01:09 (join) vu3rdd
01:09 (join) tommc
01:26 (quit) tommc: Ping timeout: 265 seconds
02:12 PascalHunger: so I was thinking about using racket for my web development but what is up with this http://hop.inria.fr/usr/local/lib/hop/2.1.1/weblets/home/articles/hop-perfs/article.html
02:12 (notice) rudybot: http://tinyurl.com/262z7y5
02:13 PascalHunger: that chart says its way slower then even PHP ?
02:53 (join) neilv
03:36 PascalHunger: anyone? hehe this is really getting to me
03:43 neilv: that article says something about the process growing to 1GB and being swapped out
03:44 neilv: you would have to look at how they wrote their test program
03:44 neilv: they might be doing something like storing thousands of continuations indefinitely
04:02 neilv: the plt web server is used for some real production systems
04:03 neilv: one of them is a class scheduling system for a university. reportedly it gets lots of traffic and has to be reliable
04:14 neilv: i have asked the author of that paper for the test code he used
04:39 (join) masm
04:39 masm: a
04:46 PascalHunger: neilv: I mean thats a craaaazy benchmark
04:46 PascalHunger: that made java interpreterd basic look faster then scheme implemented web server
04:49 neilv: i will be very interested to see the test program
04:50 PascalHunger: i am learning racket with the main goal of making websites with it
04:50 PascalHunger: hopefully tthat program was heavily flawed
04:53 neilv: i can make a guess what happened, but i'll wait to see the program
04:53 neilv: i wouldn't worry about using the plt web server
04:55 PascalHunger: neilv: you mean worry about in a good way or bad way?
04:56 PascalHunger: like go for it? heh
04:56 neilv: i mean that i suspect that the performance of the plt web server is good, for real applications that have been written properly
04:57 PascalHunger: neilv: yea from the looks of it that hop thing blows out apache :\
04:57 neilv: also, i have a backup plan of using scgi if i ever have a wildly popular web site and find a problem with the plt web server
04:58 neilv: http://www.neilvandyke.org/scgi-scheme/
04:58 PascalHunger: did you create this?
04:58 neilv: there is a popular protocol between web servers and apps, called scgi
04:58 neilv: i implemented it for plt
04:58 PascalHunger: yea i know about scgi hehe
04:59 PascalHunger: i use it in python every now n then
04:59 PascalHunger: neilv: i will use the racket but i saw the chart and was like -_-
05:00 PascalHunger: racket scheme is like 50x faster then PHP from most benchmarks I look at
05:00 PascalHunger: and the web server is not as complete as apache so i would heavily assume it would be faster then an apache+ php config
05:01 neilv: apache is pretty fast
05:01 neilv: php is slow
05:01 neilv: the speed of racket programs varies widely depending on who wrote them
05:02 PascalHunger: hmmm
05:02 PascalHunger: you can see why my head is spinning neilv hehe
05:02 neilv: all i mean is that, if you find something in racket that seems slow, it might be possible to make it fast :)
05:04 PascalHunger: yea
05:04 PascalHunger: its amazing how one chart can annoy you *that* much
05:22 neilv: aha. he's sent me the data already. i will look at this shortly
05:23 PascalHunger: neilv: fast responder heh
05:23 neilv: timestamps say 21 minutes after request
05:29 PascalHunger: yea, that chart still has me nervous as hell :)
05:59 PascalHunger: only reason i am worried because some of the things I might use this for are very high traffic
06:00 neilv: if you are making google or twitter or facebook or something, nothing will work for you out of the box
06:00 neilv: even moderately high traffic, you need to think about load-balanced servers and such
06:02 PascalHunger: neilv: looks like that article is from 2004, has much changed performance wise since then in racket?
06:02 neilv: the racket web server has changed a lot
06:03 neilv: it might have been rewritten entirely. it's certainly more comprehensive now
06:04 PascalHunger: I see
06:04 PascalHunger: well I will continue with my racket/ scheme learning
06:04 PascalHunger: seems like all the scheme work pretty much goes into racket
06:04 PascalHunger: racket seems the most popular
06:07 neilv: there is a lot of scheme work going on. it's used in a bunch of research and there are several practical platforms
06:08 neilv: but racket is the most popular platform
06:13 PascalHunger: why doesn't racket just provide a fcgi interface natively
06:13 PascalHunger: seems like a better idea then re implementing a web server?
06:17 neilv: the racket web server does some stuff that almost no one else does
06:17 neilv: continuation-based
06:18 neilv: and also, fastcgi is a heinous protocol
06:19 neilv: i have actually implemented it, but when i saw how heinous it is, and rumors of apache bugs, i implemented the stupid-easy scgi instead
06:19 neilv: http://docs.racket-lang.org/web-server/index.html
06:19 neilv: http://docs.racket-lang.org/web-server-internal/index.html
06:23 neilv: this tutorial gives one introduction, although you don't necessarily have to do things the exact way they did them: http://docs.racket-lang.org/continue/index.html
06:41 (join) mceier
06:49 PascalHunger: neilv: scgi is good too
06:50 PascalHunger: neilv: get the scgi more mature and so the rest of the racket libraries work with it good and it should be a better way?
06:52 neilv: the racket web server is probably the better way in general
06:52 neilv: scgi is a backup plan, and also i had to speed up a large legacy cgi app
06:52 PascalHunger: I will have to look more into it. I refuse to use PHP anymore :( I need to use something better
06:53 PascalHunger: But I am not a fan of a web server being written in a high level language for obvious reasons
06:53 PascalHunger: higher*
06:54 neilv: a lot of people use java-based web servers and servlet runners
06:55 PascalHunger: neilv: apache and tomcat?
06:55 neilv: you can use the plt web server fronted by apache, just like tomcat, if you want
06:56 PascalHunger: yea hmm
06:56 PascalHunger: thats one thing php did get right
06:56 PascalHunger: tight integration with the web server
06:57 PascalHunger: that tight integration allows for really good performance and scalability
06:58 neilv: depends, but i get your point
07:01 PascalHunger: hmm yea, i will just stick with racket, bigloo is sorta :\
07:01 PascalHunger: although looks like performance could be better
07:02 PascalHunger: the french inria are good at making many projects but at the same time they are always sorta "iffy"
07:02 PascalHunger: ocaml is nice but it could be way faster then it is
07:02 neilv: bigloo has been around for a long time too, and has had some neat stuff
07:02 neilv: i decided to go with plt/racket, and i haven't regretted it
07:23 PascalHunger: neilv: looks like thats what I am going with as well
07:23 PascalHunger: you ever use ocaml neilv ?
07:26 neilv: i looked at ocaml and was interested, but decided to go with racket
07:27 neilv: if i were going to a more functional language right now,i'd do haskell instead of ocaml
07:42 (join) hanDerPeder
07:52 (quit) hanDerPeder: Read error: Connection reset by peer
07:52 (join) hanDerPeder
08:18 (join) MayDaniel
08:26 (quit) lajla: Ping timeout: 245 seconds
08:36 (quit) MayDaniel: 
08:42 (quit) neilv: Quit: Leaving
09:11 (quit) hanDerPeder: Quit: hanDerPeder
09:18 (quit) mceier: Quit: leaving
09:19 (join) anachreon
09:19 anachreon: hiya
09:20 anachreon: got a couple of questions on racket if anyone's free to answer
09:21 anachreon: can it create a stadalone exceutable (all dependencies taken care of). i've read a little on raco, but don't know how this applied to OS X and linux
09:21 anachreon: and
09:22 anachreon: can the executables be used in a commercial environment (ie: sell the executable) if so, are there any requirements re the licensing?
09:24 anachreon: ?!
09:25 anachreon: hmmm, no answer?
09:26 anachreon: any pointers?
09:26 bremner: anachreon: in general, you have to wait for answers on IRC
09:26 anachreon: i get that, but not even a 'hi'
09:26 bremner: people are away from keyboard sometimes...
09:27 anachreon: must be my unlucky day. out of 30 on irc, one is not on the toilet.. :)
09:28 anachreon: mind you, being a 'chat' system you'd assume people are ready to 'chat' (as in next to the keyboard...)
09:28 anachreon: maybe that's alittle old fashioned
09:28 bremner: anachreon: that is what I tried to tell you. IRC doesn't work like that.
09:30 anachreon: no point me comparing it to another irc board then...
09:31 anachreon: maybe next time (look, i'm going away from irc and the keyboard)
09:31 (quit) anachreon: 
09:36 bremner: hmm. That went well.
09:41 ttuttle: what the hell is an "irc board"?
09:41 ttuttle rather dislikes it when people wander into the channels for free software projects, ask a question, and then get angry that nobody jumps to help them.
09:45 bremner: mostly frustration stemming from inexperience, I think
09:45 ttuttle: yeah.
09:46 ttuttle: and it's listed as "support" so they expect tech support reps to be standing by :P
09:58 PascalHunger: i just can't get over that racket web server
09:58 PascalHunger: i reallly feel performance is going to be a major issue
10:01 bremner: PascalHunger: Why is that?
10:04 (join) Lajla
10:08 PascalHunger: bremner: its a web server written in scheme
10:09 bremner: Are web servers CPU bound in general?
10:09 PascalHunger: http://hop.inria.fr/usr/local/lib/hop/2.1.1/weblets/home/articles/hop-perfs/article.html
10:09 PascalHunger: take a look at those charts bremner
10:09 (notice) rudybot: http://tinyurl.com/262z7y5
10:09 PascalHunger: well dont if you don't want to be disappointed
10:14 mario-goulart: PascalHunger: there's http://wiki.call-cc.org/egg/spiffy
10:15 bremner: So a more reasonable answer would be that the plt server did poorly in one set of benchmarks, rather than "because scheme"
10:15 mario-goulart: And http://wiki.call-cc.org/egg/awful, which runs on top of spiffy
10:17 mario-goulart thought this was #scheme
10:18 (join) b-man_
10:22 (join) MayDaniel
10:30 (quit) MayDaniel: 
10:31 PascalHunger: bremner: sure yea
10:31 PascalHunger: bremner: right now the plt web server is scheme to me :) because its the only way I can make web apps
10:32 (quit) offby1: Quit: Reblooting
10:32 PascalHunger: This is so dumb though, writing web servers in scheme
10:32 PascalHunger: its not for this
10:35 (join) offby1
10:39 (join) hanDerPeder
10:47 (quit) hanDerPeder: Ping timeout: 240 seconds
10:51 (quit) b-man_: Remote host closed the connection
10:51 (join) hanDerPeder
11:00 (quit) vu3rdd: Remote host closed the connection
11:03 bremner doesn't want to play "generalizations and clarifications" any more
11:06 (join) danbrown
11:26 (join) MayDaniel
11:28 (join) anRch
11:32 samth: PascalHunger, Racket is very much good for writing web applications and servers
11:33 samth: jay-mccarthy, you should look at that Hop link above, and see if the Racket server still has problems w/ that application
11:45 chandler: PascalHunger: Mind telling me what Scheme is for so I make sure I use it only in approved ways?
11:45 chandler: I wouldn't want to accidentally write a web server or do something else that Scheme isn't "for".
11:45 ttuttle: wait
11:45 ttuttle: I'm writing an IRC client
11:46 ttuttle: is that okay?
11:46 jay-mccarthy: figure 6?
11:47 chandler: Don't ask me. I haven't the foggiest clue. Obviously I've been Racketeering in all the wrong ways.
11:50 samth: jay-mccarthy, i think there are two figures, and in both PLT does very poorly
11:51 samth: the text says that it's b/c the memory use goes over 1 GB and starts swapping
11:51 jay-mccarthy: ya just making sure there isn't some code app
11:52 samth: all i know is from skimming the thing looking for a chart
12:07 (quit) hanDerPeder: Quit: hanDerPeder
12:10 (join) sstrickl
12:12 stamourv: samth: and isn't this paper from 2004? racket has changed a whole lot since then
12:12 stamourv: adding the jit comes to mind
12:17 (join) hanDerPeder
12:18 (quit) anRch: Quit: anRch
12:25 (join) anRch
12:30 eli: stamourv: it has cites at least at 2007, and in 2004 racket was very far from the reported 4.x.x version.
12:34 stamourv: eli: oops, had the wrong paper in mind
12:34 samth: eli, a note about TR1 - 'plt-tr1' isn't a great name for a bibtex entry
12:34 samth: since sometimes - doesn't work
12:36 Lajla: Oh wow bunch of scheme enthusiasts on their toes now.
12:37 Lajla: chandler, it's a reasonable position. Scheme, unlike Java or C# was never designed with the foremost goal to write applications in it.
12:37 Lajla: Maybe racket was.
12:37 Lajla: But it's still hindered in that by its scheme ancestry.
12:37 jay-mccarthy: samth: i can trivially get over 3k reqs per sec out of the server
12:38 (join) shofetim
12:38 jay-mccarthy: that puts it in the yaws/hop/php league of dynamic
12:38 bremner: Lajla: it is not a reasonable position from someone who admits the don't know scheme.
12:38 samth: Lajla, the aspects of Racket that it takes from Scheme are entirely a help in writing applications
12:38 jay-mccarthy: and a bit better than yaws in static
12:38 chandler: jay-mccarthy: Those numbers are dated; have you tested this against newer versions of those servers?
12:38 chandler: And on the same hardware?
12:39 jay-mccarthy: of course not, samth sent it to me 20 minutes ago
12:39 Lajla: bremner, that is also a reasonable position.
12:39 jay-mccarthy: also they don't report their httperf arguments and their text doesn't really describe it
12:39 jay-mccarthy: i have no idea if plt was that bad back then, but i doubt it
12:39 Lajla: samth, they are indeed, but in the end, to write applications thins like Java and C# might be a better base.
12:40 samth: Lajla, i suggest that in #racket, you try to restrict yourself to talking about things you know something about
12:40 Lajla: samth, you know what, you do a poll amongst all people who wrote applications and see if they favour java or racket.
12:41 Lajla: You can't deny that racket is quite 'perticular' and that for most people it doesn't work or flow natural.
12:41 stamourv: lajla: this will pretty much end up being the same result as asking: who knows java / racket
12:41 stamourv: you won't learn much
12:42 Lajla: stamourv, maybe, but you can wonder why Java has penetrated the industry and racket has not.
12:42 chandler: I would suggest not saying anything at all, honestly. You don't usefully answer questions about how to use Racket. You just ramble and annoy the other channel participants.
12:42 eli: samth: that's not my text there, but what's the problem with a dash?
12:43 Lajla: chandler, and it's still a very reasonable position that scheme was never designed to write applications in, what the original poster tried to make.
12:43 Lajla: And I answered quaestions from people who got here often enough.
12:43 samth: Lajla, answering questions is good
12:43 samth: suggesting that Racket and/or Scheme is a bad language for writing applications is less good
12:44 Lajla: Well, let me rephrase it in a more appropriate way then.
12:44 eli: Lajla: ... in particular when you're using a medium that is shared by more than a few people who have spent decades doing just that.
12:44 samth: originally, Scheme was designed as a language for understanding the semantics of Hewitt's actors in a single-threaded context
12:44 Lajla: At least, I saw it more as coming to the defence of the person who claimed that scheme was not made to write applications in who was met hostile for that opinion. It is a reasonable opinion.
12:45 samth: however, it's mostly used for other purposes now ;)
12:45 samth: and Racket was designed originally for writing a large GUI application
12:45 Lajla: Yes, but the original person whose position I defend is that it was originally not designed to write applications.
12:45 stamourv: yeah, but who cares?
12:46 eli: Lajla: Perhaps a more useful view is that the hostile responses he got (both here and on #scheme) were justified?
12:46 Lajla: Well, I'm just pointing out that being so hostile against that person doesn't strike me as fair, the position of that person was at least 'reasonable'.
12:46 Lajla: Well, as I conceded to bremner, the position is perhaps not that reasonable from someone who doesn't know scheme, as in 'right by accident'.
12:46 eli: Lajla: I guess that when some bigot says something about how women should be kept in the kitchen you're the kind of rebel that instead of joining the booing you help the guy?
12:47 Lajla: Well, I would point out that there are evolutionary arguments that indeed indicate that to a certain degree the human species has natural gender roles as any social species, yes.
12:47 eli sighs
12:48 samth: let's please keep the evo-psych out of #racket
12:48 Lajla: I'm just answering Eli's quaestion, I don't agree with it myself. I find gender roles myself an abhorant thing.
12:48 eli: Lajla: It's a good thing this thing is vitual, so you should just consider yourself slapped rather than feel it.
12:48 samth: Lajla, here's the deal:
12:49 samth: please keep this channel on topic: ie, discussion of how to use and develop Racket
12:49 samth: discussion of how you don't think it's good for thing X is *not* on topic
12:49 eli: In any case, since you made your point that you're siding with some view just because it was met with hostility, you have also proved to be an unproductive participant.
12:49 stamourv: samth: +1
12:49 chandler: Lajla: Rarely do I see you *usefully* answer questions, which is of course expected from someone with a shallow knowledge of the language. Usually I see you babbling about irrelevant nonsense such as this or about how much you love various different channel participants. It does not contribute to an environment which is focused on discussion of how to use or improve Racket.
12:57 (quit) anRch: Quit: anRch
13:04 (join) tommc
13:04 (join) jonrafkind
13:11 (quit) hanDerPeder: Quit: hanDerPeder
13:16 (quit) danbrown: Quit: danbrown
13:27 Fare: eli: I've got it basically working, but need to handle some corner cases.
13:31 Fare: eli: when @|{blah}| reads as ("blah") you rely on {} being a paren in Racket, right?
13:39 Fare: also, when you say @;{ foo } is nestable, what does it mean?
13:40 Fare: if you're reading forms, then a ;} is comment that doesn't end it
13:40 Fare: if you're reading text, then with normal @ escape?
13:54 Fare: I assume the latter
13:55 (quit) tommc: Ping timeout: 265 seconds
13:55 Fare: OK, only missing feature is the magic treatment of spaces around braces
14:05 (quit) Lajla: Read error: No route to host
14:05 (join) anRch
14:06 (join) Lajla
14:09 Fare: eli: OK, got it.
14:10 (quit) shofetim: Ping timeout: 245 seconds
14:10 Fare: I'm passing all tests from http://docs.racket-lang.org/scribble/reader.html except as doesn't apply to CL
14:17 (join) shofetim
14:18 eli: Fare: I'm not relying anywhere in the code on racket's {}s behaving like parens
14:19 eli: And the @;{...} contents is read in text mode, and the result is discarded -- so @;{... @foo ...} reads the `foo' as a sexpr, but then the whole thing is discarded as a comment.
14:20 eli: Finally, if you're done then it's time for my end of the deal, right?
14:20 Fare: yes
14:21 Fare: let me complete tests and push to a public repo before you start
14:36 (quit) Lajla: Read error: Connection reset by peer
14:36 (join) Nanakhiel
14:37 Fare: '@|{blah|@foo bleh}|
14:37 (nick) Nanakhiel -> Lajla
14:47 (join) mcarter_
15:00 (join) orukum
15:03 orukum: I'm new to racket is there a way to set a variable without using the define function?
15:04 jay-mccarthy: define binds a new identifier
15:04 jay-mccarthy: it does not 'set' a 'variable'
15:04 jay-mccarthy: set! does that
15:08 orukum: Thank you so much.
15:09 (join) danbrown
15:16 (quit) danbrown: Quit: danbrown
15:20 (quit) Fare: Quit: Leaving
15:30 (join) hanDerPeder
15:44 Lajla: orukum, it might be prudent to realize that set! only changes a binding and does not mutate data as what would happen in some languages.
15:44 Lajla: It makse the varialbe point to another thing if you like.
15:45 (quit) anRch: Quit: anRch
15:48 (quit) orukum: Quit: Page closed
15:50 shofetim: Lajla it is garbage collected though, yes?
15:51 Lajla: shofetim, well, depends if it completely comes into unreachable memory.
15:51 Lajla: shofetim, hold on, I'll make an example.
15:55 Lajla: shofetim: http://pastebin.com/gfX7jM3B
15:56 Lajla: I mean, a lot of languages work like this, but I've seen a lot of people working on the assumption that if you use set! inside a function, you update a value in memory outside of it. Though, if you use set! on x in that function, it will update the value outside of it.
15:56 Lajla: Basocally set! updates the 'nearest' variable of the same name it can see.
15:57 Lajla: To really update things in memory throughout the program you can use things like vector-set! or set-mcar!
16:02 shofetim: Not sure if I understand your example... set! inside a function can or cant update a global variable?
16:03 shofetim: I get (ie understand) lexical scoping and variables within a function hiding/shadowing variables with the same name outside of the function.
16:04 shofetim: but I thought set! could effect variables outside of the functions scope if they wheren't shadowed by a variable withing the functions scope?
16:04 shofetim: (Even though that would be a side effect and bad practice)
16:04 Lajla: shofetim, yes, they can.
16:05 Lajla: THey update the binding of the 'nearest' variable as I said.
16:05 Lajla: In this case I set! y.
16:05 Lajla: the nearest y is the function itself.
16:05 (join) danbrown
16:05 (quit) danbrown: Client Quit
16:08 Lajla: http://pastebin.com/caYRZhPE shofetim
16:08 Lajla: A lot of people have made code like this assuming that this would work like a C-like 'strcpy' function
16:08 (join) danbrown
16:09 Lajla: Not sure if it's wat you expect or not, it's just worth noting that it doesn't work like that.
16:09 Lajla: set! doesn't change some in-memory value, it only changes a binding.
16:10 Lajla: Effectively that function evaluates to #<void> in racket, and in scheme an 'unspecified value' and does nothing interesting.
16:11 shofetim: But when goes GC happen? if I (define a "contents") then (set! a "different contents") a is no pointing at a new value, and the old value now has no pointers to it and is GC at once yes?
16:11 jay-mccarthy: gc happens eventually with no guarantees
16:11 Lajla: shofetim, well, then the place in memory that held "contents" is now eligible for GC.
16:12 (quit) sstrickl: Quit: sstrickl
16:12 Lajla: But let's say you have (define a "contents") and then (set! b a) and then (set! b "different contents") you do not change a with that.
16:12 Lajla: It doesn't make a point to a string "different contents")
16:12 shofetim: ok... so I am still pretty new to scheme/racket. Since "gc ahppens eventually with no guarantees" does that meen that I don't need to care about it, or that we are likely to run out of memory?
16:14 shofetim: Lajla cool, then it behaves as I expect it would (which may be unexpected... but makes sense with what I know of the language)
16:14 jay-mccarthy: you don't need to worry about it
16:16 Lajla: shofetim, good, a lot of people who come from languages where you have to free memory yourself tend to assume you change both.
16:16 shofetim: If a program was mutating large chunks of data in memory, are there hints or anything that can be given to the GC to make it faster? or maybe just point me to rackets GC stuff... I don't know what how it handle it at all.
16:17 (join) clklein
16:19 shofetim: Just found it:
16:19 shofetim: http://docs.racket-lang.org/inside/im_memoryalloc.html?q=garbage%20collection
16:19 (notice) rudybot: http://tinyurl.com/2fyhkrn
16:20 shofetim: BTW is there a simple way to make self contained commandline executables that can be shared?
16:21 shofetim: They don't need to be small per say, just simple. I saw raco distribute but it makes a folder hiarchy...
16:22 Lajla: shofetim, what do you mean with a 'commandline executable'?
16:23 Lajla: You mean #!/usr/bin/racket or something like that?
16:23 shofetim: non-gui. in linux ./program-name
16:23 shofetim: I do little gui stuff, but regularly share scripts and small binaries.
16:23 shofetim: When I do them in PHP or python I can assume that the interpreter will be installed, but not so with racket : (
16:24 Lajla: Why not?
16:24 shofetim: so I was hoping to compile them, even if the where a little largish (since I would need to bundle the racket binary?)
16:24 Lajla: http://docs.racket-lang.org/guide/scripts.html
16:26 shofetim: Lajla python and php are installed on all our linux (and windows actually) boxes by default, but I doubt I can get racket installed on all of them to.
16:26 Lajla: shofetim, ahhh
16:26 Lajla: I see what you mean.
16:26 Lajla: Well, racket can compile
16:27 Lajla: http://docs.racket-lang.org/raco/exe.html
16:27 Lajla: http://docs.racket-lang.org/guide/intro.html 1.3 here might be more informative
16:30 shofetim: Lajla thanks : )
16:31 (quit) emma: Ping timeout: 260 seconds
16:45 (join) emma
16:53 (join) jao
16:54 (quit) danbrown: Ping timeout: 245 seconds
16:57 (part) shofetim: "ERC Version 5.3 (IRC client for Emacs)"
16:59 (join) rmitt
17:06 (quit) rmitt: Ping timeout: 265 seconds
17:11 (quit) mcarter_: Quit: Leaving
17:12 (quit) bremner: Ping timeout: 264 seconds
17:15 (join) bremner
17:32 (quit) Lajla: Read error: Connection reset by peer
17:32 (quit) MayDaniel: Read error: Connection reset by peer
17:33 (join) Lajla
18:27 (quit) Lajla: Ping timeout: 265 seconds
18:49 (join) rbarraud_
18:50 (quit) rbarraud: Read error: Connection reset by peer
18:50 (quit) rbarraud_: Read error: Connection reset by peer
19:08 PascalHunger: samth: did you see that benchmark i posted earlier
19:09 samth: the link to the paper from Hop?
19:10 PascalHunger: yes
19:10 samth: yes, i saw
19:10 samth: jay-mccarthy, (the maintainer of the Racket web server) ran the benchmark
19:11 samth: <jay-mccarthy> samth: i can trivially get over 3k reqs per sec out of the server
19:11 samth: <jay-mccarthy> that puts it in the yaws/hop/php league of dynamic
19:11 PascalHunger: samth: the one on hop?
19:11 PascalHunger: how did they get such numbers?
19:11 samth: so, the answer is, you shouldn't worry
19:11 samth: i don't know, but their paper is several years old
19:13 PascalHunger: i see
19:17 PascalHunger: samth: thats good to know, hopefully my boss wont freak out to much but when he found that he did not like the idea at all
19:18 samth: PascalHunger, your boss googles for old benchmarks of the software you use?
19:36 (join) adadglgmut
19:40 PascalHunger: samth: he just saw that and wanted me to verify that its not that way / won't be an issue
19:41 samth: but why did he see that?
19:41 samth: was he looking for performance benchmarks of the Racket web server?
19:41 samth: and how did he find that one?
19:43 PascalHunger: no idea samth
19:43 PascalHunger: Guessing he was looking for benchmarks
19:43 PascalHunger: many sites we do have like 300k req/s
19:43 PascalHunger: not all of those are humans but still request
19:47 (quit) masm: Quit: Leaving.
19:50 (join) kingless
20:00 (quit) kingless: Remote host closed the connection
20:19 (quit) hanDerPeder: Quit: hanDerPeder
20:34 (quit) jonrafkind: Ping timeout: 264 seconds
20:49 PascalHunger: samth: so jay-mccarthy is the author of the web server itself?
20:51 (join) speps
20:56 (join) Fare
20:58 speps: hi guys, trying porting a program that was for drscheme to racket and now when i try to start i get default-load-handler: cannot open input file: "/usr/lib/racket/collects/scheme/pretty.ss" (No such file or directory; errno=2)
20:58 speps: why does the app is searching for .ss instead of .rkt files?
20:59 speps: How can i make the app search for .rkt?
20:59 speps: Thanks
21:03 speps: the executed scm file halts on (require scheme/pretty)
22:41 (join) grasshops
22:42 grasshops: hi guys, im trying to learn racket and im a bit confused about the error function
22:44 grasshops: if i want to test my function and if im supposed to get error "undefined" how do i express that with check-expect?
23:05 (quit) martinhex: Remote host closed the connection
23:05 (join) martinhex
23:06 (quit) clklein: Ping timeout: 250 seconds
23:07 (quit) tonyg: Ping timeout: 276 seconds
23:07 (join) clklein
23:07 (join) tonyg
23:12 PascalHunger: how much has plt scheme / racket changed in the last few years?
23:13 PascalHunger: If i am reading  book from 04 is it still worth while?
23:34 PascalHunger: anyone know any good books on scheme which would apply to racket?
23:36 Fare: PascalHunger, htdp?
23:37 PascalHunger: htdp?
23:38 Fare: .org
23:39 PascalHunger: thats from 03
23:39 PascalHunger: is that still valid for racket as of today?
23:43 (join) jonrafkind
23:45 Fare: 2nd edition is recent
23:46 PascalHunger: is this on the whole plt scheme stuff?
23:46 Fare: no
23:46 Fare: there is no comprehensive doc on the whole plt scheme stuff, because it's big and moves fast
23:46 PascalHunger: I see
23:46 Fare: but there are plenty of docs on plenty of aspects of it
23:46 PascalHunger: so where can I fully learn it? hehe
23:47 Fare: docs.racket-lang.org
23:47 Fare: then checkout the sources
23:47 Fare: read, and explore
23:47 Fare: as with any language, really
23:49 PascalHunger: yea .....mahy langs I start with have a nice book etc etc but scheme seems more academic in that way
23:52 Fare: there are PLENTY of books about scheme.
23:52 Fare: but scheme is a small part of racket.
23:53 PascalHunger: Fare: yea isn't scheme the main language for it?
23:53 Fare: kind of
23:53 Fare: but what distinguishes schemes is their extensions, libraries, etc.
23:54 PascalHunger: I can see how scheme really scares new comers
23:56 Fare: how so?