#gemrb@irc.freenode.net logs for 13 Feb 2012 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:02:39] <-- float has left IRC ()
[00:08:28] <brad_a> so it seems our video driver goes out of its way to do clipping itself. seems if we are going to do that we could gain some small amount of performance calling SDL_LowerBlit directly
[00:08:54] <brad_a> of course there is always the possibility that i have no idea what im talking about :-p
[00:11:02] <brad_a> and the drawing methods that do manual clipping to the viewport… do they do something special im not seeing? or does SDL_SetClipRect perform poorly in comparison?
[00:13:34] <-- Yoshimo has left IRC (Quit: Yoshimo)
[01:06:03] --> joneirik has joined #gemrb
[01:17:21] <-- wrotek has left IRC (Ping timeout: 260 seconds)
[02:08:44] <lostLinSoul> Night all!
[02:09:03] <-- lostLinSoul has left IRC (Quit: Page closed)
[03:26:37] <-- CIA-125 has left IRC (Ping timeout: 248 seconds)
[03:26:50] --> CIA-28 has joined #gemrb
[03:28:43] <-- harijan has left IRC (Ping timeout: 248 seconds)
[03:28:54] --> harijan has joined #gemrb
[04:03:44] <-- joneirik has left IRC (Remote host closed the connection)
[04:59:41] <-- brad_a has left IRC (Quit: brad_a)
[05:12:34] --> brad_a has joined #gemrb
[06:11:04] <-- brad_a has left IRC (Quit: brad_a)
[06:43:16] --> edheldil_ has joined #gemrb
[06:50:03] <-- edheldil_ has left IRC (Ping timeout: 265 seconds)
[07:26:50] <-- cj_schnell has left IRC (Quit: KVIrc 4.0.4 Insomnia http://www.kvirc.net/)
[08:01:54] --> lynxlynxlynx has joined #gemrb
[08:01:54] --- ChanServ gives channel operator status to lynxlynxlynx
[09:56:06] <-- fuzzie has left IRC (*.net *.split)
[09:56:07] <-- Kiranos has left IRC (*.net *.split)
[09:59:34] --> fuzzie has joined #gemrb
[10:00:11] --> Kiranos has joined #gemrb
[10:31:46] <edheldil> lynxlynxlynx: iirc even if you have a negative con bonus, you always get at least 1hp on level up
[10:35:37] <edheldil> (re Prism)
[10:37:49] <-- Kiranos has left IRC (*.net *.split)
[10:39:17] --> Kiranos has joined #gemrb
[11:22:36] <lynxlynxlynx> you do, but the npcs don't follow it
[11:55:47] <edheldil> well, the npcs/monsters should have HitDie at least 1, which is the same. And if you create PC-class-based npcs, see PCs
[11:56:06] <edheldil> i.e. you do not add con bonus to HD for NPCs
[11:56:49] <edheldil> (maybe some wimp creater have HD 1/2 or so)
[11:56:53] <edheldil> creatures
[12:01:26] <edheldil> Quoting SRD: "Treat monster Hit Dice the same as character level for determining ability score increases. This only applies to Hit Dice increases, monsters do not gain ability score increases for levels they "already reached" with their racial Hit Dice, since these adjustments are included in their basic ability scores. "
[12:09:20] <lynxlynxlynx> but that's regular rules
[12:14:31] <fuzzie> it is likely PC-classed-based NPC with insufficient HP to account for levels, no?
[12:14:35] <fuzzie> should check what original does
[12:17:35] --> SiENcE has joined #gemrb
[12:22:58] <edheldil> the point is that you will have more HP on a higher level. The opposite does not make sense, imo
[12:25:22] <edheldil> quoting SRD again (Re PC classes): "You apply your character’s Constitution modifier to:
[12:25:22] <edheldil> Each roll of a Hit Die (though a penalty can never drop a result below 1—that is, a character always gains at least 1 hit point each time he or she advances in level)."
[12:25:38] <tomprince> Are the npcs stored without con bonuses, or with?
[12:31:15] <edheldil> I have a feeling w/o them, I vaguely remember HPs in PS:T being way off before lynx started to add the con bonus, but D&D is braindead in this regard - if you lose -1 CON mod, you have to subtract your level HP, which is not really correct if you made 1hp on levelup in the past
[12:32:35] <fuzzie> edheldil: yes, but if you gain a bunch of points and then permanently lose the points, then you are in serious trouble if that ends up below 0
[12:32:44] <fuzzie> which is effectively what this is doing
[12:33:01] <-- lynxlynxlynx has left IRC (Read error: Operation timed out)
[12:33:29] <edheldil> Although SRD states "accordingly", not an exact formula. How clever of them :)
[12:33:33] <edheldil> "If a character’s Constitution score changes enough to alter his or her Constitution modifier, the character’s hit points also increase or decrease accordingly. "
[12:34:40] <edheldil> well, pcgen did have problems on level drain due to aberrations like that
[12:36:33] <edheldil> fuzzie: the weapon it was about to wear decreases con bonus?
[12:36:52] <fuzzie> edheldil: no, it just has way too little HP for the levels
[12:37:31] <fuzzie> and the game is modded so as to have a -2 HP penalty for the con in question
[12:38:02] <fuzzie> so for each level up they should be gaining 3 pre-bonus HP
[12:38:32] <fuzzie> but of course someone has just randomly given them not enough HP, and presumably we should be not allowing the penalty to drop HP below 1 or something
[12:39:28] <edheldil> but since we do not do level drain, even so Prism should get base +level HPs
[12:39:38] <fuzzie> but how?
[12:40:00] <edheldil> 1HP for each level, Must be >= 1
[12:40:02] <fuzzie> i mean, PCs are stored on disk with pre-bonus HP
[12:40:21] <fuzzie> we can't predict what might have happened at their level ups
[12:40:28] <wjp> we should be able to test this in the real bg2, right?
[12:40:28] <fuzzie> and there are situations when you permanently lose HP
[12:40:33] <fuzzie> yes
[12:40:39] <fuzzie> that seems the only sensible approach
[12:42:03] <edheldil> so in the CRE there's base, num HD and HD?
[12:42:34] <fuzzie> not understanding what you're asking
[12:43:01] <fuzzie> the CRE has the HP and the CON, and then there's a table where you look up the con bonus/penalty and modify the HP on load
[12:43:42] <edheldil> fuzzie: the 1HP at level up is after adding the die roll. So if you e.g. have wimpy mage with 1D4 hit die and con mod -2, and you roll 1, you still get 1HP on level up
[12:44:25] <fuzzie> yes
[12:44:29] <fuzzie> i don't get it
[12:44:33] <fuzzie> i mean, this HP is *after* level up
[12:44:59] <fuzzie> and if you do a bunch of level ups and then permanently lose HP, you don't have a minimum HP of your level count
[12:45:48] <wjp> it does seem like losing levels is more painful than never having gained them in the first place
[12:45:52] <fuzzie> and since HP is stored *before* con bonus adjustment, we can't tell anything about what might have happened
[12:46:51] <fuzzie> well, if you have -2 con mod, and you level up with 1HP, you should actually get 3HP stored in the save file, so it's not disasterous on gemrb side?
[12:47:58] <tomprince> Well, so the problem is unmoded npcs, that would be killed with moded con bonus.
[12:48:37] <edheldil> ah
[12:48:40] <fuzzie> yes. and presumably the original engine is doing something less silly. but I don't know what.
[12:50:04] <edheldil> is it possible to find out through the debug interface wwhat HP Prism is in the original game?
[12:50:28] <fuzzie> shouldn't it be in ctrl-m?
[12:52:56] <edheldil> I have never played BG2, so I can't really test :)
[12:54:16] <wjp> in this case BG1, I think
[12:54:27] <fuzzie> it is a modded BGT
[12:54:27] <wjp> just for added fun :-)
[12:54:29] <fuzzie> so, nightmare
[12:56:41] <wjp> but I guess the basic concepts should be testable on most NPCs?
[12:57:53] <fuzzie> yes.
[13:01:17] --> lynxlynxlynx has joined #gemrb
[13:01:18] <-- lynxlynxlynx has left IRC (Changing host)
[13:01:18] --> lynxlynxlynx has joined #gemrb
[13:01:18] --- ChanServ gives channel operator status to lynxlynxlynx
[14:39:40] --> wrotek has joined #gemrb
[14:52:39] <-- wrotek has left IRC (Read error: Connection reset by peer)
[15:05:28] --> wrotek has joined #gemrb
[15:48:12] <-- wrotek has left IRC (Read error: Connection reset by peer)
[16:10:44] --> Yoshimo has joined #gemrb
[16:11:05] <-- SiENcE has left IRC (Quit: @all: cya)
[16:58:45] --> SiENcE has joined #gemrb
[16:58:48] <-- SiENcE has left IRC (Client Quit)
[17:30:50] --> wrotek has joined #gemrb
[17:35:54] <-- harijan has left IRC (Ping timeout: 240 seconds)
[17:44:38] --> Avenger has joined #gemrb
[17:44:38] --- ChanServ gives channel operator status to Avenger
[17:44:49] <Avenger> which game has suicide squirrel?
[17:45:12] <Avenger> tob's squirrel has 9 con which doesn't involve penalty
[17:47:18] <Avenger> ok, i modded it to have 1 con, lets see how this works
[17:49:19] --> harijan has joined #gemrb
[17:51:50] <Avenger> some trick is definitely present, based on some character flags
[17:51:56] <Avenger> not race dependent
[17:56:06] <Avenger> interesting, i set minsc's con to 1, and hp/max hp to 1. He didn't die, just became 1/7 (7 is his level)
[17:56:52] <Avenger> so, there is definitely a level based recalculation that sets the max hp to the minimum possible if it is less than the minimum possible
[17:57:30] <Avenger> and i think this is the answer to why squirrels don't die :)
[18:07:28] <Avenger> the current hp is also recalculated if it is >0 but would result in death after penalties.
[18:09:43] --> brad_a has joined #gemrb
[18:13:00] <-- Avenger has left IRC (Quit: bye!)
[18:13:44] --> kingron has joined #gemrb
[18:21:43] <-- kingron has left IRC (Ping timeout: 276 seconds)
[18:21:56] --> kingron has joined #gemrb
[18:46:47] <-- wrotek has left IRC (Read error: Connection reset by peer)
[19:16:13] <-- harijan has left IRC (Ping timeout: 252 seconds)
[19:18:17] --> harijan has joined #gemrb
[19:33:51] --> SiENcE has joined #gemrb
[21:12:09] <-- Yoshimo has left IRC (Quit: Yoshimo)
[21:47:07] <-- kingron has left IRC (Ping timeout: 276 seconds)
[22:41:29] <lynxlynxlynx> http://sprunge.us/EdQX?diff <-- any objections?
[23:22:51] <-- SiENcE has left IRC (Quit: cya)
[23:28:50] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[23:36:46] --> Gekz has joined #gemrb