<< 15-06-2014 >>
00:00:21*Kazimuth joined #nimrod
00:20:18flaviudom96: You're in charge of the builder site, right?
00:21:56*shodan45 quit (Quit: Konversation terminated!)
01:00:48*saml_ joined #nimrod
01:06:03*saml_ quit (Ping timeout: 240 seconds)
01:09:31*freezerburnv joined #nimrod
01:16:40flaviudom96: It appears you are. Check your PR box for nimbuild.
01:19:41*saml_ joined #nimrod
01:21:07flaviuAraq: Don't worry about it, but do you know where the error is in the new code? My logging statements are turning up nothing.
01:30:06*flaviu quit (Remote host closed the connection)
02:04:09VarriountAraq: ping
02:09:08Varriountdom96: Ping
02:09:23Varriountdom96: The babel package listing is invalid, one of the lines is missing a comma
02:10:26*q66 quit (Quit: Leaving)
02:10:55NimBotAraq/Nimrod devel 1354257 Varriount [+0 ±1 -0]: Update json.nim... 2 more lines
02:14:47*hoverbear joined #nimrod
02:14:52*hoverbear quit (Max SendQ exceeded)
02:16:01*hoverbear joined #nimrod
02:32:39NimBotAraq/Nimrod devel 370f06d Varriount [+0 ±1 -0]: Update categories.nim... 2 more lines
02:41:03*freezerburnv quit (Quit: freezerburnv)
03:14:16*Kazimuth quit (Ping timeout: 240 seconds)
03:53:33*springbok quit (Ping timeout: 244 seconds)
03:54:54*kemet joined #nimrod
03:57:16*BitPuffin quit (Ping timeout: 240 seconds)
05:00:22Varriount|MobileHello kemet
05:15:14*kemet quit (Remote host closed the connection)
05:15:50*kemet joined #nimrod
05:15:59*kemet quit (Remote host closed the connection)
05:28:16*BitPuffin joined #nimrod
06:15:23*io2 joined #nimrod
06:21:38*kemet joined #nimrod
06:24:56*saml_ quit (Quit: Leaving)
06:31:25*hoverbea_ joined #nimrod
06:33:03*Varriount|Mobile quit (Remote host closed the connection)
06:33:17*Varriount|Mobile joined #nimrod
06:34:41*hoverbear quit (Ping timeout: 264 seconds)
06:42:08*hoverbea_ quit ()
06:44:35*kemet quit (Quit: Instantbird 1.5 -- http://www.instantbird.com)
07:11:08*Demos quit (Read error: Connection reset by peer)
08:22:24*Simplicity joined #nimrod
08:22:35*Simplicity left #nimrod (#nimrod)
08:30:59*kunev joined #nimrod
08:37:24*eigenlicht joined #nimrod
08:45:07Araquh oh
08:45:17Araq"Simplicity has left #nimrod" ... a bad omen
09:10:33*kemet joined #nimrod
09:12:34*kemet quit (Client Quit)
09:12:53*kemet joined #nimrod
09:15:39*kemet quit (Remote host closed the connection)
09:16:59*Matthias247 joined #nimrod
09:59:41*enurlyx joined #nimrod
10:15:22*BitPuffin quit (Ping timeout: 244 seconds)
10:16:26*BitPuffin joined #nimrod
10:21:00*kemet joined #nimrod
10:36:34NimBotnimrod-code/packages master e646b42 Varriount [+0 ±1 -0]: Update packages.json... 2 more lines
10:36:34NimBotnimrod-code/packages master d23710a Dominik Picheta [+0 ±1 -0]: Merge pull request #65 from Varriount/patch-1... 2 more lines
10:36:37dom96Varriount: Nice catch.
10:38:22*Klaufir left #nimrod (#nimrod)
10:48:10*kemet quit (Quit: Instantbird 1.5 -- http://www.instantbird.com)
11:09:50*q66 joined #nimrod
11:09:50*q66 quit (Changing host)
11:09:50*q66 joined #nimrod
11:33:34dom96Well I figured out why caasdriver fails.
11:33:47dom96It uses a destructor which fires too soon
11:35:04dom96hrm, or maybe not.
11:57:13NimBotAraq/Nimrod devel 8965ee5 Dominik Picheta [+2 ±0 -0]: Add osproc stdin test.
11:57:13NimBotAraq/Nimrod devel 1658a29 Dominik Picheta [+0 ±1 -0]: Fixes #1263.
11:57:18*foodoo joined #nimrod
12:01:20Araqhow did you fix it?
12:01:28Araqtoo lazy to look at the dfif :P
12:02:33dom96there was no problem with osproc
12:02:54dom96just that caasdriver tried writing to a closed input stream
12:04:29foodooI wonder when Nimrod 1.0 will be released. Are there already plans for that?
12:05:21Araqfoodoo: christmas this year.
12:05:36Araqthough it's a really optimistic plan ...
12:06:30foodooAlright. Take your time :)
12:06:50foodooBut knowing that it's on the horizon makes me happy
12:12:10*vendethiel quit (Read error: Connection reset by peer)
12:14:54*vendethiel joined #nimrod
12:19:17*untitaker quit (Ping timeout: 240 seconds)
12:25:07*untitaker joined #nimrod
12:44:03*Varriount|Mobile quit (Quit: AndroIRC - Android IRC Client ( http://www.androirc.com ))
13:16:01*darkf quit (Quit: Leaving)
13:39:50*Trustable joined #nimrod
14:04:40*KevinKelley quit (Ping timeout: 246 seconds)
14:09:50*q66 quit (Quit: Leaving)
14:10:13*q66 joined #nimrod
14:11:31*q66 quit (Remote host closed the connection)
14:11:56*q66 joined #nimrod
14:17:31*flaviu joined #nimrod
14:19:03*kunev quit (Quit: leaving)
14:23:32flaviudom96: http://i.imgur.com/aHABpIR.png
14:24:25dom96flaviu: Nice, but it'll take a while before I can make it live.
14:42:26*foochain joined #nimrod
14:42:29dom96I wonder why there is both http://build.nimrod-lang.org/docs/system.html#getStackTrace, and http://build.nimrod-lang.org/docs/system.html#getStackTrace,_2
14:47:19*foochain left #nimrod (#nimrod)
14:51:37*foodoo quit (Quit: so long and thanks for the fish)
15:00:53Araqdom96: docgen bug ... :-/
15:01:05dom96Araq: that's what I thought
15:28:48*hoverbear joined #nimrod
15:29:36*KevinKelley joined #nimrod
15:39:59*Matthias247 quit (Read error: Connection reset by peer)
15:58:00*KevinKelley quit (Quit: Page closed)
16:02:53*enurlyx quit (Ping timeout: 252 seconds)
16:03:06*enurlyx joined #nimrod
16:27:05*flaviu quit (Ping timeout: 252 seconds)
17:25:32*flaviu joined #nimrod
17:38:00*kunev joined #nimrod
18:15:17*MayurYa joined #nimrod
18:15:32*MayurYa quit (Changing host)
18:15:32*MayurYa joined #nimrod
18:23:22Araqhi MayurYa welcome
18:24:39flaviuhttp://meta.stackoverflow.com/questions/260377/open-source-advertising-sidebar-2h-2014
18:26:44flaviuIf someone sees filwit, see if he wants to make an ad for that, I think it'd be a good way to gain exposure
18:37:02*Jehan_ joined #nimrod
18:45:08*q66 quit (Ping timeout: 252 seconds)
18:48:48VarriountAraq: So, what bugs do you think I can fix?
18:51:35*Varriount|Mobile joined #nimrod
18:57:53*q66 joined #nimrod
19:03:40VarriountAraq: Also, what kind of size does computeRecSize and friends compute? Is there anything interesting or unusual about how it computes things?
19:04:06*Puffin joined #nimrod
19:06:02*BitPuffin quit (Ping timeout: 245 seconds)
19:11:40*perturbation joined #nimrod
19:22:14*Matthias247 joined #nimrod
19:24:50dom96Why do idetools compile the file?
19:27:56Varriount|Mobiledom96: Because idetools is broken
19:29:01AraqVarriount|Mobile: easy, fix the most recently reported sigsev bug
19:29:29Araqand yes, computeRecSize is unusual in that it doesn't work at all
19:29:47Araqhence sizeof() cannot always be computed at compile-time
19:30:00Araqsomehow i got the alignment computations wrong
19:38:08dom96Varriount|Mobile: What ever happened to that explanation of compiler modules that Araq told us on Skype?
19:38:14dom96Varriount|Mobile: You said you wrote it down right?
19:40:26flaviudom96: https://gist.github.com/BitPuffin/c1d7e2c72bedf654f90f
19:41:53*kunev quit (Ping timeout: 264 seconds)
19:41:58dom96why isn't this on the wiki
19:41:58dom96?
19:42:03*Johz joined #nimrod
19:50:10flaviuI don't know, I guess no one bothered do that.
19:50:36Araqeverything except lambda lifting is self explaining anyway
19:50:54Araqlambda lifting however ... I don't understand it either
19:51:42dom96Araq: it's a good reference
20:00:22*Mat3 joined #nimrod
20:00:25Mat3hello
20:01:34NimBotAraq/Nimrod devel 79b0848 Dominik Picheta [+0 ±1 -0]: Fix idetools --def crash.
20:03:12Araqhi Mat3
20:03:38Araqso apparently <title /> is not valid HTML ... wtf
20:03:50Araqbut <title></title> is
20:04:33AraqI mean, ok, we all know html is not xml but this is just weird as fuck
20:06:05flaviuIt makes sense though, <title> is meaningless without a body
20:07:13dom96shouldn't <title> be in <head> anyway?
20:07:15flaviuAraq: <title></title> is invalid too.
20:07:50dom96flaviu: Really? Does the HTML spec state that you can't have an empty title?
20:08:46flaviuYes, the spec says so. It seems like the browsers let it work anyway
20:09:57Mat3Hi Araq
20:10:21Araqflaviu: ok,, but browsers accept <title></title> but not <title />
20:10:39Araqwhich is the real problem of course, the html spec is worthless
20:11:48Mat3as I know every HTML tag (with some exceptions) must end in a </..> tag
20:18:29Mat3and because empty tags are not specified, ther should at least surround a whitespace character
20:18:57Mat3^there
20:19:35AraqMat3: so what? maybe it becomes valid html 6?
20:20:06Araqwith an ever evolving spec and browsers all doing their own thing, why should one care about the spec?
20:20:42Araqless than 1% of all html documents adhere to the spec anyway
20:20:45flaviuMat3: from the spec: "The img element ... Content model: Empty.", "The meta element ... Content model: Empty."
20:20:49*MayurYa quit (Ping timeout: 272 seconds)
20:21:14flaviuThat sounds an awful lot like empty tags are specified.
20:22:36Mat3well as written "as I know"
20:22:38dom96I'm starting to wonder whether there is much point to CAAS.
20:23:15flaviudom96: There is
20:23:16Mat3Araq: I think you are right about these HTML specs, one can simply ignore them
20:23:16dom96It doesn't seem like it will make things much faster.
20:23:22flaviuoh
20:23:25dom96and getting it to even work is challenging.
20:23:32Jehan_Just because people deviate from it doesn't mean that a standard is worthless.
20:24:53Jehan_Without the standard, you'd get even more deviation.
20:25:28AraqJehan_: how so? a new browser need to render things close enough to the existing browsers to have any chance
20:25:59Jehan_Now imagine that there weren't a standard.
20:26:03Araqeffectively the existing implementations already provide a defacto standard
20:26:14Araqyes, I can imagine that
20:26:20Jehan_That worked out so well during the Netscape/IE wars. :)
20:26:29Araqthe situation would be exactly the same
20:26:50Jehan_I very much doubt that.
20:27:07flaviuAraq: You can't blame anyone when things don't work. For example, I can objectively say that IE is crap by pointing out all the standards it does not meet
20:27:09Mat3more dynamic, more choices, more freedom, more chaos -> better standards (sounds nice to me)
20:28:03flaviuJust a scapegoat, for the one person here who uses IE11 or something
20:31:59Mat3IE11 = Internet Explorer 11 ?
20:33:02AraqJehan_: FPC / Delphi, Wine / Windows API, Dosbox / dos, there are lots of examples where not having a standard didn't stop anything
20:33:16Matthias247dom96: i tried caas for the nimlime plugin. It seemed slower than starting the compiler again and again for hints. But most probably because it is buggy and compiles the source even when it's not required
20:33:42flaviuMat3: Yes, but I mean the latest, supposedly good, Internet Explorer version is.
20:34:28Jehan_Araq: These are really examples with a dominant vendor who can define a de facto standard via the power of the market.
20:35:42Araqyes, my point. compatibility is a political thing really, standards don't prevent any incompatibilities
20:37:00Araqjust write a brower that adheres to the html spec, but doesn't do it close enough to firefox/chrome and IE and see how far you'll get
20:37:09Jehan_No, but they lay out a common ground.
20:37:36Jehan_Araq: Depends on what your purpose is. The problem with the web in that regard is less the HTML and more that everything is a multi-megabyte webapp these days.
20:39:47Jehan_Oh, that first sentence may have been a bit oblique. Few people are going to write another web browser for the web. But writing something that processes HTML for internal documentation etc. is a different story.
20:40:30Araqwell you better be very permissive or you won't be able to parse much "HTML"
20:41:18Jehan_Araq: With internal tools, that is not a problem.
20:42:17Jehan_The web is really an absurdly chaotic environment.
20:43:13Araqwell I wrote web crawlers ... what do you think? that I looked at the HTML spec? ;-)
20:43:54Jehan_As a starting point, you probably did. Because reverse-engineering that from scratch would be a bit rough. :)
20:43:59Mat3Jehan_: this situation results because of HTML (even roff is a better and more flexible content-description language)
20:44:48Jehan_Mat3: If you want to tell me that XML/HTML is a disaster, you're preaching to the choir.
20:44:54AraqJehan_: nah, you simply "know" HTML and use that as a starting point
20:45:50flaviuIts amazing the effort the PyPy guys go through to optimize things.
20:46:21flaviuThey manually wrote the tokenizer using state machines instead of regexes
20:46:53Jehan_flaviu: I get not using regexes, but state machines?
20:47:29Araqstate machines for lexers are stupid, been there done that
20:47:32flaviuThey are effectively regexes, but are directly implemented in state machines
20:47:35flaviuhttps://bitbucket.org/pypy/pypy/src/b84ebcce7ebe37e222f75352b2e2c1274dc5f6e0/pypy/interpreter/pyparser/pytokenize.py?at=default
20:47:35Jehan_The Python syntax isn't that complex.
20:47:54Jehan_flaviu: Have to look at that, but that's what flex/lex do?
20:48:26EXetoCAraq: what have you not done?
20:48:29Jehan_Ah, they may need to do that bcause they're writing in RPython.
20:48:30EXetoCgone to the moon probably
20:48:55flaviuJehan_: Parser generator toolkit. You give it a grammar, and it gives you a parser
20:49:16Jehan_flaviu: Yeah, I know.
20:49:21Araqflaviu: # Automatically generated DFA's
20:50:02flaviuAraq: The comments lied!
20:50:06flaviu"The regular expressions have been replaced with hand built DFA's"
20:50:53Jehan_Maybe so that they consume less memory?
20:51:22Jehan_But unless have a complex syntax -- which Python doesn't -- it's generally easier and faster to write a lexer by hand.
20:51:59flaviuThey do do it by hand. Their regexes are just in a really weird format.
20:52:22Jehan_With "by hand" I mean code, not regexes.
20:53:05flaviuRegexes do make the code much more compact, although slightly less readable.
20:55:16Jehan_flaviu: That also depends on how you write regexes.
20:55:50Jehan_You don't have to do it Perl4 style as a string of characters mashed together.
20:56:09flaviuI guess so, although even extended regexps will likely be more compact
20:56:35flaviuAnd you don't really need extended regexps for the stuff in python, its very straightforward
20:57:01Jehan_flaviu: I'm talking about something like EBNF.
20:57:34AraqEXetoC: I never wrote a 3D engine ...
20:57:52flaviuI thought we're talking about tokenizers. EBNF is a bit overkill for that...
20:57:59Matthias247Araq: oh. I thought that's the one thing everybody does ;)
20:58:02Matthias247I did ;)
20:58:48Matthias247although it was far from a complete and usable engine
20:59:21Mat3by the way, does one of you know a way do bypass HTML for a web design ?
20:59:52Jehan_flaviu: EBNF is pretty simple?
21:00:36Jehan_Modern regular expressions are actually pretty complicated.
21:00:46Matthias247Mat3: do everything in flash - and everybody will hate you ;)
21:00:54Mat3t should be possible to render the whole design in Javascript and embed texts as plain UTF8 encoded texts I think
21:01:35AraqMat3: just adapt some premade free design
21:01:35Jehan_Mat3: You can do that. I'm not sure you WANT to do that, though.
21:01:38flaviuJehan_: It is, but regexps are much more concise, especially for simple things. And I know about modern regexps, insane..
21:01:45Matthias247Mat3: react.js does something like this
21:01:59Mat3thanks
21:02:01Jehan_flaviu: Conciseness at the expense of readability.
21:02:02flaviuJehan_: http://nikic.github.io/2012/06/15/The-true-power-of-regular-expressions.html :D
21:02:21Jehan_This is why there exist tools that break down and explain regular expressions for you.
21:03:36Jehan_For what it's worth, if you do write tools, I'm of the Terrence Parr school of thought.
21:03:41flaviuJehan_: Sorry, but `[0-9]+`is much simpler than `number = number | 0 | 1 | 2 | 3 | 4 |5 |6 |7 |8 |9 | ε`
21:04:29Jehan_I.e. that there's little point in writing a tool specifically for regular grammars (which generally are just a bit too weak), when you can have a tool for more general CFG grammars that gives you more power at pretty much the same speed.
21:04:46Jehan_flaviu: '0' .. '9'
21:04:55*noam_ joined #nimrod
21:05:38Jehan_Look at lpeg for an example that's fairly close to what I'm suggesting.
21:05:44Jehan_Minus the cryptic function names.
21:05:52*Trustable quit (Quit: Leaving)
21:06:07flaviuSure, but that's EEBNF :P.
21:06:33*noam quit (Ping timeout: 240 seconds)
21:07:18Jehan_flaviu: That's why I said "like EBNF", not "precisely EBNF"?
21:08:19flaviuSorry, didn't read that.
21:08:56flaviuJehan_: BTW, araq ported something like LPEG to nimrod
21:08:58Mat3Matthias247: There exist some excellent Forth environments for Javascript, so my plan is to simply program my web design online (or better I write a Postscript DSL)
21:08:58flaviuhttp://build.nimrod-lang.org/docs/pegs.html
21:09:05Jehan_flaviu: Yes, I know.
21:09:38Jehan_Such a library is of limited use for a statically typed language, unfortunately.
21:09:59Jehan_At least if you want to attach semantic information to grammar rules.
21:11:29*JehanII joined #nimrod
21:11:29*Jehan_ quit (Read error: Connection reset by peer)
21:11:58*JehanII is now known as Jehan_
21:18:06*shodan45 joined #nimrod
21:28:03*shodan45 quit (Quit: Konversation terminated!)
21:31:03*Johz quit (Ping timeout: 255 seconds)
21:39:02*io2 quit (Quit: ...take irc away, what are you? genius, billionaire, playboy, philanthropist)
21:39:52Mat3a union is defined with this statement - "type anUnion = object {.union.}" ... ?
21:40:06Araqyeah ...
21:44:06Mat3thanks
21:45:23Mat3exist there the possibility to declare object types so I can write "type anUnion = union" ?
21:45:48Araqyou can do
21:45:53Araqunion foo:
21:45:59Araq var a, b: int
21:46:14Araqwith a "union" macro but that doesn't improve anything
21:49:25Mat3hmm, right
21:50:18Mat3get some sleep, ciao
21:50:23*Mat3 quit (Quit: Verlassend)
21:53:08*io2 joined #nimrod
21:54:59*io2 quit (Client Quit)
21:57:05*Jesin joined #nimrod
22:39:49*saml_ joined #nimrod
22:46:22*enurlyx quit (Quit: Nettalk6 - www.ntalk.de)
22:56:48*Matthias247 quit (Read error: Connection reset by peer)
22:58:39*Demos joined #nimrod
23:00:27*Jehan_ quit (Quit: Leaving)
23:09:11*darkf joined #nimrod
23:41:40*flaviu quit (Quit: Leaving.)
23:41:49*flaviu joined #nimrod
23:42:46*flaviu quit (Remote host closed the connection)
23:43:50*flaviu joined #nimrod
23:51:20*flaviu quit (Remote host closed the connection)
23:51:39*flaviu joined #nimrod
23:56:38*wan joined #nimrod