00:30:08 -!- LiamH [~none@96.231.218.61] has quit [Quit: Leaving.] 01:05:23 -!- tcr1 [~tcr@84-72-21-32.dclient.hispeed.ch] has quit [Quit: Leaving.] 01:15:49 -!- hydan [~udzinari@ip-89-102-13-27.net.upcbroadband.cz] has quit [Quit: Computer has gone to sleep.] 01:35:05 LiamH [~none@96.231.218.61] has joined #sbcl 01:43:29 -!- Fare [fare@nat/google/x-fsohkzeomjahhefg] has quit [Ping timeout: 248 seconds] 02:13:38 Fare [fare@nat/google/x-ailguqekkimtdtwm] has joined #sbcl 02:18:39 prxq_ [~mommer@mnhm-4d011893.pool.mediaWays.net] has joined #sbcl 02:22:12 -!- prxq [~mommer@mnhm-5f75dcb0.pool.mediaWays.net] has quit [Ping timeout: 256 seconds] 03:12:10 -!- Posterdati [~antani@host106-212-dynamic.10-87-r.retail.telecomitalia.it] has quit [Ping timeout: 252 seconds] 03:17:51 well, people are getting papers published trying to re-implement Python's type and propagation system. 03:18:04 "Type systems involving unions, intersections and negations require a 03:18:12 subtype operator which is non-trivial to implement." 03:19:01 Krystof: your headaches were warranted ;) 03:26:07 Posterdati [~antani@host201-213-dynamic.11-87-r.retail.telecomitalia.it] has joined #sbcl 04:31:07 -!- zulu_inuoe [~quassel@184.89.111.53] has quit [Read error: Connection reset by peer] 04:32:23 zulu_inuoe [~quassel@184.89.111.53] has joined #sbcl 04:33:34 -!- Dalek_Baldwin [~Adium@108-225-26-178.lightspeed.irvnca.sbcglobal.net] has quit [Quit: Leaving.] 04:45:40 -!- LiamH [~none@96.231.218.61] has quit [Quit: Leaving.] 04:45:57 Dalek_Baldwin [~Adium@71-84-34-33.dhcp.mtpk.ca.charter.com] has joined #sbcl 05:20:33 -!- Thra11 [~thrall@54.110.112.87.dyn.plus.net] has quit [Ping timeout: 245 seconds] 05:29:51 -!- Fare [fare@nat/google/x-ailguqekkimtdtwm] has quit [Quit: Leaving] 05:42:04 -!- psilord [~psilord@c-69-180-173-249.hsd1.mn.comcast.net] has quit [Quit: Leaving.] 06:38:00 edgar-rft [~GOD@HSI-KBW-149-172-63-75.hsi13.kabel-badenwuerttemberg.de] has joined #sbcl 07:03:09 snowylike [~sn@91-67-171-156-dynip.superkabel.de] has joined #sbcl 07:06:04 sdemarre [~serge@91.180.66.31] has joined #sbcl 07:23:16 stassats [~stassats@wikipedia/stassats] has joined #sbcl 07:24:19 -!- sdemarre [~serge@91.180.66.31] has quit [Ping timeout: 248 seconds] 07:32:40 hydan [~udzinari@ip-89-102-13-27.net.upcbroadband.cz] has joined #sbcl 07:41:19 easye` [~user@213.33.70.157] has joined #sbcl 07:43:11 -!- easye [~user@213.33.70.157] has quit [Read error: Connection reset by peer] 07:44:55 our random seed is really bad if /dev/urandom is not present 08:47:27 -!- flip214 [~marek@unaffiliated/flip214] has quit [Ping timeout: 260 seconds] 08:55:14 -!- hydan [~udzinari@ip-89-102-13-27.net.upcbroadband.cz] has quit [Quit: Computer has gone to sleep.] 09:33:42 -!- cmm [~cmm@109.64.206.139] has quit [Ping timeout: 276 seconds] 09:33:45 cmm- [~cmm@bzq-109-64-206-139.red.bezeqint.net] has joined #sbcl 09:34:27 tcr1 [~tcr@84-72-21-32.dclient.hispeed.ch] has joined #sbcl 09:42:37 attila_lendvai [~attila_le@87.247.13.18] has joined #sbcl 09:42:37 -!- attila_lendvai [~attila_le@87.247.13.18] has quit [Changing host] 09:42:37 attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has joined #sbcl 09:48:34 flip214 [~marek@86.59.100.100] has joined #sbcl 09:48:34 -!- flip214 [~marek@86.59.100.100] has quit [Changing host] 09:48:34 flip214 [~marek@unaffiliated/flip214] has joined #sbcl 10:28:48 -!- tcr1 [~tcr@84-72-21-32.dclient.hispeed.ch] has quit [Quit: Leaving.] 11:33:09 pkhuong: I hope they're all citing Baker at least 11:36:55 leoc [~leoc.git@p5DDB8866.dip.t-dialin.net] has joined #sbcl 12:28:23 -!- attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has quit [Quit: Leaving.] 12:41:11 -!- edgar-rft [~GOD@HSI-KBW-149-172-63-75.hsi13.kabel-badenwuerttemberg.de] has quit [Ping timeout: 244 seconds] 12:44:12 stassats: you mean the default initialization of *random-state*? 12:44:44 not the default, the seeding 12:45:21 lp 1102748 12:45:22 https://bugs.launchpad.net/bugs/1102748 12:49:11 ok i see 12:51:50 but unix-getpid is actually present on windows, so it can be slightly rectified by just removing #-win32 12:54:29 windows has an rng that is supposed to be like /dev/random 12:54:42 -!- snowylike [~sn@91-67-171-156-dynip.superkabel.de] has quit [Read error: Connection reset by peer] 12:54:51 iirc. 12:55:10 CryptGenRandom is what google says 12:56:44 that's one the anton patches I haven't picked yet :-( 12:56:56 can send you a diff if you like 13:01:29 -!- prxq_ is now known as prxq 13:04:58 i tried to make it use CryptGenRandom today, but i lost all patience trying to set up sbcl and slime on windows 13:07:31 i guess i should just set up a remote swank and use it from a linux machine 13:33:59 ... maybe we should pool money for an Amazon Windows instance that trusted SBCL developers can all connect to, with most stuff already set up ... 13:43:43 lichtblau: if you ask real nice they might even "donate" one 13:43:58 after all, their cost is near to zero 13:48:42 is that a standard practice? 13:51:07 we also get occasional offers for donations 13:51:51 usually I tell them to talk to other developers, but maybe we should have a wishlist (e.g. arm boards, AWS instances, whatever) that people could buy to help sbcl development 13:53:10 i'd go with food and beverages 13:53:45 arm boards, AWS instances, food, beer, and MORE TIME 13:54:03 baby sitting? 13:54:10 sheesh 13:54:31 I think it might be worth thinking about what £2/£20/£200/£2000 would be used for 13:54:35 "i'll sit with your baby while you fix lp#xxxx" 14:00:19 for me, £2 is "a beer next time we meet", £20 is "a raspberry Pi", £200 is "treat yourself" and £2000 is "buy your teaching time out and spend the time on sbcl development" 14:00:39 the interesting thing there is that £200 is less useful to me (from the point of view of sbcl development, anyway) than £20 14:00:50 but maybe £200 is "a year of Windows AWS" or something 14:00:58 then Someone Else can be useful with it 14:24:50 LiamH [~none@pdp8.nrl.navy.mil] has joined #sbcl 14:39:50 so, ideas on a postcard 14:40:14 everyone send me what they would need that they don't have for a bit of motivation/productivity 14:40:42 time 14:44:44 I hear you 14:45:04 so, how much money do you need to be able to buy a noticeable amount of time? 14:46:53 psilord [~psilord@23-25-144-217-static.hfc.comcastbusiness.net] has joined #sbcl 14:48:51 Krystof: nope, not at all. "The most notable existing system" is Typed Racket, anyway ;) (also, it's sound & complete subtyping, but not polytime. Good stuff.) 15:00:46 Thra11 [~thrall@54.110.112.87.dyn.plus.net] has joined #sbcl 15:12:20 I think SBCL devs can get free accounts on http://gcc.gnu.org/wiki/CompileFarm 15:13:54 So if you want to help port sbcl to ARM and don't have an arm machine, that could help you. :) 15:16:08 snowylike [~sn@91-67-171-156-dynip.superkabel.de] has joined #sbcl 15:19:58 not very motivating without an arm machine 15:22:09 So... I'm concerned that on RISC platforms we keep accumulating test failures that aren't even marked "expected". Also, for Windows the testers aren't developers. 15:22:58 Is there any way I could convince the new release management administration to check boinkmarks for those regressions before releasing? So that the culprits could be notified and could fix their stuff every one in a while? 15:23:29 The boinkmarks already cover Linux/ppc, Solaris/sparc, Linux/x86+amd64, plus wine. Linux/MIPS isn't set up yet, but I have the machine. 15:28:09 lichtblau: if the false error rate is low, you could have it just send emails. 15:45:04 hmm. I don't think we're in a position to add features ATM. 15:45:19 The RSS feed Krystof had asked for has been added though, I think. 15:50:50 I can do a certain amount, yes 15:51:25 where's the feed? 15:51:31 Ideally all developers would check proactively, of course, and you'd never run into issues :-). 15:52:06 RSS link at: http://src.knowledgetools.de/boinkmarks/tests 15:53:38 thanks 15:54:45 erm, it's hanging at Dec 18 :-( Alerting hlavaty as we speak! 15:57:19 Eh...developers will probably never reliably go proactively check a thing. *Ideally* an auto-test system can auto-nag people who break stuff within an hour or two of their change. (Even more ideally: before their change is merged). 15:57:40 But certainly telling people to unreliably check a thing is better than no thing at all! 15:58:24 wbooze [~wbooze@xdsl-78-35-186-218.netcologne.de] has joined #sbcl 15:58:28 foom: a wild idea brews... Do you have any sense of how google feels about sbcl? 15:59:14 Agreed. So far we're evidently lagging behind significantly though. So our infrastructure isn't ready yet for "test automatically within minutes/hours after pushing", much less "automatically test before pushing". 15:59:41 Krystof: As a whole, I'd suppose "doesn't care". In my little corner? We use it a bunch. :) 15:59:42 the thought process goes: our most lacking resource is time. Grad students have infinite time. Maybe funding PhD places is the way forward? 16:00:20 Do you know if that's a thing that Google does? 16:00:37 I don't know, though it wouldn't surprise me 16:01:03 I mean, I certainly don't have influence to start doing something like that that's not being done. 16:01:21 a quick google ( :-) finds http://googleresearch.blogspot.co.uk/2012/06/2012-google-phd-fellowships.html 16:01:38 Krystof: the next problem is finding PhD students 16:01:46 finding good ones 16:01:48 of the right kind :-) 16:01:52 exactly 16:01:52 yeah 16:03:05 that would be a nice problem to have 16:09:22 a wild idea would be to have such a grant to be paid by the stichting, which could take the money from donations. 16:10:21 I have a more fundamental problem: who'd advise? 16:18:25 it would depend on the project description. 16:18:58 attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has joined #sbcl 16:34:07 The main issue would be to come up with a research project that warrants such a setup. The PhD student should get a PhD at the end, and not feel as if he/she has to hide something. 16:38:23 An easier way to recruit contributors would be to start by adding a page "How to contribute" to sbcl.org, eventually also detailing a whishlist of things that would be good to have. 16:40:38 -!- attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has quit [Quit: Leaving.] 16:40:58 "going to the launchpad and fixing bugs" seems like an obvious things to figure out for someone who actually wants to contribute 16:42:38 part of the point of such a page would be to show that you actually want help, which IMO is not all that obvious. 16:43:14 what open-source project doesn't want help? 16:43:56 or maybe there could non-programming contributions, "draw a picture of the kitten of death" 16:48:37 stassats: just because it is obvious doesn't mean that there is no value in saying it. 16:49:14 "you can contribute by compiling the list of things to do to contribute" 16:50:32 and besides, I do not think it is all that obvious. Part of being willing to accept help is being lenient and friendly if people do not help as much as they hoped. 16:54:14 attila_lendvai [~attila_le@92.47.253.199] has joined #sbcl 16:54:14 -!- attila_lendvai [~attila_le@92.47.253.199] has quit [Changing host] 16:54:14 attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has joined #sbcl 17:00:05 i'd say that figuring out how to help would be a good test to see if the person will actually be able to provide help 17:00:58 pkhuong: I can see problems around supervising a project, sure 17:01:03 otherwise what kind of help is that if takes more resources to guide the person than doing it yourself? 17:01:09 stassats: an investment 17:01:25 hi 17:01:28 please I need help with sblc on debian wheezy amd64: it sometimes hangs during quickload 17:01:36 Krystof: what is the return rate? 17:01:46 logarithmic 17:01:51 lol 17:01:52 stassats: expecting people to work out what the bottlenecks are in a project when, frankly, we're not even sure ourselves, before they are allowed to contribute in any way, seems a poor way to get any return at all 17:01:55 :) 17:02:47 the return rate has been quite good for me so far: some people whose first attempts weren't perfect nevertheless have become useful and productive members of the team 17:03:23 (rather more productive than me personally: so you could say that the return rate was substantially above 100%) 17:04:35 "I'm not going to teach you computing; if you can't figure it out yourself you'll never be any good at it" 17:04:49 what would be your suggestion as to the content of a "how to help" page? 17:06:28 something along the lines of "The most lasting contribution that any individual can make to the project is by improving the code and documentation; there are many items of differing priorities or difficulties in the bug tracker, or you may have some issue to solve yourself." (cont'd) 17:08:18 "if you cannot make direct contributions, there are nevertheless ways in which you could help the project indirectly: communicating about its needs, indicating your own pressing needs, or directing available resources to places where they can be deployed to help. Currently, the following items are desired by the following people in order to improve their effectiveness in their sbcl-related work: ..." 17:08:22 my recipe for contribution is usually (even if i don't intent to contribute originally) is "pick a piece of software, use it, you'll inevitably will run into problems, try to diagnose it to figure a reproducible bug-report, sometimes during that process it'll be clear how to fix the problem, although a good bug-report is a good contribution already" 17:08:38 you are not the only kind of person in the world 17:09:05 hydan [~udzinari@ip-89-102-13-27.net.upcbroadband.cz] has joined #sbcl 17:09:06 there are plenty of people who might want to help a project along without devoting any time to it 17:09:34 "I have no desire to learn about compilers. But something I depend on and use every day is written in Lisp" 17:09:42 "you guys seem cool" 17:12:34 there's already HACKING 17:13:03 stassats: don't do it if you don't think it's worthwhile. 17:13:03 -!- attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has quit [Quit: Leaving.] 17:13:47 at the moment, irrespective of the shortage of hackers, we are not effectively harnessing the proven existence of people who want to give us other resources 17:14:38 a "donate" button? 17:15:03 you'd be surprised how hard it is to get such a thing going. 17:16:25 a donate button requires a bank account 17:16:33 and significant trust between all potential parties 17:17:55 and a way to avoid troubles with tax authorities 17:18:24 just open it on the cayman islands 17:19:30 and then everyone wants to know if it's a 501.(c)3 or whatever 17:19:39 and someone will want audited accounts 17:19:51 that's the kind of thing the Stichting wanted to be good for. 17:20:02 it's not impossible, but there I'd be more worried about the return on investment than just helping people find out how they can help 17:23:17 so there's something to contribute. Figure out a way to make sbcl accept and use donations. 17:23:41 heh 17:36:30 I can try to ask the Stichting without mentioning that it is sbcl. 17:37:07 would there be objections against the Stichting doing that? 17:38:07 not from me 17:40:58 pnpuff [~dioxirane@unaffiliated/pnpuff] has joined #sbcl 17:41:26 -!- hydan [~udzinari@ip-89-102-13-27.net.upcbroadband.cz] has quit [Quit: ...] 17:48:45 ASau [~user@46.115.34.175] has joined #sbcl 17:54:55 -!- pnpuff [~dioxirane@unaffiliated/pnpuff] has left #sbcl 17:57:49 it looks as if the stichting could use help themselves, judging from their web page. 18:22:17 Thra11_ [~thrall@94.8.113.87.dyn.plus.net] has joined #sbcl 18:24:33 -!- Thra11 [~thrall@54.110.112.87.dyn.plus.net] has quit [Ping timeout: 248 seconds] 18:37:21 -!- stassats [~stassats@wikipedia/stassats] has quit [Ping timeout: 248 seconds] 18:40:53 -!- Thra11_ [~thrall@94.8.113.87.dyn.plus.net] has quit [Read error: Connection reset by peer] 19:06:08 leuler [~user@p548FBA85.dip.t-dialin.net] has joined #sbcl 19:14:41 -!- snowylike [~sn@91-67-171-156-dynip.superkabel.de] has quit [Ping timeout: 248 seconds] 19:15:38 snowylike [~sn@91-67-171-156-dynip.superkabel.de] has joined #sbcl 19:16:44 seems something might have broken cheneygc builds 19:17:26 between something between 1.0.53.106.master.1-9719409 and current head 19:18:35 http://random.christoph-egger.org/sbcl.b9691ef5009d3669c4f87f4dfbd2baf4538e60f8.log 19:23:29 -!- wbooze [~wbooze@xdsl-78-35-186-218.netcologne.de] has quit [Ping timeout: 255 seconds] 20:10:44 gah. That looks like the lispobj -> sword stuff? the gc-internal.h version of gc_general_alloc's first argument is a word_t, and the cheneygc.c version has "long" 20:11:25 *lichtblau* points others at boinkmarks, doesn't test himself 20:12:26 sdemarre [~serge@91.180.66.31] has joined #sbcl 20:13:48 what architectures do cheneygc by default these days anyway? 20:14:09 MIPS is the only relevant one 20:15:05 ... need to add "give me a mips toy" to the infinite sbcl donations wishlist, then 20:15:18 and by "me" I mean "a random grad student" 20:15:39 I just bought my MIPS, and now I don't have the time for it. :-( 20:16:47 (the problem with my great grad student plan, which occurred to me over supper, is that most of the grad students who've spent lots of time on sbcl have been doing it to get _away_ from their Actual Work; 20:16:56 it's a strange world when mips is a relevant arch :-D 20:17:07 it would therefore be counterproductive to have an sbcl-related studentship :-) 20:18:30 strange world indeed: Easier to order a weird new MIPS from China than a Kindle from the US. 20:21:43 christoph_debian: so, long story short, I'm responsible, but can't fix it ATM. Would give SSH to anyone willing to take care of it though (for the little that's worth). 20:22:31 Krystof: you can still get shell access for gcc compile farm and they have mips ;-) 20:23:17 christoph_debian: can we tempt _you_ to fix it? 20:23:36 maybe I don't really know much sbcl internals but can try 20:24:14 It's a search&replace task; when two types collide in that fashion, choose the newer name (sword_t, uword_t) over the older name (long, unsigned long). 20:33:51 oh ok 20:34:00 that should be possible, I'll look into it after lunch 20:50:34 -!- Dalek_Baldwin [~Adium@71-84-34-33.dhcp.mtpk.ca.charter.com] has quit [Quit: Leaving.] 20:59:37 -!- snowylike [~sn@91-67-171-156-dynip.superkabel.de] has quit [Read error: Connection reset by peer] 21:09:16 -!- psilord [~psilord@23-25-144-217-static.hfc.comcastbusiness.net] has quit [Quit: Leaving.] 21:12:57 Dalek_Baldwin [~Adium@108-225-26-178.lightspeed.irvnca.sbcglobal.net] has joined #sbcl 21:31:58 -!- sdemarre [~serge@91.180.66.31] has quit [Ping timeout: 256 seconds] 21:36:22 what's the difference between word_t and sword_t? 21:38:25 tcr1 [~tcr@84-72-21-32.dclient.hispeed.ch] has joined #sbcl 21:43:45 word_t is older, sort of deprecated, but not yet removed 21:50:28 ok 21:50:45 anyway I have a patch that seems to be working untill now. build is still running though 21:51:17 cool 21:56:50 slyrus [~chatzilla@173-228-44-92.dsl.static.sonic.net] has joined #sbcl 22:02:35 Thra11 [~thrall@94.8.113.87.dyn.plus.net] has joined #sbcl 22:03:01 -!- Thra11 [~thrall@94.8.113.87.dyn.plus.net] has quit [Read error: Connection reset by peer] 22:04:30 Thra11 [~thrall@94.8.113.87.dyn.plus.net] has joined #sbcl 22:14:24 psilord [~psilord@c-69-180-173-249.hsd1.mn.comcast.net] has joined #sbcl 22:50:41 -!- leuler [~user@p548FBA85.dip.t-dialin.net] has quit [Quit: ERC Version 5.1.2 $Revision: 1.796.2.6 $ (IRC client for Emacs)] 22:59:01 -!- LiamH [~none@pdp8.nrl.navy.mil] has quit [Quit: Leaving.] 23:23:55 -!- tcr1 [~tcr@84-72-21-32.dclient.hispeed.ch] has quit [Quit: Leaving.]