#clojure – 2010-05-09

2010-05-09

00:02	_brian2_	I think you define java exceptions for anything
00:05	sids	that's left to the library author to decide
00:05	_brian2_	which uses a wrapper around the java (html/html-resource (java.net.URL. url))
00:06	_brian2_	well, they haven't addressed this problem
00:08	technomancy	hugod: have you gotten trunk slime to work with lein swank and M-x slime-connect, or just swank-clojure-project?
00:09	sids	_brian2_: that function uses java.net.URL's getContent which should be throwing a FileNotFoundException or something
00:10	sids	,(import 'java.net.URL)
00:10	clojurebot	java.net.URL
00:10	sids	,(.getContent (URL. "http://clojure.org/doesnt-exist"))
00:10	clojurebot	java.security.AccessControlException: access denied (java.net.SocketPermission clojure.org:80 connect,resolve)
00:13	_brian2_	so if I something like use (try ( (html/html-resource (java.net.URL. url)) (catch ... ) , the program still blows up
00:13	technomancy	_brian2_: any particular reason the call to html/html-resource is nested in two levels of parens?
00:14	_brian2_	no, its a typo
00:14	_brian2_	i mean I dont in program
00:14	sids	_brian2_: you don't what?
00:15	technomancy	_brian2_: gotcha. well I think more context is needed then; try a paste site?
00:15	_brian2_	use 2 parens
00:15	_brian2_	ok, just a second, I'm going to get it al together
00:15	_brian2_	hopefully ur still here
00:18	sids	_brian2_: I just checked, enlive's html-resource does throw FileNotFoundException for a 404 URL
00:19	_brian2_	but my problem is 400 errors
00:19	sids	_brian2_: can you give me url that throws 400, I can check it for you
00:20	sids	s/throws/returns/
00:20	sexpbot	_brian2_: can you give me url that returns 400, I can check it for you
00:20	_brian2_	ok, just a sec, this is kind of a random event, so hopefully i will catch one
00:22	tomoj	sexpbot: who is your master?
00:22	tomoj	:(
00:23	_brian2_	http://is.gd/a3eEU
00:23	sexpbot	"Streetwear Shop - 80s Retro Style Fashion Clothing & Accessories — Home"
00:23	_brian2_	but it seems to be re-routed in my browser
00:25	sids	that returns a 301 redirect and following that redirect leads to a page which returns 200
00:25	_brian2_	actually, its not the problem
00:25	_brian2_	the problem I want to solve is simple if I get 400 error to exit and move on
00:26	_brian2_	the reason I got the error is that i am
00:26	_brian2_	using crappy way to locate the url in text
00:26	sids	400 is bad request, right?
00:26	_brian2_	I guess
00:27	_brian2_	something like that
00:27	sids	enlive should throw an exception for that, give me one such url and I can check
00:27	_brian2_	the url I used was : http://is.gd/a3eEU Use discount code "gaga" for 25% off!
00:27	sexpbot	"Streetwear Shop - 80s Retro Style Fashion Clothing & Accessories — Home"
00:28	_brian2_	sexbot doesnt mind
00:28	_brian2_	that one
00:28	sids	enlive doesn't mind it either
00:28	_brian2_	or maybe just make one up
00:29	_brian2_	I get "main" java.io.IOException: Server returned HTTP response code: 400 for URL: http://is.gd/a3eEU Use discount code "gaga" for 25% off! B6d
00:29	_brian2_	at sun.net.www.protocol.http.HttpURLConnection.getInputStream(HttpURLCon
00:29	sexpbot	"Streetwear Shop - 80s Retro Style Fashion Clothing & Accessories — Home"
00:30	sids	so you get the exception, then what seems to be the problem?
00:30	sids	you are unable to catch it?
00:31	_brian2_	the program crashes
00:31	sids	can you paste the relevant code somewhere?
00:31	_brian2_	ok
00:35	_brian2_	sids : http://clojure.pastebin.com/VuTid23R
00:36	sids	on a different note, you seem to be using this entire string as the url: >>http://is.gd/a3eEU Use discount code "gaga" for 25% off! B6d<<
00:36	sexpbot	"Streetwear Shop - 80s Retro Style Fashion Clothing & Accessories — Home"
00:37	_brian2_	yea, I am using a crappy method to grab the URL
00:37	_brian2_	but I want to make the program robust to the errors
00:38	_brian2_	rather improve the pattern regex
00:38	_brian2_	becasue that will never be perfect
00:42	sids	_brian2_: your catch syntax wasn't quite alright (missing the name argument)
00:43	_brian2_	yea, I just saw that
00:43	sids	I fixed that and put it in a repl, runs without a worry: http://clojure.pastebin.com/ZBhcLj1m
00:43	_brian2_	ok
00:43	_brian2_	thanks!
00:43	sids	there is no need to use a try around your let, fetch-url will catch the exceptions and return a nil in that case
00:44	_brian2_	ok
00:44	technomancy	is anyone using slime-repl.el from trunk slime successfully with clojure?
00:47	technomancy	I'm getting regular slime evaluation to work, but the repl is busted and still thinks it's in CL-USER
00:51	_brian2_	sids : runs beautifully, thanks for your help!
00:53	sids	_brian2_: you're welcome :)
03:46	ArkRost	Hi! Can someone write a code of the parallel quicksort with using transient strucrures?
04:10	LauJensen	Morning guys
04:10	LauJensen	I'm benchmarking my site using 'siege' and it gives me everything I want, except a mean transaction time, anybody know of a simple *nix tool which does the job a little better?
06:00	herdrick	,(str (repeat 1 "foo"))
06:00	clojurebot	"clojure.lang.LazySeq@18ce5"
06:00	herdrick	,(repeat 1 "foo")
06:00	clojurebot	("foo")
06:01	herdrick	Question: how do I get the usual string representation from that lazy sequence?
06:01	herdrick	I thought doall was the way, but no.
06:02	herdrick	Very basic question, yet I don't know or am having a memory lapse
06:05	herdrick	In other words, what I'm looking for is (some-fn (repeat 1 "foo")) to return a string "(foo)" or something formatted like that.
06:05	herdrick	other formats are fine: "('foo')" "(\"foo\")" whatever
06:11	herdrick	ah, gotta sign off. will be back.
06:21	LauJensen	$mail herdrick (apply str (repeat 1 "like so"))
06:21	sexpbot	Message saved.
06:25	nurv	Hi.
06:28	cschreiner	$mail
06:28	cschreiner	cool feature
06:29	cschreiner	$help
06:29	sexpbot	cschreiner: I can't help you, I'm afraid. You can only help yourself.
06:31	herdrick	Sorry, I was gone. Did i miss anything?
06:31	herdrick	$mail
06:31	herdrick	oops
06:31	Borkdude	$mail
06:40	herdrick	LauJensen: thanks. Though I would like something that preserved the tree structure of the seq. So this would work:
06:40	herdrick	,(str "(" (apply str (repeat 1 "foo")) ")")
06:40	clojurebot	"(foo)"
06:40	herdrick	...for that exact case
06:41	herdrick	but I was hoping for something more general. I guess I will write my own recursive thing. Or use the tree walker stuff in clojure.contrib
06:41	herdrick	sorry, clojure.walk
06:52	jowag	alter has its equivalent for root values in alter-var-root. Is there an ref-set equivalent for root vars?
06:58	jowag	I think for that purpose redefining root var with def should be fine, but I'm not sure
07:37	jwr7	ah-hah. So clojure.contrib.sql's insert-rows is really insert-row.
07:39	LauJensen	jwr7: yea, check out ClojureQL instead :)
07:39	jwr7	LauJensen: I would, except after spending lots of time with CLSQL I'm opposed to any additional DSLs on top of SQL…
07:41	jwr7	I might have to use something better than clojure.contrib.sql, though — there are no persistent connections and no connection pooling.
07:44	LauJensen	Oh - If you don't like a DSL ClojureQL is definitely not for you, its _the_ DSL for SQL :) Lets you extend the SQL language itself, like add fulljoins to derby etc. But what rubs you the wrong way about DSLs ?
07:45	LauJensen	@ jwr7
07:45	jwr7	LauJensen: experience. I believe they optimize for the wrong goal: db-independence. In practice, I've never encountered a situation where I would actually need to migrate to a different DB and expect it to work flawlessly.
07:45	jwr7	LauJensen: you always have to modify things when migrating to a new DB, DSL or not.
07:45	jwr7	LauJensen: so there is little advantage. And the disadvantages are many. I have to learn a new syntax, convert things into that syntax and encounter its limitations.
07:46	jwr7	It also isolates me from the db-specific features I might want to use.
07:47	LauJensen	Im not saying that you dont, but lets say you're benchmarking some app and you want results on both MySql, Oracle, Mongo and Cassandra - ClojureQL lets you do that with the same code. It always provides raw hooks into the db-specific stuff you might want you use. And the limitations are only a patch away from being removed
07:47	LauJensen	But yes, you do have to learn the syntax
07:47	jwr7	LauJensen: right — but that's a very specific use case, which isn't my use case :-)
07:47	somnium	LauJensen: eh? you support nosql too?
07:48	LauJensen	somnium: Mongo and Fleet are being implemented as we speak, and we hope to have at least one in full working order for v. 1.0
07:48	somnium	LauJensen: full joins on CouchDB?
07:49	somnium	ah, I was about to ask if it was still in flux, how far off is 1.0 looking?
07:49	LauJensen	somnium: The backend is currently multimethods, so if you implement a frontend function which returns a :fulljoin object, just add the method you want to the backend. You can see how easily we did it for Derby and Sqlite
07:49	jwr7	LauJensen: I think this discussion is similar to maps vs objects — you get way more flexibility if you put a layer of objects and multimethods on top of your data, but then again, maps get you there 90% of the way and often enough you don't need that flexibility at all.
07:49	jwr7	…and having debugged CLSQL… well, you get the point :-)
07:49	LauJensen	somnium: We have a deadline in 2 weeks for all the frontend stuff (the frontend is completely reworked), then we're assing how much we need to retrofit the backend, whereafter we can publish a release date
07:50	LauJensen	Im guessing 6 - 8 weeks, preferably 5 :)
07:50	jwr7	hey, but that lack of connection pooling worries me…
07:50	jwr7	does this mean there is no library for accessing an SQL database from a multithreaded application without establishing and tearing down connections all the time?
07:51	LauJensen	jwr7: We put in some alpha connection pooling a while ago, it does support multiple long-lived connections
07:51	somnium	mongo isnt that painful without a dsl, despite the underlying javaness of the driver
07:51	LauJensen	But thats also going to change with the release of 1.0, just not sure how much yet
07:53	somnium	LauJensen: if you really want to be heroic you could do everything from TCP/IP on up in clojure for direct BSON -> persistent datastructure :D
07:54	LauJensen	:)
07:54	somnium	someone said they were working on a prototype for a pure clojure Mongo client but they havent been around in a while
07:55	LauJensen	pure, as in rolling their own driver?
07:56	somnium	the low level bits anyway
07:56	LauJensen	cool- I'll make sure you test it again clojureql 1.0 once they're both released :)
07:56	jwr7	LauJensen: oh, cool — so all that remains is for you to remove the DSL thing, and I'll use ClojureQL :-)
07:57	somnium	heh
08:01	somnium	LauJensen: would it be better to wait for a 1.0rc to learn the api?
08:45	rhudson	Given a lazy seq of chars, how do I get a string value?
08:46	rhudson	E.g. given (concat "abc") how do I get "abc" ?
08:46	rhudson	,(str (concat "abc"))
08:46	clojurebot	"clojure.lang.LazySeq@1ecc1"
08:53	somnium	,(apply str (concat "abc"))
08:53	clojurebot	"abc"
08:54	rhudson	Thanks! This has been driving me crazy.
08:54	somnium	cheers
08:55	elguru	,(version)
08:55	clojurebot	java.lang.Exception: Unable to resolve symbol: version in this context
08:55	elguru	,(clojure-version)
08:55	clojurebot	"1.1.0-master-SNAPSHOT"
08:55	elguru	,(persistent!
08:55	clojurebot	EOF while reading
08:55	elguru	(reduce #(assoc! %1 %2 %2) (transient {}) (range 0 10))))
08:56	elguru	,(persistent! (reduce #(assoc! %1 %2 %2) (transient {}) (range 0 10)))
08:56	clojurebot	{0 0, 1 1, 2 2, 3 3, 4 4, 5 5, 6 6, 7 7, 8 8, 9 9}
08:57	elguru	,(let [result (transient {})]
08:57	clojurebot	EOF while reading
08:57	elguru	(doseq [elt (range 0 10)]
08:57	elguru	(assoc! result elt elt))
08:57	elguru	(persistent! result))
08:57	elguru	,(let [result (transient {})]
08:57	clojurebot	EOF while reading
08:57	elguru	(doseq [elt (range 0 10)]
08:57	elguru	(assoc! result elt elt))
08:57	elguru	(persistent! result))
08:58	elguru	,(let [result (transient {})] (doseq [elt (range 0 10)] (assoc! result elt elt)) (persistent! result))
08:58	clojurebot	{0 0, 1 1, 2 2, 3 3, 4 4, 5 5, 6 6, 7 7}
08:58	elguru	ok, again:
08:58	elguru	,(persistent! (reduce #(assoc! %1 %2 %2) (transient {}) (range 0 10)))
08:58	clojurebot	{0 0, 1 1, 2 2, 3 3, 4 4, 5 5, 6 6, 7 7, 8 8, 9 9}
08:58	somnium	elguru: youre not supposed to bash them in place
08:58	elguru	in the doseq version?
08:59	somnium	elguru: yes
09:00	elguru	why is that? and is it mentioned somewhere? my problem is, that the result is quite unexpected and is sometimes correct and somtimes not.
09:02	elguru	just failing would be better of course.
09:02	somnium	elguru: I dont know the answers to one and two, but the solution is to use the transient versions the same way as the immutable ones
09:03	somnium	I agree, but I suspect since they only exist for speed, they dont waste time on checking for undefined usage
09:03	somnium	into uses them under the hood btw
09:04	somnium	,(into {} (map vector (range 10) (range 10)))
09:04	clojurebot	{0 0, 1 1, 2 2, 3 3, 4 4, 5 5, 6 6, 7 7, 8 8, 9 9}
09:04	somnium	should be 'fast'
09:05	elguru	ah ok, i will check it out. thank you.
09:11	LauJensen	somnium: yes, it would be better to wait, as the frontend will change completely. In the backend, just a few dbs will be added
09:12	LauJensen	See the wiki on gitorious for more info
09:13	somnium	LauJensen: any chance of supporting BDB, maybe with cupboard?
10:29	Licenser	rhickey: how is the copyright on the 'clojure design' and logo?
10:30	Licenser	aka can I use the logo /general color scheme on a website?
10:34	Borkdude	Why can't I do this inside a let?
10:34	Borkdude	(let [my-power-list (lazy-cat [x] (map #(* % x) my-power-list))]
10:34	Borkdude	It says it can't resolve symbol my-power-list
10:35	Licenser	Borkdude: do you have declared my-power-list before?
10:35	Borkdude	nope, but I can do (def my-power-list .. same stuff ..)
10:35	somnium	Borkdude: let isnt recursive, you can use letfn, or ... [my-list (fn my-list ...)]
10:36	Borkdude	somnium: fn as in identity?
10:37	Borkdude	wait
10:37	Borkdude	it just works, I made a type elsewhere
10:38	somnium	Borkdude: O_o
10:38	somnium	,(let [x x] x)
10:38	clojurebot	java.lang.Exception: Unable to resolve symbol: x in this context
10:40	Borkdude	,(let [my-power-list (lazy-cat [2] (map #(* % 2) my-power-list))] (nth my-power-list 512))
10:40	clojurebot	java.lang.Exception: Unable to resolve symbol: my-power-list in this context
10:40	Borkdude
10:40	Borkdude	hmm wtf, why does it work on my machine now
10:41	somnium	Borkdude: guessing you defined a var named my-power-list somewhere
10:42	Borkdude	I think I did a declare somewhere yes
11:00	Borkdude	So the more general question is now: how can I close over a recursively defined lazy sequence?
11:03	Borkdude	so: (let [myseq (lazy-cat [0] (.... myseq .....))] (fn [] (something with myseq)))
11:05	somnium	Borkdude: let isnt recursive, it only shadows previous bindings
11:05	Borkdude	so I can't never do this?
11:05	Borkdude	s/never/ever
11:05	somnium	you can call a function that produces a recursive sequence
11:06	Borkdude	give an example of such a function?
11:07	Borkdude	(fn [] (let [my-seq [0]] (lazy-cat my-seq (.... my-seq)))) ?
11:07	Borkdude	this won't work because I need to lazy-cat over what I am defining
11:09	somnium	,(let [f (fn foo [xs] (lazy-seq (when (seq xs) (cons (take 3 xs) (foo (drop 3 xs)))))) v (f (range 20))] v)
11:09	clojurebot	((0 1 2) (3 4 5) (6 7 8) (9 10 11) (12 13 14) (15 16 17) (18 19))
11:09	somnium	is one (utterly pointless) example :)
11:11	Borkdude	somnium: what does it mean when you do (fn foo [x] ...) instead of just (fn [x] ...)?
11:12	somnium	Borkdude: it provides a self-reference to an anonymous fn
11:12	somnium	Borkdude: otherwise we would need the Y-combinator, and no one wants that :)
11:13	Borkdude	is it similar to recur?
11:13	Licenser	what is a Y-wombinator?
11:13	Licenser	*combinatrr
11:13	Borkdude	Wombinator... is that like a really really Mad Wombat?
11:14	Licenser	:P
11:14	Licenser	yes
11:14	Borkdude	Licenser: this is something from logic or lambda calculus I think
11:17	Borkdude	And also a company by Graham
11:18	somnium	Licenser: its a riddle, what is the result of ((fn [x] (x x)) (fn [x] (x x)))
11:18	Licenser	a tack overflow?
11:18	Licenser	*stack
11:18	Borkdude	Yes, write that down and you have proved a theorem ;-)
11:19	Licenser	it is wombinator for you good sir!
11:21	somnium	http://www.fatvat.co.uk/2009/04/understanding-y-combinator.html
11:21	sexpbot	"Fatvat: Understanding the Y Combinator"
11:22	Borkdude	brb
11:23	Borkdude	will try your (fn name [args]) solution soon
12:00	AWizzArd	~seen hoeck
12:00	clojurebot	hoeck was last seen quiting IRC, 395 minutes ago
12:06	Licenser	tach leute
12:09	bsteuber	dpes lein also download a maven executable? and if, where does it put it? :)
12:10	Licenser	raynes und ich haben etwas gespielt: http://tryclojure.licenser.net/ (meinungne?)
12:11	Licenser	ich will es noch nicht auf #clojure veröfentlichen da noch nicht fertig genug, aber ihr seid ein guter test-kreis :P
12:11	Licenser	sozusagen meine private beta-tester armee
12:13	bsteuber	Licenser: english channel :)
12:13	Licenser	... schesse
12:21	LauJensen	Gents - Anybody know of a similar tool to siege, but with more stats (like mean transaction time) ?
12:30	LauJensen	Ok - JMeter is huge, but ApacheBenchmark is in apache2-utils and does exactly what I want
13:26	LauJensen	Licenser: 'ihr seid' ?
13:37	The-Kenny	LauJensen: "(plural) you are". It was meant for #clojure.de
13:37	LauJensen	okay, thanks
13:37	LauJensen	My german is rusty :\|
14:05	Borkdude	somnium: is there maybe a recursive lazy-cat in contrib, like fn takes a name to make it recursive?
14:27	Borkdude	somnium: ah rec-cat it is called
14:27	Borkdude	,clojure.contrib.seq-utils/rec-cat
14:27	clojurebot	java.lang.Exception: Can't take value of a macro: #'clojure.contrib.seq-utils/rec-cat
14:30	Borkdude	(let [my-lazy-pow (clojure.contrib.seq-utils/rec-cat lz (lazy-cat [2] (map #(* 2 %) lz)))] (nth my-lazy-pow 512))
14:30	Borkdude	,(let [my-lazy-pow (clojure.contrib.seq-utils/rec-cat lz (lazy-cat
14:30	clojurebot	EOF while reading
14:30	Borkdude	[2] (map #(* 2 %) lz)))] (nth my-lazy-pow 512))
14:30	Borkdude	hm?
14:33	Borkdude	,(let [my-lazy-pow (clojure.contrib.seq-utils/rec-cat lz (lazy-cat
14:33	Borkdude	[2] (map #(* 2 %) lz)))] (nth my-lazy-pow 8))
14:33	clojurebot	EOF while reading
14:33	Borkdude
14:33	Borkdude	$(let [my-lazy-pow (clojure.contrib.seq-utils/rec-cat lz (lazy-cat
14:33	Borkdude	[2] (map #(* 2 %) lz)))] (nth my-lazy-pow 512))
14:34	sexpbot	EOF while reading
14:34	Blackfoot	the lines are split
14:34	Borkdude	,(let [my-lazy-pow (clojure.contrib.seq-utils/rec-cat lz (lazy-cat [2] (map #(* 2 %) lz)))] (nth my-lazy-pow 512))
14:34	clojurebot	26815615859885194199148049996411692254958731641184786755447122887443528060147093953603748596333806855380063716372972101707507765623893139892867298012168192
14:34	Borkdude	ah, right
14:48	mfex	Borkdude, are you also attending the amsterdam clojurians meeting?
14:56	Borkdude	ah mfex
14:56	Borkdude	I saw you're post on the Google Group
14:56	Borkdude	s/you're/your
14:56	Borkdude	I don't think so, but I am planning on going there some time
14:57	mfex	I have never been there either
14:57	Borkdude	I read you are finishing your MSc in Twente, I have studied there too
14:58	Borkdude	What is your subject?
14:58	mfex	I'm half way through my final thesis on action semantics for model driven engineering
14:59	mfex	how long ago were you at ut?
14:59	Borkdude	I finished in 2005
15:01	Borkdude	A fellow student of mine is still there, maybe you know him ;-)
15:02	mfex	I started in 03
15:02	Borkdude	Hmm, maybe talk in #clojure-casual? As not to generate too much noise here
15:03	mfex	sure
15:20	giacecco	mmarczyk: Thanks for the hint the other day about apply vs some
15:23	mmarczyk	giacecco: yw
15:23	giacecco	I'm still disappointed apply doesn't work with macros
15:24	Borkdude	giacecco: there is an apply for macros in contrib
15:24	Borkdude	http://clojure.org/libraries#Clojure%20Libraries-Category:%20xml-clojure.contrib.apply-macro
15:25	giacecco	Borkdude: ho thanks, but is there a theoretical reason for that?
15:25	giacecco	Rich must have designed apply that way for a reason
15:26	mmarczyk	actually apply was designed that way about 50 ways ago :-)
15:26	mmarczyk	um
15:26	mmarczyk	years
15:27	mmarczyk	although you could be nitpicky and argue that LISP 1.5 had no macros (I think...?), so maybe it's a bit less than that
15:27	giacecco	Agree, but RH changed stuff when he reckoned it was useful. I'm an idiot, but I can't see the value in not making apply work for both functions and macros in the core language
15:28	mmarczyk	apply simply can't work with macros, for fundamental reasons
15:29	Borkdude	,(type defn)
15:29	clojurebot	java.lang.Exception: Can't take value of a macro: #'clojure.core/defn
15:29	mmarczyk	note that the apply-macro thingamabob in contrib is itself a macro
15:30	mmarczyk	whereas apply is a function
15:30	Raynes	A macro is not a function.
15:30	Raynes	Apply applies a function.
15:30	mmarczyk	hi Raynes :-)
15:30	Raynes	Hi
15:30	giacecco	Probably I just need to study more and I would get it
15:31	giacecco	Someone may then wonder:
15:31	Borkdude	a macro works at macro processor time, it's a different dimension so to say
15:31	giacecco	Why something as basic as logical operators are implemented as macros, e.g. or
15:32	mmarczyk	giacecco: would you want (and false (println :foo)) to print :foo?
15:32	Borkdude	giacecco: it's because macro's don't evaluate their args
15:33	Borkdude	which makes it possible to 'short circuit'
15:33	mmarczyk	actually you could take the view that it is precisely the basic building blocks of your programmes which should be implemented as macros
15:33	mmarczyk	unless they are already built into the language as special forms, like, say, if
15:33	giacecco	Wow that's a strong statement!
15:34	mmarczyk	s/should/must
15:34	mmarczyk	/
15:34	mmarczyk	in fact
15:35	mmarczyk	unless you're using call-by-name pervasively, in which case an "and" function would be perfectly fine
15:35	mmarczyk	the only language I've used which does that is Haskell
15:36	mmarczyk	I suppose PLT's Lazy Scheme might be similar, but I've never used that
15:36	mmarczyk	s/might/is likely to/
15:36	sexpbot	I suppose PLT's Lazy Scheme is likely to be similar, but I've never used that
15:36	giacecco	I believe you're flying too high for me, I'm a humble learner :-)
15:36	mmarczyk	thanks, sexpbot :-)
15:37	mmarczyk	well, try playing around with macros, maybe implement giacecco-and as a function first, then as a macro
15:37	mmarczyk	try emulating the built-in and's behaviour
15:37	giacecco	Last quick one before I need to leave: why can't I use comp on Java functions, e.g. (def integer-sqrt (Math/floor Math/sqrt)) ?
15:38	mmarczyk	with some luck, you might be able to see why and is a macro and why apply doesn't work with macros
15:38	giacecco	I meant (def integer-sqrt (comp Math/floor Math/sqrt))
15:38	mmarczyk	Java methods are not first class objects in Clojure (nor are they in Java, mind you)
15:39	mmarczyk	you can wrap them if you like, though
15:39	mmarczyk	(comp #(Math/floor %) #(Math/sqrt %))
15:40	giacecco	A ha!
15:40	giacecco	I expected something like that
15:40	giacecco	Thanks!
15:40	mmarczyk	np
15:41	giacecco	What's the level of the discussion in the forum: are my questions too naive to be asked here?
15:41	mmarczyk	Licenser: das ist ein schoenes REPL, hoffentlich koennt ihr spaeter def*-Ausdruecke zulassen :-)
15:41	Borkdude	#(...) means "tame this Java code"
15:42	mmarczyk	:-)
15:43	mmarczyk	giacecco: not at all, ask away
15:43	giacecco	Great, I must spend Much
15:43	giacecco	much more time here then
15:43	Licenser	mmarczyk: thanks
15:44	mmarczyk	you're likely to receive more extensive answers on the ggroup, simply because of the format
15:44	giacecco	Thank you all a lot, and goodbye for today! I'll be on the ggroup, too.
15:44	mmarczyk	Licenser: is the code up anywhere? or am I asking a bit too early?
15:45	Licenser	mmarczyk: my gitup page, it is a form of Raynes web stuff but It's really not done yet
15:46	mmarczyk	right, I somehow missed the fine print below the box the first time round
15:46	Licenser	:)
15:47	Raynes	Mine is almost perfectly usable. mmarczyk I'll pass you a link to my own personal tryclojure thingy in a little while.
15:47	mmarczyk	please do, thanks! :-)
15:47	LauJensen	Is there some wonderful lib out there, which could coerce this to a Date or something IComparable "2010-03-30 22:26:45" ?
15:48	mmarczyk	LauJensen: Date/valueOf ?
15:49	mmarczyk	um, actually it wouldn't be Date...argh
15:50	chouser	LauJensen: looks like a comparable string to me
15:50	mmarczyk	,(java.sql.Timestamp/valueOf "2010-04-08 23:11:59")
15:50	clojurebot	#<Timestamp 2010-04-08 23:11:59.0>
15:50	chouser	oh my
15:50	chouser	I wouldn't have thought of looking in java.sql
15:51	mmarczyk	:-)
15:57	LauJensen	mmarczyk: wow, you impressed me :)
16:04	Raynes	mmarczyk: It didn't break yet, did it? :p
16:07	MadWombat	Hello
16:08	MadWombat	I am trying to figure out clojure.contrib.accumulators, what is wrong with (add (empty-counter) "asdf") ?
16:09	MadWombat	oops, got it, sorry, me silly
16:09	MadWombat	empty-counter is a var, not a func
16:09	MadWombat	b
16:18	eshira	is there a non-log(n) version of contains? to be used with vectors?
16:20	eshira	well i guess i'm thinking about the contains? semantics wrong. I want to know if some element x is in some vector v.
16:25	mmarczyk	LauJensen: :-)
16:26	mmarczyk	Raynes: you mean clj-sandbox...? if so, not to my knowledge :-)
16:26	Raynes	mmarczyk: I mean anything. :p
16:26	mmarczyk	Raynes: oh, in that case, I'm not sure
16:26	Raynes	clj-sandbox wont break.
16:27	Raynes	Unless _ato breaks it, that is.
16:27	Raynes	;)
16:27	mmarczyk	something's bound to have been broken recently ;-)
16:27	Raynes	mmarczyk: If it broke, I'm fairly certain you'd know it.
16:27	mmarczyk	well I sure hope I can beat him to it, I was promised a cookie if I manage that :-P
16:28	mmarczyk	the chances seem to be slim, though
16:28	Raynes	Now you have something that isn't sexpbot to break. :D
16:28	mmarczyk	yeah! that's true :-)
16:28	Raynes	I like tryclj because I can write bits of code on my cellphone without having to deal with JavaScript. It's just HTML.
16:30	mmarczyk	at some point having a private environment to do def* in would be nice
16:30	chouser	eshira: no, vectors are indexed by the position of their elements, not by the element values.
16:30	chouser	eshira: so the only way to find an element by value is to walk the whole vector. O(n)
16:30	Raynes	mmarczyk: Do you approve? ;) I think for a first web development project in any language, it's pretty good. :D
16:30	chouser	eshira: if you want to look things up by value, you need a set or a map
16:31	mmarczyk	Raynes: you mean it's your first? in that case, I approve very strongly indeed :-)
16:31	Raynes	:D
16:31	Raynes	<3
16:32	Borkdude	Raynes: I think I missed out on the link
16:32	mmarczyk	Raynes: also, that means you're ahead of me on this one -- no functioning Web project on my record as of yet :-(
16:32	Raynes	mmarczyk: Haha.
16:33	mmarczyk	Borkdude: look for Licenser speaking German above
16:34	Borkdude	mmarczyk: I did, and then what I do?
16:35	mmarczyk	fish out a link
16:35	Raynes	Borkdude: I sent you the link a second ago.
16:35	Raynes	:o
16:35	Borkdude	Raynes: where?
16:36	Raynes	In -casual.
16:36	drewr	should `mvn install` work from a fresh contrib checkout?
16:36	Raynes	I'd rather not unveil it to the entire world until It's closer to finished.
16:36	Borkdude	Raynes: http://joyofclojure.com ?
16:36	sexpbot	"The Joy of Clojure"
16:36	Raynes	Er.
16:37	Raynes	Look in -casual. >.>
16:37	pdk	come to think of it would it still be worth it to sit on my practical clojure preorder
16:37	pdk	now that it got delayed again
16:37	Borkdude	Ah ok
16:37	Borkdude	nice :)
16:39	AWizzArd	Hello Lein experts: how does Lein produce Überjars? a) it first unpacks the existing jars (the dependencies) into a temp folder and includes them in the Überjar, or b) it uses some custom classloader that allows it to put jars into the Überjar directly (without unpacking them first)?
16:42	hiredman	AWizzArd: $5 says the unpacking thing
16:42	AWizzArd	k
16:42	AWizzArd	maybe Ant can do such a thing too, hmm...
16:42	hiredman	sure
16:43	hiredman	ant has a number of jar tasks
16:43	AWizzArd	ok, i will look into those
16:43	AWizzArd	Thx hiredman.
16:44	hiredman	http://ant.apache.org/manual/CoreTasks/unzip.html
16:44	sexpbot	"Unzip Task"
16:45	mmarczyk	yup, it unpacks them
16:46	AWizzArd	ok good, i will try to teach this to my ant script
16:46	mmarczyk	their just funnily named zip files anyway, you can look inside with the archive manager of your choice :-)
16:51	AWizzArd	too bad that there is no native 7zip implementation for Java
16:51	mmarczyk	they're, oh my
17:00	chouser	rhickey: can we rename the "failed" state of an agent to "stopped"?
17:01	chouser	you recover from this state by calling restart-agent, after all, not unfail-agent.
17:01	hiredman	uh, don't they only stop when failed?
17:01	chouser	and then it is less likely to be confused with the :fail error-mode
17:02	chouser	hiredman: right. "stop" isn't currently a word used by the docs. I think I'd like to use "stopped" instead of "failed"
17:05	eshira	chouser: thanks, i'll just do (set my-vector) i suppose
17:08	Raynes	mmarczyk, Borkdude, Licenser: Hah. I forgot to remove the reload middleware before putting up tryclj. It actually works like a normal REPL.
17:14	Licenser	for those who were interestind in swing: http://blog.licenser.net/2010/05/09/the-ease-of-guis#
17:14	sexpbot	"lice! : The ease of gui's."
17:14	mmarczyk	Raynes: you caught it yourself, though, so you get to keep your own cookie :-(
17:15	Raynes	:p
17:15	MadWombat	I am playing with a markov generator again, here my function for generating sentences http://pastie.org/952885, for some reason instead of a list of words, I get a list of letters starting with a capital and ending with a period. What am I doing wrong? Where do the words get split up?
17:15	mmarczyk	Licenser: cool, I'm hoping clj-swing will make it possible for me to create reasonable GUIs at some point
17:15	mmarczyk	for the first time since I my Delphi days
17:15	Licenser	mmarczyk: I really hope that it will too :P
17:15	Licenser	delphi had a incredible editor
17:15	mmarczyk	yup
17:16	mmarczyk	that was totally amazing
17:16	mmarczyk	in fact, if I wanted to do Windows desktop programming, I'd very likely research the possibility of using Delphi to programme the front-end
17:18	mmarczyk	just thinking about a Clojure process doing the actual work with a Delphi-constructed GUI makes me weep tears of hope and joy
17:20	mmarczyk	MadWombat: shouldn't you do (conj result next-word)
17:20	mmarczyk	rather than concat?
17:21	mmarczyk	also, do you use 1.1 or 1.2? you could switch your random choice to rand-nth with the latter
17:21	MadWombat	mmarczyk: no, result is a list of words
17:21	MadWombat	mmarczyk: 1.1
17:21	MadWombat	mmarczyk: considering switching though
17:21	mmarczyk	if result is a list of words and you concat the word onto it
17:21	mmarczyk	then it becomes a list of words with some characters at the end
17:21	MadWombat	hmm
17:22	mmarczyk	so maybe (concat sentence [next-word])
17:22	mmarczyk	but then don't do this
17:22	mmarczyk	you don't want to append to the end of the list repeatedly
17:22	MadWombat	mmarczyk: ?
17:22	mmarczyk	use a vector instead
17:22	mmarczyk	and conj
17:22	mmarczyk	,(conj ['a-list-of-words-goes-here] "next-word")
17:22	clojurebot	[a-list-of-words-goes-here "next-word"]
17:23	MadWombat	mmarczyk: I thought conj was for maps?
17:23	mmarczyk	not at all
17:23	mmarczyk	it's the universal add-to-collection thingamajig
17:23	mmarczyk	for sequences it even figures out where best to put the added item
17:23	mmarczyk	beginning for lists, end for vectors and queues
17:24	MadWombat	ah, conj adds item, concat adds collection
17:24	MadWombat	why vectors instead of lists? I thought adding at the end of a list is efficient
17:24	mmarczyk	assoc is for maps
17:24	mmarczyk	btw
17:24	mmarczyk	just the opposite, actually
17:24	MadWombat	yes, silly wombat, trix is for kids
17:25	mmarczyk	lists are just linked lists
17:25	MadWombat	and vectors are arrays?
17:25	mmarczyk	so you can add to the front at the cost of constructing the front link
17:25	mmarczyk	but if you add to the end, you have to copy
17:25	mmarczyk	(no destructive updates in Clojure)
17:25	mmarczyk	um, no
17:26	mmarczyk	vectors are high-branching-factor trees of arrays with an extra tail portion
17:26	MadWombat	cool, lemme try this stuff out
17:26	mmarczyk	actually I believe the underlying abstract structure would be a trie
17:27	bmason	red black tree iirc
17:27	mmarczyk	you should probably watch one of the cool presentations Rich has done on Clojure data structures :-)
17:27	bmason	those are cool :)
17:27	mmarczyk	indeed :-)
17:34	MadWombat	hmm... seems like the following, surprisingly, works as expected (defn f [x y] (let [x x y y] ... ))
17:34	hiredman	surprising how?
17:34	MadWombat	hiredman: I would expect it to be confused at some point
17:35	bmason	redefining the variables?
17:35	hiredman	bmason: no
17:35	hiredman	you are creating a new scope
17:36	MadWombat	yes, but I am assigning llocal variables with same names as global ones to value of globals
17:36	bmason	right... inner most scope takes precedence... and defining a var to equal itself wouldn't have any effect
17:36	hiredman	locals are not vars
17:36	bmason	what do you call them for ease of terminology?
17:36	hiredman	MadWombat: function are arguments are not globals
17:36	hiredman	bmason: they are locals
17:37	MadWombat	hiredman: yes, sorry, not globals, but they are in the scope
17:37	hiredman	locals have a fixed nested lexical scope
17:37	hiredman	at the top(bottom?) of that scope is the dynamic scope of vars
17:38	MadWombat	basically I expected it to be confused at the point of "the second x, is it the local he just named or the func argument?"
17:38	hiredman	MadWombat: that is just not possible
17:39	hiredman	at the time when the expression whose value will be bound to x inside the scope of the let is being evaluated, the let's scope cannot exist
17:40	hiredman	because the binding of x to that value is part of the scope
17:40	hiredman	because the let's scope doesn't exist, how could you evaluate an expression in it?
17:41	bmason	basically how it gets interpreted is "create a new local called 'x' and assign it the value of the function argument 'x' "
17:41	hiredman	bind
17:41	hiredman	:(
17:42	bmason	ok binding... within the scope of the let, the 'x' defined there takes precendence
17:42	hiredman	scope is like a linked list of pairs
17:42	technomancy	hiredman: did you see my further hacks on serializable-fn?
17:43	hiredman	no, I didn't
17:43	technomancy	hiredman: it can serialize lexical context now =)
17:44	hiredman	cute
17:44	mmarczyk	technomancy: did you get slime-repl from HEAD to work?
17:44	hiredman	as long as you run in the same vm, yeah?
17:45	technomancy	mmarczyk: no, I was hoping I could get hugod to tell me how he fixed it. =)
17:45	mmarczyk	technomancy: funny, it "just works" for me
17:45	technomancy	hiredman: well, as long as it's something that prints decently
17:45	hiredman	let cons a new pair on to the linked list of environments
17:45	technomancy	hiredman: I don't think it would work with reference types, etc.
17:45	technomancy	haven't really searched for edge cases yet
17:46	hiredman	when you resolve a name to a value you walk the linked list and stop at the first entry
17:47	hiredman	,(first (filter #(= 'x (first %)) '((x 1) (x 2))
17:47	clojurebot	EOF while reading
17:47	hiredman	,(first (filter #(= 'x (first %)) '((x 1) (x 2)))
17:47	clojurebot	EOF while reading
17:47	hiredman	bah
17:47	hiredman	,(first (filter #(= 'x (first %)) '((x 1) (x 2))))
17:47	clojurebot	(x 1)
17:49	technomancy	mmarczyk: could you try with the "elpa-2010" branch from my slime fork?
17:49	technomancy	maybe you've got a slightly older/slightly newer version than I happened to grab
17:53	mmarczyk	technomancy: sure
17:53	mmarczyk	will do it now
17:53	technomancy	mmarczyk: oh, and the latest clojure-mode as well
17:54	technomancy	I moved a few things around so now you can use clojure-mode with slime.el directly and skip swank-clojure.el entirely
17:54	mmarczyk	oh, cool
17:54	mmarczyk	does this mean that the whole "coloured REPL" business could go into clojure-mode
17:54	mmarczyk	including the advice stuff for slime-repl?
17:55	technomancy	oh, I haven't looked into the coloured repl code yet
17:55	mmarczyk	in this case, please don't :-)
17:56	mmarczyk	I'll make it fit in well with the new clojure-mode
17:56	technomancy	sorry, I'm slow
17:58	mmarczyk	I was also going to patch leiningen to accept (and add to the pom) website, mailing list and licence info for projects
17:58	technomancy	oh, that sounds nice
17:58	technomancy	do the clojure-mode stuff first though if you don't mind; it's closer to a release
17:58	mmarczyk	ok
17:59	technomancy	not to boss you around or anything =)
17:59	mmarczyk	:-)
18:00	Raynes	mmarczyk: It should be working properly now, if you want to try it.
18:01	Raynes	mmarczyk: Also, you don't have to click the text field every time you submit your code now. That was bugging me.
18:02	Raynes	mmarczyk: I'm going to add syntax highlighting soon. :o
18:02	mmarczyk	Raynes: cool :-)
18:02	mmarczyk	Raynes: I believe I used Enter to submit code, though
18:02	Raynes	mmarczyk: Didn't matter.
18:02	mmarczyk	technomancy: um, what's package-activated-list
18:02	mmarczyk	?
18:03	Raynes	It refreshes the page every time you submit.
18:03	mmarczyk	technomancy: guess I'd have to install ELPA to use your fork, right?
18:03	Raynes	Well, not refreshes. Sort of.
18:04	mmarczyk	Raynes: are you planning paredit-like facilities? ;-)
18:05	Raynes	No. :p
18:05	Plouj	who here takes care of swank-clojure? :)
18:05	Raynes	Plouj: technomancy
18:05	Plouj	oh yeah, right
18:06	Plouj	I'm having some weird connection issues with localhost when running M-x slime
18:07	mmarczyk	technomancy: you might want to make the (when (member 'slime package-activated-list) ...) thing check for the presence of a variable by that name
18:08	mmarczyk	technomancy: with the condition replaced by nil, I can start a SLIME REPL with your fork, with a nice user=> prompt :-)
18:08	mmarczyk	technomancy: which changes to foo=> after (ns foo)
18:25	technomancy	mmarczyk: thanks, will do
18:27	technomancy	mmarczyk: hmm... wonder what's the difference then; it's busted here
18:27	technomancy	mmarczyk: what Emacs version?
18:27	technomancy	Plouj: what's up?
18:27	technomancy	Plouj: do you know that M-x slime is not meant for general use?
18:28	Plouj	technomancy: I get this error a lot of the times "Lisp connection closed unexpectedly: connection broken by remote peer"
18:28	Plouj	technomancy: I thought that M-x slime is good for just opening a repl without associating it with a project. What do you suggest?
18:28	technomancy	well, right, that's what I mean
18:28	Plouj	I'm not familiar with slime so... :
18:28	Plouj	:)
18:29	technomancy	anyway, that is a common problem. I have no idea what causes it, but I know that generally using lein swank or mvn clojure:swank gets around it, which is what I have started recommending
18:29	Plouj	oh
18:29	technomancy	Plouj: I can't reproduce that problem on my own machine, so I can't do much to debug it
18:30	Plouj	how do I use mvn clojure:swank?
18:30	Plouj	just run it in some directory?
18:30	DuneMan	maybe some nice soul will debug it for you :-)
18:30	technomancy	Plouj: run it from within a mavenized clojure project, then use M-x slime-connect
18:30	technomancy	DuneMan: that would be nice. =)
18:31	Plouj	ok, that means I need to create a clojure project
18:31	DuneMan	Too bad don't know how to use emacs at all
18:31	technomancy	Plouj: you can do it from a contrib checkout, I think
18:32	Plouj	what's a "contrib checkout"? :)
18:32	technomancy	unfortunately clojure-maven-plugin has a bug; it doesn't let you set the port
18:32	technomancy	so you can only have one mvn repl running per machine
18:32	Plouj	heh
18:32	technomancy	Plouj: a checkout of clojure-contrib
18:32	DuneMan	that sounds fixable.
18:32	technomancy	DuneMan: it is, but clojure-maven-plugin is ... "lightly-maintained"
18:33	technomancy	it has quite a few of these small easily-fixable bugs that have been around for the better part of a year
18:33	DuneMan	Someone should fork it.
18:33	Plouj	maybe this is a timing issue
18:34	technomancy	Plouj: it smells a bit like a race condition
18:34	DuneMan	I have a feeling I'm gonna keep writing clojure now that I'm up and running
18:34	mmarczyk	technomancy: "GNU Emacs 23.1.50.1 (i486-pc-linux-gnu, GTK+ Version 2.18.0)\n of 2009-09-27 on palmer, modified by Debian"
18:34	technomancy	but yeah, I don't have many details about it
18:34	Plouj	technomancy: do you know what actually prints the "Lisp connection closed" error. Is it your swank code or some slime thing?
18:35	technomancy	mmarczyk: interesting. I guess I will try a fresh .emacs init; maybe there's something in my personal config messing it up
18:35	technomancy	mmarczyk: you don't use the starter kit, do you?
18:35	mmarczyk	technomancy: I could put up a Gist with my slime-related config if you feel that could be useful to you
18:35	technomancy	Plouj: pretty sure it's slime itself, so not my code
18:35	technomancy	mmarczyk: that'd be handy
18:35	mmarczyk	technomancy: no, except as an awesome repository of ideas :-)
18:35	technomancy	mmarczyk: stealing from it is just as good as using it outright. =)
18:36	mmarczyk	technomancy: that's what I figure also :-)
18:36	mmarczyk	technomancy: it's so amazingly clean that it's absolutely my no. 1 elisp learning resource & the thing to emulate stylistically
18:42	Plouj	is swank something separate from clojure and slime?
18:49	Plouj	technomancy: have you tried creating a throw-away user on your unix machine and testing swank-clojure in that? :)
18:52	mmarczyk	technomancy: hopefully this contains all the relevant stuff: http://gist.github.com/395481
18:52	mmarczyk	technomancy: if not, I'll add in whatever's missing, of course
19:02	technomancy	Plouj: it's a client-server model: slime is the client and swank is the server.
19:03	technomancy	most swank servers are written in CL, but obviously swank-clojure is not
19:08	Plouj	oh
19:50	AWizzArd	I created a .jar file and need to start it like: java -jar my.jar one.of.my.Deftypes
19:51	AWizzArd	is there a way to add this one.of.my.Deftypes as a command line arg into the manifest file?
19:52	hiredman	no
19:52	AWizzArd	ok, so if I want to get a double-clickable file I will have to do some other little magic here
19:59	hiredman	gen-class
19:59	AWizzArd	yes
19:59	hiredman	it's the only way to generate a static main, which you need for a clickable jar
19:59	hiredman	you also need to have the right information in your jar manifest
20:00	AWizzArd	the correct Main-Class
21:02	_brian2_	anyone know if/how its possible to use congomongo on a network, ie run a clojure program on my mac using a remote mongo installation?
21:20	_brian2_	I guess I need to read about java sockets
21:31	hugod	technomancy: slime-repl from HEAD works for me, both s-c-p and s-c with lein swank. (slime-fancy doesn't because of slime-autodoc)
21:37	hugod	technomancy: although, sometimes I have to call (setq slime-dispatching-connection nil)(setq slime-default-connection nil) to make connecting reliable
22:45	_brian2_	hi, I would like to give clj-ssh a try. I put [clj-ssh "0.0.1-SNAPSHOT"] in my project file, ran lein deps, open a repl and (use 'clj-ssh) and get : FileNotFoundException: Could not locate clj_ssh__init.class
23:02	hugod	_brian2_: try (use 'clj-ssh.ssh)
23:02	hiredman	brixen: as single segment namespaces are frowned on I would be surprised if that did work
23:12	_brian2_	ok
23:12	_brian2_	thats whats on the website
23:14	_brian2_	yes, that works, thanks!
23:25	hugod	_brian2_: fixed on the website, thanks for the feedback
23:26	defn	hello all
23:29	_brian2_	ok, thanks for your work hugod
23:31	_brian2_	i dont know much about ssh, I want to log into another computer from my mac to access data in a clojure program, do I need to generate a key, fingerprint and all that?
23:32	DuneMan	That's generally the best way to use ssh
23:32	DuneMan	ssh-keygen
23:32	DuneMan	then set up your .ssh-config to use the correct key on the remote computer
23:32	_brian2_	ok so clj-ssh talks to that
23:32	DuneMan	and add the key to the remote computer's access list.
23:33	_brian2_	well, i use ssh routinely to access the computer from my command line
23:33	DuneMan	Anything that uses the unix ssh libs in the background will look into the .ssh directory and use that stuff. Not sure about clj-ssh, guessing hugod will tell you :-)
23:33	_brian2_	ok
23:33	DuneMan	I'd guess at most you'd need to point it to the rsa/dsa file?
23:34	_brian2_	there is some indication of that
23:34	DuneMan	if you needed to do anything more than that, I'd say the lib is broken.
23:34	DuneMan	heh
23:34	hugod	clj-ssh will pick up .ssh/id_rsa by default
23:34	_brian2_	ok, thanks
23:34	hugod	though if you have a passphrase, you'll need to add it explicitly
23:35	_brian2_	is that the same as password?
23:35	technomancy	iirc jcsh does not respect DSA keys, I wonder if clj-ssh has the same problem
23:35	DuneMan	should probably not use dsa anymore anyway
23:36	hugod	technomancy: I've not tried that
23:36	hugod	clj-ssh uses jsch
23:37	technomancy	hugod: thanks for the tip about slime-default-connection etc; will try that
23:37	technomancy	no idea why it's giving me the CL-USER prompt
23:37	_brian2_	hugod : i see you say you are trying to imitate something in the contrib library, why is that?
23:37	_brian2_	is your's more up to date?
23:38	hugod	_brian2_: contrib.shell is just for local sh
23:38	_brian2_	ok
23:39	_brian2_	hugod : if I have more questions, do you mind if I post them to the issues area?
23:39	hugod	technomancy: I have seen that in the past - if I remember correctly, it was some sequencing issue, and disappeared after the first comman
23:41	hugod	_brian2_: sure, if they are issues, otherwise general usage questions can be answered here
23:41	_brian2_	ok
23:42	hugod	if you don't get a response here, feel free to escalate to the issues
23:43	_brian2_	what does that mean?
23:43	_brian2_	exclamation marks?
23:48	DuneMan	technomancy: http://github.com/hugoduncan/clj-ssh/issues
23:48	DuneMan	erps
23:48	DuneMan	_brian2_: http://github.com/hugoduncan/clj-ssh/issues
23:49	_brian2_	hugod : so if I just want to log into a computer is it -> (ssh "user@something.com) but where do I put the password?
23:49	_brian2_	also that doesn't work
23:49	DuneMan	really, you should generate an rsa key and do passwordless login.
23:49	_brian2_	ok
23:49	DuneMan	even if the lib makes it easy to do it the insecure way.
23:50	_brian2_	the error is java.net.UnknownHostException
23:51	technomancy	I heard RSA had patent issues, which is why I always stuck with DSA
23:52	technomancy	that was a while ago; wouldn't be surprised if they expired
23:52	DuneMan	looks like username "blah" as an option.
23:52	hugod	_brian2_: (ssh "somthing.com" "ls" :username xx :password yy)
23:52	DuneMan	technomancy: Ah, I'm not actually aware.
23:53	hugod	_brian2_: or whatever command you want to run - it's not going to give you an interactive shell
23:53	_brian2_	ok
23:54	_brian2_	I would like to access a database and manipulate it using a clojure program, is that possible, or am I on the wrong track
23:54	technomancy	hugod: do you do those 2 setqs before it's connected?
23:55	hugod	technomancy: I run them whenever I have trouble making a connection, and then try again
23:55	DuneMan	tunneling a connection of ssh?
23:55	DuneMan	over*
23:56	DuneMan	One would think the right thing to do would be to use your databases java library to manipulate it
23:56	technomancy	hugod: remind me again why I'm running trunk slime? =)
23:56	DuneMan	and, at most, tunnel the connection over ssh.
23:56	DuneMan	And if you're using a database that doesn't have a good java driver, then you should probably think about joining 2010 and using a real database.
23:56	hugod	technomancy: because you like using sbcl ;-)
23:57	DuneMan	unless there's some reason why that's impractical.
23:57	technomancy	oh right, I am a child of the '80s.
23:57	_brian2_	DuneMan : Im using mongo, the new kid on the block
23:57	DuneMan	Then use congomongo.
23:57	_brian2_	i am
23:57	DuneMan	Then you don't need to do anything with ssh.
23:58	DuneMan	unless you don't want to open the port and want to tunnel the connection over ssh
23:58	DuneMan	I spoke at MongoSF last week
23:58	DuneMan	love it :-)
23:59	_brian2_	well, where does congo mongo sit when I use it that way?
23:59	DuneMan	Tunneling it?
23:59	_brian2_	I mean is it on the client or server
23:59	DuneMan	Congomongo is a client library for speaking to mongodb.
23:59	DuneMan	It wraps the java mongodb library.