00:00:00 awesome :D 00:04:41 -!- MarvinPA-2 is now known as MarvinPA 00:48:49 -!- murphy_slaw_ has quit [Quit: Colloquy for iPhone - http://colloquy.mobi] 00:49:02 -!- Cryp71c has joined ##crawl-dev 00:49:07 Hello hello 00:49:26 karmatic: It's called portal projectile. 00:54:43 -!- murphy_slaw_ has joined ##crawl-dev 00:54:48 -!- murphy_slaw has quit [Read error: Connection reset by peer] 00:54:49 -!- murphy_slaw_ is now known as murphy_slaw 01:09:39 oh really?\ 01:57:38 -!- Spads has quit [Ping timeout: 256 seconds] 02:03:13 -!- karmatic has quit [Ping timeout: 260 seconds] 02:12:05 White draconians don't get frost resistance (http://crawl.develz.org/mantis/view.php?id=1028) by ______ 02:26:16 :/ 02:26:18 ______ 02:29:43 sorear! <3 02:30:19 I am <3 worthy today? 02:30:47 probably? 02:30:52 did you do something bad? 02:31:13 not really, but I haven't done anything really good either 02:31:55 s'ok 02:32:00 as long as you didn't screw up, I <3 you 02:55:37 -!- Cryp71c has quit [Read error: Connection reset by peer] 02:57:46 -!- raydarken has quit [Read error: Connection reset by peer] 03:11:10 -!- Spads has joined ##crawl-dev 03:17:53 -!- [1]GrimmSweeper has joined ##crawl-dev 03:17:53 -!- GrimmSweeper has quit [Disconnected by services] 03:17:55 -!- [1]GrimmSweeper is now known as GrimmSweeper 03:49:53 -!- purge has joined ##crawl-dev 03:51:45 -!- ais523 has joined ##crawl-dev 03:56:24 -!- by has joined ##crawl-dev 03:56:34 good morning 03:57:31 03by * r470eb601b659 10/crawl-ref/source/l_food.cc: Add lua wrapper for is_fruit. 03:57:31 03by * raa06484de6a8 10/crawl-ref/source/ (items.cc stash.cc): Revert "Add ability to ^f and set annotations for fruit" 03:57:32 03by * ra5304f29b2e3 10/crawl-ref/source/dat/lua/stash.lua: Add ability to ^f and set annotations for fruit via Lua. 04:03:08 keithburgun: I love the look on the 100 Rogues youtube Crusader preview. :) Instant wannaplay. 04:04:48 @whereis !78291 04:04:48 78291 the Brawler (L16 DrIE), a worshipper of Jiyva, saved on Vault:2 on 2010-03-08 after 43811 turns. 04:15:07 03by * r30fb3bd0042e 10/crawl-ref/source/religion.cc: Revert Jiyva god gift commit that slipped into a different one. 04:15:08 03by * r60b9351a1e25 10/crawl-ref/source/religion.cc: Fix Jiyva prayed_for check in do_god_gift. 04:16:01 !tell doy Please pick 60b9e51a for 0.6; that part of the Jiyva mutations change slipped into the wrong commit. 04:16:01 by: OK, I'll let doy know. 04:17:04 !tell N78291 Found the problem with Jiyva mutations 0.6: the gifting change was merged to 0.6 incompletely because I hid part of it in an unrelated commit. Thanks! 04:17:04 by: OK, I'll let N78291 know. 04:40:37 -!- ais523 has quit [Remote host closed the connection] 04:47:17 http://crawl.develz.org/morgues/trunk/rob/rob.map 04:47:26 anyone know which layout this is? it looks faintly buggy 04:47:56 is that trapped dead-end room at the bottom right a vault? 04:49:09 -!- dpeg has joined ##crawl-dev 04:50:01 -!- ais523 has joined ##crawl-dev 04:50:06 dpeg: hi; http://crawl.develz.org/morgues/trunk/rob/rob.map 04:50:24 it looks a little bit buggy; do you know if that's a vault at the bottom right? 04:57:52 looks like it 04:57:52 dpeg: You have 1 message. Use !messages to read it. 04:57:56 are the walls rock? 04:58:20 dpeg: yes, I think so; level is magic mapped 04:58:48 could be an altar vault 04:59:16 -!- jld has quit [Read error: Operation timed out] 05:29:47 Moin! 05:29:57 unpacked crawl-ref git repository? 05:29:58 > du -sk crawl-ref.git/ 05:29:58 1568064 crawl-ref.git/ 05:30:28 hi Napkin 05:30:38 Hey Rob :) 05:30:39 that must be what caused lag a bit ago? :) 05:30:55 yeah.. a corruption in a pack.. had to unpack everything 05:35:05 Napkin: Hallo! 05:35:22 and lag again.. sorry guys 05:35:26 hey David :) 05:36:48 lunchtime! 05:41:14 Mahlzeit! 05:42:52 -!- Textmode has joined ##crawl-dev 05:52:43 -!- Mu_ has joined ##crawl-dev 05:54:36 morning all 05:57:56 I did not succeed. 05:58:26 Have I got time for final loot balancing? 05:58:54 03kilobyte * r40663473c6f8 10/crawl-ref/source/mon-project.cc: Fix IOODs disappearing above submerged monsters. 06:00:54 -!- MarvinPA has quit [Ping timeout: 245 seconds] 06:10:39 -!- MarvinPA has joined ##crawl-dev 06:13:23 was that corrupted file fixed? 06:17:51 due: just do it? doy can decide whether it should be merged 06:18:22 anything need doing with troves, etc. code-wise? 06:20:31 I think they're done. 06:20:37 I just kinda lost it today. 06:20:43 yay, ? 06:20:50 The bad kinda lost it :) 06:25:53 -!- felirx has quit [Quit: leaving] 06:26:15 -!- syllogism has joined ##crawl-dev 06:30:44 !tell Keskitalo Thanks for your reply; please have a look at http://crawl.develz.org/wiki/doku.php?id=dcss:brainstorm:interface:menu and http://crawl.develz.org/wiki/doku.php?id=dcss:brainstorm:interface:combochoice 06:30:44 dpeg: OK, I'll let Keskitalo know. 06:32:14 due: what did you lose? 06:36:33 http://www.dict.cc/englisch-deutsch/to+lose+it.html 06:37:08 by: danke :) 06:41:19 -!- Zaba_ has quit [Ping timeout: 240 seconds] 06:45:35 -!- Zaba has joined ##crawl-dev 06:48:42 was feeding by vampiric weapons meant to stay with the hunger change? 06:49:47 -!- MarvinPA-2 has joined ##crawl-dev 06:51:03 by: this has never been discussed, as far as I know :) 06:51:17 I find myself eating quite rarely; but it's probably not really an issue since where you get food through the weapon, there's also plenty of chunks around 06:51:44 -!- MarvinPA has quit [Ping timeout: 245 seconds] 06:55:40 03kilobyte * r71bc66cc1926 10/crawl-ref/source/dat/descript/spells.txt: Mention that Poison Weapon doesn't work on blunt stuff. 06:55:41 03kilobyte * r933fa59c0d73 10/crawl-ref/source/dat/descript/spells.txt: Don't lie that Ozocubu's Armour "protects from cold". 06:59:32 by: yes, also it's a compensation for the wield restriction 07:50:53 -!- MaryPoppins is now known as Adeon 07:59:04 03by * r4e9f819f47c6 10/crawl-ref/source/player.cc: Simplify tiles code in forget_map. 08:11:22 03kilobyte * r2c0449dd6702 10/crawl-ref/source/spl-data.h: Lower the levels of Poisonous and Freezing Cloud. [Mantis 166] 08:46:10 -!- MarvinPA-2 has quit [Ping timeout: 265 seconds] 08:46:46 -!- MarvinPA has joined ##crawl-dev 08:52:04 -!- eith has joined ##crawl-dev 09:13:32 03by * r0fcc8160b5a7 10/crawl-ref/source/player.cc: Restore forget_map minimap update. 09:13:35 03by * rddf5071a81ab 10/crawl-ref/source/ (show.cc showsymb.cc showsymb.h): Make get_mons_glyph return real colour by default. 09:13:55 !tell doy 0fcc8160 belongs with 4e9f819f 09:13:56 by: OK, I'll let doy know. 09:16:19 -!- Cryp71c has joined ##crawl-dev 09:20:55 dpeg, finally got the 'Approved Muts' up 09:24:47 Cryp71c: yes, I saw. 09:25:02 Wanted to ask you about it here, but never met you :) Many thanks! 09:25:09 Cryp71c: Should I react in some way? 09:26:11 dpeg, lol that sounded almost pretentious :P No, I'm building the list from what I know, if you wanted to inform me of what was and wasn't approved to be implemented, I'll copy / paste it into the new section. 09:26:45 dpeg, hrm, pretentious wasn't what I was looking for...I can't think of the word though.. 09:27:02 Cryp71c: okay, so I won't make "yes, that's fine" comments, but only "perhaps not this one yet". Okay? 09:27:33 Thats fine, whatever is most convenient for you 09:28:13 Obviously the Fire / Ice ones are going in as completed, all the notes under your ideas are appropriately flagged as approved or approved for removal 09:29:13 hey guys 09:29:21 Hi! 09:29:28 -!- [1]GrimmSweeper has joined ##crawl-dev 09:29:28 -!- GrimmSweeper has quit [Disconnected by services] 09:29:29 -!- [1]GrimmSweeper is now known as GrimmSweeper 09:29:57 how would one add a script to start "crawl.exe -sprint" in windows nowadays? .bat? general .lnk? 09:30:12 one thing I was going to mention was that Fire II's Level 3 still seems a bit weak (as its described on the wiki, anyways) 09:30:12 Napkin: .lnk probably 09:30:12 doy: You have 2 messages. Use !messages to read them. 09:30:16 !messages 09:30:16 (1/2) by said (5h 14m 15s ago): Please pick 60b9e51a for 0.6; that part of the Jiyva mutations change slipped into the wrong commit. 09:30:18 !message 09:30:19 !messages 09:30:19 (1/1) by said (16m 23s ago): 0fcc8160 belongs with 4e9f819f 09:30:37 Napkin: you can just add -sprint to the command line for the .lnk file 09:31:11 but doesn't .lnk need absolute path 09:31:13 ? 09:31:50 -!- raydarken has joined ##crawl-dev 09:31:53 -!- raydarken has quit [Client Quit] 09:32:07 shrug! 09:32:10 you add it in the properities, in the target box 09:32:28 with relative pathes, eith? 09:32:32 .bat is just kinda messy 09:32:45 absolute id guess 09:32:55 i don't see a different way 09:33:10 really, we just need to get a startup screen to let you choose the game type 09:33:32 doy: will come 09:33:40 but we need something else until then 09:33:57 Sprint is not part of 0.6, right? 09:34:00 no 09:34:14 Napkin: the menu will come rather soon. I wouldn't worry about it. 09:34:58 "Windows does not support relative paths in shortcuts." 09:35:23 nah, it should be there - especially since we got asked already in the comments of the blog post 09:35:42 someone already described how to make a shortcut for it in the somethingawful thread 09:35:52 lol 09:36:37 we shouldn't link them to that 09:37:24 a place that's full of adverts and requires registration? 09:37:42 but yes, you can just put arguments in the same field as the program name, on Windows 09:37:56 http://forums.somethingawful.com/showthread.php?threadid=3155652&userid=0&perpage=40&pagenumber=170#post373457494 09:38:39 yes, ais523 - but lnk don't work with relative paths 09:38:47 argh, my eyes! my beautiful eyes! 09:38:59 Napkin: ah, I was assuming there'd be an existing .lnk file to copy and modify 09:39:33 * dpeg pokes bhaak in the eye. 09:40:12 making fun of a blind man! 09:40:32 * dpeg pokes bhaak in the other eye, and makes fun of the blind man. 09:41:20 -!- purge has quit [Ping timeout: 258 seconds] 09:43:12 it looks like weapon acquirement gives almost only 2-handers unless you have a great shield apt 09:43:43 kilobyte: is that new or old behaviour? 09:44:08 that's been around for a while at least 09:44:10 not sure how long 09:44:20 it's because it tries to give you a good quality weapon 09:44:30 and good quality weapons are almost all two-handers 09:45:04 kilobyte: why does it like at shields aptitude rather than skill? 09:45:12 this has the side effect of making races that can't wield two-handers get almost no axes or long blades or maces/flails from acquirement 09:45:21 no, 1 handers get damage weighted 1.5* 09:45:44 I don't see any problems for MD or Mi 09:46:05 the bug is in shield weighting code I introduced 09:46:24 -!- Siber has joined ##crawl-dev 09:46:34 kilobyte: can you repair it? 09:46:35 odds are: SK_SHIELDS + 10 : xl - SK_SHIELDS + 10 09:46:39 Or does it need discussion? 09:47:10 to make me look worse, dpeg did warn that xl shouldn't be used there :p 09:47:53 kilobyte: true! 09:48:26 I just gathered some data: http://sprunge.us/fXfH -- and it looks that both casters and fighters, both good and bad with shields would be satisfied by: 09:48:47 shields*apt + 10 : best_weapon*apt + 10 09:49:01 or perhaps: 09:49:06 shields*apt + 10 : best_weapon*apt - shields*apt + 10 09:49:33 I can see why one wants to sue aptitudes (and these are sane to use, in contrast with XL), but shouldn't skill matter, too? 09:50:43 "shields*apt" -- I mean skill multiplied by apt 09:51:51 ah, that's good 09:52:22 although somebody who got high shields skill despite a terrible aptitude may feel cheated :) 09:53:24 terrible aptitude has a higher number 09:53:45 ok, you convinced me 09:54:17 or actually, chars with bad apts on the data I posted at sprunge have: 7sh 18m&f (both bad), 9 sh 27 lb (bad sh) and so on 09:54:53 I don't understand the table :O 09:55:20 I don't get it, quite... why would all chars with bad aptitudes have low skill at shields even if they use exclusively a 1-hander for which they have as bad aptitude? 09:55:45 it's: race xl handedness skills 09:56:22 1s = 1-hander with a shield, 1- = 1-hander but not shield on when died/won, 2 = 2-hander 09:57:41 kilobyte: maybe they all turned off shield skill? 09:58:52 -!- Adeon is now known as MaryPoppins 10:00:05 shields are quite good even with low skill (at least by displayed SH value) 10:01:46 no, I looked at these 10:02:17 Summoned Butterflies stop auto-travel (http://crawl.develz.org/mantis/view.php?id=1029) by GrimmSweeper 10:02:31 and my recent spriggan with 7sh 18m&f tried to victory dance shields as much as possible 10:03:09 Shields train too slow? 10:03:21 !apt shields 10:03:22 Shields: Dr=100, Dr[black]=100, Ce=180*, DD=90, DE=140, DG=110, DS=110, Gh=110, Dr[green]=100, Dr[grey]=100, Ha=130, HE=110, HO=80, Hu=100, Ke=100, Ko=130, Mf=100, Mi=80, Dr[mottled]=100, MD=70!, Mu=140, Na=140, Og=120, Dr[pale]=100, Dr[purple]=100, Dr[red]=100, SE=130, Sp=180*, Tr=150, Vp=110, Dr[white]=100, Dr[yellow]=100 10:06:12 look at this: MDs with apts Sh=70 M&F=70 get shield skills just below weapon, Spriggans with Sh=180 M&F=160 less than one third 10:07:35 kilobyte: because Spriggans evade so much, whereas Dwarves (in heavy armour) block more? 10:07:57 Are the checks done like this? (EV first, SH next, AC last) 10:09:55 SH first, EV next 10:10:38 bucklers block less 10:10:48 shield-using centaurs seem to have the same problem 10:11:18 does it get trained only on a successful block 10:14:10 is the repo clear of currupted files again? is it safe to fetch? 10:14:17 I made new formulas for Armour and Dodging training. (In a nutshell, train whenever you block/evade, with chance depending on skill vs damage.) 10:15:52 -!- [1]GrimmSweeper has joined ##crawl-dev 10:15:52 -!- GrimmSweeper has quit [Disconnected by services] 10:15:54 -!- [1]GrimmSweeper is now known as GrimmSweeper 10:17:36 ok, windows builds from master branch now have a crawl-sprint.bat inside them 10:17:48 cool! 10:17:50 off home - cheers! 10:17:52 o/ 10:18:01 \o 10:18:06 :D 10:18:06 Napkin: Thanks! 10:18:18 That is awesome. 10:18:21 Keskitalo: Ha! You exist! 10:18:26 btw, I actually like Sprint, and have been playing it 10:18:37 we all do 10:18:50 dpeg: it's more surprising in my case, due to not really liking Crawl 10:19:13 Sprint runs opposite to its design philosophies in many ways, though 10:19:21 like promoting jelly farming, and the general lack of randomness 10:19:44 by the way, is getting teleported by Sonja's distortion weapon a bug? IIRC, the level's generally no-teleport 10:20:18 teleportitis works too 10:20:22 dpeg: Only a little. :> Thanks for the wiki edits! 10:20:24 and teletraps 10:20:49 tele traps shouldn't exist 10:20:56 -!- ais523 has quit [Remote host closed the connection] 10:21:02 doy: yes, Chapayev is aware of that. 10:21:16 lucky teleports to the Orb happen often, and once there you can unwield distortion (kindly provided by Sonja 50% of the time) 10:21:58 Rob's idea of walls/statues/signs with hints about treasure in Cerebov's chamber is great (instead of secret doors). 10:22:24 I can try to make some. 10:26:13 That would be great! 10:31:50 -!- karmatic has joined ##crawl-dev 10:40:36 03kilobyte * r79cfc3cb1197 10/crawl-ref/source/effects.cc: Fix acquirement giving nearly solely 2-handers to those without a great Shields apt. 10:40:57 Keskitalo: thanks 10:41:50 Keskitalo <--> kilobyte 10:41:55 also, the xl is gone :) 10:42:00 -!- jld has joined ##crawl-dev 10:42:18 a high-level char with skills indicating he's a 100% shield user got 12 1-handers 3 2-handers, one with no shield skill but high weapon got 2 1-handers 8 2-handers. 10:42:27 Praying to Trog when standing on items - generates wrong message (http://crawl.develz.org/mantis/view.php?id=1030) by ______ 10:42:54 kilobyte: very good! 10:44:54 !seen Chapayev 10:44:54 I last saw Chapayev at Thu Mar 4 01:15:11 2010 UTC (4d 15h 29m 43s ago) quitting with message Client Quit. 10:47:36 !tell Chapayev Teleportitis/rings of teleport give you a good chance of putting you near the Orb. That's abusable (even if getting back is risky: teleport back or Abyss yourself). What about plain out disabling all long range teleports? 10:47:36 kilobyte: OK, I'll let Chapayev know. 10:50:35 hey dpeg, I'm responding to your lovely email now 10:52:41 stone of earth elementals -- lowering the skill by 5 (or more) for every friendly elemental on level would be an easy fix to abuse, at least until evokables get overhauled 11:01:15 keithburgun: oh, you are here. Nice! 11:01:38 heh yessir. concocting a big ol' email for yas 11:01:44 thanks! 11:02:05 It's really great to be able to talk to you guys. It's quite like meeting a rock star or other personal hero for me 11:02:06 kilobyte: true. The stone nerf did not seem as urgent (as the staff of earth nerf, say) b/c it didn't seem to catch on. 11:02:21 With the staff, we had players planning for that by turn 1. 11:02:52 keithburgun: I don't know about the others, but given abilities and approach, it looks more like crusty punk than rock stars, at least to me. 11:03:39 unprofessional, DIY, devotion :) 11:04:07 Yeah. 11:04:43 Then again, Keith is getting emails from you guys. 11:04:53 hehe, well, that may be the case, either way I'm not worthy. 11:04:53 I'm just sitting here lurking on IRC. 11:05:29 I've contributed one idea, that's somewhat minor, but it was considered, and that makes me happy. 11:07:35 for Cerebov's loot, we can use glass renamed and recoloured to "iron grate". That would have a side effect of letting Cerebov firestorm and balrugs smite you when inside. 11:07:42 Speaking of whitch, how is Fred doing? 11:08:00 kilobyte: would that be hard to do? 11:09:52 dpeg: I have one idea also actually, which I didn't put in my email. One issue which bothers me a bit as a designer is how similar some classes are to each other (a gladiator is a fighter with nets?) Why not take out some of the copycat classes, or failing that, how about 'grouping' the classes into groups like 'fighters' 'mages' 'sneakies' etc? 11:10:35 gladiators are ev fighters, fighters are heavy armour fighters 11:10:48 keithburgun: it is true that we never really dusted off the classes (we call them "backgrounds" now). One reason is that species are much more important and we try to differentiate them -- if need be, by removal. 11:11:22 We will address the classes, hopefully for 0.7. 11:12:46 keithburgun: the dev community here is really nice, compared to other OSS projects i've been a part of. 11:13:00 They all embraced what I wanted to do with little belligerence :D 11:13:01 ixtli: I am curious. What projects and how do they do it? 11:13:10 I see. So it's something you plan on getting to eventually? 11:13:25 yes 11:13:29 dpeg: Well I was part of the first iPhone Dev Team, right? 11:13:40 dpeg: It was ok until people started to hear about it. 11:13:48 keithburgun: given the facts that Crawl is Big, Old and Free, nothing really is very fast (by your standards, at least :) 11:13:57 Then people became paranoid and obsessed with getting "credit" for things. 11:14:05 ixtli: how big is the team? 11:14:11 I would think that's essential. 11:14:12 And they got overly protective of their source code and letting others see pre-releases. 11:14:23 ixtli: and that's the death spell :) 11:14:24 dpeg: It was first 5, then went to 10 core devs. 11:14:37 10 devs and already jealousy? I am impressed :) 11:14:38 It was terrible. I hate hearing about george hotts. 11:14:49 -!- dpeg is now known as GeorgeHotts 11:14:53 hahahaha 11:14:54 <3 11:15:06 ixtli: Your code was so ugly, I had to redo every single line. 11:15:12 -!- GeorgeHotts is now known as dpeg 11:15:19 :) 11:15:24 actually, thats was why /i/ joined them. 11:15:28 because I was the only programmer. 11:15:32 hehe 11:15:37 The rest of them were engineers of other kinds. 11:15:39 ixtli: did the project fail eventually? 11:15:55 nope, it blew up. i mean 11:16:06 http://blog.iphone-dev.org/ 11:16:13 but they're almost totally different people now 11:16:24 geohot is on his own now. 11:17:30 The information, statements, articles, and releases here are official. They have the approval of the combined Dev/Elite teams. 11:17:33 dpeg: http://code.google.com/p/iphuc/ <-- that was the source i wrote. 11:18:05 dpeg: The people at the beginning were professionals. The guy 'nightwatch' credited in the source did most of the reverse engineering. 11:18:13 I wouldn't want to be called "Dev/Elite". 11:18:13 Apparently he co-wrote the TIF exploit on the PSP. 11:18:17 xD 11:18:27 ixtli: is that stuff legal? 11:18:35 dpeg: No one has gotten sued yet. 11:18:56 :D 11:19:14 the supreme court in the US seems to take its cues from the FCC 11:19:34 and the FCC has said pretty straight out that they want the owner of the device to be able to use his or her device in whatever way he or she chooses. 11:19:42 In this way, I am in love with the obama administration. 11:19:52 They moved the FCC out of the 70s. 11:20:02 dpeg, ...muwahahaha, I just had a brilliant idea! Crawl Application for the Android OS 11:20:20 Cryp71c: Should be easy once my ipad branch gets added to master :) 11:20:32 dpeg: Also, it stands that if apple was to sue little college student ixtli it would not look good for them. 11:20:41 what are the performance requirements for crawl? 11:20:44 yes, one would hope so 11:20:51 Textmode, a keyboard. processor optional 11:21:04 Textmode: a lot, unfortunately 11:21:15 Textmode, wait, you talking about a server or? 11:21:17 Cryp71c: I'm thinking of porting to a portable, going to need details. 11:21:18 you'd be surprised though. it's not doing a lot in the way of drawing. 11:21:23 all the stashing and travel and LOS seems to be resource hogs 11:21:35 Textmode: im porting to the ipad :D 11:21:52 yeah, crawl does a surprising amount of calculations each turn 11:21:53 ixtli: dingux, gods willing. 11:22:03 Textmode, it runs on my Droid just fine 11:22:06 via SSH 11:22:17 but thats just rendering, obviously. 11:22:18 stashing isn't slow, but is a memory hog (as far as i'm aware) 11:22:21 Textmode: nice. here, clone my branch 11:22:25 the dingoo isn't really all that powerful, though. 11:22:28 Textmode: http://github.com/ixtli/crawl/tree/ipad 11:22:33 but monster ai and los calculations are relatively slow 11:22:40 Textmode, regardless, it will play..its just how well it would play :P 11:22:52 Textmode: the files i've added modularize drawing code and library-specific ui interaction 11:22:56 to make porting easier. 11:22:59 ixtli, what changes are you making in the ipad branch? 11:23:04 I'm more worried about ram. 11:23:05 ah, nvm. 11:23:06 Textmode: criticism greatly appreciated. 11:23:23 Cryp71c: I modularized dependency on opengl and sdl 11:23:27 Cryp71c: and sdl_image 11:23:30 I think I heard some numbers brandied about this channel that simply won't fit on the dingoo... 11:23:49 Textmode: such as? 11:23:50 Textmode: Can you link me to stats for the device? 11:23:55 ixtli: bookmarked, I'll take a closer look later, if I decide its viable... 11:24:00 Cryp71c: And I changed the way drawing primitives works a bit 11:24:17 http://en.wikipedia.org/wiki/Dingoo#Specifications 11:24:41 RAM 32MB at 133MHz 11:24:47 yeah, that's going to be a problem 11:24:48 yeah that's pretty low 11:24:55 yeah... 11:25:00 you might have to disable stash tracking entirely 11:25:05 to get that to work 11:25:49 doy: would it be suffcient to have it only track "flagged" stashes, or is the entire thing just too much of a hog? 11:25:55 doy: allowing manual stashign might work? 11:26:05 sure, manual stashing might be okay 11:26:12 dpeg: yeah, thats the word I was looking for. 11:26:22 does that code still exist? 11:26:25 or did we get rid of it? 11:26:27 Textmode: that thing looks cool 11:26:55 Ok ok. 11:26:59 doy: I think we got rid of it, but it lives in the history. 11:27:04 ixtli: yeah, not the most powerful, but the price is decent and does the job. 11:27:06 This might sound crazy, but how about writing stashes to flash memory? 11:27:13 :D 11:27:38 ixtli: not really workable, you'd have to load it all into memory every time you searched anyway 11:27:41 Textmode: does it have a graphics processor? 11:27:44 ixtli: I was thinking something similar, page mapping to the "drive" 11:27:57 doy: Is that really going take > 16mb of ram? 11:27:59 ... guess it could ... 11:28:01 Hmm. 11:28:01 ixtli: no, no acceleration, if thats what you mean. 11:28:13 or rather, i guess it's doable, but would be a pretty big pain 11:28:23 Textmode: Yeah, cause you have to write out ... at least 8mb of the 32 for tiles graphics. 11:28:32 Unless you want only text mode. 11:28:41 doy: Yeah. 11:28:53 doy: Is crawl capable of profiling its own memory usage 11:28:54 ? 11:29:02 I can do it right now using instruments on os x 11:29:07 actually, hold on. 11:30:07 ixtli: larger issue there, though, is that dingux still has an outstanding bug where drive write occassionally freak out and write garbage. its rare, but if it happens while updating the allocation table... 11:31:07 hahah oh boy 11:31:40 yeah, thats a fun one... 11:31:54 Textmode: Another crazy idea. 11:32:03 Run it in a garbage collected environment. 11:32:06 ;-) 11:32:37 An advanced garbage collector would allow it to at least run. 11:32:41 For you to see. 11:32:46 No matter how much ram you do/dont have. 11:33:14 not if the live set gets too big 11:33:58 True. 11:34:05 But do you really think it would? 11:34:14 * ixtli profiles crawl. 11:34:15 gimme a sec 11:38:04 if worset comes to worst, I'll just have it locked in "sprint" mode :P 11:38:09 that should fit, right 11:38:11 ? 11:38:19 sure 11:38:59 -!- nrook has joined ##crawl-dev 11:41:40 it used 41 mb of real ram after running it for a bit, tile enabled at 1440x900 11:41:54 I think that if you were actually running out of ram after some tuning 11:42:16 and couldn't figure out how to proceed 11:42:22 you could actually use a garbage collector 11:42:35 Textmode: Do they build java for that platform. 11:42:57 -!- murphy_slaw has quit [Quit: murphy_slaw] 11:43:30 not that I know of, but baring memory limitations, I see no reason I couldn't compile a jvm for it. 11:43:59 though using a java runtime on a memory constrained device is not my first choice :P 11:44:15 i'm not sure how java would help here anyway 11:44:26 Using ASCII would help, no? 11:44:43 yeah, although i'm not sure by how much 11:45:07 i don't think the tiles themselves use all that much memory 11:45:16 doy: I think if all the CDO/CAO games came with 40 MB usage, we'd gotten more complaints. 11:45:19 doy: It's just about whether or not such things have been attempted. 11:45:23 There are good gcs for c++ 11:45:46 ixtli: yes, i'm aware, but they have basically nothing to do with java 11:45:48 I do prefer ascii, but but I'm not sure you could get 24x80 in a readble font... 11:46:11 you need less 11:46:28 although that will involve a lot of manual labour 11:46:43 doy: ? 11:46:44 ho hum. 11:46:53 doy: I thought you could compile jvm bytecode from c++ apps 11:46:59 s/apps/source 11:47:00 no? 11:47:05 ixtli: i've never heard of that 11:47:47 doy: http://www.axiomsol.com/pro_serv/compiler.php 11:47:51 for instance. 11:48:03 dpeg: looks like crawl processes on cdo are using 20-30mb each 11:48:07 My university does this sort of research, so I know at least /someone/ is working for it. 11:48:09 for some value of "each" 11:48:14 doy: oh 11:48:20 ixtli: its a C compiler that targets the jvm 11:48:56 which is highly unsupported, but they mostly get away with it since noone really wants to break bytecode compatibility 11:49:02 :D 11:49:21 i mean, it's probably more work than it's worth. 11:49:35 but you could always try it for kicks :D 11:49:57 doy: so we can semi-safely assume that 30mb is an "average" game? 11:50:16 Textmode: yeah, probably 11:50:43 curious to see what a 15-rune game ends up at 11:50:48 so if I tighten crawls belt, this should work. 11:53:01 Textmode: What I'm most interested in is how you would make it work with four buttons. 11:53:15 If you were to do this, you might open up the way for a PSP/DS port which would be huge. 11:53:26