#clojure log - Feb 24 2008

The Joy of Clojure
Main Clojure site
Google Group
List of all logged dates

4:22 * baggles tries svn updating now

5:12 baggles: i'm wondering if it's a case of RT.init() needing to be run in the main thread, but it's not happening...

5:22 hm

11:49 hm

11:49 this policy stuff is turning into a pita

11:54 rhickey_: can you do what you need to do with Java Web Start? (although I guess that also has permissions...)

11:56 baggles: erm

11:56 i haven't looked into that

11:56 standalone though

11:56 might as well just distribute a jar file at that rate

11:57 i would really like an applet scenario

11:57 rhickey_: I think Sun is striving to make JWS as slick as Flash

11:58 baggles: hm. but it is for standalone applications right? the target being that you essentially are just ensuring that the client has a (suitable) jvm installed.

12:00 rhickey_: it's not for in-page embedding, but also has zero-install, dynamic loading, incremental updating properties

12:00 baggles: http://java.sun.com/sfaq/#appletSM i'm wondering if this basically means clojure is not applet-able

12:03 hm

12:04 i have a thought actually. those applets for things like menalto's gallery

12:31 bgeron: baggles: if you can make one .jar out of an app, why can't you applet it?

12:31 * bgeron assumes that's possible

12:33 baggles: yeah

12:33 i would have too

12:34 http://hobbiton.homelinux.net/clojuretests/ClojureAppletTest1.html well here's my self-signed applet, but i get that java.security.AccessControlException: access denied (java.lang.RuntimePermission createClassLoader) on trying RT.init();

12:34 rhickey_: yup, it seems the prohibition on applets creating their own classloaders nixes Clojure as applet

12:35 baggles: yeah

12:35 well this thread post seems to confirm that

12:35 http://forum.java.sun.com/thread.jspa?threadID=531684&messageID=2562241

12:35 but it also mentions that maybe you can "install your own security manager (which allows classloader creation) with the signed applet"

12:35 exactly what that means though, i'm not sure.

12:36 i got that, and a few other useful results on googling: applet signed createclassloader

12:38 bgeron: do you need to sign it?

12:38 are you storing files or something?

12:38 baggles: no, but well i was trying to sign it in order to allow more permissions.

12:38 well, i don't know, but signing or not hasn't solved my problems, put it that way

12:38 bgeron: you're not getting those permissions from me o:)

12:38 the user must trust you anyway :)

12:39 baggles: sure

12:39 well i'll show you the source code if you want

12:39 well, it's all in there anyway

12:40 http://hobbiton.homelinux.net/clojuretests/clojuredispatch/DispatchApplet.java

12:40 * bgeron would swear that background is #000, not #333

12:40 baggles: heh

12:41 anyway. i just wanted it !=white so you could see that there was something there

12:41 * baggles has never heard 'nix' used as a verb

12:41 baggles: nor has dict

12:42 * bgeron wouldn't know what it meant

12:42 baggles: oh btw rhickey_ i've commented out the try and finally in RT.init to get around not getting the right error.

12:43 bgeron: could you give the sentence?

12:43 baggles: 17:31 < rhickey_> yup, it seems the prohibition on applets creating their own classloaders nixes Clojure as applet

12:43 bgeron: ah

12:43 in Dutch, 'niks' means 'nothing' so I can guess the meaning

12:44 rhickey_: http://thesaurus.reference.com/browse/nix

12:44 baggles: oh thanks

12:45 hm. this is a real shame. i was hoping i had the ultimate ... and could using clojure with processing

12:50 bgeron: processing what?

12:50 baggles: http://www.processing.org/

16:41 alpheus: when is the :test fn in the metadata of a def form invoked? "uses assert to check various operations"

16:43 rhickey: You can call it whenever, there is a utility fn test which calls it, see (doc test)...

16:44 I presume larger test harnesses will be built atop this

16:53 alpheus: doc doesn't seem to be in this build (20080213)

16:53 rhickey: no, in SVN

17:56 jonathan__: I posted up a couple of listings on jonathanwatmough.com

17:56 All the life has been sucked out of my by trying to make sound/java work in windows

17:57 The documents lie!

19:17 rhickey: jonathan: looks neat! (I'm an audio guy too)

19:42 jonathan__: thanks Rich, I stuck on creating arrays for a bit ... will 'into-array' always create an array of boxes?

19:44 rhickey: yes - the only way to create arrays of primitives is make-array

19:55 jonathan__: it'd be nice if someone could put a sound object browser together that would work x platform

19:56 more of an api explorer I guess

20:49 alpheus: (false? nil) => false

20:50 rhickey: nil is not the value false

20:51 (= false nil) -> false

20:52 alpheus: the section 'nil' at http://clojure.sourceforge.net/reference/data_structures.html leaves me confused, then

20:54 rhickey: how so

20:56 alpheus: (and nil) returns nil, when I expected false. since "nil and false representing the values of logical falsity in conditional tests"

20:59 rhickey: and is defined to return the first logical false value it encounters, in this case it's nil

21:00 alpheus: I'm sorry, I'm looking for the definition of false? I suspect I misread it and it compares for equality to false, not for logical falseness, as I thought.

21:01 rhickey: right, false? is about equality to false, since if already tests for logical truth

21:02 alpheus: thanks, and sorry for the distraction

22:20 Is anyone porting swank to clojure? I'd enjoy doing that.

22:35 rhickey_: not afaik

Logging service provided by n01se.net