers list outputting the x, y and z position for each mobj (function = P_MobjThinker). Set a memory break point for changes to leveltime (F5 then advances the game by 1 tick). Play the demo and wait until it's just about to desync.īreak into the DOSBox debugger (ALT-Pause). Locate the addresses of thinkercap, leveltime and P_MobjThinker.

Locate D_DoomLoop, from there the heretic source code can be used as a guide by following the calls. I've just downloaded the zip from the v2 branch on git hub, applied the two changes listed in this thread (nice that Quasar spotted the other one), built with msvc and both demos playback fine for me.

Dosbox debugger documentation