| --- Log opened Wed Jan 09 00:00:57 2008 |
| 00:01 | <@Vornicus> | Well, then you're stuck with Javascript |
| 00:01 | < Alek> | heh. |
| 00:02 | < Alek> | thanks. |
| 00:02 | < Alek> | I'm out. |
| 00:02 | | Alek [~omegaboot@Nightstar-28320.dsl.emhril.sbcglobal.net] has left #code [] |
| 00:12 | | Reiv [~dogmatix@Nightstar-2154.dsl.telstraclear.net] has joined #Code |
| 00:32 | | Thaqui [~Thaqui@Nightstar-14364.jetstream.xtra.co.nz] has joined #code |
| 00:32 | | mode/#code [+o Thaqui] by ChanServ |
| 00:48 | | gnolam [lenin@Nightstar-10613.8.5.253.static.se.wasadata.net] has quit [Quit: Z?] |
| 01:31 | | Vornicus [~vorn@Admin.Nightstar.Net] has quit [Ping Timeout] |
| 01:40 | | Vornicus [~vorn@Admin.Nightstar.Net] has joined #code |
| 01:40 | | mode/#code [+o Vornicus] by ChanServ |
| 02:19 | | C_tiger [~c_wyz@Nightstar-5325.nycmny.east.verizon.net] has quit [Quit: And away she goes!] |
| 02:28 | | Reiv is now known as Clank |
| 03:04 | | Attilla [~The.Attil@194.72.70.ns-11849] has quit [Connection reset by peer] |
| 03:15 | <@McMartin> | So, a sign of how leeged up the UQM code base is. |
| 03:15 | <@McMartin> | After five years, were are finally wihtin striking distance of actually being able to use malloc and free to do our memory allocation. |
| 03:16 | < Clank> | ...what. |
| 03:17 | < Clank> | I gather you've been forced to be fiddling around the edges, so to speak, rather than attacking the core? |
| 03:17 | <@McMartin> | Enormous amounts of code relied on the ability to cram usable pointer-equivalents into 16-bit values, which were worked into bitmasks etc. |
| 03:17 | < Clank> | ...owwww |
| 03:18 | <@McMartin> | There were special routines to convert between those values and usable pointers. |
| 03:18 | <@McMartin> | (And I think the problem this was originally intended to solve was unreliable compiler support for value structs back in 1991.) |
| 03:19 | <@McMartin> | Anyway, as of about four days ago I was able to successfully break everything so that these 16-bit values stand alone. |
| 03:20 | <@McMartin> | Which caught a bunch of places where they were also being used to not store pointer values. |
| 03:20 | <@McMartin> | That was fun. |
| 03:21 | <@McMartin> | Though easy to find, since narrowing conversions are always a warning on MSVC. |
| 03:22 | <@Vornicus> | mass insanity |
| 03:23 | <@McMartin> | Yeah, well, when you have a platform where you can only keep one pointer in registers at any given time, and it takes a third of them... |
| 03:23 | < Clank> | ...what else would you /use/ those things for? |
| 03:23 | <@McMartin> | The big thing is the ability to keep a pointer to a collection and an index into it as a single value. |
| 03:23 | <@McMartin> | Which is 32-bits. |
| 03:24 | <@McMartin> | There was a lot of code that was also using those 32-bit values to store resource identifiers, replacing them with a memory-handle/index pair once it was loaded. |
| 03:24 | <@McMartin> | Which isn't actually insanity as long as you actually use unions for it, but, again, returning structs from functions in 1991. |
| 03:26 | <@Vornicus> | :( |
| 03:32 | <@Vornicus> | that is pretty damn leeged up |
| 03:33 | <@McMartin> | That said, it did give us leak detection for free. |
| 03:38 | | Chalcedon [~Chalcedon@Nightstar-2443.ue.woosh.co.nz] has joined #code |
| 03:38 | | mode/#code [+o Chalcedon] by ChanServ |
| 03:39 | <@Vornicus> | what, if you threw in a wild pointer it'd explode? |
| 03:41 | <@McMartin> | No |
| 03:41 | <@McMartin> | Those 16-bit values were indices into a 65,536-element array of memory extents. |
| 03:41 | <@McMartin> | So if at points where nothing was supposed to be allocated, there were still used array eleemnts, You Had A Leak. |
| 03:42 | <@McMartin> | Also, for major leaks likes the ones that were in the Draw Queue, it's much easier to run out of 65k extents than it is to run out of 1G RAM. |
| 03:42 | <@McMartin> | Well, 64k. |
| 04:04 | | Clank is now known as Reiv |
| 05:00 | | C_tiger [~c_wyz@Nightstar-5325.nycmny.east.verizon.net] has joined #code |
| 07:43 | | Attilla [~The.Attil@194.72.70.ns-11849] has joined #code |
| 08:23 | | You're now known as TheWatcher |
| 08:36 | | Chalcedon [~Chalcedon@Nightstar-2443.ue.woosh.co.nz] has quit [Quit: Gone] |
| 09:38 | | Vornicus is now known as Vornicus-Latens |
| 10:27 | | Thaqui [~Thaqui@Nightstar-14364.jetstream.xtra.co.nz] has quit [Operation timed out] |
| 10:36 | | Thaqui [~Thaqui@Nightstar-14364.jetstream.xtra.co.nz] has joined #code |
| 10:36 | | mode/#code [+o Thaqui] by ChanServ |
| 10:48 | | AnnoDomini [AnnoDomini@Nightstar-29237.neoplus.adsl.tpnet.pl] has quit [Ping Timeout] |
| 10:55 | | AnnoDomini [AnnoDomini@83.21.14.ns-3554] has joined #Code |
| 10:55 | | mode/#code [+o AnnoDomini] by ChanServ |
| 11:02 | | Thaqui [~Thaqui@Nightstar-14364.jetstream.xtra.co.nz] has left #code [Leaving] |
| 12:23 | | gnolam [lenin@85.8.5.ns-20483] has joined #Code |
| 12:23 | | mode/#code [+o gnolam] by ChanServ |
| 13:11 | | Reiv [~dogmatix@Nightstar-2154.dsl.telstraclear.net] has quit [Quit: Laptop fall down go boom...] |
| 13:58 | <@gnolam> | <X-G> http://www.bambooweb.com/articles/r/i/Richard_Stallman.html Hum. What a difference it makes when RMS shaves a bit. |
| 14:44 | | * gnolam ponders rewriting the entire rendering system. |
| 16:25 | | Xiphias [Ameroth@Nightstar-25243.dsl.in-addr.zen.co.uk] has joined #code |
| 16:44 | | Xiphias [Ameroth@Nightstar-25243.dsl.in-addr.zen.co.uk] has quit [Quit: I was never gone] |
| 17:02 | | You're now known as TheWatcher[afk] |
| 18:14 | | Vornicus-Latens [~vorn@Admin.Nightstar.Net] has quit [Ping Timeout] |
| 18:20 | | You're now known as TheWatcher |
| 18:27 | | AnnoDomini is now known as Pete |
| 18:49 | <@Pete> | Fuck... I'm confused about something and wikipedia is down. |
| 18:50 | <@Pete> | Is saturation, in transistors, where the function is nearly a horizontal line? |
| 18:50 | <@Pete> | Or is it the part before that? |
| 19:49 | | Otto_Flick [lenin@Nightstar-10613.8.5.253.static.se.wasadata.net] has joined #Code |
| 19:50 | | gnolam [lenin@85.8.5.ns-20483] has quit [Ping Timeout] |
| 19:50 | | Otto_Flick is now known as gnolam |
| 20:06 | | gnolam [lenin@Nightstar-10613.8.5.253.static.se.wasadata.net] has quit [Ping Timeout] |
| 20:42 | <@jerith> | Pete: Where the function is horizontal, iirc. |
| 20:42 | <@Pete> | I see. Thanks. |
| 20:42 | <@jerith> | It's been several years since I did any semiconductor physics, though. |
| 21:05 | | Vornotron [~vorn@Admin.Nightstar.Net] has joined #code |
| 21:06 | | gnolam [lenin@Nightstar-10613.8.5.253.static.se.wasadata.net] has joined #Code |
| 21:06 | | mode/#code [+o gnolam] by ChanServ |
| 21:08 | | Vornotron is now known as Vornicus |
| 21:10 | | Vornicus is now known as NSGuest-4230 |
| 21:10 | | NSGuest-4230 is now known as Vornicus |
| 21:54 | | AnnoDomini [AnnoDomini@83.21.72.ns-4440] has joined #Code |
| 21:54 | | mode/#code [+o AnnoDomini] by ChanServ |
| 21:55 | | Pete [AnnoDomini@83.21.14.ns-3554] has quit [Ping Timeout] |
| 21:55 | | AnnoDomini is now known as Pete |
| 21:59 | | Pete [AnnoDomini@83.21.72.ns-4440] has quit [Connection reset by peer] |
| 22:00 | | AnnoDomini [AnnoDomini@83.21.72.ns-4440] has joined #Code |
| 22:00 | | mode/#code [+o AnnoDomini] by ChanServ |
| 22:00 | | AnnoDomini is now known as Pete_ |
| 22:03 | | Pete_ is now known as Pete |
| 22:32 | | Pete is now known as AnnoDomini |
| --- Log closed Wed Jan 09 23:46:35 2008 |
| --- Log opened Wed Jan 09 23:47:08 2008 |
| 23:47 | | TheWatcher [~chris@82.71.33.ns-26612] has joined #code |
| 23:47 | | Irssi: #code: Total of 17 nicks [7 ops, 0 halfops, 0 voices, 10 normal] |
| 23:47 | | mode/#code [+o TheWatcher] by ChanServ |
| 23:47 | | Irssi: Join to #code was synced in 47 secs |
| --- Log closed Thu Jan 10 00:00:03 2008 |