On Sat, Jul 03, 2010 at 10:05:26AM +1200, Ralph Versteegen wrote: > If you look at r3615, you'll see I included some long winded > instructions in IMPORTANT-nightly.txt, which you can't even apply if > you're playing a passworded game. But it wouldn't be hard to fix this > problem automatically: just ignore the gameover and loadgame settings > if they are both zero. This will fix vastly more saved games than it > breaks (where a script has intentionally zeroed those out), and it's > only been a few weeks, so the chances of it actually breaking anything > is near zero. However, you want to be able to identify saves that > might be saved with a version exhibiting the bug
That seems reasonable to me. > We could just increment the save file version number, but that's feels > unnecessarily heavy-handed. Does it matter? We could add fixbits to > avoid doing so. It was my intention that incrementing the save file version number was something we could do often and easily, but right now I can't remember if that is actually the case yet. > And since we embarking into a sane new world, we ought to include the > version of Game that was used to create a save file. I suggest: > > * rsav > ** ver - int > ** game_client - string (eg "OHRRPGCE") - ignore children if the > client name isn't recognized > *** codename (aka branchname) - string (eg "voxhumana", "wip") > *** revision - either integer or string (eg. svn revision number) > ... I think that is a great idea! :) --- James _______________________________________________ Ohrrpgce mailing list [email protected] http://lists.motherhamster.org/listinfo.cgi/ohrrpgce-motherhamster.org
