| 00:12:24 | * | Kooda quit (Quit: leaving) |
| 01:02:11 | * | icebattle quit (Quit: Leaving) |
| 01:51:46 | * | DAddYE quit (Remote host closed the connection) |
| 02:12:17 | EXetoC | woop |
| 02:12:47 | OrionPK | watap |
| 02:27:27 | Varriount | OrionPK: On the latest version of master? |
| 02:27:39 | OrionPK | yeah |
| 02:27:57 | OrionPK | assuming you havent updated it ;) |
| 02:28:10 | Varriount | OrionPK: Huh, maybe my change got overridden. I separated comment and doc-comment scopes |
| 02:28:35 | OrionPK | nevermind, guesss you fixe that ;) |
| 02:28:46 | OrionPK | I just hadnt gotten latest since yesterday |
| 02:30:25 | Varriount | OrionPK: Yeah, it's fixed. All it required was me splitting the comment scopes into two, and checking for the one belonging to documentation comments |
| 02:31:49 | OrionPK | works well now |
| 02:31:53 | Varriount | :D |
| 02:32:46 | OrionPK | u borked the asm highlighting though |
| 02:33:06 | Varriount | OrionPK: I had a merge conflict, and did the best I could. |
| 02:33:25 | OrionPK | if u remove "asm" from the other keywords it should work again |
| 02:34:14 | Varriount | Or I could make asm non-greedy :P |
| 02:35:56 | OrionPK | or that |
| 02:36:15 | NimBot | Varriount/Nimrod-Sublime master 896d9a2 Clay Sweetser [+0 ±3 -0]: DocContinue.py - Remove some commenting, set debug variable to false... 2 more lines |
| 02:36:33 | * | Varriount hugs NimBot |
| 02:38:29 | OrionPK | ;D |
| 02:44:18 | * | bastian_1 quit (Ping timeout: 252 seconds) |
| 02:52:51 | * | DAddYE joined #nimrod |
| 02:57:30 | * | DAddYE quit (Ping timeout: 252 seconds) |
| 03:03:00 | * | DAddYE joined #nimrod |
| 03:06:53 | * | brson quit (Quit: Lost terminal) |
| 03:07:24 | * | DAddYE quit (Ping timeout: 252 seconds) |
| 03:08:36 | * | brson joined #nimrod |
| 03:19:33 | * | DAddYE joined #nimrod |
| 03:41:12 | OrionPK | so what's next varriount |
| 03:44:21 | Varriount | Next? |
| 03:44:36 | OrionPK | did you figure out the issue with go to definition? |
| 03:44:49 | OrionPK | next for the st plugin |
| 03:44:55 | Varriount | Nope. Sublime Text just crashes. |
| 03:45:18 | OrionPK | did u try sticking in some prints |
| 03:45:23 | OrionPK | debugging it |
| 03:45:33 | Varriount | Nope. I been asleep for most of the day. |
| 03:45:37 | OrionPK | mmk |
| 03:45:38 | Varriount | *I've |
| 03:47:20 | OrionPK | I kinda want to do a light table plugin |
| 03:47:23 | OrionPK | :-) |
| 04:25:01 | * | dmac quit (Quit: Leaving.) |
| 04:55:25 | * | EXetoC quit (*.net *.split) |
| 05:07:25 | * | EXetoC joined #nimrod |
| 05:10:28 | * | OrionPK quit (Remote host closed the connection) |
| 05:12:00 | * | DAddYE quit () |
| 05:37:23 | * | dmac joined #nimrod |
| 06:10:05 | * | ddl_smurf quit (Quit: ddl_smurf) |
| 06:30:36 | * | discoloda joined #nimrod |
| 06:31:22 | discoloda | hello |
| 06:48:33 | * | brson quit (Quit: leaving) |
| 07:38:46 | * | shodan45 quit (Quit: Konversation terminated!) |
| 08:20:54 | * | Araq_ joined #nimrod |
| 08:23:22 | Araq_ | hi discoloda welcome |
| 09:07:46 | * | dmac1 joined #nimrod |
| 09:12:20 | * | dmac1 quit (Ping timeout: 252 seconds) |
| 09:28:12 | * | achim joined #nimrod |
| 09:33:39 | * | CarpNet joined #nimrod |
| 09:35:18 | * | dmac quit (Quit: Leaving.) |
| 09:37:05 | * | dmac joined #nimrod |
| 10:11:02 | * | dmac quit (Quit: Leaving.) |
| 10:14:02 | * | faassen joined #nimrod |
| 10:27:03 | * | io2 joined #nimrod |
| 10:50:27 | * | dmac joined #nimrod |
| 10:57:50 | * | dmac quit (Ping timeout: 264 seconds) |
| 11:38:56 | * | Mat3 joined #nimrod |
| 11:39:06 | Mat3 | G'Day |
| 11:53:51 | Araq_ | hi Mat3 |
| 11:54:36 | Mat3 | hi Araq, I hope your locking code works as intended |
| 11:55:19 | * | dmac joined #nimrod |
| 11:58:00 | * | faassen quit (Remote host closed the connection) |
| 12:00:05 | * | dmac quit (Ping timeout: 252 seconds) |
| 12:05:28 | Mat3 | hi dmac |
| 12:09:10 | Araq_ | you mean locks.TLock ? |
| 12:09:22 | Mat3 | yes |
| 12:13:02 | Araq_ | pretty sure it does, yeah |
| 12:14:03 | Mat3 | nice ! |
| 12:18:53 | Mat3 | ciao |
| 12:19:02 | Araq_ | bye |
| 12:19:20 | * | Mat3 quit (Quit: http://quassel-irc.org - Chat comfortably. Anywhere.) |
| 12:21:30 | Araq_ | wow win xp still has 29% market share |
| 12:28:51 | * | radsoc joined #nimrod |
| 12:30:30 | * | Araq_ quit (Quit: ChatZilla 0.9.90.1 [Firefox 26.0/20131205075310]) |
| 12:55:32 | * | dmac joined #nimrod |
| 13:00:09 | * | dmac quit (Ping timeout: 272 seconds) |
| 13:08:28 | * | darkf quit (Quit: Leaving) |
| 13:48:24 | * | [1]Endy joined #nimrod |
| 14:04:29 | * | OrionPK joined #nimrod |
| 14:20:23 | * | Araq_ joined #nimrod |
| 14:25:07 | * | Araq_ quit (Client Quit) |
| 14:32:15 | * | radsoc quit (Ping timeout: 300 seconds) |
| 14:33:52 | * | jpoirier quit (Read error: Connection reset by peer) |
| 14:43:03 | * | faassen joined #nimrod |
| 14:43:22 | * | achim quit (Quit: Computer has gone to sleep.) |
| 14:45:55 | * | achim joined #nimrod |
| 14:56:20 | * | dmac joined #nimrod |
| 14:57:26 | * | achim quit (Quit: Computer has gone to sleep.) |
| 15:00:44 | * | dmac quit (Ping timeout: 260 seconds) |
| 15:04:53 | * | achim joined #nimrod |
| 15:25:51 | * | Demos joined #nimrod |
| 15:28:05 | * | achim quit (Quit: Computer has gone to sleep.) |
| 15:29:33 | * | achim joined #nimrod |
| 15:32:36 | * | tumanovm joined #nimrod |
| 15:41:00 | * | achim quit (Quit: Computer has gone to sleep.) |
| 15:41:34 | * | achim joined #nimrod |
| 15:48:53 | * | tumanovm quit (Ping timeout: 252 seconds) |
| 15:52:21 | * | achim quit (Quit: Computer has gone to sleep.) |
| 15:53:08 | * | achim joined #nimrod |
| 15:54:02 | * | [2]Endy joined #nimrod |
| 15:56:47 | * | [1]Endy quit (Ping timeout: 260 seconds) |
| 16:04:09 | * | psquid quit (Quit: work) |
| 16:13:00 | * | [1]Endy joined #nimrod |
| 16:16:23 | * | [2]Endy quit (Ping timeout: 252 seconds) |
| 16:22:36 | * | achim quit (Quit: Computer has gone to sleep.) |
| 16:25:08 | * | aftersha_ joined #nimrod |
| 16:25:55 | * | faassen quit (Quit: faassen) |
| 16:53:42 | * | aftersha_ quit (Quit: Computer has gone to sleep.) |
| 16:55:56 | * | achim joined #nimrod |
| 16:57:03 | * | dmac joined #nimrod |
| 16:57:12 | * | [2]Endy joined #nimrod |
| 16:57:23 | discoloda | how can i return a object of T into a ref T? |
| 17:00:40 | * | [1]Endy quit (Ping timeout: 260 seconds) |
| 17:01:45 | * | dmac quit (Ping timeout: 265 seconds) |
| 17:03:43 | EXetoC | 'ref T' is the return type? there are several ways to deal with that. new(result); result.foo = ... for example |
| 17:04:37 | EXetoC | and "result[] = valueType" I think, where [] dereferences |
| 17:05:15 | discoloda | okay, so ref is pretty much allocated only, unlike ptr where you can get the addr of T |
| 17:10:02 | EXetoC | I think you can get the address in both cases by casting to uint |
| 17:12:11 | EXetoC | the Type(x: 0...) syntax should work for ref objects too. also: "For a ref object type system.new is invoked implicitly." |
| 17:13:24 | discoloda | okay, thanks |
| 17:19:18 | * | aftersha_ joined #nimrod |
| 17:20:48 | * | [1]Endy joined #nimrod |
| 17:24:35 | Varriount | Good morning guys! |
| 17:24:56 | * | [2]Endy quit (Ping timeout: 265 seconds) |
| 17:25:49 | Varriount | Araq: I'm not surprised at WinXP's market share, it's better than 2000/ME, and less crap-loaded than Vista |
| 17:34:48 | * | icebattle joined #nimrod |
| 17:52:50 | * | shodan45 joined #nimrod |
| 17:54:38 | * | brson joined #nimrod |
| 17:55:22 | * | aftersha_ quit (Quit: Computer has gone to sleep.) |
| 17:57:16 | * | dmac joined #nimrod |
| 17:57:41 | * | io2 quit (Remote host closed the connection) |
| 17:58:06 | * | io2 joined #nimrod |
| 18:01:47 | * | dmac quit (Ping timeout: 245 seconds) |
| 18:06:12 | * | [2]Endy joined #nimrod |
| 18:06:22 | * | margocha joined #nimrod |
| 18:09:51 | * | [1]Endy quit (Ping timeout: 272 seconds) |
| 18:11:22 | Demos | gah XP needs to die |
| 18:13:15 | * | DAddYE joined #nimrod |
| 18:13:20 | * | DAddYE quit (Remote host closed the connection) |
| 18:13:33 | * | DAddYE joined #nimrod |
| 18:13:43 | * | margocha quit (Remote host closed the connection) |
| 18:20:34 | * | radsoc joined #nimrod |
| 18:25:51 | * | dmac joined #nimrod |
| 18:31:46 | * | [1]Endy joined #nimrod |
| 18:32:27 | * | io2 quit () |
| 18:35:11 | * | [2]Endy quit (Ping timeout: 272 seconds) |
| 18:36:05 | * | ddl_smurf joined #nimrod |
| 18:59:04 | * | dmac quit (Quit: Leaving.) |
| 19:18:37 | * | dmac joined #nimrod |
| 19:21:03 | * | Mat3 joined #nimrod |
| 19:21:17 | Mat3 | hello |
| 19:21:41 | discoloda | howdy |
| 19:22:42 | Mat3 | hi discoloda |
| 19:23:01 | * | [2]Endy joined #nimrod |
| 19:25:56 | * | aftersha_ joined #nimrod |
| 19:26:51 | * | [1]Endy quit (Ping timeout: 272 seconds) |
| 19:45:13 | Mat3 | I take a quick look on the article in Dr'Dobbs journal mention Nimrod and remarkely the author advertised a full article later this year about our language (I doubt in combination with an article about Rust) |
| 19:46:06 | Araq | huh? |
| 19:46:21 | Araq | he told me he would release the full article in january |
| 19:46:30 | Araq | not "later this year" |
| 19:47:27 | Mat3 | well, for me the end of january for example is "later this year" in my opinion |
| 19:47:54 | Araq | sure ... this strange |
| 19:48:00 | Araq | *still |
| 19:48:13 | * | io2 joined #nimrod |
| 19:50:00 | Mat3 | obviously an example of my personal Denglish style ;) |
| 19:50:22 | Mat3 | sorry |
| 19:50:47 | discoloda | if this is the article, http://www.drdobbs.com/jvm/the-rise-and-fall-of-languages-in-2013/240165192, it says "later this month" |
| 19:50:52 | Araq | well ... I have a sweet implementation for Future[T] but it's not correct ... |
| 19:51:13 | Mat3 | discoloda: Yes it is this article |
| 19:51:37 | Araq | it's however so sweet that I'm playing with the idea to change its semantics and declare victory ... *cough* |
| 19:52:02 | Mat3 | how about bestFuture[T] ? |
| 19:52:16 | Araq | maybe I should name it CarefulHere[T] :-) |
| 19:53:12 | Araq | it is however embarrassing to claim freedom of deadlocks and then say "oh you can only have N futures or it'll deadlock" ... |
| 19:54:00 | * | gradha joined #nimrod |
| 19:54:36 | discoloda | 2 questions, have you seen my pull request for c2nim improvements and would more be welcome. my goal is to get it to convert quake 1,2,3 source |
| 19:56:17 | Araq | oh you're that guy |
| 19:56:27 | discoloda | yeah |
| 19:56:32 | Araq | I've seen your pull request |
| 19:56:39 | Araq | well I haven't yet code reviewed it |
| 19:57:08 | Araq | I can do it later though, bbl |
| 19:57:19 | discoloda | cool |
| 20:08:46 | * | Varriount|Mobile joined #nimrod |
| 20:23:24 | * | OrionPK quit (Remote host closed the connection) |
| 20:25:20 | * | OrionPK joined #nimrod |
| 20:25:52 | Araq | discoloda: did you ensure the c2nim tests still work? |
| 20:26:50 | OrionPK | small update to anyone interested (dont link to anyone off this channel) https://dl.dropboxusercontent.com/u/417554/ircfamiliar.zip |
| 20:30:59 | Mat3 | hi OrionPK |
| 20:31:07 | OrionPK | hi Mat3 |
| 20:31:28 | OrionPK | sec |
| 20:31:34 | * | OrionPK left #nimrod (#nimrod) |
| 20:31:36 | * | OrionPK joined #nimrod |
| 20:31:39 | * | [1]Endy joined #nimrod |
| 20:31:48 | OrionPK | mm |
| 20:32:07 | Araq | OrionPK: "dont link to anyone off this channel" is a bad idea, this channel is logged |
| 20:32:22 | OrionPK | well, if someone goes through the trouble of looking it up in the logs, they're ok by my book |
| 20:34:28 | * | [2]Endy quit (Ping timeout: 252 seconds) |
| 20:38:29 | discoloda | looks like i broke systest.c, matrix.h didnt work for original or my changes |
| 20:39:27 | gradha | so it never worked? |
| 20:40:02 | Araq | hmm iirc matrix.h did work for me |
| 20:40:11 | discoloda | i wouldnt say never, but from the point i started my changes it didnt |
| 20:40:33 | Araq | well you forgot to implement "new" and "delete" in c++ mode |
| 20:41:49 | dom96 | OrionPK: Does it work in Firefox yet? |
| 20:42:05 | OrionPK | no, havent had time to work on that yet |
| 20:43:07 | discoloda | ok, you can remove my pull request. ill try again when everything passes |
| 20:43:39 | OrionPK | i'll probably work on it more this weekend |
| 20:47:12 | Araq | discoloda: nud ldp and led are cool names but if even I can't guess their meaning it means you have abbreviated too much ... ;-) |
| 20:49:32 | EXetoC | ?, ?, light-emitting diode |
| 20:52:44 | Mat3 | ldp - load pointer ? |
| 20:52:46 | discoloda | gotcha, im not sure what they mean. but i guess these work: startExpression, leftBindingPower, leftExpression |
| 20:53:01 | Mat3 | brb |
| 20:53:50 | Araq | check out how compiler/parser does precedence parsing, seems to use a somewhat simpler algorithm |
| 20:54:00 | EXetoC | mnemonics is the way to go these days |
| 20:54:42 | Araq | discoloda: you need to enable c++ mode for the matrix.h test |
| 20:55:03 | discoloda | ah |
| 20:57:37 | Araq | also congrats, so you seem to have figured out how the lookahead + backtracking works :-) |
| 20:58:38 | * | OrionPK is really looking forward to improvemens to c2nim |
| 20:59:13 | * | [1]Endy quit (Ping timeout: 252 seconds) |
| 21:00:11 | discoloda | well, i have it mostly converting quake 2 |
| 21:00:42 | OrionPK | impressive |
| 21:00:52 | OrionPK | using changes to c2nim or changes to quake 2? |
| 21:00:58 | discoloda | c2nim |
| 21:01:06 | OrionPK | nice |
| 21:01:49 | discoloda | one change to quake 2 so far, where #if #endif blocks mess with control flow. cant convert those to when |
| 21:02:34 | Araq | yeah but c2nim can't read your mind and you have to learn how to use it, so it must suck. never mind it's better than h2pas or swig. |
| 21:04:57 | discoloda | its just the nature of the CPP, character based preprocessing |
| 21:05:14 | discoloda | has no concept of the code it changes |
| 21:06:08 | * | Araq is annoyed by c2nim bashing |
| 21:06:31 | discoloda | im not bashing c2nim |
| 21:06:31 | * | Araq considers it the best tool in the existance of mankind |
| 21:06:48 | Araq | no, not you, but OrionPK is |
| 21:07:20 | Mat3 | bb |
| 21:07:36 | dom96 | Is he? |
| 21:08:05 | Mat3 | Araq: well, it has its use |
| 21:10:24 | shodan45 | wasn't someone working on nimrod+llvm? |
| 21:11:10 | Araq | shodan45: not that I'm aware of, clang works fine with nimrod |
| 21:11:36 | Araq | I started an llvm backend once but lost interest |
| 21:13:51 | shodan45 | Araq: ah ok. Just curious because I just found python bindings for the "llvm library". |
| 21:14:15 | Araq | we have some old bindings somewhere too |
| 21:14:29 | OrionPK | dom96 i must have missed it :p |
| 21:14:32 | shodan45 | there's even a nice tutorial on making your own language/compiler with it |
| 21:14:46 | Mat3 | shodan45: What's the reason for your interest in LLVM ? |
| 21:15:03 | discoloda | thats what Nimrod needs, an eval function. compiles the given string into llvm bytecode then llvm runs it (compiled or interpreted) |
| 21:15:06 | shodan45 | Mat3: nothing in particular, it's just fun & interesting |
| 21:15:43 | Araq | discoloda: no, we already have that |
| 21:16:25 | Mat3 | discoloda: I work on somewhat similar and think to be finsihed end of february |
| 21:16:42 | Mat3 | ^finished |
| 21:17:42 | discoloda | nice, kinda ment that as a joke. how does it work? |
| 21:18:07 | Araq | the devel branch has a new register based virtual machine |
| 21:18:14 | * | Mat3 a bit simplificated statement but you will code in an dynamic environment with adaptive compilation beside editing |
| 21:18:21 | Araq | that's used for compile time evaluation |
| 21:18:39 | Araq | but will be exposed as "eval" |
| 21:19:14 | OrionPK | i wasnt bashing c2nim, but obviously any improvements to it should be welcomed |
| 21:20:31 | shodan45 | a compiled language with eval... that's either the best or worst of both worlds ;) |
| 21:20:52 | gradha | I'm looking for references, anybody know of recent private picture leaks since Scarlett Johansson's? |
| 21:21:41 | OrionPK | lol |
| 21:21:47 | OrionPK | wrong channel gradha |
| 21:21:58 | OrionPK | (although I wouldnt mind seeing anything new) |
| 21:22:22 | discoloda | i thought those were fake, just a look alike |
| 21:23:01 | gradha | I'm not interested in the pictures, I only want to know if somebody else more recently made the same mistake |
| 21:23:12 | gradha | basically, I'm not hipster enough to keep with the news |
| 21:23:14 | Mat3 | of course |
| 21:24:05 | gradha | the only leaks I know are about kpop idols, but I don't think people will get that when they read my paper |
| 21:24:47 | Araq | gradha: what paper? |
| 21:25:03 | gradha | I told you, I'm writing a paper |
| 21:25:22 | gradha | the title is "how to release software periodically" |
| 21:25:36 | gradha | I believe I mentioned it a few days ago |
| 21:25:41 | Araq | well tell me again please |
| 21:25:59 | gradha | it's kind of fight club style, but don't worry, I'm making it for you |
| 21:26:27 | Araq | does it include "how to implement futures efficiently in a thread local heaps setting?" |
| 21:26:39 | gradha | nope, it's precisely against that |
| 21:27:48 | Araq | ok, is an exception "too many unresolved futures, you a channel instead" acceptable? |
| 21:27:57 | Araq | *use a channel instead |
| 21:28:30 | gradha | only if you raise it during a call to echo |
| 21:31:19 | Araq | meh this future stuff doesn't work |
| 21:31:27 | gradha | ok, Scarlett Johannson it is then |
| 21:31:46 | Araq | Scarlett Johannson is certainly not a bad choice |
| 21:31:54 | Araq | I heard she likes porn |
| 21:32:06 | dom96 | Scarlett Johannson is a goddess. |
| 21:33:19 | OrionPK | we should get her to endorse nimrod |
| 21:33:22 | dom96 | Damn. Rust is moving at a pretty fast pace. |
| 21:33:33 | dom96 | 0.9 is just out. |
| 21:34:10 | bstrie | dom96: note that 0.9 will be followed by 0.10, not by 1.0 |
| 21:34:24 | bstrie | they're just regular snapshots that happen every 3 or 4 months |
| 21:34:31 | dom96 | Ahh. Interesting. |
| 21:37:00 | Mat3 | Rust rusts faster |
| 21:40:48 | * | achim quit (Quit: Computer has gone to sleep.) |
| 21:41:09 | EXetoC | 0.10? no respect for ordering :( |
| 21:48:09 | Mat3 | go to bed, ciao |
| 21:48:15 | * | Mat3 quit (Quit: http://quassel-irc.org - Chat comfortably. Anywhere.) |
| 21:51:50 | wan | I'm still looking at the frameworkBenchmarks numbers, and I just realised something. |
| 21:52:09 | wan | I can augment the results of any db tests with nawak |
| 21:52:31 | wan | I just set up more workers that connect to Mongrel2, and it magically happens: numbers go up |
| 21:53:06 | wan | (on my 3-year-old laptop, with postgres+mongrel2+workers+wrk running on it) |
| 21:54:00 | wan | The only downside is the plaintext and json results go slightly down |
| 21:55:09 | wan | It's definitely a question of being cpu-bound or db-bound, and I can't make good enough estimations with my hardware because I'm cpu bound quite fast for db tests if I fire up more workers. |
| 21:55:33 | wan | The question being: how can I determine the optimal number of workers to win the benchmark? |
| 21:57:21 | Araq | you can't |
| 21:57:40 | Araq | you can tell the guy running your code to optimize the number of workers for his machine ... |
| 21:57:53 | Araq | he'll gladly do that, I'm sure |
| 21:59:01 | wan | But he would just guess to start as many as cores on the machine, but he'll be wrong |
| 21:59:24 | Araq | yup, often its twice the number of cores |
| 21:59:41 | wan | On my laptop (4 threads), if I fire 15 workers I get better db, queries and updates results |
| 22:00:23 | wan | Just because the db is the bottleneck, then you might as well fire up the most workers you can because they can do stuff until they block on the db response |
| 22:01:29 | wan | If I'm right, we could win on every database-driven test using this trick |
| 22:02:07 | discoloda | wont others do the same thing? |
| 22:02:16 | gradha | no, they are nicer |
| 22:02:30 | wan | And because I don't have a db running on an SSD (like their dedicated db i7 machine), I don't really know for sure that the db will be the bottleneck |
| 22:02:52 | wan | But I still guess it will be |
| 22:03:07 | wan | They don't seem to currently do it, no |
| 22:03:14 | gradha | good night, honey badgers |
| 22:03:36 | * | gradha quit (Quit: bbl, need to watch http://www.youtube.com/watch?v=dEf4PJZXBxA again) |
| 22:03:41 | Araq | good night, gradha |
| 22:03:52 | wan | That's why for example everyone has crappy numbers for the updates test, the slower. It's bound to be the db the bottleneck. |
| 22:03:57 | wan | night' |
| 22:04:21 | dom96 | Are you sure the rules don't say anything against doing this sort of thing? |
| 22:04:31 | wan | I'll check |
| 22:04:40 | wan | (but I don't see why they would forbid it) |
| 22:05:12 | wan | The configuration of the framework is just supposed to reflect what you would actually run in production. If it is better for prod, it's better for the benchmark |
| 22:06:26 | wan | Most of all the other frameworks probably just have such a small pool of db connections because they don't have that many workers fired up |
| 22:07:27 | wan | But I might be wrong, maybe I'm just fooling myself and a lower number of workers would be more beneficial |
| 22:09:12 | dom96 | I had a somewhat similar problem for the Jester benchmark. |
| 22:09:47 | dom96 | I didn't have enough time to add threading to jester, so I had to spawn some number of Jester instances. |
| 22:09:50 | dom96 | Picked 8 IIRC |
| 22:09:56 | dom96 | Obviously not ideal. |
| 22:10:34 | wan | Yeah, but for /json or /plaintext, one worker for nawak obtains a higher score. |
| 22:10:59 | wan | For example, on my laptop it's 25 000 req/s |
| 22:11:15 | wan | And with 15 workers it's 22 000 req/s |
| 22:12:23 | wan | On the other hand, for the /db test, one worker: 7 000 req/s. two workers: 12 000 req/s. more: 14 000 req/s, I can't go higher, I'm cpu bound. |
| 22:12:44 | EXetoC | dom96: any plans to allow the content retrieval method in Jester to be overwritten? |
| 22:12:56 | EXetoC | is this related to "cache control" in the TODO file or is that some http thing? |
| 22:13:01 | dom96 | EXetoC: the what? |
| 22:13:09 | dom96 | elaborate please |
| 22:16:03 | OrionPK | wan, given your name I was hoping you somehow got g-wan working with Nimrod :-P |
| 22:16:04 | shodan45 | hm, I don't remember what it was about rust that I didn't like... and now that I look at it again, it's not too bad |
| 22:18:12 | EXetoC | dom96: optional user-defined GET handling: "proc(path: string ...): string = ..." for caching and what have you |
| 22:18:26 | wan | OrionPK: Nah, g-wan is useless. I don't use proprietary software if I can avoid it. Besides, monkey/duda (http://duda.io/documentation/hello_world_5_steps) is faster. |
| 22:19:11 | wan | OrionPK: And I'll make the monkey bindings at some point and try it out. |
| 22:19:14 | EXetoC | which would override the default method, which simply attempts to fetch the file |
| 22:19:36 | dom96 | EXetoC: oh. No plans for that. |
| 22:20:03 | OrionPK | I'd never heard of duda |
| 22:20:40 | EXetoC | ok maybe I'll have a go at that this weekend |
| 22:22:23 | wan | OrionPK: Me neither before a few months ago. I think I originally found it because I was looking at the FrameworkBenchmarks issues. Lots of "I've never heard of that framework before, what the hell is it?" |
| 22:28:36 | OrionPK | won't work on windows though |
| 22:28:38 | OrionPK | ? |
| 22:32:03 | wan | Yeah no. Linux is the master on servers. |
| 22:32:31 | OrionPK | right, but some people like to develop on windows |
| 22:32:32 | shodan45 | never heard of duda or monkey either... looks interesting, but a little weird |
| 22:32:38 | OrionPK | and suport multiple platforms |
| 22:34:14 | wan | Server software usually have fewer "let's run it on everything!" requirements. If you know your way around GNU/Linux, you usually don't settle for Windows or others. |
| 22:34:54 | OrionPK | whats an example |
| 22:35:04 | wan | of? |
| 22:35:29 | OrionPK | another server library that's linux only |
| 22:35:44 | OrionPK | I can think of a lot of windows only software like that |
| 22:36:31 | wan | Not only server libraries, server software as well. Do nginx, apache run on Windows? |
| 22:36:51 | Araq | apache certainly does |
| 22:36:53 | OrionPK | yeah |
| 22:36:56 | OrionPK | they both do |
| 22:37:08 | wan | Ok, I didn't know that. |
| 22:38:11 | wan | Well, Mongrel2 only runs on Linux (yeah, it's not popular at all). |
| 22:40:14 | shodan45 | I forgot how much I don't like C... stupid identifiers popping up out of nowhere :/ |
| 22:40:46 | shodan45 | although the claimed speed of this duda thing is pretty crazy |
| 22:41:40 | shodan45 | supports x86, x86_64, and arm... but no MIPS? |
| 22:42:04 | shodan45 | (they claim it's perfect for embedded environments) |
| 22:42:46 | Araq | bstrie: did Rust move to "move" semantics everywhere with 0.9.0 ? |
| 22:44:03 | bstrie | Araq: "move" has been the default since around... 0.5? |
| 22:44:10 | bstrie | for a while, basically |
| 22:44:36 | wan | OrionPK: Damn, lots of server stuff is cross-platform (rabbitmq, mongodb, postgresql, redis has a port, ...). But I'll find something linux-only, and popular. Let's see... |
| 22:44:43 | bstrie | we used to have really complicated rules about when a thing would be moved or copied, and had both "move" and "copy" as keywords to override the defaults |
| 22:44:59 | OrionPK | lol |
| 22:45:10 | OrionPK | it's not that hard to support windows w/ mingw and such |
| 22:46:52 | OrionPK | depending on your level of abstraction from hw |
| 22:46:54 | Araq | bstrie: so I can copy a mutable pointer to create cyclic structures and linked lists? |
| 22:46:54 | Araq | but it's not the default? |
| 22:47:09 | * | psquid joined #nimrod |
| 22:48:14 | bstrie | Araq: you won't be able to copy a unique pointer, for obvious reasons |
| 22:48:32 | bstrie | Araq: for graphs and stuff you'll probably want either Gc or our strong/weak refcounted pointers |
| 22:48:55 | Araq | hmm |
| 22:49:07 | bstrie | Araq: for doubly linked lists you can either use unique pointers in one direction and unsafe pointers in the other direction, or again use the strong/weak refcounted pointers |
| 22:49:35 | Araq | ok so borrowed and unique pointers do prevent that |
| 22:49:49 | bstrie | cycles are the bane of unique pointers, yep |
| 22:50:30 | Araq | cycles can also be created via closures, how do you handle that? |
| 22:51:01 | Araq | (oh congrats on the 0.9 release btw!) |
| 22:51:24 | bstrie | thanks :) |
| 22:51:34 | bstrie | not sure if I can think of a concrete example of creating a cycle out of a closure |
| 22:51:52 | bstrie | I'm sure we prevent it statically, the only question is if we're being way more restrictive than is perhaps necessary |
| 22:54:12 | * | radsoc quit (Ping timeout: 252 seconds) |
| 22:54:21 | Araq | alright, how do you support my favourite parallel programming example? (inserting strings to a hash table in parallel) |
| 22:56:43 | * | aftersha_ quit (Quit: Computer has gone to sleep.) |
| 22:58:38 | wan | OrionPK: Finally, I found some examples: Varnish cache, uwsgi, HAProxy don't run on Windows. So yeah, lots actually have a (more difficult, of course) way of running under Windows. Yay for cross-platform! |
| 22:58:52 | OrionPK | :P |
| 22:59:04 | Araq | wan: the best example still is valgrind |
| 22:59:17 | Araq | not exactly "server software" though |
| 22:59:33 | wan | Yes, I was searching for that specifically |
| 23:00:18 | Araq | and mongodb is so broken on windows it's not funny and I would count it as "doesn't exist on windows" |
| 23:00:44 | Araq | but then maybe it's just as broken on linux ... |
| 23:01:17 | Demos | there are alternitives to valgrind that run on windows |
| 23:01:36 | Araq | Demos: they are all not free afaik |
| 23:01:53 | Demos | clang is |
| 23:02:10 | bstrie | Araq: if you want to share some memory among several tasks, we have an atomically reference-counted pointer type for that called "Arc" |
| 23:02:13 | bstrie | the name is bad |
| 23:02:21 | bstrie | everyone thinks it's the same as objC's arc... |
| 23:02:26 | Demos | not quite the same, but it can check for memory leaks and bad/uninitialized reads |
| 23:03:38 | Varriount | Araq: Downloading Whispered World right now. |
| 23:03:57 | Araq | bstrie: alright but how do you prevent races on that shared array? |
| 23:04:08 | Varriount | <3 Point and click adventures |
| 23:04:26 | Araq | Varriount: play indiana jones 4 instead ;-) |
| 23:04:45 | Araq | but whispered world is good too |
| 23:04:47 | bstrie | Araq: I'm not an expert here. have some really old documentation that should still be vaguely relevant: http://winningraceconditions.blogspot.com/2012/09/rust-4-typesafe-shared-mutable-state.html |
| 23:05:09 | discoloda | how about a FPS / Point and Click hybrid |
| 23:05:30 | * | filwit joined #nimrod |
| 23:05:45 | filwit | hey guys |
| 23:06:01 | bstrie | Araq: the tldr is that it statically prevents data races |
| 23:06:15 | Demos | http://eigen.pri.ee/shooter/ even better :D |
| 23:06:15 | bstrie | ...I think |
| 23:06:27 | Varriount | Hi filwit! |
| 23:06:29 | bstrie | I don't do much parallel stuff |
| 23:06:49 | filwit | hey Varriount :) |
| 23:08:18 | filwit | i saw araq's lock articles, looked pretty cool |
| 23:08:50 | filwit | i don't know enough about threading to comment on the article much, but i was easy to follow and well written |
| 23:09:06 | Demos | my only concern atm is that having manually managed memory that contains pointers to GC'd memory gets even more painful |
| 23:09:38 | Araq | bstrie: no need for tldr, I already read it. and it uses my solution or something very close to it :-) |
| 23:10:28 | Araq | Demos: define "more painful" please |
| 23:10:45 | Araq | currently it's allowed but so unsafe that it can't seriously be used |
| 23:10:59 | Demos | yeah, that was my point |
| 23:11:57 | Araq | hi filwit |
| 23:12:14 | filwit | hey |
| 23:12:21 | Demos | anyhow I gotta go. |
| 23:12:25 | * | Demos quit (Quit: Leaving) |
| 23:13:29 | filwit | Araq: so, monday your video comes out |
| 23:14:17 | filwit | been waiting for awhile |
| 23:14:22 | Araq | apparently yeah |
| 23:14:27 | dom96 | we should all discuss it in the VNUG |
| 23:14:36 | filwit | VNUG? |
| 23:14:45 | dom96 | Virtual Nimrod User Group |
| 23:14:57 | filwit | this is something i don't know about |
| 23:14:58 | bstrie | Araq: you should write a blog post about it :) |
| 23:15:42 | filwit | dom96: is that a fancy way to say 'forums' or is this some kind of teamspeak thing? |
| 23:15:54 | Araq | bstrie: I already did ... but please don't reddit it |
| 23:16:01 | dom96 | filwit: teamspeak |
| 23:16:26 | filwit | dom96: cool :) how do i join it? |
| 23:16:39 | filwit | is it actually teamspeak? |
| 23:16:45 | bstrie | Araq: linky linky |
| 23:16:52 | filwit | or some other software? |
| 23:17:03 | filwit | http://forum.nimrod-lang.org/t/334 |
| 23:17:10 | filwit | ^ bstrie |
| 23:17:17 | dom96 | Look at the topic "VNUG: join us on teamspeak at nimrod-lang.org" |
| 23:17:25 | * | dom96 bets no one is there though heh |
| 23:17:41 | dom96 | It's BitPuffin's idea and he's playing Dota now most of the time. |
| 23:18:17 | filwit | lol, i noticed you two on Dota over the past week |
| 23:18:29 | dom96 | Yeah. I may have been playing too lol |
| 23:19:14 | filwit | not really my kind of game or i would play with you guys sometime |
| 23:19:30 | * | darkf joined #nimrod |
| 23:19:35 | dom96 | I kind of suck at it tbh |
| 23:20:10 | filwit | idk even how it works really, i'm just not into those top-down RTS/raid-style games much |
| 23:20:53 | filwit | anyways, i'm having trouble finding that page you where talking about |
| 23:21:13 | filwit | one sec |
| 23:21:32 | dom96 | This one? https://thestrangeloop.com/news/strange-loop-2013-video-schedule |
| 23:21:42 | filwit | no for teamspeak |
| 23:22:00 | dom96 | oh |
| 23:22:45 | filwit | i don't see it on the site at all, did you mean the forums? |
| 23:23:00 | dom96 | I meant the IRC topic |
| 23:23:28 | filwit | idk what that means really |
| 23:23:59 | filwit | i only started using IRC a couple years ago, never really learned how it worked all the way |
| 23:25:34 | Araq | use an irc client that shows the topic |
| 23:26:06 | filwit | have a recommendation? i use xchat now |
| 23:26:49 | dom96 | see that textbox at the top? |
| 23:26:56 | dom96 | that's the topic |
| 23:27:12 | * | Varriount|Mobile quit (Ping timeout: 245 seconds) |
| 23:27:18 | Araq | I use chatzilla fwiw |
| 23:27:43 | dom96 | chatzilla sucks |
| 23:27:52 | OrionPK | x/y-chat is decent |
| 23:27:56 | OrionPK | x-chat or y-chat |
| 23:28:01 | Araq | dom96: why? |
| 23:28:08 | filwit | dom96: i'm still looking up irc-topics online, but i don't see how the top bar is useful |
| 23:28:37 | filwit | dom96: i can see the text you quoted, but i don't see anything else related to how to setup teamspeak or anything |
| 23:28:53 | dom96 | Setting up teamspeak is simple |
| 23:28:55 | dom96 | just download it |
| 23:29:28 | filwit | i see, so some string of text up there is what i need for teamspeak, gotit |
| 23:37:06 | fowl | filwit, /topic |
| 23:37:45 | filwit | thnx |
| 23:51:10 | * | io2 quit () |
| 23:52:36 | discoloda | irssi ftw |
| 23:53:09 | dom96 | Araq: It requires Firefox to be running |
| 23:55:21 | Araq | dom96: yeah and you shouldn't browse the web |
| 23:55:22 | Araq | it's too distracting |
| 23:55:29 | filwit | btw, is there a better way to test if a function exist than system.compiles ? (one that doesn't require object instantiation |
| 23:55:39 | filwit | ) |
| 23:56:31 | filwit | an object to be allocated*** |
| 23:59:10 | Araq | when defined(strutils.toUpper) |
| 23:59:37 | Araq | curious you know about "compiles" but not about "defined" |
| 23:59:46 | Araq | when "compiles" is about to be removed |
| 23:59:52 | filwit | oh duh |