#clojure – 2011-08-15

2011-08-15

00:03	leo2007	is there something like macrolet for clojure?
00:03	amalloy	leo2007: in c.c.macro-utils, which i think is now org.clojure/tools.macro or something
00:03	leo2007	thanks
00:09	fmeyer	exit
00:32	Drakeson	is it possible to `require' relative to the current namespace, similar to python's "from .submodule import Things" ?
00:32	amalloy	no
00:34	Drakeson	but would such a feature make sense?
00:40	Drakeson	I mean, would it be possible to modify #'clojure.core/load-lib to support this feature?
00:42	amalloy	it's certainly possible. bet you ten dollars a patch wouldn't be accepted, though
01:03	jsoft	In java its got what looks like some kind of typcast, ie NameClasspair foo = (NameClassPair) blah.next(). How do I do the typecast in clojure?
01:04	amalloy	you don't have to
01:05	jsoft	Hmm.
01:20	jsoft	:(
01:20	jsoft	Argh
01:20	jsoft	Why does my brain not get along with java.
01:35	chewbranca	anyone successfully following a CouchDB _changes feed with the clutch clojar?
01:39	chewbranca	I'm running into the same issue as here: https://groups.google.com/forum/#!topic/clojure-clutch/vqkzG1aqpUw
01:58	Pupeno	How do you add a second library, both with :as, to the require part of ns?
02:00	Pupeno	found it.
03:02	dsantiago	How do you set a property in leiningen in a file so that it's available with System/getProperty?
04:34	lnostdal	maybe i should just compile emacs 24 from source myself .. no one else is seeing leaks using the build from http://emacs.naquadah.org/ ? (natty)
07:14	joshc_	Does anyone know why when I'm using defn and defining a docstring, the doc command works but there does appear to be any doc metadata on the function?
07:22	morphling	joshc_: try (meta #'your-function) instead of (meta your-function)
07:24	joshc_	That worked, thanks. I'm trying to use Marginalia, but it seems to be unable to detect any of my docstrings.
07:24	morphling	something with meta data on the function and not the var, I think, but I'm not sure, better ask again when someone more knowlegable is around
07:26	joshc_	Thanks, will do
07:28	algernon	marginalia has some troubles detecting docstrings at times. from what I've gathered so far, it might have to do something with clojure 1.3 (but then, I've been using clojure for like two and a half weeks so far, on and off :P)
07:33	joshc_	I'm running 1.2.1 and for some reason when I run marginalia against itself, it works no problem, but will only detect regular comments when run against my code
07:41	manutter	I would have thought Marginalia would extract the docstrings directly from the source rather than from function meta
07:41	manutter	Are your docstrings formatted the same way as Marginalia's as far as line breaks in the source code, etc?
07:43	kencausey	http://groups.google.com/group/marginalia-clj/browse_thread/thread/23a424ed9fbedb4a
07:43	kencausey	related?
07:43	algernon	(and my issue is related to the above link, I filed issue #55)
07:45	joshc_	I tried all three styles of line breaks, and tried copying in a function from the marginalia source, still didn't work
07:46	algernon	it's probably not the formatting of the function itself, that breaks marginalia.
07:47	algernon	In my case at least, it wasn't. I trimmed down my (ns) and marginalia began to work.
07:48	joshc_	algernon I think we have the same problem (issue 55) I have a bunch of requires in my code
07:49	algernon	joshc_: try removing them and see if that unbreaks marginalia. If nothing else, we'll know that it's the same underlying issue then
07:49	joshc_	Will do
07:52	joshc_	algernon: I cleaned out my ns and everything worked
08:05	joshc_	algernon: I added a comment to issue #55
09:12	PPPaul	test
09:32	TimMc	PPPaul: Your test worked.
09:32	PPPaul	awesome!
09:34	TimMc	"Thank you for experimenting with Wikipedia. Your test worked, and it has been reverted or removed. Please use the sandbox for any other tests you may want to do. Take a look at the welcome page to learn more about contributing to our encyclopedia."
09:34	TimMc	^ what I think of whenever someone writes "test"
10:22	kephale	namespace question, i am refactoring what was a single file into separate files, but i would like 'use to bind symbols from the separate file to the original name (for backwards compatibility), how does one do this?
10:24	manutter	do you actually need/want separate namespaces? If you're just breaking up the files for easier maintenance, you can use (in-ns) to load the symbols directly into their old namespace
10:25	zmaril	If I were using noir and mongo, how should I set up the database? I have very little prior experience with backend and the MVC model is new to me
10:25	manutter	if you want separate namespaces though, I don't see why you can't use/refer them from the original file
10:25	kephale	well, i kind of do want the namespaces separated, in-ns is the best alternative
10:25	manutter	zmaril: what do you mean by set up the db?
10:26	manutter	kephale: (use) should be ok as long as the separated-out file isn't using or requiring the original file
10:26	zmaril	manutter: Like I see how to create a new db with mongo, but I am unsure of where to create it and how to pass it around in a functional way.
10:26	manutter	or is that what the problem is
10:27	kephale	well, use isn't binding the symbols from the new smaller NSs the way i am looking for, use case:
10:27	manutter	zmaril: ok, that's not too hard. Create a namespace for your db, and make a create-db function that does all the initial setup and returns some kind of db object that you can pass around
10:28	kephale	(ns foo) had (defn botmaker [] …) now botmaker was moved to (ns foo.bots), now calling (use 'foo) does not provide botmaker
10:28	zmaril	Let me give that a shot then
10:28	kephale	given (ns foo (:use [foo.bots]))
10:28	manutter	zmaril: you can probably find some tutorial examples if you google around
10:29	zmaril	manutter: most of them are single page things about how to get up and running on herkou
10:29	manutter	kephale: wonder if you would have better luck calling an explicit (refer 'foo.bots.botmaker)?
10:30	kephale	manutter: well thats what i was thinking, but some of the NS's have a looot of things defined
10:30	raek	kephale: you are pretty much stuck with either using (load-file) + (in-ns) or releasing a new major version...
10:30	manutter	kephale: I'm kinda out of my depth here, so I'm groping around for ideas
10:30	kephale	i was thinking about a hack that looked for all symbols in a namespace and aliased them
10:31	manutter	kephale: that's kind of the way I was leaning (in my blissful ignorance)
10:31	kephale	manutter & raek: ah, ok, i might putter at that aliasing hack
10:31	raek	kephale: there is a macro called 'immigrate' in contrib, but it doesn't really work that well
10:31	kephale	raek: oh cool!
10:32	kephale	it just has to work well enough : P
10:32	raek	if backwards compability is essential, using load-file to "include" additional files into the namespace is probably the best idea
10:32	raek	binding won't work reliably, since you have to vars for the same value (only one of them will be rebound)
10:33	kephale	well, that makes it a bit harder to become forwards compatible, since all future code should really switch to use/require on the new namespaces
10:33	manutter	zmaril: the one-page examples are probably not a bad start, but if your app gets more complicated then separating out the db stuff will be easier to maintain
10:33	kephale	with the opening 'ns call and all
10:33	manutter	zmaril: the basic approach should remain the same though
10:34	raek	well, new major version is what I would recommend
10:34	kephale	raek: heh, i'd love to… unfortunately its not my call
10:35	kephale	but immigrate (which seems to be require + a loop of interning) does look like it might hold me over, i'll poke around
10:35	raek	compojure used immigrate initially, but stopped later, IIRC
10:36	raek	load-file + in-ns will allow you to structure the implementation better, but doesn't expose the "new API"
10:38	kephale	ooo immigrate seems to be working like a charm thus far
10:38	duck1123_	There's also potemkin, but that works on a var by var basis
10:39	kephale	well, i'm hoping i can just leave this "depricated" file that used to have everything, and just use it for the immigrate call, then anything new can be written with the new API
10:39	duck1123_	but provided you pass it a var, it works with 1.3
10:40	kephale	potemkin looks like it wraps intern?
10:40	leo2007	How to configure lein to run the latest snapshot of clojure?
10:41	raek	heh, then any problems that might be caused by 'immigrate' will only affect the users of the old API...
10:41	duck1123_	leo2007: just edit your project.clj and ref the vewer version
10:41	raek	an incentive to upgrade
10:41	leo2007	duck1123_: thanks.
10:42	kephale	raek: yeah, perfect : P
10:43	leo2007	duck1123_: where is clojure installed by lein?
10:43	duck1123_	It's in your mvn repo ~/.m2
10:44	manutter	leo2007: or, depending what you mean by "installed," in your project's lib/ directory
10:44	duck1123_	lein just uses maven's dependency resolution under the hood, which makes it handy
10:45	duck1123_	manutter: Good point, as a Maven user, I tend to forget about lib/
10:45	raek	leo2007: they are downloaded to the local maven repo in ~/.m2 . the ones you use in your particular project are copied over to the lib/ directory. (in the future, they might be symlinks instead)
10:46	raek	so you have one environment per project. (but the jars are cached in your local repo)
10:46	leo2007	raek: thanks.
10:46	mattmitchell	what's the best way to select non-empty values from a map... kinda like select-keys except selecting only non-blanks ?
10:47	raek	mattmitchell: one way is to use the (into (empty m) (processing-function m)) pattern, where processing-function might be filter, for instance
10:48	leo2007	what's the version for clojure upstream from git repo?
10:48	raek	the filter predicate would receive one argument which is a vector (a map entry, actually) of the form [key value]
10:48	leo2007	I need it so that lein can pull it in
10:48	duck1123_	(into {} (filter (fn [[k v]] (identity v)) {:m 1 :n nil}))
10:48	mattmitchell	raek: cool let me try that out
10:50	raek	duck1123_: you could replace (identity v) with just v there... >/(
10:50	raek	*:-)
10:50	mattmitchell	very cool, thanks
10:50	duck1123_	good point.
10:50	raek	gah. accidentally switched keyboard layout in the middle of the smiley...
10:51	duck1123_	I guess that comment about "if you type (identity you don't know what you're doing" applies
10:51	raek	,(into {} (filter val {:m 1 :n nil}))
10:51	clojurebot	#<ExecutionException java.util.concurrent.ExecutionException: java.lang.IllegalStateException: replace already refers to: #'clojure.string/replace in namespace: sandbox>
10:51	raek	,filter
10:51	clojurebot	#<ExecutionException java.util.concurrent.ExecutionException: java.lang.IllegalStateException: replace already refers to: #'clojure.string/replace in namespace: sandbox>
10:51	duck1123_	clojurebot is still messed up
10:52	manutter	oh, sadness
10:52	raek	,((resolve (symbol "ns-unmap")) ns 'resolve)
10:52	clojurebot	#<ExecutionException java.util.concurrent.ExecutionException: java.lang.IllegalStateException: replace already refers to: #'clojure.string/replace in namespace: sandbox>
10:52	manutter	~bot-bandaid
10:52	clojurebot	Huh?
10:52	manutter	well, I tried.
10:53	Hodapp	:(
10:53	raek	anyway... (into {} (filter val {:m 1 :n nil})) => {:m 1}
10:56	raek	leo2007: the latest beta version of clojure is [org.clojure/clojure "1.3.0-beta1"]
10:57	raek	I think the nightly build version is "1.3.0-master-SNAPSHOT"
11:03	PPPaul	is there a simple way to to see if 2 arrays share a certain value somewhere in them?
11:04	PPPaul	arrays = seq
11:04	duck1123_	first question, is there anyway you can represent them as sets?
11:06	lucian	PPPaul: i think you could make them sets and intersect them, as duck1123 suggests
11:06	duck1123_	if you know what the value is, you could test that they each contain the value
11:07	PPPaul	actually, the values i'm looking for are in a seq too
11:07	PPPaul	i guess i want to test if 3 seqs have the same values
11:08	duck1123_	Then I would definitely looking into making them sets
11:08	PPPaul	ok
11:08	PPPaul	i haven't played with sets yet]
11:09	duck1123_	Lots of fun stuff http://clojuredocs.org/clojure_core/clojure.set
11:11	PPPaul	:D
11:13	raek	(let [a (set [1 3 2]), b #{1 2 3}, c #{3 2 1}] (= a b c)) => true
11:13	manutter	,(println "test")
11:13	clojurebot	#<ExecutionException java.util.concurrent.ExecutionException: java.lang.IllegalStateException: replace already refers to: #'clojure.string/replace in namespace: sandbox>
11:14	manutter	##(println "test")
11:14	lazybot	⇒ test nil
11:15	PPPaul	thanks a lot!
11:33	timvisher	hey all
11:34	timvisher	i'm looking for reading i can do on functional code smells.
11:34	timvisher	anyone have any suggestions?
11:34	wastrel	what does it mean functional code smells
11:35	timvisher	code smells in the sense of Fowler's Refactoring
11:35	timvisher	hopefully that wasn't a reference to Silence of the Lambs. :)
11:36	timvisher	wastrel: are you familiar with Fowler's book?
11:36	wastrel	no
11:37	timvisher	ah
11:38	timvisher	well, code smells are just things in code (structure, method length, variable names, coupling, etc.) that, when done in a certain way, should 'look ugly' or 'smell bad' to you when you see them and encourage you to refactor
11:38	timvisher	a simple example that is cross-cutting accross all languages and idioms would be methods or functions that do many things. You probably want every function to do only one thing, do it well, and be tested for that
11:39	timvisher	so that when you compose it later on, you can easily understand what you're doing
11:39	timvisher	however, a lot of the smells in fowler's book are more geered toward OO, i was wondering if anything like that had been done for functional
11:39	timvisher	ejackson: sorry to hear it! ;)
11:40	ejackson	i regard it as I do cheese
11:40	timvisher	hah
12:01	dnolen	damn
12:02	dnolen	,(subvec [1 2 3] 1)
12:02	clojurebot	#<ExecutionException java.util.concurrent.ExecutionException: java.lang.IllegalStateException: replace already refers to: #'clojure.string/replace in namespace: sandbox>
12:02	dnolen	&(subvec [1 2 3] 1)
12:02	lazybot	⇒ [2 3]
12:02	dnolen	no more rest on vectors for me.
12:03	chouser	oh? why's that?
12:03	dnolen	subvec 1 does the same thing while preserving the vector type (yes, i was being a little over the top :)
12:04	chouser	yeah, subvec is really nice.
12:05	dnolen	ugh, http://blog.dhananjaynene.com/2011/08/cperformance-comparison-languages-styles-and-vms-java-scala-python-erlang-clojure-ruby-groovy-javascript/
12:06	dnolen	people who don't know the language they're benchmark should not benchmark
12:08	upwardindex	Does clojure run on openjdk?
12:08	st3fan	dnolen: looks like they forgot to compile the clojure code
12:10	dnolen	st3fan: no those numbers are accurate for that kind of code.
12:10	dnolen	calling count on range is LINEAR
12:10	upwardindex	dnolen: I think the key part of the article is "I can’t write the fastest possible code across all this languages."
12:10	chouser	ugh. wasn't always.
12:10	st3fan	dnolen: oh oops. i did not look at the code.
12:10	dnolen	upwardindex: "I don't know Clojure so I'm going to write some crap"
12:11	st3fan	well, correct his code and ask to run the bench again :)
12:11	upwardindex	dnolen: I found this to be generally more accurate: http://shootout.alioth.debian.org/
12:12	dnolen	upwardindex: you don' t need to tell me, I consider myself a minor Clojure perf expert :)
12:39	lucian	upwardindex: except it's extremely biased
12:39	lucian	and mostly useless
12:40	upwardindex	lucian: biased towards what?
12:40	lucian	upwardindex: towards numeric processing code
12:40	lucian	it has extremely small synthetic benchmarks
12:41	dnolen	lucian: numeric processing is good indicator of the performance of much higher level benchmarks
12:41	lucian	it's very hard to benchmark worth-while things between languages (like a template language compiler)
12:41	dnolen	you can't implement Clojure in Clojure without very, very good numeric perf
12:41	lucian	dnolen: sure you can
12:41	upwardindex	lucian: true, however i found it matches my experience with the 2-3 languages im most proficient with
12:41	lucian	dnolen: i think it's not a good indicator at all, although it may be useful
12:41	dnolen	lucian: you can and lose performance, Clojure is performance oriented.
12:42	lucian	dnolen: my point is that the benchmark tests something almost entirely irrelevant to what mosts of us actually do
12:42	dnolen	lucian: because you benefit from it.
12:42	dnolen	lucian: but someone needs to do the work that you benefit from
12:43	lucian	dnolen: really, read the code on those benchmarks for several languages. you'll change your opinion
12:43	dnolen	lucian: I have, and I do stuff like that all the time.
12:57	timvisher	anyone know if anything's been written about common lisp/functional anti-patterns and smells?
12:57	timvisher	something along the lines of the first few chapters of Fowler's Refactoring?
12:58	ejackson	timvisher: the next Clojure book, authored by you.... ?
12:58	timvisher	lol
12:58	timvisher	a bunch of people just told me the same thing at lunch
13:02	timvisher	i'm actually thinking i'm just going to have to work my way through MM. fogus and houser's book again
13:05	ejackson	yeah, i guess you're looking for (map not (extract-practices joy-of-clojure))
13:07	timvisher	ejackson: basically. :)
13:08	timvisher	in a completely unrelated turn of events, anyone watch any of my VimGolf in Emacs videos?
13:14	jfields	wjat
13:15	jfields	sorry...
13:16	jfields	what's the easiest way to convert {"bob" {"accounts" {"NGM1" {"lynx-available" true, "account" "acct"}}, "default-account" "xyz"}, "joe" {"accounts" {"NGM1" {"lynx-available" true, "account" "acct"}}, "default-account" "asd"}} to {"bob" {:accounts {"NGM1" {:lynx-available true, :account "acct"}}, :default-account "xyz"}, "joe" {:accounts {"NGM1" {:lynx-available true, :account "acct"}}, :default-account "asd"}}
13:16	jfields	note the 2nd and 4th level keys are keywords in the version I'm looking for.
13:26	TimMc	jfields: You want all the non-top keys to go string -> keyword?
13:26	TimMc	some kind of tree-walk inside an (into {} (map ...))
13:26	jfields	no, unfortunately, only the keys at depth 1 and 3.
13:27	jfields	yeah... I'm looking through the source of clojure.walk now... thanks. =)
13:27	lobotomy	hmm, would get-in work?
13:49	jfields	my solution, feedback welcome: http://www.pastie.org/2376311
13:53	amalloy	i'd call it depths instead of indexes
13:58	amalloy	and probably pull the duplication of (hv v) outside of the (if) so you don't have to repeat it. neat little corecursive solution, though
13:59	amalloy	duck1123_: if it makes you feel better, clojure.core has (identity in it too
13:59	amalloy	&(macroexpand '(.getClass Class))
13:59	lazybot	⇒ (. (clojure.core/identity Class) getClass)
14:14	jfields	amalloy, thanks for the feedback.
14:50	gtrak`	is there a way to do some custom global lein config that overrides what's in a project.clj? for instance, I want to ignore a certain dev-dependency (so I don't have to ask everyone else to change it)
14:52	gtrak`	I see there's a ~/.lein/init.clj file I can stuff code in
14:53	ejackson	gtrak: i know that can be used to store your binary repository credentials and lein will fetch them from there
14:54	gtrak`	I'd like to globally override a version of a dev-dependency just for my box... maybe there's some api call to hook into?
14:55	amalloy	interesting. is there any clear reason that (for [x (range 10) :when (even? x)]) is illegal? specifically, it requires a body instead of allowing an implicit nil like (fn) does
14:56	amalloy	i found myself wanting to use for just to count a sequence, and nil would have been fine
15:01	TimMc	Wait, what are you trying to do?
15:02	raek	gtrak`: maybe you could install a new custom version of the dependency locally
15:03	raek	or install it as a plugin
15:03	amalloy	TimMc: me or gtrak?
15:03	raek	which dependency is it?
15:03	gtrak`	lein swank
15:03	TimMc	amalloy: You. You want (nil...) ?
15:03	amalloy	TimMc: (count (for [x some-seq, y some-seq, :when (some-condition x y)] ...))
15:03	TimMc	so you can (count..) it?
15:03	TimMc	ah
15:04	amalloy	i don't care about the elements, i just want to see how many pass a condition
15:04	amalloy	and was a bit annoyed to have to put in a body, is all
15:04	raek	don't know a way to add the :excludes option globally
15:04	gtrak`	everyone else uses lein 1.5, I use 1.6, which is broken on swank-clojure 1.2, I'd like to workaround it automatically for myself instead of fiddling with project dependencies or asking them to change
15:04	ibdknox	amalloy: why not filter?
15:05	amalloy	ibdknox: because (for) does the nice nested-iteration of xs,ys
15:05	ibdknox	amalloy: gotcha
15:07	raek	gtrak`: so you want something like :excludes but global?
15:08	oich	I am new to clojure. On http://webnoir.org the getting started and tutorial pages show running the web server and then show some lisp code, but don't say where (what file? what path?) to put the code. do you know where?
15:08	gtrak`	yes, generically, it'd be cool to intercept the map created in a project.clj that lein uses and munge it up a bit
15:08	ibdknox	we need to just convince everyone to remove lein-swank from their project dependencies
15:08	ibdknox	and use the wonderful lein plugin install to do it locally for themselves
15:09	amalloy	(inc ibdknox)
15:09	lazybot	⟹ 1
15:09	amalloy	or the global cake project, which is what i do
15:09	ibdknox	or that :)
15:09	gtrak`	raek, I want to write a bit of code that says, if dev-dependencies has swank-clojure, then override its version..
15:09	ibdknox	gtrak: the simple solution would be to remove the jar from lib
15:10	amalloy	gtrak`: why are you still using swank-clojure 1.2?
15:10	amalloy	make everyone upgrade, problem solved?
15:11	gtrak`	sure
15:11	gtrak`	still surprised I can't do it with some code
15:14	raek	gtrak`: if your project depends on for example intancter, you can exclude its swank-clojure dependency: [incanter "1.2.3" :exclusions [swank-clojure]]
15:15	raek	you have to do this for every library that pulls in the bad dependency, though
15:15	gtrak`	right, that's not really what I'm talking about though, would like to override it globally without touching the project.clj
15:16	ibdknox	We should log issues against all the libs that have swank-clojure as a dep
15:16	gtrak`	or having to remember to run some command each time, etc..
15:16	ibdknox	just to start removing this problem entirely
15:18	upwardindex	I'm trying to do "lein deps" with [ring/ring‑core "0.3.9"] in my project and I errors POM for 'ring:ring‑core:pom:0.3.9:compile' is invalid. Any idea what I'm doing wrong?
15:19	gtrak`	I think I would have to mess with the defproject macro to do it
15:20	upwardindex	http://clojars.org/repo/ring/ring-core/0.3.9/ seems to exist and everything
15:20	gtrak`	ah, this should be simple, I think I can provide a function in a var that filters the project, any chance they would pull in something like that?
15:31	sleepynate	upwardindex: is there a conflicting dependency, by chance? i know frequently i've had to move back from 1.2.1 to 1.2.0 to satisfy a lot of deps
15:31	upwardindex	sleepynate: That might be it, i'll try reverting to 1.2.0
15:32	sleepynate	upwardindex: especially contrib. the website says contrib is still 1.2 :)
15:32	upwardindex	sleepynate: Unfortunately, no difference on 1.2.0
15:34	upwardindex	sleepynate: I'm doing a lein search ring to see if it can find it
15:37	sleepynate	upwardindex: the version on clojars is 0.3.11
15:39	upwardindex	sleepynate: Same thing with 0.3.11, something very basic must be broken with my setup
15:39	upwardindex	Is there a way to test a clojure/leiningen installation?
15:40	upwardindex	To make sure that everything is working as it should?
15:41	upwardindex	Ah one thing I wasn't sure about: Is OpenJDK fine or should I use another VM?
15:41	sleepynate	i use openjdk generally without issue
15:42	sleepynate	the easy way to test is to create a new blank project with "lein new derp", then add contrib as a dependency, and lein deps
15:43	sleepynate	if those download and compile properly, clojure itself and leiningen should be functioning properly
15:43	upwardindex	sleepynate: exactly what I did except instead of adding contrib I added ring/ring-core
15:43	MasseR	upwardindex: Otherwise openjdk seems to work, but clojurescript is unusable with it (and while it is so, I refrain from clojurescript)
15:43	upwardindex	ill try with contrib
15:44	sleepynate	oh yea. i should add that i've never touched clojurescript yet
15:44	sleepynate	so if that's what you're shooting for i have no idea
15:45	sleepynate	https://github.com/weavejester <-- this fellow is in charge of packaging ring-core among a number of others. perhaps he knows more (assuming your standard setup is working)
15:46	upwardindex	sleepynate: Yay, clojure-contrib works!
15:49	xandrews	is there a way to list an agent's watch functions?
15:49	sleepynate	fwiw, ring/ring-core builds find for me against org.clojure/clojure 1.2.1
15:50	sleepynate	java version "1.6.0_22"
15:50	sleepynate	OpenJDK Runtime Environment (IcedTea6 1.10.2) (6b22-1.10.2-0ubuntu1~11.04.1)
15:51	sleepynate	0.3.9 and 0.3.11
15:51	raek	upwardindex: maybe you could try deleting everything in ~/.m2
15:51	upwardindex	raek: I'll try that right away
15:52	sleepynate	yea. that will force maven to fetch everything anew
15:56	upwardindex	raek: same behavior, let me paste the output I'm not too familiar with maven maybe something will catch your eye
15:58	upwardindex	Here is the output i get http://pastebin.com/XbyRhHWX
15:59	upwardindex	And that is on: Leiningen 1.6.1 on Java 1.6.0_22 OpenJDK Client VM
15:59	sleepynate	yea this totally looks like a dependency thing
16:00	sleepynate	you only have org.clojure/clojure and ring/ring-core ?
16:00	sleepynate	in project.clj?
16:00	upwardindex	sleepynate: no I have clojure-contrib also
16:00	upwardindex	sleepynate: that I added earlier to test
16:01	upwardindex	sleepynate: should I paste my project.clj?
16:01	sleepynate	can't hurt
16:03	upwardindex	http://pastebin.com/NVRD5R89
16:04	sleepynate	that builds for me
16:05	upwardindex	something must be wrong with my environment
16:06	sleepynate	rm project/libs/*.jar
16:06	sleepynate	then rm ~/.m2/repository/*
16:06	sleepynate	then lein deps
16:06	sleepynate	if that doesn't work, no idea :)
16:07	sleepynate	(though i'm by no means an expert)
16:08	upwardindex	ah there is no "libs" folder in my project folder, could that be the problem?
16:08	sleepynate	lib, sorry
16:09	amalloy	&(doc when-first) is a neat function that i always forget exists. just mentioning it here as a sort of public service announcement
16:09	lazybot	⇒ "Macro ([bindings & body]); bindings => x xs Same as (when (seq xs) (let [x (first xs)] body))"
16:09	upwardindex	sleepynate: same no lib folder, do i have to create it?
16:09	sleepynate	upwardindex: shouldn't
16:10	sleepynate	leiningen should create it
16:10	sleepynate	i assume you're on linux?
16:11	amalloy	ah, irc, one of the few places where "i assume you're on linux" isn't some sort of weird joke
16:11	upwardindex	sleepynate: yes, kubuntu
16:12	upwardindex	never thought going through the clojure tutorials would be so hard
16:12	amalloy	upwardindex: have you tried 4clojure.com? it'll let you get started learning without having to install stuff
16:13	sleepynate	amalloy: well, he mentioned he was using openjdk :)
16:14	upwardindex	amalloy: no, looks like a nice set of exercises, however it will only delay the inevitable :D
16:14	xandrews	if anyone was interested in how to list watch fns of an agent, this will do the trick: (.getWatches a)
16:14	sleepynate	upwardindex: what's your java -version ? and what version of kubuntu ?
16:14	amalloy	upwardindex: you can be our 1,001st user. just missed the chance to be #1000, i guess
16:15	upwardindex	java version "1.6.0_22"
16:15	upwardindex	OpenJDK Runtime Environment (IcedTea6 1.10.2) (6b22-1.10.2-0ubuntu1~11.04.1)
16:15	upwardindex	OpenJDK Client VM (build 20.0-b11, mixed mode, sharing)
16:15	sleepynate	huh
16:15	sleepynate	same as me
16:17	sleepynate	amalloy: this is fun. it's like the thing those ruby dudes do
16:17	amalloy	sleepynate: what is? debugging someone's openjdk, or starting on 4clojure?
16:17	sleepynate	4clojure
16:18	sleepynate	c'mon, the ruby guys don't know how to debug a stack trace ;) </playful>
16:18	dsantiago	Is there some place you can stick a Java properties file to have it automatically be loaded and have its properties available via System/getProperty?
16:18	amalloy	yeah, it's genuinely a good learning tool for all levels of experience. i've been humbled by a number of really-cool tweeted solutions
16:19	dsantiago	I thought this was possible, but I can't seem to google up any docs for that.
16:19	sleepynate	amalloy: love when that happens :D
16:20	sleepynate	amalloy: i will play more later. good work and thanks for the community asset
16:27	sjl	Anyone know if Slimv has an issue tracker or mailing list somewhere?
16:28	upwardindex	ring/ring-core doesnt work but ring/ring-jetty-adapter does. Closing in on the problem....
16:33	upwardindex	Wow, something is definitely broken, I can lein deps [ring "0.3.11"] which in turn requires ring/ring-core but not ring/ring-core directly
16:33	upwardindex	Am I going crazy or what?
16:36	amalloy	upwardindex: well, that version of ring is depending on ring-core 0.3.11, not 0.3.9
16:36	amalloy	and while 0.3.9 appears to exist, i can imagine it might be broken
16:36	upwardindex	amalloy: thats the thing i only switched to 0.3.9 because 0.3.11 was not working
16:36	upwardindex	amalloy: but not the issue seems to be resolved
16:37	upwardindex	amalloy: not sure what the solution has been
16:38	upwardindex	sleepynate: thanks for your help the issue is finally resolved, not exactly sure why or how...
16:58	eskatrem	quick question: what is the clojure equivalent of for(int i=0;i<10;i++) { for(int j=0;j<10;j++) { //do stuff}} ?
16:59	lobotomy	,(for [i (range 2) j (range 3)] [i j])
16:59	clojurebot	#<ExecutionException java.util.concurrent.ExecutionException: java.lang.IllegalStateException: replace already refers to: #'clojure.string/replace in namespace: sandbox>
16:59	lobotomy	hmm, what
17:00	lobotomy	&(for [i (range 2) j (range 3)] [i j])
17:00	lazybot	⇒ ([0 0] [0 1] [0 2] [1 0] [1 1] [1 2])
17:00	amalloy	lobotomy: clojurebot has been on a bender for about 24 hours now
17:00	eskatrem	thanks lobo
17:01	TimMc	eskatrem: "do stuff"?
17:01	amalloy	eskatrem: lobotomy is right as long as you realize he means "// return stuff", not "//do stuff"
17:01	TimMc	or compute stuff?
17:01	amalloy	TimMc: cooperative ninja?
17:01	TimMc	:-)
17:02	lobotomy	what's the proper way to do stuff then? doseq?
17:02	lobotomy	e.g. if there's print or whatever
17:02	amalloy	yes. better, of course is to try not to do stuff
17:03	schlecht	is there an idiomatic way to make the following less clunky? "(- {:key map1} {:key map2})"
17:03	amalloy	&(let [stuff-not-done (for [x (range 10)] (println x))] 10) ;; doesn't print anything
17:03	lazybot	⇒ 10
17:03	schlecht	I'm looking at something like (- map1 map2 key)
17:03	amalloy	schlecht: that will throw an exception anyway
17:04	eskatrem	sorry, what's the & ford?
17:04	amalloy	do you mean to use ( instead of {
17:04	eskatrem	*does
17:04	eskatrem	dont know how I typed "ford"
17:04	schlecht	amalloy: oups
17:04	schlecht	amalloy: yes, but that was not my problem
17:04	amalloy	eskatrem: your left hand was one letter too far right :P
17:05	eskatrem	amalloy: hmm true :s
17:05	schlecht	amalloy: I mean something like: (- {:a 1 :b 2} {:a 1 :b 3} :b) == -1
17:06	dnolen	,(apply - (map :key [{:key 10} {:key 5}]))
17:06	clojurebot	#<ExecutionException java.util.concurrent.ExecutionException: java.lang.IllegalStateException: replace already refers to: #'clojure.string/replace in namespace: sandbox>
17:06	dnolen	&(apply - (map :key [{:key 10} {:key 5}]))
17:06	lazybot	⇒ 5
17:07	dnolen	&(reduce - (map :b [{:b 10 :a 9} {:b 5 :a 3}]))
17:07	lazybot	⇒ 5
17:08	schlecht	dnolen: hey, kinda obvious once you think about it :-) Thanks
17:10	TimMc	eskatrem: The & is so the bot will pay attention.
17:11	eskatrem	TimMc: Ok, I was wondering if it was a clojure command of some kind...
18:14	sleepynate	eskatrem: you can put your "do stuff" in that last square brackets
18:15	sleepynate	well, you don't need the squre brackets unless you want vectors
18:16	sleepynate	&(for [i (range 3) j (range 3) k (range 3)] (+ i j k))
18:16	lazybot	⇒ (0 1 2 1 2 3 2 3 4 1 2 3 2 3 4 3 4 5 2 3 4 3 4 5 4 5 6)
18:17	sleepynate	&(apply - (for [i (range 3) j (range 3) k (range 3)] (+ i j k)))
18:17	lazybot	⇒ -81
18:17	sleepynate	etc
18:18	eskatrem	sleepynate sorry, I am not an english native speaker. square bracker is [ ] ?
18:18	sleepynate	yea
18:18	sleepynate	that would be if you wanted vectors
18:19	eskatrem	ok so basically its
18:20	eskatrem	(for [i (range 2) j (range 3) k (range 4)] (;do stuff))
18:20	sleepynate	(for [implicit let statement] (do this) )
18:20	sleepynate	right
18:20	ShreeMulay	Hello everyone! Quick question: If I want to build restful services in clojure, what are the recommendations? Got links?
18:21	tremolo	https://github.com/mmcgrana/ring
18:21	ibdknox	ShreeMulay: http://www.webnoir.org
18:21	eskatrem	and if I want to put several stuff inside my loop its gonna be
18:21	eskatrem	(for [i (range 2) j (range 3) k (range 4)] (do (stuff1) (stuff2)))
18:22	eskatrem	?
18:23	ShreeMulay	checking out Noir -> I'ven't come across this before (that I remember)...
18:23	sleepynate	yes, but the return will always be a list of the last result returned from your do
18:23	sleepynate	erm
18:23	ibdknox	ShreeMulay: it's built on top of Ring and Compojure
18:24	ShreeMulay	@ibdknox sweet!
18:24	ibdknox	ShreeMulay: and should you need any help, I'm the author :)
18:24	amalloy	sleepynate: whoa what, implicit let statement? fors bindings are pretty different from a let
18:25	amalloy	eskatrem: do not attempt to perform any side effects in a (for); as i said earlier it will just make you sad later
18:25	ShreeMulay	ibdknox: thanks!
18:25	eskatrem	amalloy: yes you said it yersterday, but...
18:25	amalloy	ShreeMulay: there's also http://github.com/flatland/wakeful
18:26	amalloy	use a doseq if what you want is side effects
18:26	ShreeMulay	ibdknox: Does it do html5 and support including javascript easily?
18:26	eskatrem	Man, it's hard to get a grisp on clojure, I could do those euler problems so easily in a non functional language, but with clojure ... :'-(
18:27	ibdknox	ShreeMulay: it's a full service webstack. To my knowledge it can do anything the web can do.
18:27	tremolo	ShreeMulay: I suggest looking into hiccup and ClojureScript also
18:27	ibdknox	tremolo: Noir uses hiccup by default :)
18:27	amalloy	it's hard to imagine an http server that somehow forbade html5 in its output
18:27	tremolo	idbknox: Ahh, cool. Didn't know that
18:28	ibdknox	amalloy: pfft, who wants html5? ;)
18:28	ShreeMulay	hiccup's cool!
18:29	amalloy	(defn output [input] (apply str (remove #{\5} input))) ;; first draft of html5 un-support
18:30	ibdknox	for all your retro-web needs! On the roadmap is support the blink tag and doubling your margins in all browsers...
18:30	amalloy	(assoc-in resp [:headers "Content-Type"] "ISO 8859")
18:30	ibdknox	oh man
18:31	ibdknox	amalloy: I can't tell you how long I fought with character encoding shit back in the day when we were writing PHP websites
18:31	eskatrem	amalloy: yes, somehow I can use def inside a doseq more easily
18:31	tremolo	idbknox: how about auto generation of "under construction" animated gifs?
18:31	amalloy	ibdknox: i did the same, back in the day when i was working at a php shop
18:32	ibdknox	tremolo: I'll accept a patch as long as it actually has a pickaxe in it somewhere. ;)
18:32	tremolo	haha
18:33	amalloy	i still get emails from them with "you<big mess>ve got new messages!" or whatever
18:33	ibdknox	hahaha
18:34	ibdknox	the worst part about it is we were working on a site that dealt with wines... all those special characters and multiple database formats each with a different encoding...
18:34	ibdknox	such a disaster
18:34	amalloy	jeez
18:34	ibdknox	that was my last project with that company ;)
18:36	sleepynate	amalloy: sorry i know nothing about the inner workings of the for macro... just that it works like a let
18:36	eskatrem	here is my code for the 4th euler problem: https://gist.github.com/1148056 it works but I'd like to know if my code is "proper clojure" or just some random crap that happens to return a result, can someone have a quick look?
18:36	ibdknox	sleepynate: there are lots of things you can do in a for that you can't do in a let
18:36	amalloy	sleepynate: it has the same structure as a let (plus some special features like :when)
18:36	ibdknox	sleepynate: like :when
18:38	sleepynate	ok
18:38	sleepynate	i'm coming from clisp :)
18:38	amalloy	eskatrem: https://gist.github.com/1148071
18:39	amalloy	any time you find yourself writing (do___) or (def) inside a function, stop and think really hard about whether you need to
18:39	sleepynate	amalloy: juxt is like haskell's flip?
18:39	amalloy	&((juxt inc dec) 1)
18:39	lazybot	⇒ [2 0]
18:39	amalloy	(ie, no)
18:40	ordnungswidrig	sleepynate: (jux inc dec) => (fn [x] [(inc x) (dec x)])
18:40	eskatrem	amalloy: for the second time your neat code is like a slap in my face
18:40	ibdknox	juxt is awesome
18:40	ibdknox	underutilized I thikn
18:40	ibdknox	think*
18:40	amalloy	ibdknox: i use it enough to average out for everyone
18:40	sleepynate	ordnungswidrig: ahh ok
18:41	ibdknox	haha
18:41	ibdknox	outliers ftw
18:41	ShreeMulay	what about netty vs jetty vs mongrel2? any thoughts anyone?
18:41	sleepynate	amalloy: so that's largest palindrome you gist'd ?
18:41	amalloy	sleepynate: yeah
18:41	sleepynate	ok cool. thanks
18:41	ibdknox	netty for persistent connections, jetty for anything else, I can't comment on mongrel
18:42	amalloy	eskatrem: maybe a language thing, but it sounds like you find my code insulting rather than helpful
18:43	ibdknox	lol
18:43	sleepynate	haha
18:43	ordnungswidrig	g
18:43	ibdknox	I think it's a language thing ;)
18:43	eskatrem	amalloy: not at all, I mean: your code is so neat in comparison mine seems stupid, it makes me feel how much I have to learn
18:43	ShreeMulay	ibdknox: would you expand on "persistent connections"?
18:43	amalloy	k, just checking
18:43	eskatrem	actually I am really admirative
18:44	ordnungswidrig	eskatrem: just keep practicing. thinking functional is an ongoing journey.
18:44	eskatrem	and, OK, I dont know what juxt is...
18:44	ibdknox	ShreeMulay: if you need to have websockets or long polling
18:44	ibdknox	ShreeMulay: where you keep an open socket, you want netty because it's nonblocking io
18:44	amalloy	eskatrem: well, it's arguable whether juxt makes my code actually more readable; i was going for "short to write"
18:45	ibdknox	ShreeMulay: Jetty would force you to have a thread for every single connection, which is very limiting
18:45	eskatrem	quick question, do you think I should write a blog about "noob point of view: my discovery of clojure and functional programming"
18:45	ibdknox	ShreeMulay: Jetty will outperform Netty in normal http traffic though
18:46	ordnungswidrig	eskatrem: do it.
18:46	jamiltron	eskatrem: yes please
18:46	eskatrem	inside a for, :let is equivalent to "(let ... " in lisp?
18:46	ordnungswidrig	next blog will be "semi-noob point of view: how to teach point-free-style from point-less-style"
18:46	ShreeMulay	ibdknox: as I understand it, it would be easy to switch from jettty to netty and vice versa via simple line changes, right? I'm planning on building restful services that would allow mobile apps to access backend data (I thinks)...
18:46	sleepynate	ahh i see. hyperpolyglot says clj's (for) is like scheme's for*/list
18:47	eskatrem	yeah actually it's good, it will force myself not to give up
18:47	sleepynate	ordnungswidrig: ooh, you vicious ;)
18:47	ordnungswidrig	clojure's for smells like common lipsps loop
18:47	ordnungswidrig	sleepynate: g
18:47	sleepynate	ordnungswidrig: yes and no
18:47	eskatrem	what's point-free-style?
18:47	amalloy	ordnungswidrig: yes, it's closest to loop. the difference is that loop is heavily imperative and for is heavily functional
18:47	ibdknox	ShreeMulay: Ring's abstraction makes that relatively painless, though I've never tried to run noir on top of aleph, it should work
18:48	ibdknox	ShreeMulay: aleph is built on top of netty and provides a generic asynchronous net library
18:48	ibdknox	ShreeMulay: Noir uses Jetty by default
18:48	sleepynate	amalloy: yea, my instinct for clisp's "loof for" is control monad
18:48	amalloy	eskatrem: i advise not to worry about point-free style. it is a siren song to distract people like me from using let statements
18:49	amalloy	sleepynate: i don't know enough haskell to understand that
18:49	ShreeMulay	ibdknox: how do we send these private messages in irc?
18:49	sleepynate	amalloy: basically what you said
18:49	amalloy	ShreeMulay: /msg ibdknox you're my hero
18:49	sleepynate	amalloy: and why i say "yes and no" to ordnungswidrig's "code smell" comment
18:52	ibdknox	amalloy: lol
18:52	ibdknox	oh the boost to the ego
18:52	ibdknox	lol
18:53	amalloy	ibdknox: you have to get used to the adulation, with noir and brepl
18:53	ibdknox	haha
18:53	ibdknox	amalloy: I'm more excited about Pinot to be honest
18:54	ShreeMulay	what's brepl?
18:55	amalloy	man, i remember when i was new to clojure, i scoffed at (for...) as pointless sugar over map
18:55	amalloy	now everything i write has a (for) in it
18:56	sleepynate	(= __ true)
18:56	sleepynate	do it. DO EET!
18:56	amalloy	(apply > (for [x (range 1000)] (- x)))
18:57	amalloy	&(apply > (for [x (range 1000)] (- x)))
18:57	lazybot	⇒ true
18:57	ibdknox	ShreeMulay: it's a browser connected repl for ClojureScript
18:57	sleepynate	\o/
18:57	amalloy	chuckle
18:57	ShreeMulay	does aleph require netty?
18:58	ibdknox	yes
18:58	ShreeMulay	:P
19:02	amalloy	incidentally, what do people think about adding some euler problems to 4clojure? it's a bit tricky because we tell you the answer and ask for code, whereas euler asks for the answer
19:03	mdeboard	eh.
19:04	ibdknox	amalloy: euler problems are much more math-centric I think
19:04	eskatrem	what's 4clojure?
19:04	mdeboard	I agree with ibdknox
19:04	ibdknox	eskatrem: http://www.4clojure.com/
19:05	amalloy	anyway, eskatrem, i think what would help you most at the moment is a mastery of (let) and/or (loop). you're doing all these (def)s in places where they're just evil
19:05	eskatrem	I remember when I was 17 or so I wrote in basic a program to break down a integer into a product of prime numbers. it took me 1 hour, and now I am struggling with the same in clojure
19:06	ibdknox	eskatrem: this is a different way of thinking
19:06	ibdknox	you should try 4clojure
19:06	eskatrem	Now I get that a def inside a loop is bad, but I dont see why
19:06	amalloy	eskatrem: it's a side effect: you're telling the computer how to do something instead of what to do
19:07	eskatrem	hmmm
19:07	eskatrem	I need to ponder that, amalloy
19:07	tufflax	amalloy about adding project euler problems to 4clojure: how would that work exactly? I don't get it
19:08	eskatrem	4clojure is cool I think
19:09	duck1123	ibdknox: do you know what it was about pinot that causes the "can't recur here"? Giving it another shot today
19:09	ibdknox	an older version of clojurescript has a problem with my macros
19:09	amalloy	tufflax: well, problem 4 could be something like "given a min, max, and filtering function, return the largest palindrome number which satisfies the filter and is a product of two numbers between min and max"
19:10	duck1123	Ok, cause I re-enabled the middleware again, and that's what I'm getting
19:11	eskatrem	how long have you guys been programming in functional programming?
19:11	amalloy	for example, my gist works out to something like "create the products of all 3-digit numbers, and find me the largest number which satisfies the following criteria", while yours is more like "set i to 0; set j to 0; set tmp to i*j; check if it's a palindrome and is larger than our largest palindrome; if so reset our largest palindrome. now set j to 1...."
19:12	ibdknox	duck1123: it's hard for me to reason about your setup
19:12	ibdknox	duck1123: one thing to try would be just create a noir project and see if it even works there
19:12	ibdknox	if not, it'll be easier for me to figure out why
19:13	duck1123	I got errors on OSX with that sample project, but it passed after an explicit clean
19:13	duck1123	Haven't tried it on Ubuntu yet
19:14	ibdknox	yeah i had to clean a couple of my projects too
19:14	amalloy	mutability is contagious in that way; because you handle things like (def) yourself, you can't use the built-in (max) to compute the maximum, and instead have to maintain it yourself
19:15	tufflax	amalloy so you would try do steer people towards functional programming by modifying the problem just a litte bit?
19:15	amalloy	tufflax: no, it's just there should be multiple test cases, where euler tends to have a single problem and a single answer
19:15	tufflax	oh right
19:17	eskatrem	4clojure is getting hard... and then it becomes so tempting to use the repl
19:17	amalloy	eskatrem: please, use the repl! it's good for you
19:18	eskatrem	hmm, if the question if (= __ (conj '(2 3 4) 1)), I can just enter to the repl (con '(2 3 4) 1) to see what it produces
19:18	amalloy	so? you've succeeded, by learning what that form evaluates to
19:18	amalloy	nobody demands that you "figure it out"
19:19	jhulten	its only cheating if you don't learn anything
19:19	amalloy	(inc jhulten)
19:19	lazybot	⟹ 1
19:19	eskatrem	hmm ok
19:19	sleepynate	jhulten: which is why i scripted a bot in clojure to automatically answer the questions for me :o
19:19	sleepynate	wait...
19:19	sleepynate	shit.
19:19	duck1123	That's one of the nice things about functional programming, it's easy to just run a fn, no heavy objects to set up to get it to run
19:19	eskatrem	&(conj '(2 3 4) 1)
19:19	lazybot	⇒ (1 2 3 4)
19:19	Raynes	amalloy: Integrate try-clojure into 4clojure.
19:19	amalloy	Raynes: do it yourself, in cljs
19:19	ibdknox	lol
19:20	jhulten	sleepynate: If you managed that, you probably solved all the hard ones already.
19:20	ibdknox	good luck :-p
19:20	amalloy	ibdknox: it'll be a breeze with brepl, amirite?
19:20	tufflax	Hm, I kinda like project euler where they only require an answer. some problems can be done completely by hand, that's one reason. But also sometimes your code is just a mess after you have experimented. I just like that getting the answer is the only thing that matters. But anyway, I don't really see the point of making pe problems into 4c problems unless you really want more problems.
19:20	Raynes	amalloy: Seriously, it probably wouldn't be that difficult. Might be a good project for Carin.
19:20	ibdknox	amalloy: easy as pie. :D
19:20	eskatrem	hmm not cool, the stuff returned bu lazybot makes 4clojure return an error
19:22	amalloy	##(1 2 3 4) is not the same as ##'(1 2 3 4)
19:22	lazybot	(1 2 3 4) java.lang.ClassCastException: java.lang.Integer cannot be cast to clojure.lang.IFn
19:22	lazybot	(quote (1 2 3 4)) ⇒ (1 2 3 4)
19:23	eskatrem	oh I get it
19:24	sleepynate	#(..) creates a macro?
19:24	tufflax	no a function
19:24	eskatrem	actually I knew the first one but I failed the second one all along, and 4clojure didnt tell me my answer for the first one was right
19:24	sleepynate	nvm reading :)
19:24	amalloy	eskatrem: it should turn the light green for each test you pass
19:26	eskatrem	amalloy: the question I was trying to solve (solved now) has two sub questions, when I put the first answer right and the second one wrong it doesnt make the first light green
19:31	eskatrem	is a set automatically sorted?
19:31	ibdknox	no
19:32	eskatrem	but a set is not ordered then?
19:32	tufflax	no... should it be? :P
19:33	tufflax	there are sorted sets
19:33	tufflax	,(sorted-set 12 6 1 2)
19:33	clojurebot	#{1 2 6 12}
19:36	eskatrem	, (= #{1 2 3} #{3 2 1})
19:36	clojurebot	true
19:36	amalloy	hey, someone woke up clojurebot
19:37	eskatrem	so when clojure compares a set, it doesnt take into account the order
19:37	eskatrem	amalloy (ooops)
19:37	amalloy	eskatrem: there is no order
19:37	amalloy	i mean, yes, internally they are necessarily stored in some order
19:38	amalloy	but as an API, sets don't have an order
19:38	amalloy	&#{3 2 1}
19:38	lazybot	⇒ #{1 2 3}
19:39	eskatrem	ok, so as a convention clojure orders a set to show it, but that's all?
19:39	amalloy	no, it only gets ordered by coincidence
19:39	amalloy	&(set (range 35))
19:39	lazybot	⇒ #{0 32 1 33 2 34 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31}
19:43	amalloy	eskatrem: small integers happen to come out sorted, but that's just luck
19:43	eskatrem	,#{0.3 -1 2.5 0.1}
19:43	clojurebot	#{-1 0.3 2.5 0.1}
19:44	eskatrem	ok, double values are not coming out sorted
19:44	Raynes	&(sorted-set 3 2 4 1023 30)
19:44	lazybot	⇒ #{2 3 4 30 1023}
19:45	eskatrem	,(= (sorted-set 0.3 1.2 0.1) #{0.3 1.2 0.1})
19:45	clojurebot	true
20:01	eskatrem	Guys, 4clojure is coool!!
20:01	amalloy	laugh glad you like it
20:01	amalloy	source is at http://github.com/dbyrne/4clojure if you're interested
20:02	eskatrem	i bookmark that link
20:03	eskatrem	a lady made that! I am impressed
20:05	tremolo	xb
20:05	amalloy	eskatrem: new to github? she's the most recent committer, but it's not a one-person project
20:05	tremolo	hah, emacs fail... sorry about that
20:06	amalloy	https://github.com/dbyrne/4clojure/contributors
20:07	eskatrem	surprising to find two of the top contributors here...
20:09	tufflax	it is? :p
20:10	eskatrem	good work guys, I find myself impressed by my progress
20:11	eskatrem	here is my code for Nth:
20:11	eskatrem	(fn [x n] (loop [y x, i 0]
20:11	eskatrem	(cond (= i n) (first y)
20:11	eskatrem	:else (recur (rest y) (inc i)))))
20:11	eskatrem	(ok, I copied it from a sample code amalloy gave me yesterday)
20:12	amalloy	hey, i didn't write nth for you. now you know how loop works, so you wrote a good implementation of nth
20:13	eskatrem	yeah, but I feel I blantantly copied the loop mechanism from your code
20:13	st3fan	why cond instead of if?
20:13	eskatrem	still I am proud of my nth
20:13	amalloy	st3fan: probably because he copied the loop from an example i gave that had three branches :)
20:13	eskatrem	st3fan: because amalloy did so in a code he sent me yesterdasy
20:14	st3fan	:-)
20:14	amalloy	stealing is a tradition of open source. help yourself to my stuff
20:14	st3fan	where do you live?
20:14	eskatrem	uk
20:14	st3fan	i'll come and pick up your pc ok? :)
20:14	tufflax	lol
20:14	eskatrem	gonna go to sleep soon actually
20:15	amalloy	har har
20:17	amalloy	st3fan: when i open-source my computer itself, you'll be the first person i call
20:19	st3fan	with your open source phone?
20:20	eskatrem	that's neat, my count function is timing out...
20:20	st3fan	i have some code too that hangs the thread
20:20	st3fan	for no obvious reason
20:21	amalloy	your end condition is probably broken, because you're using rest and never seqing the result
20:22	eskatrem	(fn [x] (loop [y x, i 0]
20:22	eskatrem	(cond (= y nil) i
20:22	eskatrem	:else (recur (rest y) (inc i))
20:22	eskatrem	)))
20:22	eskatrem	that's my code, I guess (= y nil) must be somehow screwed uo
20:22	eskatrem	up
20:22	amalloy	&(rest [])
20:22	lazybot	⇒ ()
20:22	amalloy	&(= nil ())
20:22	lazybot	⇒ false
20:23	amalloy	&(= nil (seq ()))
20:23	lazybot	⇒ true
20:23	eskatrem	ohhh
20:23	st3fan	eskatrem: oh don't but those closing braces there please
20:23	st3fan	baaaaaaad style
20:23	amalloy	st3fan: it's true, but there are only so many things you can focus on at a time
20:23	eskatrem	you mean, a bunch of closing brackets in the same time is bad?
20:24	amalloy	i figure he'll eventually see enough code with parens right that he'll figure it out
20:24	st3fan	ok
20:24	amalloy	eskatrem: just the opposite. a line that contains only closing brackets is bad
20:25	eskatrem	hmm I thought closing brackets where like } in C, in C it's cool to have a line containing }, AFAIK
20:26	tufflax	but, at least as far as i'm concerned, i don't want my code to remind me of c or anything like it
20:26	amalloy	chuckle
20:32	eskatrem	when I submit an answer, is my code stored somewhere?
20:39	st3fan	i'm working on a little dsl for web services
20:39	st3fan	to do APIs very similar to Amazon Web Services
20:41	mdeboard	eskatrem: no, just whether you solved it or not
20:41	eskatrem	alright, thank you
20:43	amalloy	mdeboard, eskatrem: untrue. we save your most-recent correct solution
20:44	eskatrem	ooh, neat
20:44	amalloy	eskatrem: try going back to a problem you've already solved - your answer is there
20:45	sritchie	is there some way to catch an exception thrown for an undefined var?
20:46	sritchie	I'm trying to add a prepend-tasks hook in project.clj for native-deps, but native-deps requires deps to be run first
20:47	amalloy	sritchie: those happen at compile time, don't they? you can't really catch them because they happen before your code runs
20:48	amalloy	instead you probably want to do like (when-let [my-var (resolve 'my-var-name)] ...)
20:50	sritchie	it works with eval, though I'm not sure that's the best form:
20:50	sritchie	(try (eval `(do (require 'leiningen.native-deps) (append-tasks #'deps leiningen.native-deps/native-deps))))
20:50	sritchie	with an appropriate catch statement
20:51	amalloy	sritchie: eval definitely looks wrong. does a resolve-based solution not work?
20:52	sritchie	amalloy: hmm, had trouble with it before, since I still needed to send the value of the var I obtain into append-tasks
20:53	sritchie	though that's probably easy -- I'm blanking on the way to get at a value, given a var
20:53	amalloy	@
20:53	sritchie	ah
20:53	amalloy	a one-character solution to all your problems - who could ask for more
20:54	sritchie	haha, it's quite a panacea
20:54	sritchie	trying it out now
20:55	sritchie	yup, all's well… definitely beats eval
20:55	sritchie	amalloy: thanks a lot
21:03	livingston	I wrote a bunch of tests with clojure.test and now realize the work equally well for another set of my code, it seems dumb to cut and paste when the only thing that changes is a star variable...
21:03	livingston	any good way to reference a set of tests and run them twice in different binding scopes?
21:05	eskatrem	I finished the 1st page of 4clojure. I am going to bed now. thanks everybody!
21:05	lucian	livingston: why not factor out a fn that takes that variable as an argument?
21:07	livingston	I could seemed odd of even have to repeat the tests if I could help it. then I'd have (deftest foo-test (is (foo A))) in one file and (deftest foo-test (is (foo B))) in another
21:11	lucian	livingston: i'd make a (defn is-foo ...) to import in both test files
21:12	livingston	lucian: I could imbed the "is" too it doesn't need to be inside deftest?
21:12	lucian	livingston: don't really know
21:13	livingston	the problem here is remembering to update both files ... I'll probably just macro around it... just was hoping for something slicker.
21:13	lucian	livingston: wouldn't a fn be enough? does it really need a macro?
21:14	lucian	depends on the context of is i guess
21:14	livingston	if I want them to register as a lot of different tests so it's easier to tease out where the failure is. I should look at clojure.test documentation more..
21:15	livingston	is doesn't need to be in deftest
21:24	technomancy	livingston: deftest generates function
21:24	technomancy	*functions
21:25	technomancy	if you (deftest my-foo-test) you can just call (my-foo-test) from another test
21:25	hiredman	~ping
21:25	clojurebot	PONG!
21:25	amalloy	deftest is absurdly macro-heavy. an (is) that's not inside a deftest won't work
21:26	technomancy	amalloy: not actually true
21:26	technomancy	maybe you're thinking of thrown? outside an is?
21:27	technomancy	because that one's annoying
21:28	amalloy	technomancy: i don't think you're right. `is` is a symbol that's parsed specially by the deftest macro, not a macro/function that's actually defined
21:28	amalloy	also, i just checked and i'm wrong about that
21:28	amalloy	so never mind, technomancy
21:28	technomancy	,(ns-resolve (doto 'clojure.test require) 'is)
21:28	clojurebot	#'clojure.test/is
21:28	technomancy	yay doto abuse
21:29	amalloy	$sed technomancy s/abuse/magic
21:29	lazybot	<technomancy> yay doto magic
21:29	technomancy	amalloy: thrown? is like that
21:29	amalloy	nod
21:29	amalloy	i must have gotten them confused in my head somewhere
21:29	cemerick	Bots with sed. I'm going to go stock up on canned goods.
21:30	amalloy	cemerick: is that a sign of the apocalypse?
21:31	cemerick	amalloy: Close enough for me! :-D
21:33	livingston	sorry spaced out for a second...
21:34	livingston	I could call all the test from the second file but then I would still have to make sure I keep them synchronized.
21:35	amalloy	(ns a) (defn run-all-the-damn-tests [] (...)) (ns b (:require a)) (a/run-all-the-damn-tests)
21:35	livingston	basically what I have is two implementation of an API. I'd like to write all the API tests once, and then have both implementations check themselves against it.
21:36	livingston	so I found run-tests that allows me to call tests in another namespace... I think I can bind over my star-var and then call that, I was just about to test that.
21:36	cemerick	Bugger, nil-safe threading ops should be in core by now.
21:38	livingston	ideally I'd like the tests defined in the API test-namespace, but not run (how could they there's no implementation there) and then run for each of the implementations.
21:39	livingston	pwd
21:39	livingston	^ wrong window
21:40	amalloy	livingston: c'mon, correcting it is no fun. you want to give someone a chance to jump in there with "#clojure"
21:40	livingston	sorry didn't realize that was protocol
21:40	st3fan	any compojure experts here? i have troubes getting a basic route to work
21:41	amalloy	~anyone
21:41	clojurebot	Please do not ask if anyone uses, knows, is good with, can help you with <some program or library>. Instead, ask your real question and someone will answer if they can help.
21:41	livingston	amalloy: lazybot or some bot should have yelled at me.
21:41	livingston	lol, like that
21:41	st3fan	i have this route: (GET "/ws" {params :params} (dispatch-action params))) .. i can call it with /ws?action=Foo but params is always an empty map
21:46	amalloy	st3fan: (a) there's a #compojure room, but i don't know how much traffic it gets; (b) do you have a middleware in place that will parse out the params for you?
21:47	st3fan	yeah i guess that is what i need
21:48	amalloy	if you don't have something like wrap-params intervening, nobody's going to put anything in that :params map
21:49	amalloy	also (c) i can never remember the compojure syntax for (GET) stuff, so there might be something wrong with your {params :params} binding too
21:52	ShreeMulay	Anyone around that can help me solve a 4clojure problem? @ http://www.4clojure.com/problem/19#prob-title
21:53	ShreeMulay	I can't use "last" so I'm at a loss, eh...
21:53	jeremyheiler	What have you tried?
21:53	ShreeMulay	last
21:53	ShreeMulay	they don't want me using that, though...
21:54	ShreeMulay	and I really don't know enough, to be able to think of what next to try unfortunately... :(
21:55	jeremyheiler	$((peek [1 2 3]))
21:56	jeremyheiler	damn. what's the clojurebot syntax?
21:56	jeremyheiler	for running code
21:56	amalloy	,
21:56	clojurebot	#<ExecutionException java.util.concurrent.ExecutionException: java.lang.RuntimeException: EOF while reading>
21:56	amalloy	or & for lazybot
21:57	jeremyheiler	,((peek [1 2 3]))
21:57	clojurebot	#<ClassCastException java.lang.ClassCastException: java.lang.Long cannot be cast to clojure.lang.IFn>
21:57	livingston	do namespace names with hyphens get munged to underscore too? e.g. does foo.foo-core.bar-code become foo/foo_core/bar_code.clj ?
21:57	amalloy	ShreeMulay: the point is to reimplement last yourself
21:57	jeremyheiler	,(peek [1 2 3])
21:57	clojurebot	3
21:57	amalloy	livingston: yes
21:57	amalloy	because they're java package names
21:58	jeremyheiler	Shree: As you can see, peeking on a vector gives you last
21:58	livingston	amalloy: thanks. great, time to mess with what in subversion...
21:58	jeremyheiler	however
21:58	jeremyheiler	,(peek '(1 2 3))
21:58	clojurebot	1
21:58	jeremyheiler	The same thing on a list doesn't work
21:58	jeremyheiler	this is because lists are designed for adding/removing from the front, where vectors were designed for adding/removing from the back.
21:58	jeremyheiler	Play with that thought for a bit
22:00	jeremyheiler	,(peek (vec '(1 2 3)))
22:00	clojurebot	3
22:00	jeremyheiler	It's not good code, but it works
22:01	amalloy	meh. i'd classify that more as a "trick" than a solution to the problem - being able to implement `last` yourself is valuable
22:01	amalloy	not that i'm against tricks
22:01	jeremyheiler	Yes, I agree
22:02	jeremyheiler	I was more-or-less playing golf with 4clojure lol
22:02	ruggeri	Any advice for newbies on defrecord? It's a shame print-dup doesn't produce something that can be you can call read-string on.
22:03	ruggeri	There's this defrecord2 guy
22:03	amalloy	ruggeri: it does in 1.3
22:03	ruggeri	Oh snap
22:03	ruggeri	Thanks; will check it out now
22:03	amalloy	jeremyheiler: the shortest solution for problem 19 is 8 characters long, and it's hysterical. any guesses?
22:05	cemerick	Everyone just uses snapshots, alphas, and betas anyway; Clojure should follow the path of chrome, version up to Ω-SNAPSHOT, and everyone can just ride HEAD. :-P
22:06	jeremyheiler	hmmm
22:07	ruggeri	yeah, records being second-class was about to super-depress me
22:07	amalloy	ruggeri: that said, "newbie" and "record" don't belong in the same sentence
22:08	amalloy	they're an advanced feature, however appealing they look when you're coming from an OOP background
22:08	ruggeri	why do you say so? I want to implement me some multimethods
22:08	amalloy	ruggeri: so write some multimethods; you don't need protocols for that
22:08	amalloy	(or records)
22:08	ruggeri	but what's my dispatch function? I would call something like class or type, right?
22:09	cemerick	records are a proper replacement of structs, which are reasonable for any "level" of usage, IMO.
22:09	ShreeMulay	peek is interesting - but can't figure out how to "reimplement last" cause, I frankly don't know enough clojure - haven't used it enough...
22:09	amalloy	(defmulti pluralize count) (defmethod pluralize 1 [objs] (str "1 thing: " (first objs)))
22:10	amalloy	(defmulti pluralize 0 [_] "No things at all!") ;;etc
22:11	ruggeri	Fair enough; but what if I have two different dog objects, and they bark differently?
22:11	ShreeMulay	peek works for the vector... somehow, I've got to create a function... Is "peek" gonna' be part of the function I need to put together, guys?
22:11	ShreeMulay	on problem #19?
22:12	jeremyheiler	It's not
22:12	ruggeri	I want (bark big-dog) ; -> growl
22:12	ruggeri	(bark lil-dog) ; -> yip-yip
22:12	jeremyheiler	ShreeMulay: Look up how to use recur
22:13	amalloy	ruggeri: shrug you can make a record for that, and define a protocol for barking. or you can use a plain map, and a dispatch function that looks up the :kind-of-dog key in the map. or you can do any number of things
22:13	jeremyheiler	amalloy: I give lol
22:13	amalloy	jeremyheiler: rand-nth, and hitting the submit button ~40 times
22:14	jeremyheiler	hahaha
22:16	ruggeri	amalloy: yeah, definitely could manage a canonical mapping of dogs to bark behaviors
22:16	st3fan	hmm Clojure in Action has an compojure example that import compojure using (ns foo (:use compojure)) but when I try that I get an import error
22:17	ruggeri	amalloy: I'll start there; it'd be a pain for one-place to centralize every type of dog in existence for a huge project, but good idea to keep things simple maybe
22:17	amalloy	i think that the concept of "but someday i might have a thousand different kinds of dogs!" is overemphasized in OO code
22:18	ruggeri	amalloy: I hear you; makes sense
22:18	amalloy	i'm a fan of splitting things up when they need to be; i'm addicted to defmulti and i write protocols when they make sense
22:18	amalloy	but they're not the sorts of things that give you a feel for how the language is different from (say) java
22:19	amalloy	do some sequence manipulation, recursion, and function composition. those are things that you don't get much of in java
22:19	livingston	amalloy: I think that the concept that everything needs a class is over emphasized, why have one function and some configuration when you could have the land of 1000 classes
22:20	amalloy	anyway, i'm headed home. ta-ta
22:20	ruggeri	amalloy: night
22:20	livingston	later
22:21	jeremyheiler	later
22:24	ShreeMulay	KILL
22:24	ShreeMulay	QUIT
22:42	pcavs	how do people do reading files from the classpath?
22:49	technomancy	pcavs: clojure.java.io/resource
22:49	pcavs	technomancy: thanks
22:50	technomancy	sure
22:55	st3fan	is there a better way to update nested maps than nested assoc calls?
22:58	cemerick	st3fan: assoc-in and update-in
22:58	st3fan	thanks
23:00	st3fan	oh assoc-in is nice
23:04	cemerick	st3fan: they work with any 'associative' as well…not just maps
23:04	cemerick	,(update-in [5 6 {:a [\b \c \d]} 7] [2 :a 1] str)
23:04	clojurebot	[5 6 {:a [\b "c" \d]} 7]
23:08	st3fan	yeah very cool
23:08	st3fan	how do i test if a condition is true for each element in a vector?
23:09	amalloy	&(doc every?)
23:09	lazybot	⇒ "([pred coll]); Returns true if (pred x) is logical true for every x in coll, else false."
23:09	st3fan	argh i was looking for each? :-/
23:13	livingston	stupid question, what's fncall in clojure ie. I have a list of functions I just want to map through them and call them.
23:13	livingston	(map (fn [x] (x)) functions) seems dumb
23:13	amalloy	$findfn even? [2 4 6] true
23:13	lazybot	[clojure.core/not= clojure.core/some clojure.core/distinct? clojure.core/every?]
23:13	amalloy	livingston: (
23:13	amalloy	oh, i see
23:14	amalloy	yes, i wish invoke would make it into core also. it doesn't really exist
23:14	livingston	so there's nothing?
23:14	amalloy	livingston: fwiw, #(%) is the shortest way to write it :P
23:15	livingston	I really despise that % syntax, I find it lending itself to too much confusion
23:15	amalloy	but (fn [f & args] (apply f args)) is a useful thing to have defined somewhere. i call it invoke because of IFn.invoke, but i guess funcall is more traditional
23:16	amalloy	st3fan: you caught the use of $findfn for your benefit, right?
23:16	livingston	I don't need apply, they are thunks, with no arguments
23:17	livingston	in common-lisp is LISP2 so you need a fncall there was no (foo) to call the function in foo like you can in LISP1
23:18	livingston	is there anyway to deftest in a namespace but not have it called with clojure:test and only have the functions there to call with another namespace?
23:19	amalloy	livingston: i'm aware. if you look at it, you'll see i just wrote funcall
23:20	amalloy	&(let [funcall (fn [f & args] (apply f args))] [(funcall +) (funcall + 1 2)])
23:20	lazybot	⇒ [0 3]
23:20	amalloy	ie, it works for thunks or with args
23:21	cemerick	Just adding a single-arg arity to apply would be equivalent, no?
23:22	amalloy	cemerick: not equivalent to my definition, no, because (apply + 1 2) wouldn't work
23:22	amalloy	perhaps equivalent to what livingston needs at the moment
23:22	livingston	yeah. it's equivalent, except for the list cons-ing
23:24	livingston	... all this work just to call some tests twice sigh 4 hours in...
23:28	amalloy	hey livingston, can you point me at some public, preferably not-totally-trivial code you've written? i had the idea recently to look at how ye olde lispers write clojure
23:29	livingston	um, hum I don't know that I have any - but I am in the process of preparing a large library of code for dealing with RDF.
23:31	livingston	amalloy: I'm not exactly an old lisper but I was trained by them and programmed my entire dissertation in it.
23:32	livingston	the biggest thing that's kicking me in the pants is dynamic not lexical scope. there are a ton of cool tricks with lexical scope that get fubared by lazy-list / deferred evaluation in dynamic scope
23:34	amalloy	livingston: CL has the same scope rules as clojure, doesn't it? it's just that clojure has a lot more lazy bits
23:35	cemerick	How is the CL world doing these days in general? Anyone keep tabs on it?
23:35	livingston	amalloy: no dynamic vs. lexical scope for special variables that can be bound ove is a huge difference
23:35	livingston	cemerick: it's fine. AI people still use it as always.
23:35	amalloy	livingston: but clojure's lexical variables are the same as CL's, and clojure's vars are the same as CL's specials, i thought
23:35	livingston	every time you buy a plane ticket there's lisp under the hood ;)
23:36	cemerick	yeah, but every time I swipe my credit card, an AS/400 gets its wings. Not really a good barometer, perhaps. :-)
23:36	cemerick	I wonder about its vector more than anything else. Growing, shrinking, etc.
23:36	cemerick	I would think that the heat around Clojure and other FPLs must help.
23:37	amalloy	cemerick: functional? CL isn't exactly functional
23:37	livingston	lol
23:38	livingston	CL is plenty functional when coded well, although there are the people that call it hybrid or whatever.
23:38	cemerick	Exactly, no — but before state was a big concern, Lisps were always considered functional in the functions-are-first-class-values sense.
23:40	amalloy	i was just reading that in let-over-lambda, actually
23:40	cemerick	This floated by in the transom a few days ago; I'm not sure whether to interpret the results as positive or negative overall, tho. http://www.indiegogo.com/SBCL-Threading-Improvements-1
23:42	livingston	you have to remember that CL was standardized before file-systems/sockets/threads had any major standardization, consequently those things have been ideosyncratic in lisps
23:42	dnolen	cemerick: why negative?
23:43	livingston	but there are generally libraries that will present a layer over your specific implementation
23:43	cemerick	livingston: Right, I'm mostly looking at community participation / size / engagement / commitment.
23:44	livingston	depends on what's necessary for you to define a community.
23:44	cemerick	dnolen: I can see it both ways. (a) it's clearly a vital community ready to step up and support key contributions with hard cash vs. (b) The premier open source CL impl can garner "only" $9K to solve probably its most long-standing and most-griped-about problem?
23:45	grant_	man, how do you debug errors that have no source code line?
23:45	grant_	any tips or tricks aside from weeping and gnashing your teeth?
23:46	dnolen	grant_: what dev environment?
23:46	grant_	emacs, swank, clojure 1.3beta (i updated to see if it would make a difference... it didnt :( )
23:46	livingston	symbolic AI guys like it. (and I'm one) do I care that few games, websites, or mcdonalds ordering systems are written in it, not really. frankly if those guys all had their way everything would look like java, and we know what kind of fun that is ;)
23:46	dnolen	grant_: are you evaling fn's one at time or the whole file?
23:46	grant_	whole file
23:46	grant_	i stepped through evalling each one and they all evalled fine
23:47	dnolen	grant_: what is the exception?
23:47	cemerick	livingston: Willingness to lay out cash seems like a pretty unassailable metric of engagement. Fundraisers like that are one manifestation, another might be attendance figures at yearly confs, etc.
23:47	amalloy	grant_: that's not a whole file at once; that's one-at-a-time, over-and-over
23:47	grant_	clojure.lang.Cons cannot be cast to clojure.lang.IFn
23:47	grant_	i'm saying i did both that and whole-file evaluation
23:47	amalloy	okay
23:47	dnolen	&((cons 1 nil) 1)
23:47	lazybot	java.lang.ClassCastException: clojure.lang.PersistentList cannot be cast to clojure.lang.IFn
23:47	livingston	SBCL is a popular open source lisp true. But Franz makes the Allegro compiler and they are doing quite nicely as a company.
23:48	dnolen	&((cons 1 ()) 1)
23:48	lazybot	java.lang.ClassCastException: clojure.lang.Cons cannot be cast to clojure.lang.IFn
23:48	dnolen	grant_: ^
23:48	grant_	right
23:48	grant_	so i should be on the loop out for extra parens
23:48	livingston	with their lisp they created AllegroGraph which is as a far as I know one of the best of breed triple/quad stores (RDF/OWL etc.)
23:48	amalloy	grant_: or something quoted where it shouldn't be
23:49	grant_	but is there is a faster way to narrow this down besides manual inspection?
23:49	amalloy	grant_: read the stacktrace to see which of your functions are in it
23:49	cemerick	livingston: right — maybe all the real users of CL in industry are on Allegro and Lispworks. If they released sales figures, that'd make these sorts of judgements a lot easier… ;-)
23:49	dnolen	grant_: if there's no source line after C-c C-k it's happening during compilation so very certain to be a syntax error.
23:49	amalloy	and see if there are any clojure.lang.Compiler methods - if so it's a problem with macroexpansion
23:49	grant_	a syntax error??
23:49	cemerick	The fellow who wrote AllegroGraph lives a few miles from me. Very nice guy.
23:50	grant_	my only function appearing in the trace is my top-level main fn
23:50	grant_	oh yeah, i do see a Compiler method
23:51	livingston	cemerick: I think there are several "fellows" including the CEO of Franz who work on it.
23:51	pcavs	has anyone played around with noir?
23:52	livingston	cemerick: the people that use Franz lay out cash, the licenses are not cheap.
23:52	amalloy	&(let 1 2)
23:52	lazybot	java.lang.IllegalArgumentException: let requires a vector for its binding
23:52	cemerick	livingston: IIRC, he was the tech lead, anyway. shrug
23:52	amalloy	that one would be at compile-time as well
23:52	cemerick	memory fades, too :-P
23:52	livingston	cemerick: he's a nice guy too, John, right?
23:54	cemerick	livingston: Gary King is the one I know.
23:54	cemerick	Man, it's been a while since I talked to him, too. :-(
23:54	livingston	also a nice guy, I've worked with him some too.
23:56	grant_	amalloy, dnolen: thanks guys, that helped!
23:56	dnolen	grant_: what was it?