| 2015-11-19 | * | ben_vulpes is seeing the rot in webtech takeover of cl | #trilema |
| 2015-11-12 | ben_vulpes | i've included some barebones instructions on how to set up a cl for testing, shinohai. | #trilema |
| 2015-11-09 | trinque | phf: thanks for the example code for cl-irc; works well. I'm now porting over the deed handling code from my wad of python. | #trilema |
| 2015-11-09 | asciilifeform | phf: and i specifically didn't go with cl because this thing CANNOT cons | #trilema |
| 2015-11-05 | ascii_field | that's what ansi cl means. | #trilema |
| 2015-11-01 | phf | huh, libgpg-error and gpgme come with common lisp bindings (and only cl) in the standard distribution.. | #trilema |
| 2015-10-29 | phf | vulpes_a_hopital: edi weitz's code only recently got its own github repos when others took over the maintenance (mostly to ensure that it can be packaged into quicklisp). but a canonical place for cl-ppcre (and a handful of other lisp packages for doing "modern" programming) is http://weitz.de/lisp.html | #trilema |
| 2015-10-29 | assbot | CL-PPCRE - Portable Perl-compatible regular expressions for Common Lisp ... ( http://bit.ly/1MVvWuT ) | #trilema |
| 2015-10-29 | assbot | cl-ppcre/index.html at master · edicl/cl-ppcre · GitHub ... ( http://bit.ly/1MVvNYb ) | #trilema |
| 2015-10-27 | phf | trinque: adlai: tried cl-irc, turned off my wifi until the test nick was disconnected on freenode's end, and connectedp still returns T on the cl-irc end << cl-irc is a thin layer on top of irc protocol, so connection state is something you have to handle yourself. connectedp simply queries underlying stream for whether it's still formally open. you have to setup mechanism (say a separate thread) to send PING to the server at a regular | #trilema |
| 2015-10-25 | gabriel_laddel | hence why cl etc etc | #trilema |
| 2015-10-24 | trinque | adlai: tried cl-irc, turned off my wifi until the test nick was disconnected on freenode's end, and connectedp still returns T on the cl-irc end | #trilema |
| 2015-10-24 | trinque | I've only been farting around yet, so I'll see if I like cl-irc | #trilema |
| 2015-10-24 | trinque | yep, I am going to use CL | #trilema |
| 2015-10-24 | adlai | cl-irc is good, it doesn't support sasl but you don't necessarily need that | #trilema |
| 2015-10-17 | asciilifeform | if you feel like doing it as an exercise, do it. if not, use one of the published items. (unlike what is customary in the barbarian world, cl libraries generally... work) | #trilema |
| 2015-10-17 | ben_vulpes | any point to using a cl wrapper/bridge to gnuplot or just pipe out? | #trilema |
| 2015-10-17 | asciilifeform | (and cl-dot) | #trilema |
| 2015-10-17 | asciilifeform | ben_vulpes: cl-graphviz also works | #trilema |
| 2015-10-17 | ben_vulpes | asciilifeform: any suggestions on time series plotting from cl? | #trilema |
| 2015-10-14 | phf | and then there's always bitcoind source, which is infinite rabbit hole. i'm studying the source by rewriting chunks in cl, and the more i learn, the more i know i don't know | #trilema |
| 2015-09-15 | assbot | Logged on 14-09-2015 21:24:02; phf: i suppose one half of b-a should start on a linux distro and the other work on cl-emacs naggum always wanted *ducks* | #trilema |
| 2015-09-14 | phf | i suppose one half of b-a should start on a linux distro and the other work on cl-emacs naggum always wanted *ducks* | #trilema |
| 2015-09-13 | asciilifeform | this preoccupation with format is a disease of the non-cl world, aha | #trilema |
| 2015-09-13 | trinque | asciilifeform: is there something respectable to look at in the CL world in regards to RPC? | #trilema |
| 2015-09-09 | BingoBoingo | <gabriel_laddel> mircea_popescu: the people who authored CL thought of your use case. << AND yet you Hearnia around the details | #trilema |
| 2015-09-09 | gabriel_laddel | mircea_popescu: the people who authored CL thought of your use case. | #trilema |
| 2015-09-04 | trinque | looks like it's formatted like a guy who likes CL might | #trilema |
| 2015-09-04 | ascii_field | as for cl, a unix tool with multi-second startup times is not usable here. | #trilema |
| 2015-09-04 | trinque | two options I see are ball o' bash or CL | #trilema |
| 2015-09-03 | ben_vulpes | cl is such a delightfully crazy mindset | #trilema |
| 2015-09-03 | trinque | I was addressing "own the machine" not CL | #trilema |
| 2015-09-03 | ben_vulpes | how is /etc/hosts a cl data structure? | #trilema |
| 2015-09-03 | asciilifeform | ben_vulpes: partly because cl is a 'heavy' runtime and not fit for use as a 'run ten times a minute' shell script | #trilema |
| 2015-09-03 | asciilifeform | ben_vulpes: partly because there is no handy gpg callout lib for cl | #trilema |
| 2015-09-03 | ben_vulpes | so why python and not cl, out of curiosity? | #trilema |
| 2015-08-13 | asciilifeform | ^ or to program in cl without a button that evaluates the current set of parens your cursor is in | #trilema |
| 2015-08-09 | ben_vulpes | williamdunne: what's your cl development flow like? | #trilema |
| 2015-08-09 | trinque | for CL itself, hyperspec.el | #trilema |
| 2015-08-07 | asciilifeform | and 3) great student exercise, where the kids won't demand to hear why both 'elt' and 'nth' exist (perennial retardation example from cl, which was born in a 'no vendor left behind' multi-turdculturist compromise) | #trilema |
| 2015-08-07 | assbot | Logged on 07-08-2015 23:30:34; trinque: whereas CL appears to value completeness, scheme heads for the minimal set of primitives from which higher functionality can be derived | #trilema |
| 2015-08-07 | trinque | whereas CL appears to value completeness, scheme heads for the minimal set of primitives from which higher functionality can be derived | #trilema |
| 2015-08-06 | trinque | rather than strap myself to golang I'm going to jettison it at some point and reimplement the services model in CL | #trilema |
| 2015-07-31 | ben_vulpes | phf: we're drinking and talking about how to persist data appropriately in cl proggies as we speak | #trilema |
| 2015-07-31 | trinque | what I mean is, say you want to work on n slots, perhaps you just bail to regular CL data structures from there | #trilema |
| 2015-07-31 | ben_vulpes | and i'm not quite sure what the "select * from Y" equivalent in a cl proggie would be | #trilema |
| 2015-07-29 | trinque | I'm gonna eventually write one that works in CL | #trilema |
| 2015-07-26 | ben_vulpes | (i happen to enjoy the amount of study that writing cl takes) | #trilema |
| 2015-07-26 | gabriel_laddel | I can't wait for the inevitable blog post about how "CL isn't really all that good" and "the language has serious problems and a lack of tooling" | #trilema |
| 2015-07-26 | assbot | Logged on 26-07-2015 22:16:44; ben_vulpes: the "async" cl webserver the nodebros are banging on about blocks the repl when instantiated in a cl-async 'event loop', unlike how hunchentoot starts a server and then returns on threaded sbcl. | #trilema |
| 2015-07-26 | asciilifeform | at any rate, there is a lesson: the difference between ending up like this in cl and in, e.g., python, is sorta like the difference between being covered in own shit in civilian life, vs. inside an oubliette | #trilema |
| 2015-07-26 | ben_vulpes | heh like the "modern cl" thinger? | #trilema |
| 2015-07-26 | * | asciilifeform wonders why these folks took cl from the shelf, rather than something more 'komyoonitiful' | #trilema |
| 2015-07-26 | ben_vulpes | the "async" cl webserver the nodebros are banging on about blocks the repl when instantiated in a cl-async 'event loop', unlike how hunchentoot starts a server and then returns on threaded sbcl. | #trilema |
| 2015-07-26 | asciilifeform | derps can 'colonize cl' in precisely the same way redditards colonize bitcoin. | #trilema |
| 2015-07-26 | ben_vulpes | "cl-async" | #trilema |
| 2015-07-26 | ben_vulpes | gabriel_laddel: asciilifeform: did you see the libuv folks are colonizing cl? | #trilema |
| 2015-07-26 | * | asciilifeform is aware of folks who hire for 'we don't care what you write it in', but shudders at the thought of 'cl work' - it could only result from an astonishingly gnarly pre-existing codebase | #trilema |
| 2015-07-26 | ben_vulpes | gabriel_laddel: i'd probably bite it off myself, simply because there isn't much by way of CL horsepower around the shop. | #trilema |
| 2015-07-26 | gabriel_laddel | I'm mentioning this because people say it's so goddamn hard to get CL work, which I have not found to be the case. | #trilema |
| 2015-07-26 | gabriel_laddel | asciilifeform: potential CL work? | #trilema |
| 2015-07-26 | assbot | Logged on 26-07-2015 05:24:04; decimation: unfortunately it seems that the "cl logs" are quite scattered and need to be reconstructed by a student (unlike b-a logs) | #trilema |
| 2015-07-26 | ben_vulpes | https://github.com/orthecreedence/cl-async << this reeks of "let's bring js to cl" | #trilema |
| 2015-07-26 | decimation | unfortunately it seems that the "cl logs" are quite scattered and need to be reconstructed by a student (unlike b-a logs) | #trilema |
| 2015-07-26 | asciilifeform | folks who bring in very tangible money from cl-related work, and want/need a serious support relationship, might benefit from franz or lispworks | #trilema |
| 2015-07-26 | decimation | would you recommend franz over the cl with the menorah? | #trilema |
| 2015-07-23 | gabriel_laddel | trinque: do you want CL work? | #trilema |
| 2015-07-22 | mircea_popescu | and then get clueless working cl;ass kids thinking "it's theiur calling" because totally, that's how the world works, with "Callings" | #trilema |
| 2015-07-22 | assbot | Logged on 22-07-2015 04:37:29; asciilifeform: between this and the block parser, we have maybe a quarter of a cl implementation of bitcoin here... | #trilema |
| 2015-07-22 | asciilifeform | between this and the block parser, we have maybe a quarter of a cl implementation of bitcoin here... | #trilema |
| 2015-07-21 | assbot | Logged on 20-07-2015 02:36:12; ben_vulpes: asciilifeform, phf, Adlai, gabriel_laddel: still curious about deploying CL code to running instances | #trilema |
| 2015-07-19 | ben_vulpes | asciilifeform, phf, Adlai, gabriel_laddel: still curious about deploying CL code to running instances | #trilema |
| 2015-07-19 | ben_vulpes | Adlai: once told me that the 'log reading' period for CL was well in excess of that for #b-a | #trilema |
| 2015-07-19 | ben_vulpes | i am going to spend 2 years just reading CL tooling documentation before putting anything into production at this rate. | #trilema |
| 2015-07-19 | gabriel_laddel | it is pretty easy to crash your lisp using CL-OPENGL | #trilema |
| 2015-07-19 | gabriel_laddel | Or I could use CL, PCLOS, retain my sanity. | #trilema |
| 2015-07-19 | ben_vulpes | i'm writing raw sql in my cl going forward | #trilema |
| 2015-07-16 | ben_vulpes | making persistent cl objstores sounds seriously useful | #trilema |
| 2015-07-16 | gabriel_laddel | ftr, I think the 3 canonical CL books are CLtL2, Keene and Art of the MOP | #trilema |
| 2015-07-15 | assbot | heroku-cl-example/model-utils.lisp at master · mck-/heroku-cl-example · GitHub ... ( http://bit.ly/1Smp64x ) | #trilema |
| 2015-07-15 | ben_vulpes | rewritten in cl yet trinque? | #trilema |
| 2015-07-12 | trinque | spent some time with it over the weekend; though I'm not actually writing CL on the client side it's close enough that I don't feel like I'm context switching between server and client | #trilema |
| 2015-07-12 | assbot | Logged on 12-07-2015 02:30:55; gernika: I attended a lisp meetup in SV today at which someone presented their altcoin which happened to be written in CL. This altcoin exists apparently because bitcoin too "too slow." | #trilema |
| 2015-07-11 | assbot | Logged on 11-07-2015 22:43:42; gabriel_laddel: For example, say that you find the existence of both elt & nth to be stupid. write a (meta)program that walks an arbitrary CL program and fixes all elt's to nths. Let everyone know that it's deprecated, and that they can run this program on their code to update it to CL 1.2 or whatever. | #trilema |
| 2015-07-11 | gernika | I attended a lisp meetup in SV today at which someone presented their altcoin which happened to be written in CL. This altcoin exists apparently because bitcoin too "too slow." | #trilema |
| 2015-07-11 | gabriel_laddel | I mean that someone should have paid for you to write CL, preferable packaging it up as an unchanging distro supporting only a subset of hardware. | #trilema |
| 2015-07-11 | trinque | if you mean somebody should've kicked my ass ten years ago for not writing CL, I agree. | #trilema |
| 2015-07-11 | gabriel_laddel | seriously, redirecting a trinque at CL would make the development 100x better. | #trilema |
| 2015-07-11 | trinque | and that was super powerful as well, and I don't know how to get that on the CL side | #trilema |
| 2015-07-11 | gabriel_laddel | :/ I know. Climacs is the inevitable conclusion. Doing the elisp dance around the cl package system just isn't fun. | #trilema |
| 2015-07-11 | ascii_modem | http://log.bitcoin-assets.com/?date=11-07-2015#1197814 << cl on pc only really became tolerably livable with in late '90s | #trilema |
| 2015-07-11 | assbot | Logged on 11-07-2015 22:11:04; gabriel_laddel: should have just used CL | #trilema |
| 2015-07-11 | gabriel_laddel | sbcl, ccl, allegro, lispworks - each had own << You can still take portable CL programs and run the upgrade program on them without forcing compiler changes. | #trilema |
| 2015-07-11 | gabriel_laddel | ascii_modem: there is a scheme which didn't throw away the CL compiler and start from scratch? | #trilema |
| 2015-07-11 | assbot | Logged on 11-07-2015 22:43:42; gabriel_laddel: For example, say that you find the existence of both elt & nth to be stupid. write a (meta)program that walks an arbitrary CL program and fixes all elt's to nths. Let everyone know that it's deprecated, and that they can run this program on their code to update it to CL 1.2 or whatever. | #trilema |
| 2015-07-11 | phf | i'm not one of those people you're arguing against, i like cl standard just fine. i did paid common lisp consulting on and off for the past 10 years. | #trilema |
| 2015-07-11 | gabriel_laddel | Work CL down to the core language and then when you get something new to play with you can build up that core and run (some) 'CL' programs on it. | #trilema |
| 2015-07-11 | gabriel_laddel | For example, say that you find the existence of both elt & nth to be stupid. write a (meta)program that walks an arbitrary CL program and fixes all elt's to nths. Let everyone know that it's deprecated, and that they can run this program on their code to update it to CL 1.2 or whatever. | #trilema |
| 2015-07-11 | gabriel_laddel | should have just used CL | #trilema |
| 2015-07-11 | gabriel_laddel | I can't speak to that, but using the class browser is a great boon to CL development. | #trilema |
| 2015-07-11 | gabriel_laddel | trinque: have you been using the CLIM listener in your CL hacking? | #trilema |
| 2015-07-11 | trinque | I'll take all the CL prosthetics I can get | #trilema |
| 2015-07-11 | trinque | CL | #trilema |
| 2015-07-11 | gabriel_laddel | trinque: wait, is this for CL or elisp | #trilema |
| 2015-07-10 | trinque | got something out in the world that runs cl, talks to a web api | #trilema |
| 2015-07-10 | gabriel_laddel | And you talk to the thing in CL? | #trilema |
| 2015-07-10 | gabriel_laddel | I'm noodling around with the idea of my hunting down contracts for CL people | #trilema |
| 2015-07-10 | trinque | saw you found work writing cl, pretty cool | #trilema |
| 2015-07-10 | gabriel_laddel | I got serial ports comms working in a *single line* of cl today. | #trilema |
| 2015-07-10 | gabriel_laddel | o nice, then you've pretty much the best of what emacs has to offer for CL development. | #trilema |
| 2015-07-10 | Gabriel-work | anyways, I work with lab equipment and CL | #trilema |
| 2015-07-10 | trinque | was written in CL first | #trilema |
| 2015-07-08 | gabriel_laddel | writing CL + mcclim | #trilema |
| 2015-07-04 | trinque | ben_vulpes: guess what else I'll be rewriting in cl at some point. | #trilema |
| 2015-05-31 | asciilifeform | lispworks is actually (afaik to this day) the maintainer of the online cl hyperspec! while documentation from franz was once superb - came in printed volumes, which i bought second-hand | #trilema |
| 2015-05-28 | gabriel_laddel | https://www.linkedin.com/in/richling << Impressive resume. Mitch also has a collection of CL mathematics hacks, mjrcalc. | #trilema |
| 2015-05-22 | gabriel_laddel | DE (with all the goodies MP and I discussed previously - !s paper clip), 3D CLIM (actually a rather involved project - selecting the correct hardware, reverse engineered opengl drivers etc is expensive and how exactly this gets split off into various products is complicated), a program that does run-time analysis on your CL code and optimizes it in | #trilema |
| 2015-05-22 | gabriel_laddel | I'm not planning on selling Masamune. It will be free and stocked with niceties. Where I do plan to make money is on (source-included) extensions, which amounts to contracting gigs setting up intricate lessons for large companies, militaries, private schools and homeschoolers. Aside from this, various products I'd like to make and sell: a CL only I | #trilema |
| 2015-05-21 | trinque | gabriel_laddel | story time: I ran into loper-os after a stint programming clojure as a day job. I ended up getting evicted for playing around with CL instead of paying rent. << don't ever do this; if you can't make money doing it, you're not the guy to do it. | #trilema |
| 2015-05-21 | assbot | Logged on 21-05-2015 10:04:19; gabriel_laddel: story time: I ran into loper-os after a stint programming clojure as a day job. I ended up getting evicted for playing around with CL instead of paying rent. I got another place, found trilema+#bitcoin-assets and... got evicted again. | #trilema |
| 2015-05-21 | gabriel_laddel | story time: I ran into loper-os after a stint programming clojure as a day job. I ended up getting evicted for playing around with CL instead of paying rent. I got another place, found trilema+#bitcoin-assets and... got evicted again. | #trilema |
| 2015-05-21 | gabriel_laddel | CL. Additionally, I want people who have no prior knowledge of computing to develop taste. Gedankenexperiment: spend 10k USD, purchase stock PCs, preload with something somewhat sane (Actually complete Masamune/bitcoin-assets distro) and distribute to a class of kids "learning to program". IMHO they'll play around with it, do their work and then p | #trilema |
| 2015-05-04 | assbot | Re: Engineering Envy [was: Re: CL and UML] - Naggum cll archive ... ( http://bit.ly/1JlOqV7 ) | #trilema |
| 2015-04-29 | gabriel_laddel | trinque: you're just learning CL, right? | #trilema |
| 2015-04-29 | gabriel_laddel | ^ http://dev.gentoo.org/~zmedico/portage/doc/ this is probably a good place to start for portage -> CL | #trilema |
| 2015-04-29 | gabriel_laddel | Adlai: you're piping data back and forth from CL to elisp or...? | #trilema |
| 2015-04-29 | gabriel_laddel | If someone would like to make themselves useful and move us that much closer to a source-only #-assets distribution, a CL interface to portage is desperately needed. | #trilema |
| 2015-04-28 | trinque | let it never be said that cl does not go to 11 | #trilema |
| 2015-04-28 | mircea_popescu | and where do i find a cl server ? | #trilema |
| 2015-04-28 | jurov | redo in in CL | #trilema |
| 2015-04-26 | trinque | I was for example shocked to find an object system in CL as a layman | #trilema |
| 2015-04-25 | trinque | I'm fucking 28 and learning CL *now* | #trilema |
| 2015-04-21 | ben_vulpes | and trinque: why u CL? | #trilema |
| 2015-04-21 | trinque | CL that is | #trilema |
| 2015-03-23 | assbot | Logged on 23-03-2015 16:34:10; *: Adlai is fairly certain that phuctor is written in cl | #trilema |
| 2015-03-23 | * | Adlai is fairly certain that phuctor is written in cl | #trilema |
| 2015-03-12 | gabriel_laddel | Perhaps I'll have the chance to clean up the CL interface at some point. | #trilema |
| 2015-02-28 | assbot | Re: Upper limits of CL - Naggum cll archive ... ( http://bit.ly/1DnplEL ) | #trilema |
| 2015-02-24 | asciilifeform | u that since it knew Latitude and Longitude), you could find the sun. From the sun, and predetermined celestial positions, you could find the appropriate star in the region. Once you found that star, it was easy find the next one and go back and forth between the two, determining position. Heres the deal. How do you search for a star, though. The least inefficient way is to use a square pattern. Obviously a diagonal is a cl | #trilema |
| 2015-02-21 | asciilifeform | fluffypony: my particular favourite language (cl) is as high level as it gets. | #trilema |
| 2015-02-17 | Adlai | some mix of: lazyness, the desire to periodically load my code into a fresh image, and ansi cl portability | #trilema |
| 2015-02-13 | asciilifeform | naggum's -other- lang. the only one other than cl that had not only a standard but a -rationale document- for every page | #trilema |
| 2015-02-13 | mircea_popescu | i'd have thoughjt a cl bitcoind | #trilema |
| 2015-02-09 | mircea_popescu | and what were symbolics sources in anyway, cl ? | #trilema |
| 2015-02-08 | asciilifeform | haskell package <<< pay someone to add a sexpr backend to it so I can work it into my CL toolchain ... | #trilema |
| 2015-02-08 | gabriel_laddel | As for what language I'll be using on any given project, it will be determined entirely by what is correct for the project. If I hack something out in CL, but switch it over to C, I'll do that. If it makes sense to release a protopye that is hacked together C+CL+CFFI I'll do that. In any case, both of these come after getting the #b-a distro finished. | #trilema |
| 2015-02-08 | gabriel_laddel | My current plan for dealing with these projects: The haskell package Language.C.AST parses C99 entirely, taking into account all GCC extensions. I'm going to pay someone to add a sexpr backend to it so I can work it into my CL toolchain. | #trilema |
| 2015-02-08 | asciilifeform | mircea_popescu: you can write cl without 'cons' but this is like english without letter 'e' | #trilema |
| 2015-02-08 | mircea_popescu | asciilifeform cl can't do these ? :D | #trilema |
| 2015-02-08 | gabriel_laddel | asciilifeform: gossipd requires C. PGP, CL (at least, this is where I'd start). | #trilema |
| 2015-02-08 | gabriel_laddel | mircea_popescu: CL + C when necessary. | #trilema |
| 2015-02-08 | mircea_popescu | gabriel_laddel i take this is going to be lisp, as in not scheme, not clojure, cl ? | #trilema |
| 2015-02-06 | asciilifeform | mircea_popescu: the cl rewrite of emacs, 'climacs', was almost usable | #trilema |
| 2015-02-06 | mircea_popescu | asciilifeform: trinque: how do you propose to have the browser end run cl? << haha this i wanna see. clonqueror. | #trilema |
| 2015-02-06 | asciilifeform | trinque: so you're actually using js. you will notice this, because the cl subset that can be thus eaten/shat is microscopic | #trilema |
| 2015-02-06 | trinque | doesn't the parenscript compiler eat cl and poop js? | #trilema |
| 2015-02-06 | asciilifeform | trinque: how do you propose to have the browser end run cl? | #trilema |
| 2015-02-06 | trinque | asciilifeform: does cl fall under $retardlang too? | #trilema |
| 2015-02-06 | asciilifeform | parenscript is just a lib for any ordinary cl | #trilema |
| 2015-02-05 | gabriel_laddel | asciilifeform: lol. The idea of a closure doesn't need to be stretched out across several chapters, but e.g., the cl-ppcre chapter is good for the non-lisper to read. | #trilema |
| 2015-02-04 | mircea_popescu | not like cl has that many usable apps out there :D | #trilema |
| 2015-01-25 | ben_vulpes | asciilifeform: what about a cl node? | #trilema |
| 2015-01-21 | ben_vulpes | stumpwm (a cl window manager) handles multiple monitors better than OS X. | #trilema |
| 2015-01-13 | assbot | cl-async/tcp.lisp at master · orthecreedence/cl-async · GitHub ... ( http://bit.ly/1CcX6cI ) | #trilema |
| 2015-01-13 | assbot | cl-async/tcp.lisp at master · orthecreedence/cl-async · GitHub ... ( http://bit.ly/1CcX6cI ) | #trilema |
| 2015-01-03 | mircea_popescu | ben_vulpes: your cl is oddly object oriented << o.O | #trilema |
| 2014-12-26 | adlai | the Right Thing™ would be to use cl-json instead of st-, and schematize everything | #trilema |
| 2014-12-26 | assbot | Logged on 26-12-2014 09:05:31; ben_vulpes: your cl is oddly object oriented | #trilema |
| 2014-12-26 | ben_vulpes | your cl is oddly object oriented | #trilema |
| 2014-12-16 | asciilifeform | i own computers that do not run emacs,slime,cl. | #trilema |
| 2014-12-16 | gabriel_laddel | asciilifeform: while we're on the subject, is there anything you consider absolutely essential for computing outside of CL+SLIME? | #trilema |
| 2014-12-02 | assbot | Re: Upper limits of CL - Naggum cll archive ... ( http://bit.ly/1vjzLFu ) | #trilema |
| 2014-11-19 | asciilifeform | decimation: anything short of cl's programmable condition system is retarded. | #trilema |
| 2014-11-07 | Adlai | damnit python, where's cl:change-class when you need it | #trilema |
| 2014-11-05 | decimation | asciilifeform: it's certain that there was some government meeting where the cl-http server was regarded as 'obsolete' because it wasn't compatible with frontpage or whatever | #trilema |
| 2014-11-05 | assbot | CL-HTTP History | #trilema |
| 2014-11-02 | gabriel_laddel | in CL this is utterly fucking trivial (apropos ""). | #trilema |
| 2014-10-31 | assbot | Re: Engineering Envy [was: Re: CL and UML] - Naggum cll archive | #trilema |
| 2014-10-31 | adlai | - is amazing, because CL actually has the concept of a quine programmed into it's specification | #trilema |
| 2014-10-30 | asciilifeform | ~6k lines of cl. | #trilema |
| 2014-10-30 | mircea_popescu | good lord, he wants to rewrite cl as part of rewriting bitcoin | #trilema |
| 2014-10-30 | ben_vulpes | note that asciilifeform says "my cl". | #trilema |
| 2014-10-30 | asciilifeform | my cl often runs 1.5x or so of naive 'c' | #trilema |
| 2014-10-28 | asciilifeform | ben_vulpes: interpreting << my cl system has no interpreter at all. | #trilema |
| 2014-10-27 | Adlai | it's kinda sad that going from CL to Python is a step backwards in concurrency support | #trilema |
| 2014-10-24 | bounce | hm? plenty glasses are marked with dl, not cl and marking with ml would be misleading | #trilema |
| 2014-10-24 | mircea_popescu | bounce it may not be to you, but it's proven nonsense in practice. your beer is not 5 deciliters, but most likely 50 cl or 500ml. | #trilema |
| 2014-10-22 | asciilifeform | redo the whole thing in cl in a night already << ask me again in peacetime | #trilema |
| 2014-10-22 | bounce | redo the whole thing in cl in a night already | #trilema |
| 2014-10-22 | mircea_popescu | certainly the cl is a good antecessor. | #trilema |
| 2014-10-17 | ben_vulpes | i suspect pythongs are going to resist a cl | #trilema |
| 2014-10-17 | ben_vulpes | it's no cl, but it's nice. | #trilema |
| 2014-10-13 | Adlai | try writing the shortest possible quine in ansi cl. if you want to check the answer, it was just in #lisp and there are public logs, but try yourself before you check. | #trilema |
| 2014-10-09 | asciilifeform | problem (or not problem, depending on who you are) is that most cl work was and is done by 'naggum-like' figures - that is, behind closed doors, no public derping, no 'foss' | #trilema |
| 2014-10-09 | asciilifeform | gabriel_laddel: tricky bit here. until rather recently (2009-ish) this was debatable (sparse ecosystems of publicly-available libraries in cl to deal with unfortunately ubiquitous crud like www) | #trilema |
| 2014-10-09 | gabriel_laddel | decimation: re, the problem with lisp... this isn't CL's problem, and not a real problem anyway. Compare to writing ~ALGOL on c-machine. | #trilema |
| 2014-09-10 | assbot | Re: Upper limits of CL - Naggum cll archive | #trilema |
| 2014-08-26 | assbot | Re: Upper limits of CL - Naggum cll archive | #trilema |
| 2014-07-18 | thickasthieves | fukn instant cl;assic https://www.youtube.com/watch?v=gj39POBr-OU | #trilema |
| 2014-06-10 | asciilifeform | mallery wrote cl-http, infamous for draconian license | #trilema |
| 2014-04-08 | asciilifeform | most web crap in common lisp uses 'CL+SSL', an FFI wrapper of traditional openssl. | #trilema |
| 2014-01-27 | BingoBoingo | Do most things graphically and occasionally kick the machine in the pants by the cl | #trilema |
| 2014-01-24 | nubbins` | in general usage, web services don't have an I of any type, CL or otherwise | #trilema |
| 2013-12-13 | Apocalyptic | question is how is that related to CL | #trilema |
| 2013-12-13 | Apocalyptic | so why are you suggesting it was a CL default coverup ? | #trilema |
| 2013-12-13 | Apocalyptic | what was there to cover for CL ? | #trilema |
| 2013-12-13 | pankkake | so he took the easy route trying to cover his lies about CL | #trilema |
| 2013-12-05 | asciilifeform | well, genera comes with john c. mallery's cl-http. | #trilema |
| 2013-05-02 | ThickAsThieves | maybe Gox feels CL broke | #trilema |
| 2013-05-02 | ThickAsThieves | CL just wants a settlement to go awaya | #trilema |