2015-01-16T19:05:31Z ccl-logbot joined #lisp
2015-01-16T19:05:31Z 
2015-01-16T19:05:31Z names: ccl-logbot Ragnaroek tajjada nikki93_ Denommus Vutral hoosieree joneshf-laptop hiroakip nand1 attila_lendvai EvW duggiefresh rpg ikki nell ndrei flash- ThePhoeron yenda pjb novemberist boogie mishoo psy_ sheilong qiemem cmack scymtym isoraqathedh theseb Kruppe loke_ reb dkcl xyh gravicappa sword Karl_Dscc eMBee antonv billstclair tristero LiamH vaporatorius Mon_Ouie askatasuna oleo capcar janmuffino BitPuffin harish Jirachier jewel JuanDaugherty
2015-01-16T19:05:31Z names: urandom__ eudoxia tsumetai keen__________34 taraz` araujo hardenedapple Sgeo quazimodo mrkkrp MrWoohoo Alfr DeadTrickster emlow _5kg gingerale lieven newcup aap schjetne AntiSpamMeta theos angavrilov doomlord___ chu Grue` fridim_ grungier Shinmera zymurgy AeroNotix mrSpec ggole TristamWrk Oddity H4ns heurist`_ REPLeffect vhost- moei yeticry |3b| kanru alchemis7 roo mlamari jocuman Riviera Jesin Petit_Dejeuner DrCode stardiviner pillton antoszka Tristam
2015-01-16T19:05:31Z names: Plazma tomaw c74d brandonz kushal cibs Ober K1rk oconnore ConstantineXVI Natch kjeldahl nydel emma Intensity chameco sbryant nicdev`` jayne victor_lowther foom GGMethos segmond xk05 Guest44562 arrdem replcated qlkzy thomas smull diginet someone Guest38736 rotty vlnx stux|RC-only Xach froggey eagleflo sshirokov splittist_ loke zbigniew brucem ivan\_ CrazyEddy johs tbarletz eazar001 ahungry Faed_ frkout axion Oladon nowhere_man_ gz djinni`_ ski MightyJoe
2015-01-16T19:05:31Z names: jasom sjl alakra schoppenhauer milosn copec tessier eak xristos brent80_plow 7YUAAFK6A j0ni funnel ozzloy tokenrove faheem__ stokachu peccu2 dtw luis jsnell mtd Nshag clog ramus les phadthai ``Erik wooden Hydan scharan katco sharkz__ hitecnologys Takumo backupthrick killmaster snafuchs BlastHardcheese nisstyre ggherdov gensym agumonkey_ TrafficMan Longlius eigenlicht dfox salv0 drmeister shifty778 yorick fmu sigjuice gniourf hvxgr otwieracz guaqua`
2015-01-16T19:05:31Z names: larme abbe ircbrowse ered pchrist gluegadget gko nightshade427 jdz swflint PuercoPop cross mathrick soggybread ferada enn_ sellout- Bike Tordek_ Mandus_ Soft- gabot cmbntr s_e White__Flame capitaomorte 77CAABXCD yrdz`` fikusz honkfestival enfors samebcha1e teiresias minion ft xorpse cjmacs`` john-mcaleely stopbit SHODAN necronian _8hzp egp_ trn housel akkad TDT` rk[1] bytecrawler finnrobi gabc `micro_ tkd Colleen setheus FrostyX_ hratsimi1ah cmatei bege
2015-01-16T19:05:31Z names: aftershave Vivitron nitro_idiot misv justinmcp cpc26 clop2 hugoduncan acieroid jpanest_ BeLucid The_third_man shwouchk_ p_l|backup rvchangue dim sauerkrause josteink stepnem mood_ anunnaki_ p_l ivan4th` Amaan mmathers_ tstc` whartung joshmcmillan joga lemoinem wormphle1m kyl_____1_ cods mingvs sid_cyph1r ssake_ flip214 specbot jim87864` aksatac NhanH d4gg4d Vivitron` rtra mearnsh theBlackDragon Patzy impulse krrrcks blahzik joshe karswell easye rtoym
2015-01-16T19:05:31Z names: Jubb quasisane jtz girrig endou______ __main__ dxtr ck_ zickzackv troydm z0d alexherbo2 benny dlowe Blkt sytse lpaste drdo fe[nl]ix gregburd arrsim rvirding notty daimrod _death galdor_ j_king Krystof ecraven dan64 pendapi dmiles_afk sfa yauz nitrix butyoudonot kbtr cpt_nemo GuilOooo mikaelj stacksmith spacebat babypie eli bjorkintosh Neet Subfusc rick-monster Neptu wasamasa renard_ vert2 atgreen lifenoodles bcoburn hyoyoung edran nightfly oGMo Borbus
2015-01-16T19:05:31Z names: Zhivago Rudolph-Miller jackdaniel Posterdati sismondi yeltzooo9 clop redline6561 aerique Anarch
2015-01-16T19:06:34Z nikki93_ quit (Remote host closed the connection)
2015-01-16T19:07:35Z Denommus` joined #lisp
2015-01-16T19:08:07Z wheelsucker joined #lisp
2015-01-16T19:08:16Z yenda: any that would be considered using lattices. I don't want to define it, just see what got implemented in its name.
2015-01-16T19:08:43Z Denommus quit (Ping timeout: 245 seconds)
2015-01-16T19:09:53Z nikki93_ joined #lisp
2015-01-16T19:09:55Z Ragnaroek_ joined #lisp
2015-01-16T19:10:57Z booly-yam joined #lisp
2015-01-16T19:11:57Z Denommus` quit (Ping timeout: 245 seconds)
2015-01-16T19:11:57Z Ragnaroek__ joined #lisp
2015-01-16T19:12:04Z ehu joined #lisp
2015-01-16T19:12:19Z oleo: aye aye captn!
2015-01-16T19:12:23Z oleo shrubs the decks!
2015-01-16T19:12:24Z oleo: lol
2015-01-16T19:12:30Z Ragnaroek quit (Ping timeout: 265 seconds)
2015-01-16T19:12:30Z Ragnaroek__ is now known as Ragnaroek
2015-01-16T19:13:42Z slyrus joined #lisp
2015-01-16T19:15:03Z rhllor joined #lisp
2015-01-16T19:15:24Z Ragnaroek_ quit (Ping timeout: 265 seconds)
2015-01-16T19:16:20Z xyh quit (Remote host closed the connection)
2015-01-16T19:18:18Z taraz` quit (Ping timeout: 265 seconds)
2015-01-16T19:19:03Z Quadrescence joined #lisp
2015-01-16T19:21:21Z Adlai joined #lisp
2015-01-16T19:23:52Z Bicyclidine joined #lisp
2015-01-16T19:23:57Z manuel__ joined #lisp
2015-01-16T19:25:05Z rpg quit (Quit: rpg)
2015-01-16T19:25:21Z tajjada quit (Ping timeout: 276 seconds)
2015-01-16T19:26:25Z tajjada joined #lisp
2015-01-16T19:28:53Z otwieracz quit (Ping timeout: 244 seconds)
2015-01-16T19:30:19Z wheelsucker quit (Remote host closed the connection)
2015-01-16T19:31:45Z Adlai quit (Ping timeout: 250 seconds)
2015-01-16T19:34:07Z malbertife joined #lisp
2015-01-16T19:36:57Z hugoduncan is now known as hugod
2015-01-16T19:36:59Z Adlai joined #lisp
2015-01-16T19:37:03Z Bicyclidine quit (Ping timeout: 272 seconds)
2015-01-16T19:38:36Z EvW quit (Ping timeout: 265 seconds)
2015-01-16T19:38:43Z Vutral quit (Ping timeout: 245 seconds)
2015-01-16T19:39:33Z otwieracz joined #lisp
2015-01-16T19:39:58Z Alfr_ joined #lisp
2015-01-16T19:40:10Z williamyaoh joined #lisp
2015-01-16T19:40:25Z eudoxia quit (Quit: Leaving)
2015-01-16T19:41:51Z ndrei quit (Ping timeout: 256 seconds)
2015-01-16T19:42:10Z Bicyclidine joined #lisp
2015-01-16T19:42:51Z Guest38736 quit (Quit: Leaving)
2015-01-16T19:43:04Z Khisanth joined #lisp
2015-01-16T19:43:05Z rpg joined #lisp
2015-01-16T19:44:35Z williamyaoh quit (Ping timeout: 244 seconds)
2015-01-16T19:46:23Z malbertife quit (Quit: Leaving)
2015-01-16T19:47:18Z hiroakip quit (Ping timeout: 265 seconds)
2015-01-16T19:48:06Z blahzik quit (Quit: blahzik)
2015-01-16T19:48:20Z blahzik joined #lisp
2015-01-16T19:50:31Z Denommus joined #lisp
2015-01-16T19:51:49Z REPLeffect quit
2015-01-16T19:53:25Z Adlai quit (Ping timeout: 250 seconds)
2015-01-16T19:53:35Z REPLeffect joined #lisp
2015-01-16T19:54:23Z soggybread quit (Quit: Lost terminal)
2015-01-16T19:56:03Z isoraqathedh_l joined #lisp
2015-01-16T19:56:43Z Adlai joined #lisp
2015-01-16T19:57:27Z attila_lendvai quit (Ping timeout: 265 seconds)
2015-01-16T19:57:38Z zadock joined #lisp
2015-01-16T19:58:59Z JuanDaugherty quit (Remote host closed the connection)
2015-01-16T19:59:18Z oleo is now known as Guest39586
2015-01-16T19:59:22Z isoraqathedh quit (Ping timeout: 240 seconds)
2015-01-16T19:59:46Z Shinmera: Having quicklisp custom dists is really, really nice. Radiance can now be installed in a jiffy https://twitter.com/Shinmera/status/556178585321758721
2015-01-16T19:59:59Z attila_lendvai joined #lisp
2015-01-16T19:59:59Z attila_lendvai quit (Changing host)
2015-01-16T19:59:59Z attila_lendvai joined #lisp
2015-01-16T20:01:01Z oleo__ joined #lisp
2015-01-16T20:01:52Z Vutral joined #lisp
2015-01-16T20:02:15Z Guest39586 quit (Ping timeout: 256 seconds)
2015-01-16T20:03:05Z jlongster joined #lisp
2015-01-16T20:03:26Z Denommus` joined #lisp
2015-01-16T20:04:28Z prxq joined #lisp
2015-01-16T20:04:51Z Denommus quit (Ping timeout: 264 seconds)
2015-01-16T20:06:16Z Denommus joined #lisp
2015-01-16T20:06:18Z tajjada quit (Quit: leaving)
2015-01-16T20:07:51Z Denommus` quit (Ping timeout: 264 seconds)
2015-01-16T20:08:25Z nikki93_ quit (Remote host closed the connection)
2015-01-16T20:08:32Z PaulCapestany joined #lisp
2015-01-16T20:09:08Z Denommus` joined #lisp
2015-01-16T20:10:23Z Denommus quit (Ping timeout: 245 seconds)
2015-01-16T20:10:41Z yenda: How are you researches on painting software strategies going Shinmera ?
2015-01-16T20:11:04Z Shinmera: I haven't worked on Parasol in a while
2015-01-16T20:11:09Z EvW joined #lisp
2015-01-16T20:11:11Z rhllor_ joined #lisp
2015-01-16T20:11:23Z Shinmera: Mostly because I've had a general low motivation and inspiration phase the past months.
2015-01-16T20:11:30Z rhllor quit (Read error: Connection reset by peer)
2015-01-16T20:11:43Z Shinmera: I'm starting to get back into things, but unfortunately I also have university tasks to do, so don't hold your breath
2015-01-16T20:12:35Z jgrant joined #lisp
2015-01-16T20:12:47Z rhllor_ is now known as rhllor
2015-01-16T20:13:24Z Shinmera: Is anyone actually watching the stream atm?
2015-01-16T20:13:32Z jgrant quit (Read error: Connection reset by peer)
2015-01-16T20:13:36Z yenda: What stream ?
2015-01-16T20:13:44Z Shinmera: Wrong channel, sorry
2015-01-16T20:13:53Z Denommus` quit (Ping timeout: 265 seconds)
2015-01-16T20:13:53Z Shinmera is trying to multitask several things
2015-01-16T20:13:55Z jgrant joined #lisp
2015-01-16T20:16:28Z Denommus joined #lisp
2015-01-16T20:16:48Z specbot quit (Remote host closed the connection)
2015-01-16T20:16:48Z minion quit (Remote host closed the connection)
2015-01-16T20:17:44Z minion joined #lisp
2015-01-16T20:17:45Z specbot joined #lisp
2015-01-16T20:19:07Z zacharias joined #lisp
2015-01-16T20:21:54Z araujo quit (Ping timeout: 245 seconds)
2015-01-16T20:21:59Z Xach: Shinmera: wicked
2015-01-16T20:23:00Z Shinmera: Xach: Yeah, it's really nice. Thanks a lot for making it possible!
2015-01-16T20:23:37Z Xach: Shinmera: are you using dists that aren't ql-dist:dists?
2015-01-16T20:23:39Z Vutral quit (Ping timeout: 244 seconds)
2015-01-16T20:24:08Z Shinmera: I'm not sure what that would even mean!
2015-01-16T20:24:40Z araujo joined #lisp
2015-01-16T20:25:12Z Shinmera: What's in the screenshot is all it takes to set it up, so I'm guessing ql-dist:install-dist will install a regular ql-dist:dist ?
2015-01-16T20:25:20Z Xach: yeah
2015-01-16T20:25:35Z Xach: there's a protocol in place so you can (theoretically) make things that don't use the standard fetch-a-tarball model
2015-01-16T20:25:50Z Xach: nobody's used it yet, even me, so i wonder if it does enough
2015-01-16T20:26:06Z Shinmera: I'm quite fine with that model, though I've adapted it a bit to support more than one update in a day.
2015-01-16T20:26:06Z ggole quit
2015-01-16T20:27:27Z pnpuff joined #lisp
2015-01-16T20:28:19Z admg joined #lisp
2015-01-16T20:28:22Z Xach: i was hoping to make it possible to support something like clbuild or asdf-install entirely within quicklisp
2015-01-16T20:28:26Z Xach: don't know if it's feasible
2015-01-16T20:28:57Z Vutral joined #lisp
2015-01-16T20:29:04Z Shinmera: If anything I'd like a way to tell QL to use GIT repositories, but that seems a bit outlandish.
2015-01-16T20:29:09Z tajjada joined #lisp
2015-01-16T20:29:10Z pnpuff left #lisp
2015-01-16T20:29:23Z ndrei joined #lisp
2015-01-16T20:29:58Z innertracks joined #lisp
2015-01-16T20:31:55Z Xach: Shinmera: i think that could be done, actually.
2015-01-16T20:32:32Z Xach: Shinmera: you would push something on ql-dist:*dist-enumeration-functions*, and then have a dist-like object that responds to provided-releases and provided-systems and all the rest of the protocol
2015-01-16T20:32:46Z Xach: again, that's all in theory - in practice i might have done too little to support it well
2015-01-16T20:33:05Z Shinmera: I'll keep it in mind, if I ever feel like being exploratory :)
2015-01-16T20:33:07Z Xach: enabledp, ensure-installed, etc
2015-01-16T20:33:47Z Shinmera: Having GIT repos could be nice since the load could be taken off the server (not needing to distribute any of the actual files), as well as having the entire history on every client.
2015-01-16T20:34:19Z rpg: Xach: Is that the mechanism to support a private repository of tarballs, too?
2015-01-16T20:34:51Z Shinmera: rpg: A standard dist can be installed with just (ql-dist:install-dist address-to-dist-file)
2015-01-16T20:35:22Z rpg: OK, so I just need to spend a little time to figure out how to make a dist and I'm good....
2015-01-16T20:35:32Z Denommus quit (Read error: Connection reset by peer)
2015-01-16T20:35:46Z Shinmera: rpg: Have a look at https://github.com/Shirakumo/dist
2015-01-16T20:35:50Z rpg: Trying to decide which is easier for a new project: git submodules (ugh) or rolling my own ql dist.
2015-01-16T20:35:59Z williamyaoh joined #lisp
2015-01-16T20:36:12Z Denommus joined #lisp
2015-01-16T20:36:23Z Shinmera: I'm not sure that submodules and quicklisp dists provide the same thing...
2015-01-16T20:36:45Z rpg: Looking to have all developers using known stable versions of S/W.
2015-01-16T20:37:06Z rpg: The submodules allow us to fork the libraries if we find problems in them, so it's probably best.
2015-01-16T20:37:13Z Denommus: rpg: report program generator?
2015-01-16T20:37:27Z Shinmera: As long as it's all QL software that'll work fine, but if you have other languages you're going to get in trouble with finding the proper source location.
2015-01-16T20:37:36Z Shinmera: *all Lisp software
2015-01-16T20:37:57Z rpg: Denommus: rocket propelled grenade ;-) No -- just my initials and user id since forever. (but I'm Not That rpg)
2015-01-16T20:37:57Z Shinmera: So in that case submodules would be better.
2015-01-16T20:38:39Z rpg: Shinmera: it's all lisp libraries. Probably submodules + properly configured ASDF is what I need.
2015-01-16T20:38:55Z Shinmera: Well in that case a QL dist would work fine as well.
2015-01-16T20:40:19Z gravicappa quit (Remote host closed the connection)
2015-01-16T20:40:27Z francogrex joined #lisp
2015-01-16T20:42:20Z soggybread joined #lisp
2015-01-16T20:42:37Z kalzz joined #lisp
2015-01-16T20:43:10Z nikki93 joined #lisp
2015-01-16T20:45:05Z genii joined #lisp
2015-01-16T20:46:48Z Ragnaroek_ joined #lisp
2015-01-16T20:48:41Z Ragnaroek quit (Ping timeout: 265 seconds)
2015-01-16T20:48:41Z Ragnaroek_ is now known as Ragnaroek
2015-01-16T20:52:22Z killmaster quit (Ping timeout: 240 seconds)
2015-01-16T20:55:21Z killmaster joined #lisp
2015-01-16T21:00:28Z farhaven joined #lisp
2015-01-16T21:01:01Z Bicyclidine quit (Ping timeout: 264 seconds)
2015-01-16T21:03:59Z askatasuna quit (Ping timeout: 245 seconds)
2015-01-16T21:04:46Z oleo__ is now known as oleo
2015-01-16T21:05:42Z nikki93 quit (Remote host closed the connection)
2015-01-16T21:06:07Z JuanDaugherty joined #lisp
2015-01-16T21:06:12Z november` joined #lisp
2015-01-16T21:06:16Z nikki93 joined #lisp
2015-01-16T21:06:18Z RenRenJuan joined #lisp
2015-01-16T21:07:03Z yenda: Shinmera would you mind to discuss Radiance ? I'm wondering what part of web development you intend to handle. I don't see a clear definition of your system at the moment, maybe I need to search more.
2015-01-16T21:07:46Z Shinmera: yenda: I haven't been able to come up with one that doesn't have buzzwords out the wazoo yet, so don't worry. You haven't failed in your search :)
2015-01-16T21:07:51Z Shinmera: yenda: What do you want to know?
2015-01-16T21:10:01Z novemberist quit (Ping timeout: 264 seconds)
2015-01-16T21:10:39Z ndrei quit (Ping timeout: 276 seconds)
2015-01-16T21:10:39Z nikki93 quit (Ping timeout: 245 seconds)
2015-01-16T21:11:37Z kcj joined #lisp
2015-01-16T21:14:13Z nell quit (Ping timeout: 264 seconds)
2015-01-16T21:14:47Z ikki quit (Ping timeout: 265 seconds)
2015-01-16T21:16:23Z Davidbrcz joined #lisp
2015-01-16T21:18:10Z yenda: For instance could you explain how your system work ? What part of web development is covered by it and how ?
2015-01-16T21:18:15Z nikki93 joined #lisp
2015-01-16T21:18:52Z Shinmera: The parts that it covers are as many or few as you want
2015-01-16T21:19:17Z Shinmera: It can cover everything from webserver over database and authentication to user management, profiles and administration
2015-01-16T21:19:26Z Shinmera: And it can... not do any of that, if you don't want it to.
2015-01-16T21:19:57Z nikki93 quit (Remote host closed the connection)
2015-01-16T21:20:10Z Shinmera: "Web development" is a bit of a broad term, so I'm not sure what you want to hear.
2015-01-16T21:20:45Z williamyaoh: Perhaps it'd be simpler to be more specific about what you want to use it for?
2015-01-16T21:24:54Z Ragnaroek_ joined #lisp
2015-01-16T21:27:31Z sheilong quit (Quit: Konversation terminated!)
2015-01-16T21:27:39Z Bicyclidine joined #lisp
2015-01-16T21:27:50Z Jirachier quit (Ping timeout: 265 seconds)
2015-01-16T21:27:53Z Ragnaroek quit (Ping timeout: 272 seconds)
2015-01-16T21:27:58Z Ragnaroek_ is now known as Ragnaroek
2015-01-16T21:28:11Z Jirachier joined #lisp
2015-01-16T21:29:16Z Sol90 joined #lisp
2015-01-16T21:29:23Z mrkkrp left #lisp
2015-01-16T21:32:39Z yenda: I do not want to use it at the moment for a particular purpose as I am not sure what exactly it has to offer. When you say it covers databases for instance what do you mean exactly ? Is there a documentation where I can find such informations ?
2015-01-16T21:33:16Z Shinmera: There is currently no documentation/specification, as that's my next step. It's also not "released" yet for that exact reason.
2015-01-16T21:33:40Z Shinmera: You can get a glimpse of my efforts as well as some discussion of what a simple web-app development looks like in my articles though https://reader.tymoon.eu/tagged/radiance
2015-01-16T21:34:11Z Shinmera: (a simple example being https://reader.tymoon.eu/article/291 )
2015-01-16T21:37:16Z kons joined #lisp
2015-01-16T21:37:42Z angavrilov quit (Remote host closed the connection)
2015-01-16T21:37:44Z november` is now known as novemberist
2015-01-16T21:39:18Z ruste joined #lisp
2015-01-16T21:40:31Z Quadrescence: Are there any idiomatic approaches to not consing up the dimension argument to MAKE-ARRAY, aside from using single dimensional arrays and doing the indexing yourself?
2015-01-16T21:41:08Z Quadrescence: The only clumsy way I can think of is allocating a reusable two-element list and just mutating it.
2015-01-16T21:41:31Z cmack quit (Read error: Connection reset by peer)
2015-01-16T21:41:45Z cmack joined #lisp
2015-01-16T21:43:18Z yenda: Shinmera: but it's on Quicklisp now ?
2015-01-16T21:43:55Z Shinmera: It is not.
2015-01-16T21:44:08Z Shinmera: It's in my own quicklisp dist, which is however publicly available.
2015-01-16T21:44:33Z yenda: oh yeah I just read through that in your post sorry
2015-01-16T21:44:40Z Shinmera: No worries.
2015-01-16T21:46:51Z EvW quit (Quit: EvW)
2015-01-16T21:47:33Z isoraqathedh_l is now known as isoraqathedh
2015-01-16T21:49:23Z sivoais joined #lisp
2015-01-16T21:51:07Z MoALTz joined #lisp
2015-01-16T21:51:19Z mearnsh quit (Ping timeout: 272 seconds)
2015-01-16T21:52:09Z Adlai: Quadrescence: another clumsy way is to use hash consing, if you're often reusing the same dimensions
2015-01-16T21:52:21Z Shinmera: Quadrescence: Why is consing it up a problem?
2015-01-16T21:52:29Z TristamWrk quit (Quit: Some days you're the pigeon, some days the statue...)
2015-01-16T21:52:37Z mearnsh joined #lisp
2015-01-16T21:56:10Z Quadrescence: Shinmera, the same reason consing is a problem anywhere else
2015-01-16T21:56:48Z Shinmera: It just sounds like a thing that I can't imagine being the bottleneck.
2015-01-16T21:57:18Z TristamWrk joined #lisp
2015-01-16T21:57:41Z Quadrescence: consing a bunch of little patch arrays constantly, each of which requires consing up a little list, unless I hash cons or figure out some finite number of lists that take care of all of the dimensions I care about
2015-01-16T21:58:00Z nisstyre quit (Changing host)
2015-01-16T21:58:01Z nisstyre joined #lisp
2015-01-16T21:58:28Z Quadrescence: you're right, it's not a bottleneck, but neither are the 5000 other individual repeated-in-time individual allocations that occur everywhere
2015-01-16T21:59:01Z Shinmera: So are you just optimising for the sake of it or?
2015-01-16T21:59:32Z Quadrescence: Well I was asking because I was curious if there was an answer, and yes, usually I optimize things that don't obviously impede correctness if I can.
2015-01-16T22:00:02Z Shinmera: Sounds like a plan for madness, if you ask me.
2015-01-16T22:00:42Z Quadrescence: I know all about the common wisdom, the oft touted "premature optimization", "optimize what shows up at the top of the profiler", etc.
2015-01-16T22:01:06Z yenda: Shinmera: I got "The function RADIANCE-CORE::FLATTEN-METHOD-LAMBDA-LIST is undefined." when installing radiance
2015-01-16T22:01:06Z yenda:  
2015-01-16T22:01:21Z Quadrescence: But when I'm sculpting something, I like to clean the scraps, which is different than trying to sculpt with a toothpick.
2015-01-16T22:01:24Z Shinmera: yenda: Your quicklisp dist is too old
2015-01-16T22:01:27Z mearnsh quit (Ping timeout: 272 seconds)
2015-01-16T22:01:36Z Shinmera: yenda: (ql:update-dist "quicklisp")
2015-01-16T22:02:08Z mearnsh joined #lisp
2015-01-16T22:02:10Z Shinmera: Quadrescence: Depends on what you see as "scraps".
2015-01-16T22:02:32Z Quadrescence: Sloppy allocations?
2015-01-16T22:02:40Z jasom: Quadrescence: in this case, not allocating a small list when you are allocating a multi-dimensional array seems pointless
2015-01-16T22:02:51Z Shinmera: I see optimisation complications as scraps :)
2015-01-16T22:03:20Z williamyaoh: I agree that something that small would probably be in the noise for optimisation
2015-01-16T22:03:24Z Fare joined #lisp
2015-01-16T22:03:29Z williamyaoh: But hell, if that's what gets you your kicks, who are we to judge?
2015-01-16T22:03:34Z Quadrescence: jasom, the arrays are typically but not always within 10x10, with the mean around 5x5
2015-01-16T22:03:40Z jasom: allocating small short-lived objects is essentially free on sbcl
2015-01-16T22:03:45Z Quadrescence: they are allocated within a loop during the generation of a patched grid
2015-01-16T22:03:54Z Quadrescence: generation and updating*
2015-01-16T22:04:05Z williamyaoh: I click circles for fun and Shinmera dodges danmaku
2015-01-16T22:04:08Z Quadrescence: The loop represents execution over time.
2015-01-16T22:04:25Z williamyaoh: Reusing the same two-item list would probably be the simplest way
2015-01-16T22:05:33Z arnaudga joined #lisp
2015-01-16T22:05:43Z Shinmera: williamyaoh: I'm guessing that's a subtle hint at my Touhou art, but I'm afraid to tell you that I don't actually play the games. </ot>
2015-01-16T22:06:18Z jasom: yeah reusing it is simple inside a loop
2015-01-16T22:06:51Z williamyaoh: Shame, was gonna ask if you had any good 1cc's :P
2015-01-16T22:07:12Z Quadrescence: So, then, I'll take the answer to my original question to be "no, there is not another especially idiomatic way."
2015-01-16T22:08:23Z rhllor quit (Ping timeout: 246 seconds)
2015-01-16T22:09:50Z hoosieree quit (Remote host closed the connection)
2015-01-16T22:09:54Z Xach: Don't stop at #lisp! Maybe a wizened wizard of usenet can help
2015-01-16T22:11:13Z wasamasa: Shinmera: what touhou art?
2015-01-16T22:11:31Z Quadrescence: Yes, I would love to read a thread of 45 replies of a rapidly devolving argument!
2015-01-16T22:11:50Z Shinmera: wasamasa: http://tumblr.shinmera.com/tagged/touhou
2015-01-16T22:12:04Z wasamasa: ok, enough offtopic
2015-01-16T22:12:04Z zadock quit (Quit: Leaving)
2015-01-16T22:12:06Z wasamasa: Shinmera: thank you
2015-01-16T22:12:22Z Xach: #lisp isn't much different
2015-01-16T22:12:24Z Shinmera: yenda: Were you able to install it after all?
2015-01-16T22:12:26Z Xach: It's just more real-time
2015-01-16T22:12:38Z Quadrescence: That is true.
2015-01-16T22:12:55Z Xach: There are some quality brains still in effect on comp.lang.lisp amongst the chaff
2015-01-16T22:13:01Z Shinmera: Arguably IRC is worse because you don't have time to reply, whereas with mails/usenet you can lay back and read/skip what you want.
2015-01-16T22:13:40Z nell joined #lisp
2015-01-16T22:13:52Z Xach: IRC is better moderated, usually.
2015-01-16T22:14:15Z Xach: For topicality if not quality
2015-01-16T22:14:37Z Shinmera: Sure, but I find it is much too often almost impossible to properly debate due to the speed involved and people talking all over each other.
2015-01-16T22:15:07Z ahungry: And a million WJ posts
2015-01-16T22:15:25Z Shinmera: Praise killfiles
2015-01-16T22:16:20Z Xach: something ought to be done by Someone
2015-01-16T22:17:18Z Fare: sometime
2015-01-16T22:18:40Z nydel: oo a new drakma
2015-01-16T22:19:07Z jasom: Shinmera: in #lisp or IRC in general?
2015-01-16T22:19:20Z jasom: #lisp tends to be low-enough volume to allow for debate, imo
2015-01-16T22:19:24Z CrazyM4n joined #lisp
2015-01-16T22:19:34Z Shinmera: jasom: IRC in general. It happens much less in #lisp, which I am very, incredibly, thankful for.
2015-01-16T22:20:08Z nydel: does anone know if stassat's cl-qt-web-browser package is working
2015-01-16T22:21:04Z theseb quit (Remote host closed the connection)
2015-01-16T22:22:24Z White__Flame: Quadrescence: if your syntax is (make-array (list x y) ...) then I'd look at making the lisp implementation smart enough to extract those values straight without consing
2015-01-16T22:22:43Z White__Flame: if you're doing (make-array dims ...) then you're likely out of luck there
2015-01-16T22:23:04Z jasom: White__Flame: often it's (make-array (list x y) ...)
2015-01-16T22:23:23Z White__Flame is now known as White_Flame
2015-01-16T22:23:46Z Alfr_ quit (Quit: Leaving)
2015-01-16T22:23:57Z Quadrescence: Well, do Lisp implementations optimize that?
2015-01-16T22:25:58Z yenda: Shinmera: yes it just worked
2015-01-16T22:26:08Z Shinmera: yenda: Hooray!
2015-01-16T22:26:35Z White_Flame: Quadrescence: are you looking at your disassemblies?
2015-01-16T22:26:49Z Shinmera: yenda: If you'd like I'll remember to notify you once I have some actual documentation together.
2015-01-16T22:27:02Z White_Flame: I'm pretty sure sbcl will optimize constants that way, eg (make-array '(10 10) ...)
2015-01-16T22:27:47Z Quadrescence: '(10 10) isn't quite as bad. It's (list x y) that could potentially be optimized, but who knows it if is with the beast that is MAKE-ARRAY
2015-01-16T22:28:34Z Shinmera: You can always just keep the list, setf the first/second and then pass that. But I //really// don't see much point in this.
2015-01-16T22:28:35Z duggiefresh quit
2015-01-16T22:29:09Z Bicyclidine: from what i can see, ccl does not
2015-01-16T22:29:16Z LiamH quit (Quit: Leaving.)
2015-01-16T22:29:36Z Bicyclidine: i think most implementations only distinguish one dimensional and multidimensional arrays, and have an internal multidimensional array constructor that just takes a list
2015-01-16T22:30:06Z White_Flame: yeah, sbcl is calling #'list for (make-array (list x y)) as well
2015-01-16T22:30:17Z DrCode quit (Ping timeout: 250 seconds)
2015-01-16T22:30:23Z sivoais quit (Ping timeout: 240 seconds)
2015-01-16T22:30:33Z White_Flame: (v1.1.9)
2015-01-16T22:30:35Z Bicyclidine: so at most the list could be stack allocated
2015-01-16T22:31:18Z White_Flame: if make-array doesn't keep a copy of it along with the array metadata
2015-01-16T22:31:20Z Quadrescence: Shinmera, I shared a few reasons, but I guess I could go in depth as to why allocating 1M new multidimensional arrays to represent patch structure of a 3 dimensional surface every frame might afford looking at what gets allocated for every array instantiation
2015-01-16T22:31:24Z Bicyclidine: and i guess probably not, since the internal array structure might be a single dimensional array and a list of dims
2015-01-16T22:31:37Z scymtym_ joined #lisp
2015-01-16T22:31:44Z Quadrescence: allocating (every frame)
2015-01-16T22:31:58Z White_Flame: remember, allocations are fast. GC is slow
2015-01-16T22:32:20Z Quadrescence: Correct!
2015-01-16T22:32:26Z Shinmera: Quadrescence: Sounds to me like there'd be more to be done re-using the arrays than the dimension args.
2015-01-16T22:33:36Z Quadrescence: yes, pooling is another approach, though one I didn't want to get into because the main operations are gluing and cutting.
2015-01-16T22:35:12Z Shinmera: Well, it's just that array disposal and allocation seem much bigger horses to me than a mere cons.
2015-01-16T22:38:15Z Bicyclidine: in particular it looks like ccl represents multidimensional arrays as displaced arrays into a flat array, and displaced arrays have all the dims in separate slots of another flat array
2015-01-16T22:38:36Z Quadrescence: And you know what? Maybe use one dimensional arrays. Maybe use lazy cutting and gluting. Maybe do all those different larger scale algorithmic techniques instead!
2015-01-16T22:38:51Z Shinmera: Optimisation never ends.
2015-01-16T22:38:52Z Quadrescence: gluing*
2015-01-16T22:39:47Z Bicyclidine: i think sbcl is similar.
2015-01-16T22:40:49Z Bicyclidine: so, not only am i not sure that any implementation won't allocate a list in (make-array (list ...) ...), it might not be practical for them to do that. maybe
2015-01-16T22:41:56Z pacon joined #lisp
2015-01-16T22:42:30Z DrCode joined #lisp
2015-01-16T22:46:29Z williamyaoh quit (Remote host closed the connection)
2015-01-16T22:47:20Z munksgaard joined #lisp
2015-01-16T22:48:56Z White_Flame: Quadrescence: In any sort of game environnment, you should seek to do zero consing
2015-01-16T22:49:06Z White_Flame: the quantity of allocation is much more important than the frequency of allocation
2015-01-16T22:49:31Z White_Flame: so yes, the arrays themselves are much more a problem than the size conses
2015-01-16T22:50:07Z White_Flame: well, the bytes consed per frame is much more important than just the raw number of allocations per frame
2015-01-16T22:50:32Z White_Flame: those tiny allocations are generally irrelevant if you're doing other large allocations at the same time
2015-01-16T22:51:13Z araujo quit (Ping timeout: 245 seconds)
2015-01-16T22:51:22Z sivoais joined #lisp
2015-01-16T22:51:54Z prxq quit (Remote host closed the connection)
2015-01-16T22:51:56Z Quadrescence: Sorry, clearly asking if I can avoid consing a list as an argument to MAKE-ARRAY was a very inept question and next time I should describe the scope of the entire program so I can get wisdom from the great #lisp channel.
2015-01-16T22:52:06Z robot-beethoven joined #lisp
2015-01-16T22:53:19Z Shinmera: People just want to understand your real problem at hand because what you're asking for seemed questionable.
2015-01-16T22:53:34Z araujo joined #lisp
2015-01-16T22:53:37Z White_Flame: (and I was AFK for a bit)
2015-01-16T22:54:15Z joneshf-laptop quit (Ping timeout: 252 seconds)
2015-01-16T22:54:27Z Sol90 quit (Quit: Leaving)
2015-01-16T22:55:02Z Quadrescence: Shinmera, I dunno, 32 MB of list consing per time step doesn't seems super questionable to me, even if there's other allocation optimizations that can be done.
2015-01-16T22:55:34Z Shinmera: Quadrescence: Well, see, people don't /know/ that that's happening because you never /told/ anybody.
2015-01-16T22:55:49Z Quadrescence: I said 1M arrays.
2015-01-16T22:56:13Z Shinmera: Yes, and that leads to questioning whether the arrays can be re-used smarter, rather than the question of the conses.
2015-01-16T22:56:35Z francogrex quit (Quit: ERC Version 5.3 (IRC client for Emacs))
2015-01-16T22:56:40Z White_Flame: I think that in your context, eliminating the size array consing will drop your GC rate by less than 0.1%
2015-01-16T22:57:03Z White_Flame: but it could be as high as a couple of percent as well :)
2015-01-16T22:58:05Z White_Flame: also, the amount of GC pressure applied by the size lists are in a fractional linear relationship to the overall array llocation
2015-01-16T22:58:11Z White_Flame: allocation
2015-01-16T22:58:42Z White_Flame: so reusing arrays will eliminate both allocations
2015-01-16T22:59:08Z White_Flame: and take similar amount of programming work to reusing size lists
2015-01-16T23:00:13Z Quadrescence: Reusing a list during allocation is much simpler than setting up a memory pool.
2015-01-16T23:00:22Z Bicyclidine quit (Ping timeout: 240 seconds)
2015-01-16T23:00:30Z enitiz joined #lisp
2015-01-16T23:00:50Z White_Flame: I was meaning reusing immutable size lists, since that's safer
2015-01-16T23:01:16Z Quadrescence: yes
2015-01-16T23:01:37Z doomlord___ quit (Ping timeout: 264 seconds)
2015-01-16T23:01:45Z White_Flame: reusing a thread-local 2-element list via mutation would be easier
2015-01-16T23:02:20Z Quadrescence: that is what i was originally thinking, though I don't really know if I can rely on MAKE-ARRAY not storing that somewhere, can I?
2015-01-16T23:03:57Z Quadrescence: Without checking the spec, I would guess that I can rely on MAKE-ARRAY not storing.
2015-01-16T23:04:38Z Xach: There are issues with modifying objects used as hash table keys, but I can't recall anything about array dimensions.
2015-01-16T23:07:24Z Bicyclidine joined #lisp
2015-01-16T23:07:27Z fridim_ quit (Ping timeout: 252 seconds)
2015-01-16T23:07:51Z Davidbrcz quit (Ping timeout: 264 seconds)
2015-01-16T23:13:12Z Bicyclidine quit (Ping timeout: 265 seconds)
2015-01-16T23:14:38Z White_Flame: it appears that SBCL's make-array implementations do not hold on to the dimension list
2015-01-16T23:15:04Z Karl_Dscc quit (Remote host closed the connection)
2015-01-16T23:15:19Z White_Flame: the make-array spec doesn't mention anything about it, but there might be other sections covering parameters
2015-01-16T23:16:00Z Bicyclidine joined #lisp
2015-01-16T23:17:06Z atgreen quit (Remote host closed the connection)
2015-01-16T23:18:11Z novemberist: hello. if i want to define a generic average function. how do i get the number of optional arguments the function caller has provided?
2015-01-16T23:20:09Z Bicyclidine: You're defining this with something like (defun average (&rest args) ...)?
2015-01-16T23:20:50Z novemberist: yes
2015-01-16T23:21:20Z novemberist: (/ (+ arguments) number-of-arguments)
2015-01-16T23:22:12Z Fare: is an array of dimensions '(0 2) the same as an array of dimensions '(2 0) ? Maybe we should summon the language committee to decide...
2015-01-16T23:22:58Z Bicyclidine: novemberist: (length args) then
2015-01-16T23:22:59Z Fare: (length arguments)
2015-01-16T23:23:21Z atgreen joined #lisp
2015-01-16T23:23:35Z novemberist: yeah, i just figured out i get a list :)
2015-01-16T23:24:23Z novemberist: thanks
2015-01-16T23:26:17Z novemberist: (apply #'+ args) would be the best way to get the sum of all list elements?
2015-01-16T23:27:04Z Bicyclidine: should work, yeah. you could also use reduce.
2015-01-16T23:27:09Z arnaudga quit (Ping timeout: 276 seconds)
2015-01-16T23:27:27Z Bicyclidine: usually someone says "no, you use reduce, because of call-arguments-limit" but you're already hitting that with the &rest so no big
2015-01-16T23:27:31Z Bicyclidine: clhs call-arguments-limit
2015-01-16T23:27:31Z specbot: http://www.lispworks.com/reference/HyperSpec/Body/v_call_a.htm
2015-01-16T23:27:41Z novemberist: Bicyclidine: will check that out, thanks
2015-01-16T23:35:16Z rpg quit (Read error: Connection reset by peer)
2015-01-16T23:35:27Z rpg joined #lisp
2015-01-16T23:35:38Z yenda: Shinmera: have ever looked at antiweb ?
2015-01-16T23:37:06Z Denommus quit (Quit: ERC Version 5.3 (IRC client for Emacs))
2015-01-16T23:38:02Z Shinmera: yenda: Can't say I ever heard of it
2015-01-16T23:40:25Z munksgaard quit (Ping timeout: 256 seconds)
2015-01-16T23:41:40Z gingerale quit (Ping timeout: 255 seconds)
2015-01-16T23:46:07Z mishoo quit (Ping timeout: 245 seconds)
2015-01-16T23:46:10Z antonv quit (Ping timeout: 255 seconds)
2015-01-16T23:48:22Z Ethan- joined #lisp
2015-01-16T23:51:27Z ahungry quit (Ping timeout: 252 seconds)
2015-01-16T23:53:14Z ahungry joined #lisp
2015-01-16T23:53:49Z yenda: Shinmera: I would be interested if you could tell me how it would integrate with radiance
2015-01-16T23:54:02Z Ethan- quit (Remote host closed the connection)
2015-01-16T23:54:39Z Alfr quit (Remote host closed the connection)
2015-01-16T23:55:18Z Shinmera: yenda: You would write a server interface implementation for it -- something that implements the functions outlined in the interface specification and delegates requests to radiance's REQUEST function.
2015-01-16T23:56:46Z Shinmera: yenda: such an implementation for, say, hunchentoot is here https://github.com/Shirakumo/radiance-drivers/blob/master/i-hunchentoot/i-hunchentoot.lisp -- a general tutorial on writing implemenations for interfaces is here https://reader.tymoon.eu/article/303
2015-01-16T23:57:07Z akkad: how does hunchentoot compare with say aserve?
2015-01-16T23:57:56Z hitecnologys quit (Ping timeout: 246 seconds)
2015-01-16T23:59:59Z hitecnologys joined #lisp