-- [ Picked text/plain from multipart/alternative ] Thanks for the advice, looks like we have some work to do. :) I would say the question now is, will we end up running into any problems not having engine access?
Thanks again Ryan On Jan 29, 2008 6:23 AM, Garry Newman <[EMAIL PROTECTED]> wrote: > The problem is there's too much info in the string tables. You can see > the contents of the string tables using dumstringtables - but I've > found that to be unreliable, so here's a couple of commands I made: > > http://pastebin.ca/878546 > > Basically you want to keep the total string table size quite a bit > under 96000 bytes. > > There's a bunch of ways I've tried to do this: > > Sounds aren't precached anymore, so they don't get added to the string > table until they're actually used. > > Same with particles. > > Same with scenes. > > The models seem to take an entry even if they don't exist now. And it > doesn't correct backslashes in the model name etc (so models/error.mdl > and models\error.mdl are stored as seperate entries, doubling the > space it uses). So I hacked a function to try to correct this in > CBaseEntity:PrecacheModel(). > > That's all I can remember. I don't think most of these things would be > an issue unless you were like GMod and mounting a bunch of other > content (like sounds from cs, dod, tf2, ep2, portal etc).. but > something must be going wrong for you to be overflowing. > > Hope that helps > > garry > > On Jan 29, 2008 6:59 AM, Ryan Sheffer <[EMAIL PROTECTED]> wrote: > > -- > > [ Picked text/plain from multipart/alternative ] > > Hey > > > > I hope everyone is having fun with the new sdk, it's been ok to us so > far > > but there is a massive issue we are experiencing with the new engine. > This > > problem existed before, but now it's out of control. > > > > The dreaded! > > BING* > > Host_Error: Overflow error writing string table baseline Scenes > > > > I was attempting to make the various Breencast and Kleiner monitor > scenes > > work again, as I had fixed it in the previous codebase. When I tried to > do > > my fix again in the new engine, I receive that error upon every server > > join. The exact code I changed was in sceneentity.cpp, in the > > DispatchStartSpeak function. I made the > > > > CPASAttenuationFilter filter( actor ); > > > > into > > > > CBroadcastRecipientFilter filter( actor ); > > > > I realize that's not the most efficient way to do it, but we want users > to > > be able to play the Half-Life series' maps in our mod and scenes are > pretty > > important. I thought that this might just be too many scenes to send > across > > the wire in some cases though, so I made a more specific fix by checking > for > > microphones in the world and basing the attenuation checks off of them. > > However, having the CPASAttenuationFilter call AddAllPlayers (skipping > the > > engine call to Message_DetermineMulticastRecipients), I receive the same > > error. It seems like it's not only caused by sounds in scenes, but > sounds > > in general. I've run out of ideas as I can't really tell what's going > on > > here, it's all locked away in the engine. I would hate to have to leave > the > > monitor scenes broken in my mod, it just wouldn't be the same in those > maps > > anymore. And unfortunately, we get this error outside of those maps from > > time to time as well. > > > > I recently noticed that Garry Newman had corrected an issue like this, > and I > > hope it is possible for us to correct this without engine access. > > > > Thanks for any help :) > > -- > > Ryan > > -- > > > > _______________________________________________ > > To unsubscribe, edit your list preferences, or view the list archives, > please visit: > > http://list.valvesoftware.com/mailman/listinfo/hlcoders > > > > > > _______________________________________________ > To unsubscribe, edit your list preferences, or view the list archives, > please visit: > http://list.valvesoftware.com/mailman/listinfo/hlcoders > > -- ~skidz -- _______________________________________________ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
