#gemrb@irc.freenode.net logs for 22 Jul 2015 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:25:54] <-- phao has left IRC (Ping timeout: 244 seconds)
[00:43:24] --> phao has joined #gemrb
[04:33:54] --> edheldil_ has joined #gemrb
[04:47:13] <-- phao has left IRC (Ping timeout: 252 seconds)
[05:08:35] <-- edheldil_ has left IRC (Ping timeout: 244 seconds)
[05:15:40] --> edheldil_ has joined #gemrb
[06:48:18] <-- edheldil_ has left IRC (Ping timeout: 265 seconds)
[08:09:51] --> Eli2_ has joined #gemrb
[08:13:13] <-- Eli2 has left IRC (Ping timeout: 264 seconds)
[08:22:21] --> GeneralDuke has joined #gemrb
[08:27:34] --> trickyhero has joined #gemrb
[08:31:07] <-- norito has left IRC (Ping timeout: 244 seconds)
[11:24:20] <lynxlynxlynx> wjp: around?
[11:24:32] <wjp> yes
[11:24:44] <lynxlynxlynx> i found a wierd spritecover issue
[11:24:49] <lynxlynxlynx> https://youtu.be/ED0eygy2ljI?t=2m25s
[11:24:50] <Pepelka> Icewind Dale (HD) - Heart of Winter - Icasaracht - YouTube
[11:24:51] <Pepelka> »The Heroes defeat the White Dragon Queen, Icasaracht. For my complete guide to playing IWD, hosted by Sorcerer's Place, see: http://www.sorcerers.net/Games/I...«
[11:25:09] <lynxlynxlynx> the dragon is drawn over fog of war there
[11:25:17] <wjp> (s/wierd/weird/ )
[11:25:29] <wjp> which timestamp?
[11:25:56] <lynxlynxlynx> we had a bad animation for it, so you couldn't hit it before, but with the fix, we hit the Covers assert in DrawActorSprite
[11:26:09] <lynxlynxlynx> timestamp a few seconds later
[11:26:23] <wjp> oh, wait, the link has a timestamp
[11:26:28] <lynxlynxlynx> (no eastern facing animations before)
[11:26:56] <wjp> so the video is from the original and gemrb crashes?
[11:27:03] <lynxlynxlynx> disabling the assert just makes it boom later in the blitter
[11:27:08] <lynxlynxlynx> yes
[11:27:20] <lynxlynxlynx> I can push the animation fix if you want to check this out yourself
[11:28:38] <wjp> the assert(newsc->Covers(...)) in the part loop?
[11:28:52] <lynxlynxlynx> yep
[11:29:14] <wjp> is the comment about anims[0]->animArea there still true?
[11:29:15] <lynxlynxlynx> the first check passes (worldx), but not the corners
[11:30:32] <lynxlynxlynx> no, this bam has frames of different sizes and there doesn't appear to be a pattern
[11:31:10] <lynxlynxlynx> (just loooking at width and height)ž
[11:31:32] <wjp> my memory is a bit foggy, but I think we may be setting that animArea ourselves
[11:32:16] <lynxlynxlynx> only 10 years went by ^^
[11:32:40] <wjp> if (part > 0)
[11:32:41] <wjp> anims[0]->AddAnimArea(a);
[11:33:12] <wjp> in CharAnimations::GetAnimation
[11:33:44] <wjp> so the first thing to check would probably be if that's still going right
[11:35:38] <wjp> I can't really look myself until tonight
[11:35:50] <lynxlynxlynx> it only has one part
[11:36:26] <lynxlynxlynx> so that doesn't get hit
[11:36:56] <wjp> that's very weird
[11:37:20] <lynxlynxlynx> why?
[11:38:05] <wjp> because it's then doing a BuildSpriteCover for anim->animArea, while nextFrame is anim->GetFrame(...)
[11:38:23] <wjp> (right above the assert)
[11:38:49] <wjp> hrm
[11:39:09] <lynxlynxlynx> you mean the sprite cover shouldn't change in this case?
[11:39:28] <wjp> the check in the assert is the same check as in the if() above
[11:39:57] <wjp> so if the assert fails, the BuildScriptCover did something wrong, or animArea was wrong
[11:40:05] <wjp> BuildSpriteCover
[11:40:25] <wjp> is any sprite mirroring involved?
[11:40:38] <lynxlynxlynx> yes, double, horizontal and vertical
[11:40:55] <lynxlynxlynx> at that point in time only horizontal
[11:41:46] <wjp> I wonder why the animArea update line in Animation::MirrorAnimationVert() is commented out
[11:42:13] <lynxlynxlynx> i'll blame it
[11:42:50] <wjp> been that way since the initial commit of that function
[11:42:53] <lynxlynxlynx> like that from the start, yes
[11:44:17] <lynxlynxlynx> trying it out
[11:44:48] <wjp> I added that animArea line to the horizontal mirroring in my initial SpriteCover commit
[11:44:50] <lynxlynxlynx> maybe the projectile and vvc code needs it this way, but it's also possible nothing hit it yet and it is inverted there too
[11:44:58] <wjp> (the vertical mirroring function was added later)
[11:46:16] <lynxlynxlynx> heh, this avoids the assert, but the results are funny
[11:46:40] <lynxlynxlynx> during the fow phase, the animation is wrongly inverted and sometimes later too
[11:46:47] <lynxlynxlynx> i'll recheck my orientations
[11:48:21] <lynxlynxlynx> yeah, that's my fault
[11:49:06] <lynxlynxlynx> shouldn't be using vertical mirroring at all; the data is missing half of the orientations (after horizontal mirroring)
[11:53:36] <lynxlynxlynx> ok, i'll fix that and then check which vvcs and pros need vertical mirroring
[11:57:25] <lynxlynxlynx> thanks
[11:58:07] <wjp> np
[11:58:25] <wjp> ION, looks like KOTOR2 now has a linux version
[11:59:32] <lynxlynxlynx> cool, need to replay that with the "unfinished business" mods sometime, the original was pretty lacking
[12:09:56] * wjp spots a comment by one DrMcCoy on RPS on how to get a mod working in linux
[12:29:57] <DrMcCoy> :P
[12:37:31] <fuzzie> that troublemaker
[12:58:43] <-- GeneralDuke has left IRC (Ping timeout: 240 seconds)
[13:03:00] <-- trickyhero has left IRC (Read error: Connection reset by peer)
[13:16:52] <DrMcCoy> As a balance, I also complained about how much SWTOR disappointed me :P
[13:37:03] --> phao has joined #gemrb
[13:48:50] --> gembot has joined #gemrb
[15:43:50] <-- fuzzie has left IRC (Ping timeout: 255 seconds)
[15:43:56] --> fuzzie has joined #gemrb
[15:54:48] --> GeneralDuke has joined #gemrb
[15:57:41] <-- GeneralDuke has left IRC (Client Quit)
[18:57:50] <-- phao has left IRC (Ping timeout: 244 seconds)
[19:24:26] <lynxlynxlynx> fresh sf update
[19:24:37] <lynxlynxlynx> basically no release this week
[19:24:54] <lynxlynxlynx> but web should be back todayish
[19:32:20] --> phao has joined #gemrb
[19:34:56] <DrMcCoy> Oh, the bugtracker is back
[19:35:39] <fuzzie> time to scrape? :p
[19:36:21] <DrMcCoy> And of course, ScummVM web won't be back so soon, since the letters jkms are still missing
[19:37:07] <fuzzie> but 'g' is all good
[19:37:17] <DrMcCoy> Yes, yes, rub it in :P
[19:37:54] <fuzzie> I am looking through the scummvm bug tracker and I wish I hadn't
[19:38:26] <DrMcCoy> :(
[20:13:34] <-- phao has left IRC (Quit: Leaving)
[20:15:49] --> phao has joined #gemrb
[20:26:09] <lynxlynxlynx> grrr, all three spamming animations are of the same general type, but lacking different files
[20:31:43] <Pepelka> [commit] lynxlynxlynx: added new subtype of IE_ANI_TWO_FILES_3 without eastern files https://github.com/gemrb/gemrb/commit/fb2cb7620aa95531bfde1d0a78d9f4fd7351b48f
[20:31:44] <Pepelka> [commit] lynxlynxlynx: iwd: fixed remorrhaz animation type, they have no jabbing files https://github.com/gemrb/gemrb/commit/026eff24821dbaf340291da734bea9c96c0ae8c4
[20:35:46] <xrogaan> how dare they?
[20:36:59] <lynxlynxlynx> we know it's even worse in the original, so there's at least that
[22:32:08] <Pepelka> [commit] lynxlynxlynx: Wall_Polygon: init all members in ctor https://github.com/gemrb/gemrb/commit/de27d4e123d934177cf65977686e5c629cff3508
[22:32:09] <Pepelka> [commit] lynxlynxlynx: OpenALAudioDriver: init all members in ctor https://github.com/gemrb/gemrb/commit/394cfc1c1e923856eb61fc110dabd9c3e2f0e63a
[22:32:10] <Pepelka> [commit] lynxlynxlynx: renamed all the unhardcoded portalh projectiles to their true names https://github.com/gemrb/gemrb/commit/ba08833a29d2c7c1678c721f53afd162ce46b29d
[23:09:57] <-- phao has left IRC (Ping timeout: 265 seconds)
[23:24:57] --> norito has joined #gemrb