#clojure log - Jun 09 2008

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

7:16 lisppaste8: cgrand pasted "VerifyError" at http://paste.lisp.org/display/61971

7:34 rsynnott: finally got around to taking a look at clojure last night; actually very very impressed

9:44 cgrand: rhickey: did you see this bug: http://paste.lisp.org/display/61971?

9:45 rhickey: cgrand: no, I'll look

9:48 cgrand: thanks

10:44 rhickey: cgrand: fixed (rev 899) - thanks for the report

10:54 cgrand: rhickey: you're welcome -- thanls to you for the fix and for Clojure

14:27 nsinghal: e

15:31 meredydd: Hey...rhickey appears to be gone right now, but does anyone have an idea when the next "proper" release is likely to be? I'm happily using SVN, but it would be good to know.

15:37 Chouser_: nope, haven't heard anything.

15:38 I think the fast math stuff was the last new feature he was going to do before the next release, but of course a release needs documentation, packaging, etc.

15:38 Dunno if he's working on any of that or not.

15:48 meredydd: Yeah...it's mostly the docs I'm missing. I'd quite like gen-and-save-class in particular to live in properly-documented-land.

15:50 Chouser_: trying to describe gen-and-*-class clearly is pretty hard. he took a crack at it in (doc gen-class)

15:50 Anything I can help you with?

15:50 meredydd: Ooh. (doc) would be another one I'd love to have read about.

15:57 Chouser_: Sorry about that, did I miss anything?

16:28 la_mer: Interesting thread about Scala in the google group...

16:28 bpattison: do you have a link to it?

16:28 la_mer: http://groups.google.com/group/clojure/t/7677d76db79ef6a1

16:29 bpattison: thanks

16:29 la_mer: We use scala pretty extensively right now, but I've been watching clojure since last fall with great interest.

16:30 rhickey: la_mer: how would you compare their complexity?

16:30 * la_mer is thinking

16:31 la_mer: Taken in whole, I think they're actually pretty comparable, but in oddly unbalanced ways.....

16:32 The overall model in clojure is *very* simple....

16:32 But deploying it and managing it is much more complex (or, simply more difficult) than Scala. Part of that has to do with tooling, and part of that has to do with having a rich, dynamic runtime, etc.

16:33 Being able to code in Scala, and *boom* have a jar of classfiles is about as simple as it gets. Bootstrapping a real clojure app currently requires connecting a *lot* of dots, and rolling a lot of stuff that you get for free in a static environment.

16:35 Of course, that's balanced out by Scala's staggering syntactic complexity, implicits, and all of the arcane stuff that goes along with Type Systems (tm) in general.

16:35 That being said, I am very fond of Scala's syntax, implicits, and a lot of the declarative power you get with a rich type system.

16:36 bpattison: yeah, compared to Clojure, Scala seems syntactically heavy but there are a lot of nice constructs in it

16:39 la_mer: In the end, I'd rather be using Clojure. The open issues at this point include (a) integrating/porting our existing Scala codebase, (b) getting over/addressing the "shipping source code" concern, and (c) evaluating how using clojure will impact our recruiting, future integration efforts/ambitions, etc.

16:40 * la_mer will shut up now :-D

16:41 Chouser_: shipping source is always an interesting issue.

16:42 rhickey: thanks. I think the packaging/deploy issues are tractable

16:43 source hiding another issue - maybe some encryption could be used

16:43 not that bytecode is very secure either

16:43 la_mer: rhickey: to be clear, I'm not entirely sure what the issues *are*. Saying that things are harder than they are with Java/Scala is only one aspect -- ideally, we'd want to get to a point where clojure offers a *better* story than Java/Scala (hot code replace, remote debugging, etc), as we can expect from other lisps.

16:43 Chouser_: rhickey: are you sympathetic with the desire to dump compiled code into a .class file with the idea it's then not meant to be developed, sorta "frozen"?

16:44 la_mer: rhickey: regarding shipping source, we're going down that exact line of thought, too.

16:44 rhickey: la_mer: understood, for now I think the issue is a pull a lever and get a jar

16:44 la_mer: A good piece on the folly of obfuscation from just a few days ago: http://www.matasano.com/log/1055/de-obfuscation-for-the-impatient/

16:45 rhickey: Chouser: no, it would be more like Lisp fasl files, just precompiled but still a dynamic runtime

16:45 la_mer: rhickey: That sounds glorious -- just enough to bootstrap the thing, and then we can still do whatever majick we want to a runtime.

16:46 * rhickey days of avoiding ahead-of-time compilation are numbered

16:47 la_mer: In the meantime, I'm thinking that just scrambling the source so that casual grave-robbers aren't encouraged will be enough.

16:47 rhickey: I have thought about aot-compiling quite a bit, it may rise to the top of the todo list soon

16:47 la_mer: :)

16:48 la_mer: Things are helped a lot by the fact that our next project will primarily be hosted by (a) us, or (b) our customers under significant agreements.

17:08 Lau_of_DK: What cooking? Anyone done anything interesting in Clojure lately ?

Logging service provided by n01se.net