#gemrb@irc.freenode.net logs for 17 Sep 2009 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage

[00:08:21] <-- tombhadAC has left IRC (Read error: 148 (No route to host))
[00:09:21] <-- Gekz has left IRC (Connection timed out)
[02:35:23] --> Darki has joined #gemrb
[02:47:57] <-- Dark-Star has left IRC (Read error: 110 (Connection timed out))
[05:09:44] --> jcompton_ has joined #gemrb
[05:27:11] <-- jcompton has left IRC (Read error: 110 (Connection timed out))
[06:37:29] --- Darki is now known as Dark-Star
[06:37:31] --- ChanServ gives channel operator status to Dark-Star
[07:29:31] --> lynxlynxlynx has joined #gemrb
[07:29:31] --- ChanServ gives channel operator status to lynxlynxlynx
[08:53:29] <-- |Cable| has left IRC (Remote closed the connection)
[10:17:32] <lynxlynxlynx> weidu is so silly with paths
[11:08:29] --> Gekz has joined #GemRB
[11:08:31] <Gekz> Ohai
[11:12:25] <-- Gekz has left IRC ("Reconnecting")
[11:12:39] --> Gekz has joined #GemRB
[11:21:52] <-- Gekz has left IRC (Read error: 104 (Connection reset by peer))
[11:39:11] --> Gekz has joined #GemRB
[11:46:54] <lynxlynxlynx> fuzzie: do you know of any other way than RevealAreaOnMap to reveal and area on the world map?
[11:47:35] <lynxlynxlynx> kuldahar doesn't get revealed on the world map, so you can't travel there. ar2000 has nothing and the hermit dialog only sets Hermit_Exit. No other script or dialog references it or calls RevealAreaOnMap with kuldahar (ar2100).
[11:48:23] <lynxlynxlynx> the hermit is the one telling you how to get there
[11:48:35] <fuzzie> what's the hermit dialog?
[11:48:43] <lynxlynxlynx> dhermit
[11:50:33] <lynxlynxlynx> i created dumps of the scripts and dialogs if you need them
[11:50:36] <fuzzie> interesting
[11:51:11] <fuzzie> that would be helpful
[11:52:02] <fuzzie> the "we'll find our own way" end node does the same variable set
[11:53:34] <lynxlynxlynx> http://lynxlynx.info/ie/how.dlg.tbz2 http://lynxlynx.info/ie/how.baf.tbz2
[11:53:56] <fuzzie> so perhaps it's simply a world map bug, but i thought i fixed those :/
[11:54:23] <lynxlynxlynx> the map scripts differ
[11:54:42] <fuzzie> there's certainly a link there, east from ar2000 to ar2100
[11:54:54] <lynxlynxlynx> ah
[11:55:03] <lynxlynxlynx> so you shouldn't even get the travel screen :)
[11:55:22] <fuzzie> but no reachable flags set on any areas
[11:55:27] <lynxlynxlynx> i was just about to say that RevealAreaOnMap is used only for 2 areas total
[11:55:49] <fuzzie> well, i wonder
[11:57:20] <fuzzie> got to turn on cheats for iwd, heh
[12:00:39] <lynxlynxlynx> where do you see the area link?
[12:00:54] <lynxlynxlynx> i see only the travel regions for the subareas
[12:02:41] <fuzzie> in real iwd, you get the travel screen
[12:03:03] <fuzzie> but kuldahar is there and visible even if i teleport there right from the start of the game and scripts are sabotaged
[12:03:06] <fuzzie> and i can travel there fine
[12:03:26] <fuzzie> so maybe iwd just ignores the wmap area flags entirely?
[12:03:41] <fuzzie> and, sorry, i mean area link as in, a wmap one
[12:03:46] <-- Gekz has left IRC (Read error: 104 (Connection reset by peer))
[12:03:50] <lynxlynxlynx> in the original, if i check the map in easthaven, kuldahar is not yet on it
[12:04:16] <fuzzie> i mean, i'm checking by going east from ar2000
[12:04:57] <lynxlynxlynx> aha
[12:05:30] <lynxlynxlynx> could be an engine difference
[12:05:48] <lynxlynxlynx> it can't ignore the flags completely or you would probably be able to return to easthaven
[12:06:11] <fuzzie> well, my first guess would be that it sets the flags when encountering a direct link
[12:07:26] <fuzzie> but got to run for a bit, i'll check it later if no-one else does
[13:14:57] --> Gekz_ has joined #GemRB
[13:31:47] <lynxlynxlynx> well, it is all about the WMP_ENTRY_ADJACENT flag
[13:32:19] <lynxlynxlynx> RevealAreaOnMap unconditionally adds it along WMP_ENTRY_VISIBLE
[13:36:18] <fuzzie> yes
[13:36:26] <lynxlynxlynx> commenting the adjacency test in WorldMap::UpdateAreaVisibility was enough
[13:37:51] <fuzzie> but you probably want to look at which flags iwd actually sets, by peering at an original WMP file
[13:38:13] <lynxlynxlynx> just WMP_ENTRY_VISITED
[13:38:44] <lynxlynxlynx> but you probably meant after the entry
[13:39:01] <fuzzie> well, i mean after the interaction with the travel point..
[13:39:13] <fuzzie> which i think is where it fiddles with flags
[13:57:44] <lynxlynxlynx> in the original, kuldahar is on the wmap as soon as i get into the pass
[14:07:58] <CIA-22> gemrb: 03lynxlupodian * r7184 10/gemrb/trunk/gemrb/GUIScripts/iwd/GUIREC.py: iwd::guirec: show the effective thac0 too
[14:09:17] <fuzzie> hum, i was rather hoping i'd solved the mysteries of the worldmap after fixing it for bg2
[14:10:28] <lynxlynxlynx> bg2 still has some problems with it
[14:10:52] <fuzzie> just visibility problems, though?
[14:11:23] <lynxlynxlynx> pretty much
[14:11:31] <lynxlynxlynx> 3 issues
[15:10:41] <lynxlynxlynx> heh, there's an actual item called "evil spider crusher of doom"
[15:14:16] --> Avenger has joined #gemrb
[15:14:20] --- ChanServ gives channel operator status to Avenger
[15:14:23] <Avenger> hello
[15:16:11] <fuzzie> hi
[15:20:58] <Avenger> is there any action/trigger you need to know more about?
[15:21:28] <Avenger> bg2
[15:21:46] <-- Gekz_ has left IRC (Remote closed the connection)
[15:21:57] <Avenger> is there anything we should implement?
[15:22:33] <lynxlynxlynx> there's plenty of missing actions in tob
[15:23:02] <lynxlynxlynx> the spellcast bunch comes to mind
[15:29:38] <CIA-22> gemrb: 03zefklop * r7185 10/gemrb/trunk/gemrb/plugins/ACMImporter/readers.cpp: Use correct type so that ACMImporter compiles on every platform.
[15:30:13] <Avenger> woo zefklop is back
[15:30:43] <Avenger> ok, i check the spellcast triggers, so the cowled wizards can do their job
[15:30:52] <-- Avenger has left IRC ("bye!")
[15:45:48] --> kingron has joined #gemrb
[15:47:01] <lynxlynxlynx> seggy
[15:49:16] <lynxlynxlynx> a trap set off and fx_damage got called with a null Owner
[15:53:35] <lynxlynxlynx> actually it's already null in AddAllEffects
[16:15:38] <CIA-22> gemrb: 03lynxlupodian * r7186 10/gemrb/trunk/gemrb/plugins/Core/Actor.cpp: fixed a trap-triggered crash
[16:16:18] <jcompton_> You guys are such diligent committers.
[16:16:21] <jcompton_> It's inspiring.
[16:16:48] <lynxlynxlynx> you should see the good days :)
[16:17:30] <lynxlynxlynx> and unlike with real obligations, the amount of work done varies a lot
[16:17:37] <jcompton_> Sure, of course.
[16:18:18] <jcompton_> But I'm thinking of the fact that I have (on more than one occasion) had to *beg* a TBH implementor to do a commit after I know he's been working on something for, like, a week, and I'm sitting here saying "if you suffer a sudden loss of data we will both be very, very, very, unhappy, just commit already!" and it goes on for a while.
[16:18:41] <jcompton_> So then I get enormous commits of dozens of files with no comment.
[16:19:12] <lynxlynxlynx> yeah, ideally you want plenty of small commits that are self-contained
[16:19:45] <lynxlynxlynx> avenger also tends to commit various stuff at once and it bites you when you are debugging later
[16:20:28] <fuzzie> but mostly it seems to work pretty well
[16:20:41] <lynxlynxlynx> the splitting stuff in small parts is not easy when the basic framework is not there though and some people don't like to commit stuff they know is not good enough
[16:24:43] <-- jcompton_ has left IRC (Remote closed the connection)
[16:25:01] --> jcompton has joined #gemrb
[16:26:54] <jcompton> Are any of you XML-parsing Java (preferably)-writing masters, or know anybody who is? I'd like to commission someone to do a Near Infinity-style browser for us. (just browsing, no editing.) I have had all I can take of our previous debugging system.
[16:27:20] <jcompton> Java preferred so it actually runs on all of our platforms, but I'd accept an alternative approach.
[16:29:32] <lynxlynxlynx> nice to hear you're still thinking multiplatform :)
[16:29:48] <jcompton> Yeah, that hasn't changed. Wes only runs Linux unless I expressly tell him "load this up in Windows if you're not seeing what I see", etc.
[16:29:58] <lynxlynxlynx> hehe :)
[16:30:14] <jcompton> I admit I have been too afraid to see what Apple has done to X11 support in OSX lately, but I'm sure it can be worked around one way or another.
[16:30:33] <jcompton> Or, hey, if this takes us a few more years, maybe that native GTK port will finally get done.
[16:31:30] <lynxlynxlynx> aren't there any good free viewers you could use? From what you said it appears you just need a browser
[16:32:44] <fuzzie> the native GTK port is kind of annoyingly stalled
[16:34:46] <jcompton> lynx: Honestly, it would be worth it to me to pay someone a modest sum of money to slap a nicer front-end or some prebuilt reports or whatever on an existing XML reporting tool (if in fact one exists which would handle our output) than to search through the zillion hits on "xml reporting tool" or whatever.
[16:35:30] <lynxlynxlynx> i'm looking into it
[16:39:32] <lynxlynxlynx> http://www.codeplex.com/xmlexplorer <-- the first candidate
[16:40:25] <jcompton> Let's see...
[16:41:42] <jcompton> Chokes on the file with a bunch of .Net call debugging info.
[16:42:17] <lynxlynxlynx> is your xml standards compliant?
[16:42:27] <jcompton> A fine question!
[16:42:43] <jcompton> (which I am not qualified to answer!)
[16:42:45] <lynxlynxlynx> try loading it in your browser
[16:43:06] <lynxlynxlynx> opera nicely prettifies xml :)
[16:43:20] <jcompton> Ahah. They are complaining about the file in the same place. So we are in fact doing something noncompliant.
[16:48:16] <lynxlynxlynx> how does the offending line look like?
[16:48:24] <jcompton> It had an ampersand.
[16:48:40] <lynxlynxlynx> not quoted, i presume
[16:48:48] <jcompton> No. <Name><value>orbit 2 (a b & c)</value>
[16:49:17] <jcompton> I fix that and then it hates this line:
[16:49:17] <jcompton> <recolor_sprite recolor=skin rgba=red/></value>
[16:49:26] <jcompton> so obviously there are some issues that will need to be addressed.
[16:50:00] <lynxlynxlynx> does the game already load it fine?
[16:50:38] <jcompton> Oh, yes, the engine processes this fine. This is from the massive 12 MB "thing report" debug database we output in the hopes of reporting on it. (we also output a massive HTML table, which is what we had been using, but I hate it.)
[16:53:01] <lynxlynxlynx> hmm
[16:53:52] <lynxlynxlynx> if it wont be too intrusive, it is probably better to fix the format and then use standard apps on top
[16:54:03] <lynxlynxlynx> in the end modders would appreciate that too
[16:54:54] <jcompton> Changing the effect syntax (which is the second example) isn't feasible, I think we're looking more at "escape some characters when generating the XML debug database."
[16:55:44] <lynxlynxlynx> that would fix the ampersand, sure
[16:55:47] <fuzzie> well, adding some quotes shouldn't be too hard :)
[16:55:57] <lynxlynxlynx> it wasn't really clear what was wrong with the second example
[16:56:20] <fuzzie> it should be <recolor_sprite recolor="skin" rgba="red"/>
[16:56:31] <jcompton> XML Explorer says "skin is an unexpected token."
[16:57:19] <lynxlynxlynx> nice, it is pretty strict
[16:57:42] <jcompton> (Firefox dislikes it in the same place, so they're in agreement.)
[16:58:03] <jcompton> Oh, you know, I think that actually is a malformatted effect.
[16:59:01] <lynxlynxlynx> http://www.java2s.com/Code/Java/XML/XMLTreeView.htm hehe
[17:01:15] <jcompton> Yeah, syntactically it should have quotes or single-quotes around it. But we don't get compile-time warns about effects, and it's in a test item, so it wasn't obvious.
[17:10:07] <jcompton> Hmm. Well, removing the two ampersands and fixing the effect gets me to a point where XML Explorer and Firefox now disagree about the next bad thing (XMLE is complaining about a character encoding now) so this will take some additional research. Thanks, though. If we get the output in order, maybe this app will in fact do the trick.
[17:30:14] --> barra_library has joined #gemrb
[17:56:02] <lynxlynxlynx> fuzzie: do you see anything bad in vstherik.baf? The cutscene mode starts, i get the mpsync message, but no dialog
[17:56:22] <lynxlynxlynx> suspect: ClearAllActions
[18:03:08] <fuzzie> hm, i don't know
[18:03:17] <fuzzie> the Dialogue() should still run in any case
[18:04:03] <fuzzie> but that SetDialogueRange is basically a null-op
[18:04:26] <fuzzie> so it could just be that the actor gets stuck trying to get to the target?
[18:05:57] <lynxlynxlynx> he does
[18:06:13] <lynxlynxlynx> i get the cutscene mode for a sec and then it is back to normal
[18:06:19] <lynxlynxlynx> i can repeat by teleporting
[18:06:31] <lynxlynxlynx> can't initiate the dialog manually with any pc
[18:06:52] <fuzzie> hm, if the cutscene mode ends then i thought that meant the dialog was successful
[18:07:23] <lynxlynxlynx> would it be considered successful if the creature has no dialog?
[18:07:29] <fuzzie> yes
[18:07:40] <lynxlynxlynx> the debug dump doesn't show any, but the .d exists
[18:07:50] <lynxlynxlynx> dtherik.d
[18:08:16] --> |Cable| has joined #gemrb
[18:08:40] <fuzzie> i assume there's no other script referencing that? maybe we don't load the cre correctly or something
[18:08:51] <lynxlynxlynx> no script or dialog
[18:09:01] <lynxlynxlynx> he also has gnundead, but that only changes the race
[18:09:20] <lynxlynxlynx> and the current script
[18:09:26] --> tombhadAC has joined #gemrb
[18:10:18] <lynxlynxlynx> jcompton: this one is for weimer: http://github.com/halhen/xmlfs
[18:27:00] <jcompton> Hmm. Interesting.
[18:55:07] <lynxlynxlynx> fuzzie: eh, he doesn't have the dialog set in the cre either
[18:55:25] <lynxlynxlynx> too bad the patch notes are practically nonexistant
[18:57:14] <fuzzie> which cre is it?
[18:57:37] <fuzzie> therik, i guess
[18:58:37] <fuzzie> which area?
[18:59:11] <lynxlynxlynx> 3301, no script
[18:59:36] <lynxlynxlynx> checking if totl has a fix
[19:00:07] <fuzzie> there is a dialog set, then
[19:00:12] <fuzzie> ar3301 has DTHERIK
[19:01:10] <fuzzie> AREImp reads the Dialog resref but doesn't set it in the actor
[19:01:16] <lynxlynxlynx> ah, didn't know maps can override them
[19:02:20] <fuzzie> try AREImp.cpp:939, if (Dialog[0]) { ab->SetDialog(Dialog); }
[19:04:42] <lynxlynxlynx> in a sec, now i get a fresh segfault
[19:04:49] <lynxlynxlynx> i think i know why too
[19:07:12] <CIA-22> gemrb: 03lynxlupodian * r7187 10/gemrb/trunk/gemrb/plugins/GUIScript/GUIScript.cpp: fixed an optional crash in HasResource
[19:08:12] <lynxlynxlynx> slow svn
[19:08:20] <CIA-22> gemrb: 03lynxlupodian * r7188 10/gemrb/trunk/gemrb/GUIScripts/iwd/GUIWORLD.py: iwd: remove the spam from GetWindowPack
[19:09:43] <lynxlynxlynx> yep :)
[19:18:49] <CIA-22> gemrb: 03lynxlupodian * r7189 10/gemrb/trunk/gemrb/plugins/AREImporter/AREImp.cpp: apply the map dialog override too, thanks fuzzie
[19:20:21] <fuzzie> heh
[19:20:27] <fuzzie> there might be a reason why it wasn't there already..
[19:20:34] <fuzzie> but i guess Avenger will notice, if so
[19:21:50] <lynxlynxlynx> it was the only missing field
[19:22:13] --> braj has joined #gemrb
[19:24:31] <-- braj has left IRC (Client Quit)
[19:25:38] --> braj has joined #gemrb
[19:28:45] <lynxlynxlynx> heh
[19:29:03] <lynxlynxlynx> you can't ^y the undead in iwd
[19:29:27] <lynxlynxlynx> looks like applydamage isn't available
[19:30:38] <fuzzie> i think i discussed this on irc, i think we concluded that just constructing the effects and applying manually in C++ should be ok
[19:31:55] <lynxlynxlynx> yes, i remember
[19:32:11] <lynxlynxlynx> i was completely in bg2 mode, so i didn't really care though
[19:42:38] <-- braj has left IRC ("Quitte")
[19:49:05] <-- kingron has left IRC (Remote closed the connection)
[20:45:32] <CIA-22> gemrb: 03lynxlupodian * r7190 10/gemrb/trunk/gemrb/plugins/Core/EffectQueue.cpp: fixed fx_disable_spellcasting to only be special for iwd2
[20:51:34] <lynxlynxlynx> turns out most uses of GF_IWD2_SCRIPTNAME for detecting iwd2 are bad
[20:51:57] <fuzzie> oh?
[20:52:14] <lynxlynxlynx> iwd and how have that flag set too
[20:52:20] <fuzzie> we should maybe rename it, then
[20:52:34] <lynxlynxlynx> we just shouldn't abuse it
[20:52:45] <fuzzie> or at least fix the header file
[20:52:57] <lynxlynxlynx> the name it has is for the only place where its use is correct
[20:58:45] <CIA-22> gemrb: 03lynxlupodian * r7191 10/gemrb/trunk/gemrb/plugins/Core/Actor.cpp: removed the last missuses of GF_IWD2_SCRIPTNAME
[21:14:53] <CIA-22> gemrb: 03lynxlupodian * r7192 10/gemrb/trunk/gemrb/override/ (bg1/modal.2da how/modal.2da iwd/modal.2da pst/modal.2da): added the modal.2da table to all of the games
[21:15:26] <lynxlynxlynx> now you can detect traps again
[21:15:47] <lynxlynxlynx> too bad this dungeon i'm in has them with nondetectable bit set
[21:16:54] <fuzzie> you think the flag is used differently in iwd, or is that intended?
[21:17:21] <lynxlynxlynx> no idea, it has been to long since i played it
[21:17:22] <jcompton> Being able to recompile the game on the fly is hugely useful. If you guys ever get around to "Native GemRB Extra Awesomeness Mode", I highly recommend it.
[21:18:08] <lynxlynxlynx> you mean the game scripts?
[21:18:33] <lynxlynxlynx> and we already have some extra awesomeness :)
[21:18:35] <fuzzie> unfortunately the underlying design is rather spectacularly unhelpful for trying to reload anything but scripts
[21:18:57] <jcompton> To fix your problem, you'd save the game, quit, change the traps to remove that bit in an editor, reload the engine, reload the game, and then walk around, right?
[21:19:21] <lynxlynxlynx> unless the bug is in the loader
[21:19:25] <jcompton> Heh, true.
[21:19:31] <jcompton> But assuming it wasn't.
[21:19:50] <fuzzie> actually, it's worse than that
[21:19:59] <fuzzie> because the areas are rather unfortunately bundled in the saves
[21:20:06] <jcompton> Ooof, I forgot about that.
[21:20:08] <fuzzie> so by reloading a save, you'd get the *old* trap bits
[21:20:25] <fuzzie> it is quite spectacularly annoying when trying to debug these things :)
[21:21:21] <lynxlynxlynx> ni can edit stuff in saves
[21:21:55] <lynxlynxlynx> these iwd traps are deadly, the second reproducible crash today
[21:21:58] <jcompton> Yeah. What I was getting at is that I can just hit pause, change a script or XML property, use a debug command to force the game to recompile, and see the change. (for 99.9% of things, anyway. There are a few things which either require the area to be re-shown and like one thing that requires an actual "start a new game or reload a save" maneuver.)
[21:22:07] <fuzzie> yes, but if you're actually trying to edit an area, you have to make sure you tweak both the area you're working on and the area in the save..
[21:22:23] <fuzzie> jcompton: yes, if you're designing from scratch i imagine that kind of thing would be great :)
[21:22:35] <jcompton> And yeah, I know you have some cooler stuff, but not a true "sixth game mode" yet, right?
[21:23:06] <lynxlynxlynx> not yet, nobody wants to make a new game bad enough yet
[21:23:12] <jcompton> I can understand that.
[21:23:29] <lynxlynxlynx> there were a few starts though :)
[21:23:31] <jcompton> They want to keep their money and hair, no doubt.
[21:23:34] <fuzzie> the prerendered background design is a bit of an obstacle
[21:23:51] <lynxlynxlynx> all you need is a blender export filter
[21:23:52] <fuzzie> because it requires so much graphical work, and everyone seems to want to use isometric tiles to minimise that
[21:24:21] <jcompton> fuzz: It was actually far easier for me to find competent area designers than it was to find *sprite* guys. That was by far the worst/most obscure/most difficult task.
[21:24:24] <fuzzie> lynxlynxlynx: and something to do the wall polys and doors and all the maps.. :/
[21:25:03] <lynxlynxlynx> the map can be done the same, the rest are just pretty simple overlays
[21:25:05] <jcompton> Wall polygons would be your biggest problem, yeah. Infinity makes it brutal because the engine was designed with (seriously) "we will just have interns trace shapes for 18 months" in mind.
[21:25:14] <fuzzie> yes :)
[21:25:38] <fuzzie> lynxlynxlynx: i mean, search maps and height maps and etc
[21:25:40] <jcompton> But you guys would be welcome to steal our process. I've been trying to convince someone to adopt our process for a WED converter for years, but nobody's taken me up on it.
[21:25:58] <jcompton> fuzz: Search/height are fairly easy concept for area artists.
[21:26:05] <jcompton> I was shocked.
[21:26:12] <fuzzie> well, i have little hope for anyone actually making a new game, but i'm sure there'd be a lot of interest in tools to help with modding the existing games, even now
[21:26:16] <jcompton> I had about six legitimate and affordable candidates to choose from.
[21:26:19] <lynxlynxlynx> you just draw on a transparent layer
[21:26:36] <fuzzie> i quite enjoy painting them in DLTCEP but it's certainly more work :)
[21:27:44] <jcompton> Drawing wall polygons is a horrible thing you should never make anybody do. Let me find the thread where I explain our process...
[21:28:15] <fuzzie> my problem at the moment is simply time: once i have more, i'd certainly be interested in implementing anything to help with the polygon drawing
[21:28:16] <lynxlynxlynx> i remember it
[21:28:21] <fuzzie> although i have a huge gemrb list too :)
[21:28:29] <jcompton> lynx: Oh, I've already bothered you about this? Sorry. :)
[21:28:42] <lynxlynxlynx> no, but i have a peculiar memory
[21:28:52] <lynxlynxlynx> wasn't it covered in the ml?
[21:30:55] <jcompton> http://www.shsforums.net/index.php?showtopic=32402&st=20 ... the example of "this is how we do it" post is http://www.shsforums.net/index.php?s=&showtopic=32402&view=findpost&p=395312 .
[21:31:43] <jcompton> We have a front-end to the Triangle program that creates "accurate enough for gaming" XML polygon data, which I'm really very sure could be processed into a WED file.
[21:32:24] <jcompton> And the area artist generates the white-on-black silhouette map by just shutting off everything but the objects, basically.
[21:35:19] <jcompton> But, yeah, you can explain to a competent environment artist how to make an "Infinity-style" area with a minimum of fuss. It's sprites that are brutal.
[21:35:31] <jcompton> Particularly recolorable ones. Yikes.
[21:36:05] <CIA-22> gemrb: 03lynxlupodian * r7193 10/gemrb/trunk/gemrb/plugins/Core/Spell.cpp: Spell::GetEffectBlock was breaking commandment #2
[21:37:14] <fuzzie> hm
[21:37:27] <lynxlynxlynx> found something for avenger
[21:37:42] <lynxlynxlynx> one of the hardcoded projectiles
[21:38:40] <fuzzie> huh, 'git blame file.cpp' is just sitting there, is it slow?
[21:38:59] <lynxlynxlynx> big file?
[21:40:24] <fuzzie> i guess so. slower than svn is impressive though :)
[21:42:24] <lynxlynxlynx> you timed it?
[21:51:42] <lynxlynxlynx> ok, i tried actor.cpp
[21:51:53] <lynxlynxlynx> did one run first to get it into the cache
[21:52:23] <lynxlynxlynx> the next runs git took about 5s while svn did 12s
[21:52:36] <lynxlynxlynx> the cold run was roughly double
[21:53:30] <fuzzie> yes, i zapped the cache first for both
[21:53:45] <fuzzie> it makes sense i guess, svn is going to have less data to poke through
[21:54:56] <lynxlynxlynx> it stores revisions per-file
[21:55:08] <lynxlynxlynx> but this didn't help it in my case :)
[22:00:04] <-- lynxlynxlynx has left IRC (Remote closed the connection)
[22:21:54] <CIA-22> gemrb: 03avenger_teambg * r7194 10/gemrb/trunk/gemrb/plugins/ (14 files in 2 dirs): SpellCast* triggers
[22:49:29] <-- Dark-Star has left IRC ()
[23:09:39] --> Maighstir has joined #gemrb