So I managed to grab/hassle Brian and got a reply, sent me in the right direction and I managed to fix the issue in our case. I'll quote the reply I got in case anyone else comes across this issue and is searching for a fix.
"It's related to the number of unique shader states (alpha blending on, etc.) your materials are using in your app. I don't believe we ship the source code for shaderapidx9.dll, which is where you could up the limit (let me know if I'm wrong and I'll try to find the line of code to change). Killing support for fixed-function graphics hardware (dx7-level) in your shaders will help, and may well fix it (if you have support for them). Try modifying your materials to use less of the space of your shader states, or modifying your shaders to eliminate some of the features that are setting states like alphablending / depth test, etc etc. I'm afraid we don't have any real tools to help identify which materials are generating problems though. I'd suggest precaching the content to help you know when you've reduced the shader state usage enough though, at least while you're working on this problem." Honestly, I'm not competent enough with the engine/shaders to know how to go in and mess around with the states, I fixed it in our case by going through and fixing up the odd texture that was using a different shader when we could have used something else so we'll probably run into the issue again at a later date, hopefully by then I'll either know enough to actually do a more permanent fix. On 23 August 2012 12:08, J. McBroom <jimbo...@gmail.com> wrote: > Yep, checked it out and everything looks normal, not that close to > hitting the limit of any of the stringtables and can't see any > duplicate entries. Thanks though. > > On 23 August 2012 11:49, Tobias Kammersgaard > <tobias.kammersga...@gmail.com> wrote: >> Did you try these commands to check for duplicates etc. in your string >> tables? >> >> http://pastebin.ca/878546 >> >> (Thanks to Garry for posting the snipets) >> >> - ScarT >> >> >> >> On 23 August 2012 12:46, J. McBroom <jimbo...@gmail.com> wrote: >>> >>> Gave it a shot, no luck unfortunately. >>> >>> It definitely seems like it's somehow tied to the amount of content >>> loaded, if we only precache half the models then we're fine... until >>> the rest of them are loaded at a later time and the error pops right >>> back up. It seems like both Garry's Mod and Obsidian Conflict has ran >>> into the issue in the past, being fixed by an engine update and >>> letting the user mount content "statically" respectively but still no >>> closer to finding out how to avoid running into this issue without >>> just plain loading less content. >>> >>> On 23 August 2012 10:30, Adam "amckern" McKern <amck...@yahoo.com> wrote: >>> > I came across this bug in a lost coast source when working on NightFall >>> > many >>> > years ago, a clean rebuild of the code base seemed to fix things >>> > >>> > >>> > -------- >>> > Owner Nigredo Studios http://www.nigredostudios.com >>> > >>> > ________________________________ >>> > From: J. McBroom <jimbo...@gmail.com> >>> > To: Discussion of Half-Life Programming >>> > <hlcoders@list.valvesoftware.com> >>> > Sent: Thursday, 23 August 2012 3:45 PM >>> > Subject: [hlcoders] Transition table overflow. >>> > >>> > We've ran into this issue, currently having it happen in our 2007 based >>> > mod. >>> > >>> > It seems to be tied to both the map we load and the precaching of the >>> > large amount of content in our mod (~35 player models, 30+ weapons ). >>> > It's happening 100% of the time when loading straight to dm_overwatch, >>> > it doesn't happen when we load up a small testing map that doesn't >>> > have much content. >>> > >>> > As soon as we get in game, there's a large amount of artifacting >>> > (along with a Transition table overflow. Grab Brian error), things >>> > rendering in the wrong order etc (http://i45.tinypic.com/255up10.jpg). >>> > >>> > We tried delaying the precaching of certain playermodels until we knew >>> > they were going to be used (which really isn't an ideal situation >>> > anyway as there's noticeable hitching when the model is loaded) and >>> > while this lets the map load and function normally, the error happens >>> > and it returns back to the broken state once we precache more content. >>> > >>> > Has anyone ran into this issue or know of any way to fix/avoid it? >>> > >>> > _______________________________________________ >>> > To unsubscribe, edit your list preferences, or view the list archives, >>> > please visit: >>> > https://list.valvesoftware.com/cgi-bin/mailman/listinfo/hlcoders >>> > >>> > >>> > >>> > >>> > _______________________________________________ >>> > To unsubscribe, edit your list preferences, or view the list archives, >>> > please visit: >>> > https://list.valvesoftware.com/cgi-bin/mailman/listinfo/hlcoders >>> > >>> > >>> >>> _______________________________________________ >>> To unsubscribe, edit your list preferences, or view the list archives, >>> please visit: >>> https://list.valvesoftware.com/cgi-bin/mailman/listinfo/hlcoders >>> >> >> >> _______________________________________________ >> To unsubscribe, edit your list preferences, or view the list archives, >> please visit: >> https://list.valvesoftware.com/cgi-bin/mailman/listinfo/hlcoders >> >> _______________________________________________ To unsubscribe, edit your list preferences, or view the list archives, please visit: https://list.valvesoftware.com/cgi-bin/mailman/listinfo/hlcoders
