Go to the bottom | Channel Index

first << < 33 34 35 36 37 38 39 40 41 42 43 > >> last
Tue Jun 23 13:06:29 2009 [dgd] Subversion@Way of the Force: Nullinfinite committed Gurbalib revision 293 to svn://wotf.org/gurbalib : Player input not matching command or verb will result in a random error message instead of "What?".  See /daemons/cmd_error_d.c  
Sat Jun 27 20:40:04 2009 [dgd] Cratylus@Dead Souls Dev: outta frickin nowhere, aidil 
Sat Jun 27 20:40:12 2009 [dgd] Cratylus@Dead Souls Dev: nice shot 
Sat Jun 27 20:54:11 2009 [dgd] Aidil@Way of the Force: hmm? :) 
Sat Jun 27 20:54:32 2009 [dgd] Cratylus@Dead Souls Dev: i liked the hawking thing 
Sat Jun 27 20:54:36 2009 [dgd] Aidil@Way of the Force: the Steven Hawking one? had been waiting to make that point for a little while :) 
Mon Jun 29 19:09:30 2009 [dgd] Subversion@Way of the Force: Aidil committed Gurbalib revision 294 to svn://wotf.org/gurbalib : Slightly reformatted the auto-game object, fixed a bug that broke compatibility with 'vanilla' dgd, added preliminary support for kernel maintained linked lists (not enabled yet)  
Mon Jun 29 20:22:06 2009 [dgd] Cratylus@Dead Souls Dev: :) "That is in theory also roughly what communism is about, but it doesn't work that way. " 
Mon Jun 29 20:35:29 2009 [dgd] Aidil@GurbaDev2: hehe 
Mon Jun 29 20:35:54 2009 [dgd] Aidil@GurbaDev2: I'm somewhat entertaining myself with that thread :) 
Tue Jun 30 11:24:37 2009 [dgd] Aidil@Way of the Force snores. 
Tue Jun 30 11:24:44 2009 [dgd] Aidil@Way of the Force: its way too warm here. 
Tue Jun 30 13:58:33 2009 [dgd] Kalinash@Fire and Ice: move to england 
Tue Jun 30 14:04:45 2009 [dgd] Silenus@Dead Souls Dev: move to Hong Kong! 
Tue Jun 30 14:04:50 2009 [dgd] Silenus@Dead Souls Dev: em.... 
Tue Jun 30 14:07:00 2009 [dgd] Silenus@Dead Souls Dev: btw I think I am mildly insane but I may be actually making progress with this for a change. 
Tue Jun 30 14:09:22 2009 [dgd] Silenus@Dead Souls Dev: i guess hiding in heavily air conditioned flats doesnt make HK too bad 
Tue Jun 30 18:48:20 2009 [dgd] Aidil@The Zone: both sound like a possibility. 
Tue Jun 30 18:48:33 2009 [dgd] Aidil@The Zone: (mildly insane and making progress) 
Tue Jun 30 18:48:38 2009 [dgd] Quixadhal@Bloodlines: Our high temperature today is supposed to be 64F, last week it ws 96F.  Move to Michigan!  Then you can enjoy every kind of weather with no job to hinder said enjoyment. :) 
Tue Jun 30 18:48:42 2009 [dgd] Aidil@The Zone: better still, they often go together. 
Tue Jun 30 18:48:57 2009 [dgd] Aidil@The Zone: heh. 
Tue Jun 30 18:49:12 2009 [dgd] Aidil@The Zone: I can hide most of the day in a noisy but cool datacenter :P 
Tue Jun 30 18:49:28 2009 [dgd] Aidil@The Zone: just have to think up an excuse to go fiddle with hardware :) 
Tue Jun 30 18:51:22 2009 [dgd] Aidil@The Zone: hmm. any of you played with the security stuff in gurbalib yet? 
Tue Jun 30 19:00:05 2009 [dgd] Silenus@Dead Souls Dev: hmm not yet. I just got back from vacation and was getting over jetlag while implementing some of my driver design ideas. 
Tue Jun 30 19:00:30 2009 [dgd] Aidil@Way of the Force: hmm. :P interesting way for getting over your jetlag. 
Tue Jun 30 19:02:39 2009 [dgd] Silenus@Dead Souls Dev: maybe i should upgrade to definitely insane :D 
Tue Jun 30 19:06:44 2009 [dgd] Quixadhal@Bloodlines: I don't think I can use the security system, as I don't trust myself. *grin* 
Tue Jun 30 19:09:21 2009 [dgd] Aidil@Way of the Force: thats just a very good reason to use it :) 
Tue Jun 30 19:10:05 2009 [dgd] Aidil@Way of the Force: but heh, if you updated to the latest svn version, it will be enabled by default 
Tue Jun 30 19:11:19 2009 [dgd] Quixadhal@Bloodlines: I guess I'm using it then.... oh, and I actually remembered to give the statedump parameter this time, so it restored state.  I was told there'd be punch and pie? 
Tue Jun 30 19:14:15 2009 [dgd] Aidil@Way of the Force: heh. 
Tue Jun 30 19:15:06 2009 [dgd] Aidil@Way of the Force: did you do a warmboot? 
Tue Jun 30 19:15:57 2009 [dgd] Aidil@Way of the Force: if so, there was no reason to stop/start the driver for this:) 
Tue Jun 30 19:17:28 2009 [dgd] Quixadhal@Bloodlines: Nah, warmboots are for winter.  Besides, my task scheduler needed something to do. 
Tue Jun 30 19:17:49 2009 [dgd] Quixadhal@Bloodlines: Hehehehe, check this out.... http://news.bbc.co.uk/2/hi/uk_news/magazine/8117619.stm 
Tue Jun 30 19:25:55 2009 [dgd] Aidil@Way of the Force: :) 
Tue Jun 30 19:26:03 2009 [dgd] Aidil@Way of the Force: I still have one :) 
Tue Jun 30 19:28:00 2009 [dgd] Quixadhal@Bloodlines: I will admit to being tempted to sell my vinyl Thriller album on ebay... there's one with 12 bids without a jacket going for $51 at the moment. 
Tue Jun 30 19:29:24 2009 [dgd] Aidil@Way of the Force: heh. 
Tue Jun 30 19:29:32 2009 [dgd] Quixadhal@Bloodlines: Wow... $130 for an unused victory tour ticket.  What, do people think they can use it now? 
Tue Jun 30 19:29:50 2009 [dgd] Aidil@Way of the Force: I have 2 copiesof it, but, they are meant formixing, so won't sell it :) 
Tue Jun 30 19:30:06 2009 [dgd] Aidil@Way of the Force: hmm. add a few spaces in there. 
Tue Jun 30 19:31:56 2009 [dgd] Quixadhal@Bloodlines: I kept thinking of plugging my turntable into my sound card to rip them to mp3's... but nowadays it's easier to just download them.  That SHOULD BE (but isn't) legal, since I do own the physical media. 
Tue Jun 30 19:37:52 2009 [dgd] Aidil@Way of the Force: heh. 
Tue Jun 30 19:39:23 2009 [dgd] Aidil@Way of the Force: I have 2 of them connected to a nice modern digital mixer/audio player 
Tue Jun 30 19:39:56 2009 [dgd] Aidil@Way of the Force: one of them has a builtin a/d converter and coaxial digital out. 
Tue Jun 30 19:53:20 2009 [dgd] Quixadhal@Bloodlines: Heh, my turntable is part of an ancient entertainment system.  It also has an AM/FM, a cassette, and... wait for it... a recording 8-track. 
Tue Jun 30 19:54:15 2009 [dgd] Quixadhal@Bloodlines: Not an 8 track recorder, mind you... I'm talking the ugly little 8-track looped cartridges that were 4 "programme" stereo. :) 
Tue Jun 30 20:24:54 2009 [dgd] Aidil@Way of the Force: oh.. a recording 8 track is a nice obscure device :) 
Tue Jun 30 20:25:18 2009 [dgd] Aidil@Way of the Force: players are pretty rare outside the USA btw. 
Tue Jun 30 20:27:11 2009 [dgd] Aidil@Way of the Force: I only ever saw pictures of those devices, and the cartridges for them. 
Tue Jun 30 20:27:32 2009 [dgd] Aidil@Way of the Force: I do however have a 8 track recorder, 2 actually. 
Tue Jun 30 20:27:40 2009 [dgd] Aidil@Way of the Force: a revox and an akai. 
Tue Jun 30 20:27:55 2009 [dgd] Aidil@Way of the Force: both in storage, no room for such stuff where I am living now 
Tue Jun 30 22:10:27 2009 [dgd] Quixadhal@Bloodlines: Not only were they super-popular back in the 70's and early 80's, but a friend of mine in college had an 8-track in his car.  I couldn't resist recording Metallica's black album to 8-track so he could play it. :) 
Tue Jun 30 22:11:16 2009 [dgd] Quixadhal@Bloodlines: But yes, they've gone the way of Betamax here now... in fact, so have VHS tapes.  I'm not even sure it's easy to get blank cassettes anymore. 
Tue Jun 30 22:12:50 2009 [dgd] Raudhrskal@Dead Souls Dev: Haven't looked for audio tapes in years, but VHS cassettes are still available around here. 
Tue Jun 30 22:13:26 2009 [dgd] Raudhrskal@Dead Souls Dev: Fact, it's usually half of shelf next to whole aisle full of modern recordable media, but it's still there. 
Tue Jun 30 22:14:06 2009 [dgd] Tim@TimMUD: local walmart has blank audio tapes and VHS... 1 offering each, with 5 or 6 brands of CD-R and all the various DVD types 
Thu Jul  2 13:42:44 2009 [dgd] Frutsel@fruts: hmm 
Thu Jul  2 13:42:58 2009 [dgd] Raudhrskal@Dead Souls Dev: hmm? 
Thu Jul  2 13:43:03 2009 [dgd] Frutsel@fruts nods. 
Thu Jul  2 13:44:05 2009 [dgd] Silenus@Dead Souls Dev nods in agreement with Frutsel. 
Thu Jul  2 19:04:50 2009 [dgd] Aidil@The Zone lowers his head, thinks deeply, and hmmmms. 
Thu Jul  2 23:01:02 2009 [dgd] Quixadhal@Bloodlines starts to hmmmm, but then thinks of an improvement and has to reboot his Diku to implement it. 
Thu Jul  2 23:53:38 2009 [dgd] Aidil@The Zone: in-place recompilation ftw. 
Thu Jul  2 23:54:18 2009 [dgd] Raudhrskal@Dead Souls Dev: The type we had since forever, or the type where it preserves the variables? 
Thu Jul  2 23:55:03 2009 [dgd] Aidil@The Zone: both is better then having to rebuild a binary and restart the mud :) 
Thu Jul  2 23:55:11 2009 [dgd] Aidil@The Zone: both are. 
Thu Jul  2 23:55:38 2009 [dgd] Raudhrskal@Dead Souls Dev: Sounds quite right. 
Thu Jul  2 23:55:40 2009 [dgd] Aidil@The Zone: obviously, the one preserving the variables is best.. :) 
Thu Jul  2 23:56:06 2009 [dgd] Raudhrskal@Dead Souls Dev: if you have some way to explicitly erase them when needed... yea. 
Thu Jul  2 23:56:53 2009 [dgd] Aidil@The Zone: you dest the object and create a new clone if thats what you want :) 
Thu Jul  2 23:57:56 2009 [dgd] Raudhrskal@Dead Souls Dev: Ah, of course... so simple... wait, that's exactly what reload does. Ah, well. 
Thu Jul  2 23:58:18 2009 [dgd] Aidil@The Zone: not when you want the variables preserved :) 
Thu Jul  2 23:58:21 2009 [dgd] Quixadhal@Bloodlines: Amusing question... what happens if you load an object and then remove the source file.  Will the existing instances live on as long as you keep your statedump intact? 
Thu Jul  2 23:58:36 2009 [dgd] Raudhrskal@Dead Souls Dev thinks so. 
Thu Jul  2 23:58:47 2009 [dgd] Aidil@The Zone: yes. 
Thu Jul  2 23:59:05 2009 [dgd] Aidil@The Zone: the source file is completely irrelevant to the object existing. 
Thu Jul  2 23:59:28 2009 [dgd] Quixadhal@Bloodlines: That is similar to a virtual object, which may not have ever had a source file, but I was curious what would happen when  one tried to update things (such as things it might have inherited). 
Fri Jul  3 00:00:24 2009 [dgd] Raudhrskal@Dead Souls Dev: well, on fluff it wouldn't change the obj, it'd keep refs to old version of the code. In dgd, the new versions might snap into place, iirc? 
Fri Jul  3 00:00:30 2009 [dgd] Aidil@The Zone: it will fail to compile that object, and cause the old inherited code to stay referenced (and hence loaded) 
Fri Jul  3 00:00:59 2009 [dgd] Aidil@The Zone: no, you need to recompile the object to make it inherit the new inheritable. 
Fri Jul  3 00:01:34 2009 [dgd] Aidil@The Zone: that does not involve destructing the object, but it does require source for the object in one way or another. 
Fri Jul  3 00:01:42 2009 [dgd] Quixadhal@Bloodlines: Which you could only do if you had source, either cached or generated.  Makes sense. 
Fri Jul  3 00:01:46 2009 [dgd] Raudhrskal@Dead Souls Dev: Ah, like here. Why i thought that inherits are dynamically propagated in dgd? 
Fri Jul  3 00:02:25 2009 [dgd] Aidil@The Zone: heh, you asking me how your logic works? :) 
Fri Jul  3 00:02:28 2009 [dgd] Quixadhal@Bloodlines: That would be tricky... possible, but any API change would break things until you recompiled them. 
Fri Jul  3 00:02:29 2009 [dgd] Raudhrskal@Dead Souls Dev: probably misunderstood some thing. Thanks for clarification. 
Fri Jul  3 00:03:10 2009 [dgd] Raudhrskal@Dead Souls Dev: my logic not only doesn't work, it doesn't exist, at least on this plane. 
Fri Jul  3 00:03:14 2009 [dgd] Aidil@The Zone: indeed, technically, dgd could do it dynamically (would require a change to the driver, but one that is relatively easy to make) 
Fri Jul  3 00:03:36 2009 [dgd] Aidil@The Zone: however, you really do need control over when dependent objects get recompiled. 
Fri Jul  3 00:03:48 2009 [dgd] Aidil@The Zone: or get the new code in the inheritable. 
Fri Jul  3 00:04:09 2009 [dgd] Quixadhal@Bloodlines: I'd hate to see the errors when function foo(int x) becomes foo(float x, int y), and old code tries to use it. 
Fri Jul  3 00:05:16 2009 [dgd] Raudhrskal@Dead Souls Dev: well, that's quite obvious - that's why i was wondering how it managed to do it when i thought it did. 
Fri Jul  3 00:05:19 2009 [dgd] Raudhrskal@Dead Souls Dev: or something. 
Fri Jul  3 00:05:30 2009 [dgd] Raudhrskal@Dead Souls Dev: guess i should get some sleep. 
Fri Jul  3 00:05:37 2009 [dgd] Raudhrskal@Dead Souls Dev: have a nice night. 
Fri Jul  3 00:05:42 2009 [dgd] Aidil@The Zone: the way in which it could do it.. 
Fri Jul  3 00:05:45 2009 [dgd] Quixadhal@Bloodlines: night! 
Fri Jul  3 00:05:57 2009 [dgd] Raudhrskal@Dead Souls Dev: wait, let him finish! 
Fri Jul  3 00:06:01 2009 [dgd] Raudhrskal@Dead Souls Dev listens. 
Fri Jul  3 00:06:05 2009 [dgd] Aidil@The Zone: the inheritable is a pointer in the control block struct of the object 
Fri Jul  3 00:06:06 2009 [dgd] Quixadhal@Bloodlines chuckles. 
Fri Jul  3 00:06:12 2009 [dgd] Aidil@The Zone: it is really easy to change :) 
Fri Jul  3 00:06:46 2009 [dgd] Raudhrskal@Dead Souls Dev: but resolving API changes isn't so easy... 
Fri Jul  3 00:06:57 2009 [dgd] Aidil@The Zone: but, it would have to check for new code for any depending object, and if so, somehow deal with that :) 
Fri Jul  3 00:06:59 2009 [dgd] Aidil@The Zone nods. 
Fri Jul  3 00:07:25 2009 [dgd] Aidil@The Zone: that is very likely why felix never implemented it, at least not in a public version. 
Fri Jul  3 00:07:30 2009 [dgd] Quixadhal@Bloodlines: That was my point... recompiling will catch it and yell at you... just flipping blocks, and maybe the VM pops too many or too few arguments from the stack and BOOM. 
Fri Jul  3 00:07:40 2009 [dgd] Aidil@The Zone: gurbalib's warmboot is as close as it will get probably :) 
Fri Jul  3 00:07:46 2009 [dgd] Raudhrskal@Dead Souls Dev nods solemnly. 
Fri Jul  3 00:07:59 2009 [dgd] Raudhrskal@Dead Souls Dev: Thanks. 
Fri Jul  3 00:08:06 2009 [dgd] Aidil@The Zone: sleep well :) 
Fri Jul  3 00:08:11 2009 [dgd] Aidil@The Zone: good plan.. 
Fri Jul  3 00:08:15 2009 [dgd] Raudhrskal@Dead Souls Dev: Impossible. 
Fri Jul  3 00:10:22 2009 [dgd] Aidil@The Zone: very few things are truely impossible 
Fri Jul  3 00:12:55 2009 [dgd] Quixadhal@Bloodlines: Using a windows system for heavy gaming and viewing random flash-based web sites for a year without rebooting? 
Fri Jul  3 02:12:05 2009 [dgd] Silenus@Dead Souls Dev: i am debating in place recompilation 
Fri Jul  3 02:12:29 2009 [dgd] Silenus@Dead Souls Dev: i wonder how bad it would break a fluffos mudlib tho :/ 
Fri Jul  3 02:31:24 2009 [dgd] Kalinash@Fire and Ice: forget fluff 
Fri Jul  3 02:31:28 2009 [dgd] Kalinash@Fire and Ice: blaze new trails 
Fri Jul  3 02:34:24 2009 [dgd] Silenus@Dead Souls Dev: problem is building a library takes more time than do a driver engine :( 
Fri Jul  3 15:15:53 2009 [dgd] Aidil@The Zone: building a library is a problem? 
Fri Jul  3 15:16:12 2009 [dgd] Aidil@The Zone: rather, it is a matter of sitting down and writing code :) 
Fri Jul  3 15:16:49 2009 [dgd] Aidil@The Zone: it might be more work, but, there are many more people capable of doing that work. 
Fri Jul  3 15:28:24 2009 [dgd] Silenus@Dead Souls Dev: evening Aidil :-) 
Fri Jul  3 15:34:00 2009 [dgd] Silenus@Dead Souls Dev: i suppose that is true but having a complete fresh platform 
Fri Jul  3 15:35:16 2009 [dgd] Silenus@Dead Souls Dev: means that coders would have to be able to learn how to use it then code a library which adds up. i suspect i will work on this close to compat system and maybe save my more incompat type ideas for another project. but i am not really sure how bad an in place recompilation system would be. 
Fri Jul  3 15:48:29 2009 [dgd] Aidil@The Zone: you could play with gurbalib a bit to get an idea :P 
Fri Jul  3 15:49:17 2009 [dgd] Silenus@Dead Souls Dev: *nod* 
Fri Jul  3 15:49:19 2009 [dgd] Aidil@The Zone: gurbalib is closer to lima then to ds, but the consequences of in-place recompilation will be rather similar. 
Fri Jul  3 15:50:25 2009 [dgd] Silenus@Dead Souls Dev: if you recompile the auto object does it result in a far amount of lag on a modern machine? 
Fri Jul  3 15:56:29 2009 [dgd] Aidil@The Zone: if you rebuild everything that depends on it, and happen to have a seriously huge lib, then probably yes. 
Fri Jul  3 15:57:35 2009 [dgd] Aidil@The Zone: on my pIII 800, gurbalib takes a few seconds for a warmboot (which forces a recompile of everything) 
Fri Jul  3 15:57:44 2009 [dgd] Silenus@Dead Souls Dev: *nod* something to play with later i guess. i will try getting this working the fluffos way. 
Fri Jul  3 15:58:01 2009 [dgd] Aidil@The Zone: something in the order of 4-8 secs. 
Fri Jul  3 15:58:55 2009 [dgd] Aidil@The Zone: part of that time is due to a 2 sec call_out, so actual time spent on compiling is less then that. 
Fri Jul  3 15:59:06 2009 [dgd] Silenus@Dead Souls Dev: have you tried something like that on the larger wotf lib? 
Fri Jul  3 15:59:32 2009 [dgd] Aidil@The Zone nods. 
Fri Jul  3 16:00:03 2009 [dgd] Aidil@The Zone: the wotf lib is far less efficient in this, and may take upto a minute for a full rebuild. 
Fri Jul  3 16:00:17 2009 [dgd] Silenus@Dead Souls Dev nods. 
Fri Jul  3 16:01:05 2009 [dgd] Aidil@The Zone: if I optimize it to work as efficiently as gurbalib, that would likely be somewhere between 10-20 secs on the current server. 
Fri Jul  3 16:04:31 2009 [dgd] Silenus@Dead Souls Dev: i think dgd's coding is quite intricate and i dont think i will try to emulate it on my first draft 
Fri Jul  3 16:05:03 2009 [dgd] Silenus@Dead Souls Dev: some aspects of it i dont really understand the issues involved in arriving at a particular design. 
Fri Jul  3 16:06:21 2009 [dgd] Silenus@Dead Souls Dev: implementing a good diskbased system + transactions and persistence is a fair amount of work even without STM dgd/MP type code. 
Fri Jul  3 16:08:46 2009 [dgd] Aidil@The Zone: esp. without stm alike code :) 
Fri Jul  3 16:09:47 2009 [dgd] Aidil@The Zone: btw, a rebuild of the wotf lib on a modern machine (2.4ghz quad core, lots of memory etc) takes 7 seconds. 
Fri Jul  3 16:12:41 2009 [dgd] Silenus@Dead Souls Dev: hmm. i am just wondering if it could be done lazily without the extra call hooks dgd has. 
Fri Jul  3 16:12:44 2009 [dgd] Aidil@The Zone: but, while it is a good idea to allow making use of the latest and greatest in hardware, I think that for a mud 'engine' it is a bad idea to require the latest and greatest in hardware 
Fri Jul  3 16:13:43 2009 [dgd] Silenus@Dead Souls Dev: yeah i think it's nice to have it work relatively well on modest hardware. 
Fri Jul  3 16:14:09 2009 [dgd] Aidil@The Zone: the reason why dgd doesn't allow doing it lazily is because recompilation (as opposed to compilation) happens at the end of a transaction. 
Fri Jul  3 16:14:48 2009 [dgd] Aidil@The Zone: so, recompile lazily when referencing an object will result in referencing the old code initially, regardless of the recompile. 
Fri Jul  3 16:15:23 2009 [dgd] Aidil@The Zone: if you forgo atomic functions, or provide some way to roll back code replacement, you may be able to get around that. 
Fri Jul  3 16:15:42 2009 [dgd] Silenus@Dead Souls Dev: *nod* i doubt i will implement atomics 
Fri Jul  3 16:15:53 2009 [dgd] Aidil@The Zone: for all I can tell, it is a limitation of the implementation, not a theoretic limitation. 
Fri Jul  3 16:16:06 2009 [dgd] Raudhrskal@Dead Souls Dev: You need a clearance for atomic stuff... 
Fri Jul  3 16:16:10 2009 [dgd] Raudhrskal@Dead Souls Dev hiccups. 
Fri Jul  3 16:16:12 2009 [dgd] Silenus@Dead Souls Dev: but if i go as far as 
Fri Jul  3 16:16:29 2009 [dgd] Aidil@The Zone: atomic functions are somewhat needed for making this work reliably :) 
Fri Jul  3 16:16:35 2009 [dgd] Silenus@Dead Souls Dev: implementing recompilation i might be tempted to talk to hamlet a bit more 
Fri Jul  3 16:16:43 2009 [dgd] Silenus@Dead Souls Dev: and try stm 
Fri Jul  3 16:17:00 2009 [dgd] Silenus@Dead Souls Dev: and i would also have to be able to rollback recompiles. 
Fri Jul  3 16:18:01 2009 [dgd] Aidil@The Zone nods. 
Fri Jul  3 16:18:11 2009 [dgd] Silenus@Dead Souls Dev: sounds expensive tho :/ 
Fri Jul  3 16:18:23 2009 [dgd] Aidil@The Zone: while serving different purposes, atomic functions and stm require very similar technology. 
Fri Jul  3 16:19:51 2009 [dgd] Silenus@Dead Souls Dev nods 
Fri Jul  3 16:19:53 2009 [dgd] Aidil@The Zone: at any rate, the main reason for needing atomics in this is to deal with compilation failures. 
Fri Jul  3 16:20:34 2009 [dgd] Silenus@Dead Souls Dev: *nod* i guess that would make sense. 
Fri Jul  3 16:20:37 2009 [dgd] Aidil@The Zone: when you don't rollback a recompile of some critical objects on failure, you are bound to end up with a non functional environment and no way to repair it. 
Fri Jul  3 16:22:12 2009 [dgd] Silenus@Dead Souls Dev: sounds a bit too intricate for me for now 
Fri Jul  3 16:22:15 2009 [dgd] Aidil@The Zone: when you look at gurbalib's warmboot command, you'll notice it consists of 2 'phases'. The 'critical' phase, which is atomic, and the 'non critical phase'. 
Fri Jul  3 16:22:25 2009 [dgd] Silenus@Dead Souls Dev: maybe version two. 
Fri Jul  3 16:22:47 2009 [dgd] Silenus@Dead Souls Dev: i will take a look sounds like interesting code. 
Fri Jul  3 16:23:12 2009 [dgd] Aidil@The Zone: failures in the critical phase result in a complete rollback, failures in the non critical phase result in an abort. 
Fri Jul  3 16:24:42 2009 [dgd] Aidil@The Zone: if you ever implement in-place recompilation with dgd's klib, you'll quickly find that you have to do the same thing. The critical phase at least includes the driver and auto objects, and your object daemon (or compiler daemon as gurbalib calls it). 
Fri Jul  3 16:28:12 2009 [dgd] Aidil@The Zone: at any rate.. the recompile at end of transaction does make life easier. It ensures that you can recompile a set of objects that call eachother and have the changes show when there can't be any calls between those objects. 
Fri Jul  3 16:28:48 2009 [dgd] Aidil@The Zone: and with that, prevent problems due to api changes. 
Fri Jul  3 16:29:10 2009 [dgd] Silenus@Dead Souls Dev: *nod* makes sense 
Fri Jul  3 16:30:31 2009 [dgd] Aidil@The Zone: and on another note, replacing code in a running system is always a somewhat intricate process. 
Fri Jul  3 16:32:33 2009 [dgd] Silenus@Dead Souls Dev: yeah i thought about it a bit and it seems like it could be pretty tricky to get right 
Fri Jul  3 16:32:53 2009 [dgd] Silenus@Dead Souls Dev: which is why atm i am not thinkign too much about STM 
Fri Jul  3 16:33:14 2009 [dgd] Aidil@The Zone: but stm is cool! :) 
Fri Jul  3 16:34:52 2009 [dgd] Silenus@Dead Souls Dev: it is :-). 
Fri Jul  3 16:34:59 2009 [dgd] Silenus@Dead Souls Dev: but maybe for later :P 
Fri Jul  3 16:35:33 2009 [dgd] Silenus@Dead Souls Dev: just trying to get a simple fluffos runtime (well not quite the same) + compiler system working for now. 
Fri Jul  3 16:35:43 2009 [dgd] Silenus@Dead Souls Dev: i think thats enough on my plate :D 
Fri Jul  3 16:36:13 2009 [dgd] Aidil@The Zone: iirc, Hamlet was looking at using a toolchain that provides stm support instead of implementing an stm alike system in the driver. 
Fri Jul  3 16:36:56 2009 [dgd] Aidil@The Zone: I don't know how efficient that would be, but it would make it easy to get around this issue of rolling back recompiles. 
Fri Jul  3 16:38:14 2009 [dgd] Silenus@Dead Souls Dev: yeah intel compiler has an stm system 
Fri Jul  3 16:38:27 2009 [dgd] Silenus@Dead Souls Dev: but it costs money so 
Fri Jul  3 16:38:32 2009 [dgd] Aidil@The Zone: dgd can't really do that because its stm implementation deals with lpc data only. 
Fri Jul  3 16:38:34 2009 [dgd] Silenus@Dead Souls Dev: ppl might not want to use it 
Fri Jul  3 16:39:00 2009 [dgd] Silenus@Dead Souls Dev: though it would be an interesting experiment. 
Fri Jul  3 16:39:17 2009 [dgd] Silenus@Dead Souls Dev: i guess Hamlet could supply binaries to ppl. 
Fri Jul  3 16:39:38 2009 [dgd] Aidil@The Zone: it is possible to use dgd's stm system from within the driver, but that requires using lpc variables that are placed on the 'proper' execution plane. 
Fri Jul  3 16:40:55 2009 [dgd] Aidil@The Zone: or just hope other compilers will do the same thing over time, which is what he seems to be expecting if I recall the discussion about this correctly. 
Fri Jul  3 16:42:21 2009 [dgd] Silenus@Dead Souls Dev: hmm that is a possibility. 
Fri Jul  3 16:42:59 2009 [dgd] Silenus@Dead Souls Dev: i havent seen him in a while was hoping to ask him which library he was using for his experiment with STM and fluffos 
Sat Jul  4 16:54:44 2009 [dgd] Subversion@Way of the Force: Aidil committed Gurbalib revision 295 to svn://wotf.org/gurbalib : Fixed shutdown bug in channel_d  
Sat Jul  4 21:35:57 2009 [dgd] Aidil@GurbaHub snores. 
Wed Jul  8 17:12:55 2009 [dgd] Kalinash@Fire and Ice: Aidil? 
Wed Jul  8 17:14:22 2009 [dgd] Kalinash@Fire and Ice: *wpr doesn't like me anymore :( 
Wed Jul  8 18:54:36 2009 [dgd] Aidil@Way of the Force: it doesn't? 
Wed Jul  8 18:56:27 2009 [dgd] Kalinash@Fire and Ice: it wasn't earlier... i had to switch back to crat's... it would connect to *wpr and send packets but i never got anything back 
Wed Jul  8 18:57:24 2009 [dgd] Kalinash@Fire and Ice: i can try again if you want 
Wed Jul  8 18:59:08 2009 [dgd] Aidil@The Zone: odd. lemme enable some debugging stuff before you try again 
Wed Jul  8 19:04:07 2009 [dgd] Aidil@The Zone: ok, try if you want :) 
Wed Jul  8 19:43:55 2009 [dgd] Kalinash@Fire and Ice: ok, just a sec 
Wed Jul  8 19:44:55 2009 [dgd] Kalinash@Fire and Ice: foo 
Wed Jul  8 19:45:07 2009 [dgd] Kalinash@Fire and Ice: working now... got the startup reply and obviously channels too 
Wed Jul  8 19:45:10 2009 [dgd] Kalinash@Fire and Ice: wierd 
Wed Jul  8 19:46:44 2009 [dgd] Kalinash@Fire and Ice: why is element 3 of the startup reply "yatmin"?  The spec says it's null 
Wed Jul  8 19:49:52 2009 [dgd] Raudhrskal@Dead Souls Dev: isn't that the list of alternative routers? 
Wed Jul  8 19:50:11 2009 [dgd] Kalinash@Fire and Ice: no, that's element 6 and is an array 
Wed Jul  8 19:53:17 2009 [dgd] Aidil@The Zone: it is yatmim? hmm. lemme see. :) 
Wed Jul  8 19:53:54 2009 [dgd] Kalinash@Fire and Ice:   "startup-reply", 
Wed Jul  8 19:53:55 2009 [dgd] Kalinash@Fire and Ice:   5, 
Wed Jul  8 19:53:56 2009 [dgd] Kalinash@Fire and Ice:   "*wpr", 
Wed Jul  8 19:53:58 2009 [dgd] Kalinash@Fire and Ice:   "yatmim", 
Wed Jul  8 19:53:59 2009 [dgd] Kalinash@Fire and Ice:   "Fire and Ice", 
Wed Jul  8 19:54:00 2009 [dgd] Kalinash@Fire and Ice:   0, 
Wed Jul  8 19:54:01 2009 [dgd] Kalinash@Fire and Ice:   ({ /* sizeof() == 1 */ 
Wed Jul  8 19:54:02 2009 [dgd] Kalinash@Fire and Ice:     ({ /* sizeof() == 2 */ 
Wed Jul  8 19:54:03 2009 [dgd] Kalinash@Fire and Ice:       "*wpr", 
Wed Jul  8 19:54:04 2009 [dgd] Kalinash@Fire and Ice:       "195.242.99.94 8080" 
Wed Jul  8 19:54:06 2009 [dgd] Kalinash@Fire and Ice:     }) 
Wed Jul  8 19:54:07 2009 [dgd] Kalinash@Fire and Ice:   }), 
Wed Jul  8 19:54:08 2009 [dgd] Kalinash@Fire and Ice:   342998783 
Wed Jul  8 19:54:09 2009 [dgd] Kalinash@Fire and Ice: whoa... that was cool 
Wed Jul  8 19:56:29 2009 [dgd] Aidil@The Zone: you were right. fixed now. 
Wed Jul  8 19:56:36 2009 [dgd] Kalinash@Fire and Ice: OMG! 
Wed Jul  8 19:56:43 2009 [dgd] Kalinash@Fire and Ice: I FOUND A BUG IN AIDIL'S CODE! 
Wed Jul  8 19:56:48 2009 [dgd] Aidil@The Zone: and heh. what was so cool? :) 
Wed Jul  8 19:56:49 2009 [dgd] Kalinash@Fire and Ice: I will remember this day always 
Wed Jul  8 19:56:52 2009 [dgd] Aidil@The Zone laughs out loud. 
Wed Jul  8 19:57:00 2009 [dgd] Aidil@The Zone: it is code, it has bugs :) 
Wed Jul  8 19:57:09 2009 [dgd] Kalinash@Fire and Ice: oh, pasting that multi-line packet output sent it as one message per line 
Wed Jul  8 19:57:13 2009 [dgd] Kalinash@Fire and Ice: i hadn't been expecting that 
Mon Jul 13 15:07:09 2009 [dgd] Subversion@Way of the Force: Aidil committed revision 37 to svn://wotf.org/dgd-devel-net : Removed kernel lib (a new repository will be created for this)  
Fri Jul 17 13:08:16 2009 [dgd] Aidil@Way of the Force: hmm. if I have to believe vmware server, I have a 1.1 thz cpu. 
Fri Jul 17 22:00:10 2009 [dgd] Aidil@Way of the Force: lovely router resync. 
Fri Jul 17 22:00:41 2009 [dgd] Kalinash@Fire and Ice: i blame crat 
Fri Jul 17 22:01:00 2009 [dgd] Aidil@Way of the Force: you'd be right, for once. 
Fri Jul 17 22:12:35 2009 [dgd] Kalinash@Fire and Ice: i marked my calendar :) 
Sat Jul 18 00:13:01 2009 [dgd] Aidil@Way of the Force: paging cratman 
Sat Jul 18 00:17:21 2009 [dgd] Cratylus@Dead Souls Dev: hi 
Sat Jul 18 00:23:02 2009 [dgd] Aidil@Way of the Force: hey :) 
Sat Jul 18 00:23:55 2009 [dgd] Aidil@Way of the Force: was just wondering, saw *i4 and such sending updates about listeners during the 'big resync'. is that on purpose? 
Sat Jul 18 00:24:34 2009 [dgd] Cratylus@Dead Souls Dev: hmm 
Sat Jul 18 00:24:47 2009 [dgd] Cratylus@Dead Souls Dev: i thot i changed that 
Sat Jul 18 00:24:54 2009 [dgd] Aidil@Way of the Force nods. 
Sat Jul 18 00:24:57 2009 [dgd] Aidil@Way of the Force: me too :) 
Sat Jul 18 00:25:19 2009 [dgd] Aidil@Way of the Force: I'm not seeing them with normal changes to channels, but I do see them during a resync. 
Sat Jul 18 00:25:52 2009 [dgd] Aidil@Way of the Force: no biggie for me, but thought you'd want to know :P 
Sat Jul 18 00:38:37 2009 [dgd] Cratylus@Dead Souls Dev: ya i should prolly do a code refresh soon of all nodes 
Sat Jul 18 00:38:42 2009 [dgd] Cratylus@Dead Souls Dev: thank you 
Sat Jul 18 00:41:10 2009 [dgd] Aidil@Way of the Force: :) 
Sat Jul 18 17:07:23 2009 [dgd] Subversion@Way of the Force: Aidil committed revision 39 to svn://wotf.org/dgd-devel-net : Imported dgd 1.3.1, updated documentation, bumped network package version  
Mon Jul 20 07:22:18 2009 [dgd] Subversion@Way of the Force: Aidil committed Gurbalib revision 297 to svn://wotf.org/gurbalib : Imported dgd 1.3.1, imported network package 0.7, removed some obsolete files  
Mon Jul 20 14:15:00 2009 [dgd] Aidil@GurbaMP: whee :) 
Mon Jul 20 14:44:42 2009 [dgd] Kalinash@Fire and Ice: ? 
Mon Jul 20 14:46:33 2009 [dgd] Raudhrskal@Dead Souls Dev: hint: "MP" 
Mon Jul 20 14:51:18 2009 [dgd] Kalinash@Fire and Ice: Military Police? 
Mon Jul 20 14:52:14 2009 [dgd] Raudhrskal@Dead Souls Dev: No, Microsoft Power. 
Mon Jul 20 15:15:04 2009 [dgd] Kalinash@Fire and Ice: Ahh, the magical powers of MS-DOS 
Mon Jul 20 21:25:03 2009 [dgd] Aidil@GurbaMP mumbles something. 
Mon Jul 20 21:25:39 2009 [dgd] Raudhrskal@Dead Souls Dev sends the recording for analysis. 
Mon Jul 20 21:25:51 2009 [dgd] Kalinash@Fire and Ice: "NON-HUMAN ENTITY DETECTED" 
Mon Jul 20 21:26:40 2009 [dgd] Aidil@GurbaMP: I've been distributed over multiple cores, having some concurrency issues here. 
Mon Jul 20 21:27:01 2009 [dgd] Kalinash@Fire and Ice: I use fiber to help with that. 
Mon Jul 20 21:27:21 2009 [dgd] Aidil@GurbaMP: I need something along the lines of 'the giant lock'. 
Mon Jul 20 21:27:41 2009 [dgd] Aidil@GurbaMP: which isn't gonna mean much to a mickysofty like you :P 
Mon Jul 20 21:27:58 2009 [dgd] Kalinash@Fire and Ice: heh, you'd like to think that :P 
Mon Jul 20 21:29:01 2009 [dgd] Aidil@GurbaMP: call_out related concurrency issues are fun :P 
Mon Jul 20 21:29:35 2009 [dgd] Kalinash@Fire and Ice: I find your definition of fun ... disturbing. 
Mon Jul 20 21:29:59 2009 [dgd] Aidil@GurbaMP: that might be due to you being disturbed. 
Mon Jul 20 21:30:12 2009 [dgd] Kalinash@Fire and Ice: Oh, I'm WAY beyond distrubed. 
Mon Jul 20 21:30:23 2009 [dgd] Aidil@GurbaMP: well, there is your explanation. 
Mon Jul 20 21:30:24 2009 [dgd] Kalinash@Fire and Ice: I'm working on setup right now :( 
Mon Jul 20 21:30:34 2009 [dgd] Kalinash@Fire and Ice: if that doesn't warp you mind, nothing will 
Mon Jul 20 21:30:36 2009 [dgd] Aidil@GurbaMP: setup? 
Mon Jul 20 21:31:10 2009 [dgd] Aidil@GurbaMP: as in.. installshield or whatever its called now? 
Mon Jul 20 21:31:14 2009 [dgd] Kalinash@Fire and Ice: Yeah, the program that installs Zune onto your desktop...  Does pre-req checks, patches, etc too 
Mon Jul 20 21:31:19 2009 [dgd] Kalinash@Fire and Ice: WIX 
Mon Jul 20 21:31:37 2009 [dgd] Kalinash@Fire and Ice: InstallSHIELD was evil 
Mon Jul 20 21:31:45 2009 [dgd] Raudhrskal@Dead Souls Dev: still is. 
Mon Jul 20 21:31:52 2009 [dgd] Kalinash@Fire and Ice: aye :) 
Mon Jul 20 21:32:00 2009 [dgd] Kalinash@Fire and Ice: but it's no longer the standard 
Mon Jul 20 21:32:09 2009 [dgd] Aidil@GurbaMP: anything installing microsoftware is by definition evil. 
Mon Jul 20 21:32:19 2009 [dgd] Kalinash@Fire and Ice: but we also have a setup program that's outside of the MSI portion of setup 
Mon Jul 20 21:33:36 2009 [dgd] Aidil@GurbaMP: ah well :) 
Mon Jul 20 21:33:53 2009 [dgd] Kalinash@Fire and Ice: trade? 
Mon Jul 20 21:34:21 2009 [dgd] Aidil@GurbaMP: heh. 
Mon Jul 20 21:34:35 2009 [dgd] Aidil@GurbaMP: I rather keep to my unixy job :) 
Mon Jul 20 21:34:37 2009 [dgd] Kalinash@Fire and Ice: I can implement a kernel lock in dgd 
Mon Jul 20 21:34:53 2009 [dgd] Aidil@GurbaMP: thats pretty easy :) 
Mon Jul 20 21:35:16 2009 [dgd] Kalinash@Fire and Ice: i didn't say it was a good trade for you ;) 
Mon Jul 20 21:35:42 2009 [dgd] Aidil@GurbaMP: /daemons/event_d.c line 44  
Mon Jul 20 21:36:04 2009 [dgd] Aidil@GurbaMP: makes dgdmp behave as if it has a giant lock. 
Mon Jul 20 21:37:08 2009 [dgd] Aidil@GurbaMP: it does however look like I'll have to add that to the code starting callouts. 
Sun Aug  2 05:23:04 2009 [dgd] Cerihan@Dead Souls Dev: Is anyone alive? 
Sun Aug  2 05:25:32 2009 [dgd] Kalinash@Fire and Ice: nope 
Sun Aug  2 05:25:45 2009 [dgd] Cerihan@Dead Souls Dev: Huzzah.  I am taking this channel over, then. 
Sun Aug  2 05:25:56 2009 [dgd] Kalinash@Fire and Ice: aidil is a zombie, though 
Sun Aug  2 05:26:05 2009 [dgd] Kalinash@Fire and Ice: so undead counts, right? 
Sun Aug  2 05:26:14 2009 [dgd] Cerihan@Dead Souls Dev: Hmm.  Not sure. 
Sun Aug  2 05:26:38 2009 [dgd] Cerihan@Dead Souls Dev: So... yeah.  I kind of fell off the grid for a while there.  Did Aidil think I'd expired? 
Sun Aug  2 05:32:38 2009 [dgd] Cerihan@Dead Souls Dev: I carefully plotted to be here at a time when Aidil would be asleep, so that he could not thwack me for disappearing. 
Sun Aug  2 05:51:48 2009 [dgd] Cerihan@Dead Souls Dev: Oh well. 
Sun Aug  2 05:52:34 2009 [dgd] Cerihan@Dead Souls Dev: I'll check back in the next day or two.  If Aidil shows up let him know I will try to start hanging around again although I still don't have a server at the moment. 
Sun Aug  2 12:58:54 2009 [dgd] Aidil@Way of the Force: oh wow. Cerihan found back this channel ? :) 
Sun Aug  2 12:59:00 2009 [dgd] Aidil@Way of the Force boggles at the concept. 
Sun Aug  2 12:59:13 2009 [dgd] Aidil@Way of the Force grins. 
Sun Aug  2 12:59:24 2009 [dgd] Raudhrskal@Dead Souls Dev: Good *yawn*ing, Aidil. 
Sun Aug  2 12:59:41 2009 [dgd] Aidil@Way of the Force: morning or something. 
Sun Aug  2 13:00:02 2009 [dgd] Raudhrskal@Dead Souls Dev: Something. 
Sun Aug  2 22:25:47 2009 [dgd] Aidil@Way of the Force smiles. 
Mon Aug  3 02:59:40 2009 [dgd] Cerihan@Dead Souls Dev: *whistle* 
Mon Aug  3 11:51:21 2009 [dgd] Cerihan@GurbaHub: /history 
Mon Aug  3 11:51:32 2009 [dgd] Cerihan@GurbaHub: Huh. 
Mon Aug  3 12:01:54 2009 [dgd] Aidil@The Zone: hmm. think gurbahub didn't get that update or no warmboot after that update yet :) 
Mon Aug  3 12:05:43 2009 [dgd] Aidil@The Zone: and now it does. 
Mon Aug  3 19:21:16 2009 [dgd] Aidil@Way of the Force: right :) step 1 of fitting gurbalib in the 'railsmud' framework done. I have a working json parser and message generator. 
Mon Aug  3 19:21:45 2009 [dgd] Aidil@Way of the Force: and for that matter.. json looks like a very good idea for a message format for a next generation intermud protocol 
Mon Aug  3 19:22:22 2009 [dgd] Aidil@Way of the Force: structured data like xml, but without the parsing headache and insane verbosity :) 
Mon Aug  3 19:22:59 2009 [dgd] Raudhrskal@Dead Souls Dev: Erm... json's pretty nice but has _one small problem_ 
Mon Aug  3 19:23:09 2009 [dgd] Raudhrskal@Dead Souls Dev: deserialization via eval(). 
Mon Aug  3 19:23:11 2009 [dgd] Aidil@Way of the Force: which is? :) 
Mon Aug  3 19:23:34 2009 [dgd] Raudhrskal@Dead Souls Dev: If you want to be sure it is what it is, you need to manually scan the string to make sure that it's only data. 
Mon Aug  3 19:23:36 2009 [dgd] Aidil@Way of the Force: heh. but how important is that gonna be for this purpose? :) 
Mon Aug  3 19:24:20 2009 [dgd] Aidil@Way of the Force: well, I wrote a little parser that does that, and gives an lpc mapping as output. 
Mon Aug  3 19:24:42 2009 [dgd] Raudhrskal@Dead Souls Dev: ofc, in your case it will work. But what about using JSON in, i dunno... JS? 
Mon Aug  3 19:24:57 2009 [dgd] Aidil@Way of the Force: heh. 
Mon Aug  3 19:25:09 2009 [dgd] Raudhrskal@Dead Souls Dev: indeed. 
Mon Aug  3 19:25:29 2009 [dgd] Aidil@Way of the Force: how nice the other side of my connection isn't really my concern :) 
Mon Aug  3 19:27:33 2009 [dgd] Aidil@Way of the Force: I do believe however the problem there is javascript :) 
Mon Aug  3 19:51:23 2009 [dgd] Subversion@Way of the Force: Aidil committed Gurbalib revision 302 to svn://wotf.org/gurbalib : Added module to create/parse json messages  
Mon Aug  3 19:58:58 2009 [dgd] Kalinash@Fire and Ice: JSON is the guy with the hockey mask, right? 
Mon Aug  3 20:16:34 2009 [dgd] Aidil@Way of the Force puts his fingers to his lips and goes: blblblblblblblblblbl. 
Mon Aug  3 20:18:34 2009 [dgd] Quixadhal@Bloodlines: "Just Stupid Old Nonsense" messages, yay! 
Mon Aug  3 20:18:57 2009 [dgd] Silenus@Dead Souls Dev: hey Aidil hey Quix 
Mon Aug  3 20:19:50 2009 [dgd] Aidil@Way of the Force: :) 
Mon Aug  3 20:20:00 2009 [dgd] Quixadhal@Bloodlines: Run Aidil!  Oy, you'll not take me alive copper!... I mean, Hi Silenus! 
Mon Aug  3 20:20:23 2009 [dgd] Silenus@Dead Souls Dev: heheh 
Mon Aug  3 20:20:28 2009 [dgd] Aidil@Way of the Force: hehe. 
Mon Aug  3 20:21:08 2009 [dgd] Quixadhal@Bloodlines: People keep stealing my ideas... I'm gonna have to make the tinfoil hat thicker.  I was JUST going to sit down and write what packagemapping.com does. 
Mon Aug  3 20:21:34 2009 [dgd] Aidil@Way of the Force: Noah Gibbs (the guy who wrote phantasmal) and me got caught up into plugging gurbalib as a gameserver into a project of his. 
Mon Aug  3 20:22:03 2009 [dgd] Aidil@Way of the Force: or at least see how feasable it is. 
Mon Aug  3 20:24:53 2009 [dgd] Quixadhal@Bloodlines: Ruby annoyed me earlier.  The first time I used it, it took me about 5 minutes to get postgresql working.  This time, it is considerably more cranky. 
Mon Aug  3 20:25:24 2009 [dgd] Silenus@Dead Souls Dev: Quix are you mainly a db related guy? 
Mon Aug  3 20:25:29 2009 [dgd] Aidil@Way of the Force: heh. I'm not exactly going to deal with the ruby side of this anyway :) 
Mon Aug  3 20:25:43 2009 [dgd] Aidil@Way of the Force: ruby annoys me at the first glance of it :P 
Mon Aug  3 20:25:55 2009 [dgd] Aidil@Way of the Force: but then.. 
Mon Aug  3 20:25:59 2009 [dgd] Aidil@Way of the Force: so does perl. 
Mon Aug  3 20:27:04 2009 [dgd] Quixadhal@Bloodlines: Yeah, I had a nice job for 6 years doing perl server scripts, postgresql (using perl for stored procedures even), and a bit of php for the frontend (which I didn't touch very often). 
Mon Aug  3 20:27:19 2009 [dgd] Aidil@Way of the Force: heh. 
Mon Aug  3 20:27:27 2009 [dgd] Aidil@Way of the Force: to me that sounds like hell kinda :) 
Mon Aug  3 20:27:47 2009 [dgd] Aidil@Way of the Force: perl definitely has its uses. 
Mon Aug  3 20:27:55 2009 [dgd] Aidil@Way of the Force: but.. 
Mon Aug  3 20:28:00 2009 [dgd] Quixadhal@Bloodlines: I was happy with ruby, as it is a true OO redesign of perl in many senses... and far fewer dollar-signs. 
Mon Aug  3 20:28:12 2009 [dgd] Aidil@Way of the Force: it tends to annoy me within a second of starting to use it. 
Mon Aug  3 20:29:02 2009 [dgd] Silenus@Dead Souls Dev: does ruby let you override "primitives" ? 
Mon Aug  3 20:29:14 2009 [dgd] Silenus@Dead Souls Dev: like smalltalk i believe does... 
Mon Aug  3 20:29:25 2009 [dgd] Aidil@Way of the Force: I find it interesting to note that Noah isn't using the lib he wrote himself for this. 
Mon Aug  3 20:29:30 2009 [dgd] Quixadhal@Bloodlines: Did I mention that the perl scripts were a rather large object-oriented system, so we had things like table objects that directly reflected database queries? :) 
Mon Aug  3 20:29:49 2009 [dgd] Aidil@Way of the Force: heh. 
Mon Aug  3 20:30:14 2009 [dgd] Aidil@Way of the Force: you can write a decent application in perl, but you'd be better off writing it in a real language :) 
Mon Aug  3 20:30:44 2009 [dgd] Quixadhal@Bloodlines: Depends on how you define real, to some degree... perl does let you get very messy. 
Mon Aug  3 20:31:06 2009 [dgd] Quixadhal@Bloodlines: I'm not sure how much you can override in ruby... I know you can extend things at a pretty low level. 
Mon Aug  3 20:31:15 2009 [dgd] Silenus@Dead Souls Dev: ah ok. 
Mon Aug  3 20:31:29 2009 [dgd] Aidil@Way of the Force: ok, lemme redefine.. a language that does what I say instead of trying to guess wrongly at what I mean. 
Mon Aug  3 20:31:40 2009 [dgd] Quixadhal@Bloodlines: Everything is an object, even constants.  "hello".length or 3.times both work. 
Mon Aug  3 20:31:59 2009 [dgd] Silenus@Dead Souls Dev: that sort of sounds like smalltalk/self 
Mon Aug  3 20:32:27 2009 [dgd] Quixadhal@Bloodlines: Hehehehe, so you like C and assembly? *grin* 
Mon Aug  3 20:32:41 2009 [dgd] Aidil@Way of the Force: in a way, yes. 
Mon Aug  3 20:32:56 2009 [dgd] Aidil@Way of the Force: C could use proper string handling tho :) 
Mon Aug  3 20:33:58 2009 [dgd] Aidil@Way of the Force: but I'd rather have the runtime environment barf then trying to make something of rubbish yes. 
Mon Aug  3 20:34:06 2009 [dgd] Quixadhal@Bloodlines: I've come to the conclusion that I really hate C++.  It's not dynamic enough to let me do what I want using objects, and it has too much cruft to just be a better C. 
Mon Aug  3 20:34:17 2009 [dgd] Aidil@Way of the Force agrees with that. 
Mon Aug  3 20:35:17 2009 [dgd] Aidil@Way of the Force: I think my just written json module in lpc is about as big as hello world in c++ 
Mon Aug  3 20:35:20 2009 [dgd] Quixadhal@Bloodlines: What I'd kindof like is a fully dynamic language (with introspection/reflection/whatever you call it these days), but hard types. 
Mon Aug  3 20:36:00 2009 [dgd] Quixadhal@Bloodlines: No bloody "mixed". :) 
Mon Aug  3 20:36:22 2009 [dgd] Silenus@Dead Souls Dev: i have been playing around with C++ it's pretty crufty but I guess so far I am somewhat ok with it. 
Mon Aug  3 20:36:23 2009 [dgd] Aidil@Way of the Force: mixed is useful but potentially dangerous 
Mon Aug  3 20:36:38 2009 [dgd] Silenus@Dead Souls Dev: but there are alot of things one cannot do :/ 
Mon Aug  3 20:37:00 2009 [dgd] Quixadhal@Bloodlines: mixed is lazy... if you can ask what the type of something is, you should be able to properly convert it to what you want.  If you can't, rethink your design. 
Mon Aug  3 20:37:12 2009 [dgd] Silenus@Dead Souls Dev: mixed in LPC probably is the worse performance problem in the language.... 
Mon Aug  3 20:37:15 2009 [dgd] Aidil@Way of the Force: the biggest problem I see with C++ is the people who set the supposed standards. 
Mon Aug  3 20:37:33 2009 [dgd] Quixadhal@Bloodlines: Oh, you mean Microsoft... *grin* 
Mon Aug  3 20:37:46 2009 [dgd] Silenus@Dead Souls Dev: the standards committee? 
Mon Aug  3 20:37:53 2009 [dgd] Aidil@Way of the Force: yep, them. 
Mon Aug  3 20:38:11 2009 [dgd] Silenus@Dead Souls Dev: do they make things a bit too intricate? 
Mon Aug  3 20:38:21 2009 [dgd] Aidil@Way of the Force: I know one former member, and met a few more. They all are convinced that you have to do it their way or you are wrong. 
Mon Aug  3 20:38:30 2009 [dgd] Quixadhal@Bloodlines: Actually, C++ would be a lot further along if Microsoft DID control it, rather than having a full United Nations meeting  for every little detail. 
Mon Aug  3 20:39:00 2009 [dgd] Aidil@Way of the Force: a UN meeting is quite efficient, relatively spoken. 
Mon Aug  3 20:39:10 2009 [dgd] Silenus@Dead Souls Dev: my concern is sometimes perhaps wrongly that especially with C++0x it seems like they are piling features on features 
Mon Aug  3 20:39:26 2009 [dgd] Aidil@Way of the Force: not only are there way too many people trying to steer the thing, they are also quite stubborn people. 
Mon Aug  3 20:39:39 2009 [dgd] Quixadhal@Bloodlines: I actually approve of C# and .NET, although I could wish they had a few less layers, and I still can't adjust to the GUI development tools. 
Mon Aug  3 20:40:00 2009 [dgd] Aidil@Way of the Force: as a language, C# seems ok. 
Mon Aug  3 20:40:21 2009 [dgd] Aidil@Way of the Force: it does rather feel like a ripoff with a few fixes to the original tho. 
Mon Aug  3 20:41:28 2009 [dgd] Silenus@Dead Souls Dev: i wish that languages like C# = more streamlined but i suspect outside of very orthogonal systems with small primitives which ppl dont use much  i equal dreaming. 
Mon Aug  3 20:42:00 2009 [dgd] Silenus@Dead Souls Dev: most languages that seem to see alot of use tend to be pretty complex beasts. 
Mon Aug  3 20:42:54 2009 [dgd] Quixadhal@Bloodlines: .NET tries to allow that, by offloading quite a bit standard library type stuff away from language dependance. 
Mon Aug  3 20:43:33 2009 [dgd] Quixadhal@Bloodlines: It's not perfect, but the idea of pushing that away from the language and into a place any .NET langauge can use is a good one, IMO. 
Mon Aug  3 20:43:52 2009 [dgd] Silenus@Dead Souls Dev: ah interesting. 
Mon Aug  3 20:44:14 2009 [dgd] Silenus@Dead Souls Dev: was thinking of toying with mono a bit 
Mon Aug  3 20:44:28 2009 [dgd] Silenus@Dead Souls Dev: is that reasonably compatible these days? 
Mon Aug  3 20:44:33 2009 [dgd] Quixadhal@Bloodlines: .NET works quite a bit like the JVM, it just abstracts the interface so you can mix and match languages, as long as they all know the interfaces. 
Mon Aug  3 20:44:39 2009 [dgd] Aidil@Way of the Force: its what a language independent framework must do almost by definition, but it works out relatively well I think. 
Mon Aug  3 20:44:51 2009 [dgd] Silenus@Dead Souls Dev: *nod* 
Mon Aug  3 20:45:37 2009 [dgd] Quixadhal@Bloodlines: Haven't tried in a couple years... I wrote a hello world, compiled it in visual studio (command line), copied the executable to linux and ran it in mono. :) 
Mon Aug  3 20:46:05 2009 [dgd] Silenus@Dead Souls Dev: ok. was thinking of trying to port the JIT system but not sure how much work it would be 
Mon Aug  3 20:46:31 2009 [dgd] Quixadhal@Bloodlines: decompiling it to byte code is interesting... the mono version and VS versions are identical in code, but the VS one tacks a bit of extra header data on. 
Mon Aug  3 20:46:45 2009 [dgd] Aidil@Way of the Force: yet here we are, talking on a channel about a vm that tries to offer the most minimalistic runtime environment possible, and leaves everything to the object library. 
Mon Aug  3 20:47:26 2009 [dgd] Quixadhal@Bloodlines: Yeah, if only someone would write something for the LPMUD world to do that... I wonder what it would be called... *looks around the room* 
Mon Aug  3 20:47:49 2009 [dgd] Kalinash@Fire and Ice: Sil will do it 
Mon Aug  3 20:48:09 2009 [dgd] Silenus@Dead Souls Dev: isnt dgd pretty close? ;-) 
Mon Aug  3 20:48:50 2009 [dgd] Silenus@Dead Souls Dev: my main problem right now with my server project is 
Mon Aug  3 20:49:03 2009 [dgd] Silenus@Dead Souls Dev: wrapping my sensibilities around call_other and mixed. 
Mon Aug  3 20:49:26 2009 [dgd] Silenus@Dead Souls Dev: or convincing myself they are needed. 
Mon Aug  3 20:50:18 2009 [dgd] Quixadhal@Bloodlines: mixed is like the evil brother of C's void.  It would be what you'd get if declaring a void object got you automatic memory allocation attatched to it, and the compiler/interpreter automatically tried to coerce it into something useful by context. 
Mon Aug  3 20:50:58 2009 [dgd] Silenus@Dead Souls Dev: its just that with the LPC implementation you carry a type tag 
Mon Aug  3 20:51:01 2009 [dgd] Quixadhal@Bloodlines: It's the point in LPC where Aidil cries, since the interpreter can guess wrong every so often. :) 
Mon Aug  3 20:51:03 2009 [dgd] Aidil@Way of the Force: and yet, without mixed, no polymorphism in lpc 
Mon Aug  3 20:51:20 2009 [dgd] Silenus@Dead Souls Dev: can you not have something like java's C#'s setup? 
Mon Aug  3 20:51:21 2009 [dgd] Aidil@Way of the Force: and no, the compiler can't guess wrong. 
Mon Aug  3 20:51:42 2009 [dgd] Silenus@Dead Souls Dev: it wouldnt be lpc anymore tho 
Mon Aug  3 20:51:45 2009 [dgd] Aidil@Way of the Force: the runtime environment might guess wrong :) 
Mon Aug  3 20:52:11 2009 [dgd] Silenus@Dead Souls Dev: well with type tags the data is labelled in most LPC VM's 
Mon Aug  3 20:52:25 2009 [dgd] Silenus@Dead Souls Dev: so unless there is a bug it shouldnt go wrong i dont think. 
Mon Aug  3 20:52:38 2009 [dgd] Aidil@Way of the Force: exactly. if you want to create a new language, theres a bunch of things in which you don't wanna copy lpc. 
Mon Aug  3 20:53:09 2009 [dgd] Aidil@Way of the Force: indeed. if wrong, its in the runtime environment, but since the runtime environment knows what it stuck in there in the first place, it won't go wrong. 
Mon Aug  3 20:53:52 2009 [dgd] Aidil@Way of the Force: mixed says you can stick anything in there, but what is in there is typed data still. 
Mon Aug  3 20:54:21 2009 [dgd] Silenus@Dead Souls Dev: yeah it would seem that down casting up casting + having virtual function pointers might be more suitable and easier to optimize than what lpc does. 
Mon Aug  3 20:54:24 2009 [dgd] Quixadhal@Bloodlines: I seem to remember issues with things like a + "hello" being an error, whereas "hello" + a concatonated the integer as a string without complaint... where a is mixed and happens to have an integer assigned to it. 
Mon Aug  3 20:54:33 2009 [dgd] Aidil@Way of the Force: and you are quite able to hide your own mistakes that way. 
Mon Aug  3 20:55:13 2009 [dgd] Silenus@Dead Souls Dev: call_other = a nightmare construct for any form of static analysis. 
Mon Aug  3 20:55:47 2009 [dgd] Aidil@Way of the Force: yes, Sil, which is why originally, lpmud 3.x required you to declare the return type of a call_other with a cast. 
Mon Aug  3 20:56:10 2009 [dgd] Aidil@Way of the Force: since that is the only way the compiler can know. 
Mon Aug  3 20:56:19 2009 [dgd] Aidil@Way of the Force: dgd resolves this at runtime. 
Mon Aug  3 20:56:39 2009 [dgd] Silenus@Dead Souls Dev: *nod* i deduced that to some extent as well. 
Mon Aug  3 20:56:42 2009 [dgd] Silenus@Dead Souls Dev: when working on this. 
Mon Aug  3 20:56:51 2009 [dgd] Kalinash@Fire and Ice: in soviet russia runtime resolves dgd 
Mon Aug  3 20:57:41 2009 [dgd] Silenus@Dead Souls Dev: well in alot of lpc variants Quix i think that the compile time checking might be semi porous 
Mon Aug  3 20:57:49 2009 [dgd] Aidil@Way of the Force: re: quix, that seems more about variables being converted to a string, you'll find similar behavior for ints and floats. 
Mon Aug  3 20:57:58 2009 [dgd] Silenus@Dead Souls Dev: i.e. the annotations are mainly there to help the person writing code 
Mon Aug  3 20:58:10 2009 [dgd] Aidil@Way of the Force: if the mixed would be containing an array, it wouldn't even try and throw an error inmediately. 
Mon Aug  3 20:58:11 2009 [dgd] Silenus@Dead Souls Dev: and are discarded to some extent 
Mon Aug  3 20:58:29 2009 [dgd] Silenus@Dead Souls Dev: except for operator runtime checks 
Mon Aug  3 20:59:03 2009 [dgd] Silenus@Dead Souls Dev: i.e. everything is just a struct with a type code. 
Mon Aug  3 21:00:01 2009 [dgd] Quixadhal@Bloodlines: Heh, now extend that idea to "everything is just an object with a type code" and we're back to ruby. 
Mon Aug  3 21:00:09 2009 [dgd] Silenus@Dead Souls Dev: i founnd two problems potentially with LPC and mixed 
Mon Aug  3 21:00:17 2009 [dgd] Silenus@Dead Souls Dev: which are hard to optimize 
Mon Aug  3 21:00:32 2009 [dgd] Silenus@Dead Souls Dev: one is operators like + are so heavily overloaded 
Mon Aug  3 21:00:47 2009 [dgd] Silenus@Dead Souls Dev: that you cannot really tell which function to call until runtime 
Mon Aug  3 21:01:01 2009 [dgd] Silenus@Dead Souls Dev: + runtime typing 
Mon Aug  3 21:01:07 2009 [dgd] Silenus@Dead Souls Dev: i think in tandem 
Mon Aug  3 21:01:16 2009 [dgd] Silenus@Dead Souls Dev: it makes type inference a problem. 
Mon Aug  3 21:01:26 2009 [dgd] Aidil@Way of the Force: oh and Quixadhal, on dgd, your example would work for both cases :) 
Mon Aug  3 21:01:32 2009 [dgd] Silenus@Dead Souls Dev: since the compiler doenst have much to go on. 
Mon Aug  3 21:02:12 2009 [dgd] Quixadhal@Bloodlines: Aidil - Ah, so it doesn't do strict left-to-right type evaluation?  Good. :) 
Mon Aug  3 21:03:00 2009 [dgd] Aidil@Way of the Force: it goes for 'if its an int or float and being added to a string, instead concat the string representation of that int or float'. 
Mon Aug  3 21:03:40 2009 [dgd] Quixadhal@Bloodlines: Part of the confusion, at least with some LP dialects, is the fact that types are almost viral.  They hang onto the data, not the variable... so you can get wacky things like int i; i = (float) 3.7; and i is now magically a float! 
Mon Aug  3 21:03:44 2009 [dgd] Silenus@Dead Souls Dev: yeah int float gets "promoted" to a string then concat. 
Mon Aug  3 21:04:17 2009 [dgd] Aidil@Way of the Force: heh, on many lpc dialects that is possible. 
Mon Aug  3 21:04:30 2009 [dgd] Silenus@Dead Souls Dev: well they seem to be a bit of an after thought since the system doesnt really have different boxes for differen types. 
Mon Aug  3 21:04:48 2009 [dgd] Aidil@Way of the Force: and if we weren't using typechecking 2 with gurbalib, it would be possible as well. 
Mon Aug  3 21:05:08 2009 [dgd] Quixadhal@Bloodlines: To me, that's just evil.  i = (int) 3.7 normally would make i = 3 (i = 4 in perl, which would round)... changing the type of something not mixed is just wrong. 
Mon Aug  3 21:05:31 2009 [dgd] Silenus@Dead Souls Dev: *nod* 
Mon Aug  3 21:05:57 2009 [dgd] Silenus@Dead Souls Dev: you can typecheck that more carefully probably. 
Mon Aug  3 21:06:01 2009 [dgd] Aidil@Way of the Force: on dgd that will throw an error 
Mon Aug  3 21:06:09 2009 [dgd] Aidil@Way of the Force: unless you cast to int first. 
Mon Aug  3 21:06:20 2009 [dgd] Quixadhal@Bloodlines: I'd rather see i = (float) 3.7 generate an error of "Hey moron, if you want a float, declare a float!" :) 
Mon Aug  3 21:06:32 2009 [dgd] Silenus@Dead Souls Dev: hehe 
Mon Aug  3 21:06:42 2009 [dgd] Aidil@Way of the Force: heh. 
Mon Aug  3 21:07:08 2009 [dgd] Silenus@Dead Souls Dev: anyhow i am off 
Mon Aug  3 21:07:11 2009 [dgd] Silenus@Dead Souls Dev: later guys 
Mon Aug  3 21:07:17 2009 [dgd] Aidil@Way of the Force: have a good day :) 
Mon Aug  3 21:07:19 2009 [dgd] Quixadhal@Bloodlines: cya! 
Mon Aug  3 21:07:29 2009 [dgd] Kalinash@Fire and Ice floats. 
Mon Aug  3 21:07:51 2009 [dgd] Aidil@Way of the Force: enough air in there for that yes. 
Mon Aug  3 21:08:09 2009 [dgd] Quixadhal@Bloodlines: Surprising though, given the density. 
Mon Aug  3 21:08:21 2009 [dgd] Kalinash@Fire and Ice: :) 
Mon Aug  3 21:09:18 2009 [dgd] Quixadhal@Bloodlines: Rails huh?  Silly humans, taking a perfectly good language and hanging it on a web server where it gets into all kinds of trouble. 
Mon Aug  3 21:11:20 2009 [dgd] Quixadhal@Bloodlines: That's probably why ruby-pg isn't being taken care of anymore... they expect you to install rails and just use that. 
Mon Aug  3 21:12:59 2009 [dgd] Aidil@Way of the Force: silly humans.. heh. 
Mon Aug  3 21:13:04 2009 [dgd] Aidil@Way of the Force: whats new there :) 
Tue Aug  4 10:21:06 2009 [dgd] Subversion@Way of the Force: Aidil committed Gurbalib revision 303 to svn://wotf.org/gurbalib : Added support for using json objects as values in another json object  
Tue Aug  4 12:01:32 2009 [dgd] Aidil@Way of the Force: ok. this IMC3 discussion on mudbytes seems rather pointless. 
Tue Aug  4 12:02:42 2009 [dgd] Aidil@Way of the Force: either you use a structured data format, and deal with proper escaping, and take into account that in order to not be ambigious, it will have some overhead, or you try to get around that by reinventing it. 
Tue Aug  4 12:03:13 2009 [dgd] Aidil@Way of the Force: the end result will be lots of time wasted on reinventing the weheel :) 
Tue Aug  4 12:03:19 2009 [dgd] Aidil@Way of the Force: wheel also. 
Tue Aug  4 12:04:05 2009 [dgd] Aidil@Way of the Force: and when reading this entire discussion, it exactly shows why mudmode is in fact very useful, and why you need something somewhat equivalent. 
Tue Aug  4 16:03:14 2009 [dgd] Aidil@Way of the Force snores. 
Sun Aug  9 19:59:11 2009 [dgd] Subversion@Way of the Force: Aidil committed Gurbalib revision 313 to svn://wotf.org/gurbalib : Added a misc document describing the security system, added manpage for unguarded() afun, fixed documentation for require_priv() afun  
Sun Aug  9 20:01:31 2009 [dgd] Subversion@Way of the Force: Aidil committed Gurbalib revision 314 to svn://wotf.org/gurbalib : /sys/daemons/data is no longer world readable  
first << < 33 34 35 36 37 38 39 40 41 42 43 > >> last

Go to the top | Channel Index