00:14:50 -!- leo2007 [~leo@cpc1-cmbg13-0-0-cust596.cmbg.cable.ntl.com] has left #ccl
00:29:35 -!- nurv [nurv@62.32.144.157] has quit [Read error: Connection reset by peer]
00:30:14 -!- Phoodus [foo@174-22-192-234.phnx.qwest.net] has quit [Read error: Connection reset by peer]
02:13:14 rme [~rme@pool-70-106-132-157.chi01.dsl-w.verizon.net] has joined #ccl
05:33:29 -!- rme [rme@clozure-941A5270.chi01.dsl-w.verizon.net] has quit [Quit: rme]
05:33:29 -!- rme [~rme@pool-70-106-132-157.chi01.dsl-w.verizon.net] has quit [Quit: rme]
09:32:04 leo2007 [~leo@cpc1-cmbg13-0-0-cust596.cmbg.cable.ntl.com] has joined #ccl
09:32:54 <leo2007> Is ccl going to move to git?
10:24:41 nurv [nurv@62.32.129.203] has joined #ccl
10:24:58 <nurv> Hi.
11:06:56 <billstclair> leo2007: I don't know why the pfib speedup isn't more than it is. I was just guessing.
11:07:02 <billstclair> I doubt ccl will move to git any time soon
11:07:31 <billstclair> I like git, but I think I'm the only one of the team, and I do almost entirely contract work, very little work on CCL itself
11:07:33 <leo2007> billstclair: I will weigh in parallel at a later stage.
11:08:33 <leo2007> billstclair: I rather dislike cvs/svn that place a directory .svn/.cvs in every directory.
11:09:17 <billstclair> That doesn't bother me, but I like having my own copy of the whole repository, so I can do everything but push/pull without going over the net
11:09:50 <leo2007> certainly
11:10:07 <billstclair> And git is faster and my feeling, not verified by actual experiment, is that it merges better
11:10:58 <billstclair> But I doubt the switch is beneficial enough to the CCL developers or our community to make it worth the effort
11:28:56 nunb [~nundan@59.178.209.163] has joined #ccl
11:40:01 <leo2007> I'll live with that ;)
11:44:12 <sellout> leo2007: I think the switch is beneficial, but the reason we use svn is because it's included in Apple's developer tools. When Apple starts including a dVCS, that's when we'll start looking at one. If Apple chooses hg or bzr (hah!), that's likely what we'll end up on.
11:45:51 <leo2007> although install git is almost effortless.
11:47:03 <sellout> leo2007: It's not just that, but that Mac devs also tend to be most familiar with svn because of its inclusion.
11:50:08 <billstclair> Never played with bzr. Interesting that their comparison, though obviously favoring bzr, noticeably shows git's speed: http://doc.bazaar.canonical.com/migration/en/why-switch-to-bazaar.html#high-storage-efficiency-and-speed
11:58:20 <leo2007> hehe
11:58:33 <leo2007> there's no reason to use bzr in my view.
11:59:06 <sellout> leo2007: Yeah, from what I can tell, the emacs devs aren't even happy with it  they just have to use it because it's gnu.
11:59:32 <leo2007> some are wishing bzr never existed
12:01:07 <leo2007> although it has a semi-official git mirror http://repo.or.cz/w/emacs.git
12:22:26 -!- nunb [~nundan@59.178.209.163] has quit [Read error: Connection reset by peer]
12:22:55 nunb [~nundan@59.178.214.172] has joined #ccl
12:43:02 -!- nunb [~nundan@59.178.214.172] has quit [Quit: Leaving]
12:44:44 alms [~alms@64.214.179.14] has joined #ccl
12:52:00 -!- alms [~alms@64.214.179.14] has quit [Quit: alms]
13:05:45 -!- nurv [nurv@62.32.129.203] has quit []
13:16:41 alms [~alms@64.214.179.14] has joined #ccl
13:18:19 -!- alms [~alms@64.214.179.14] has quit [Client Quit]
13:21:14 alms [~alms@64.214.179.14] has joined #ccl
13:22:24 -!- alms [~alms@64.214.179.14] has quit [Client Quit]
15:16:03 anRch [~markmilli@64.134.46.153] has joined #ccl
16:23:26 -!- anRch [~markmilli@64.134.46.153] has quit [Read error: Operation timed out]
16:30:06 anRch [~markmilli@ip-64-7-14-173.bos.megapath.net] has joined #ccl
17:16:47 -!- anRch [~markmilli@ip-64-7-14-173.bos.megapath.net] has quit [Quit: anRch]
18:50:17 anRch [~markmilli@64.134.242.124] has joined #ccl
18:58:06 -!- hypno_ is now known as hypno
19:49:20 alms [~alms@146-115-42-237.c3-0.bkl-ubr1.sbo-bkl.ma.cable.rcn.com] has joined #ccl
20:07:48 -!- anRch [~markmilli@64.134.242.124] has quit [Quit: anRch]
20:40:22 brett_h [~brett_h@99-12-194-0.lightspeed.austtx.sbcglobal.net] has joined #ccl
20:41:30 <brett_h> anyone around familiar with asdf-install and clozure?
20:42:55 <leo2007> brett_h: yes,
20:43:28 <brett_h> leo2007: I used asdf-install to install hunchentoot - and right after that I was able to (require 'hunchentoot) and start the test server
20:43:50 <brett_h> leo2007: but when I closed and opened a new session, I can't (require 'hunchentoot) --- what should I look into to understand the uh, "path" system?
20:43:53 <brett_h> I'm new to CL
20:45:13 <leo2007> brett_h: don't use require. I think it is deprecated. Use (asdf:oos 'asdf:load-op :hunchentoot)
20:45:41 <brett_h> leo2007: Error: Error component "hunchentoot" not found
20:45:54 <brett_h> do I need to do something first to init asdf?
20:46:29 <sellout> leo2007: (asdf:load-system :hunchentoot) is shorthand for that.
20:46:47 <leo2007> yeah
20:46:59 <leo2007> I should probably use that instead.
20:47:00 <sellout> brett_h: That error means asdf is loaded, but maybe your asdf:*central-registry* is missing something.
20:47:06 <brett_h> what would I use to load asdf if it weren't loaded other than require?
20:47:56 <sellout> brett_h: Definitely use (require :asdf), but don't require other asdf systems  it's less portable.
20:48:26 <brett_h> oh, ok, to make sure that asdf can use it's own registry and stuff?
20:48:33 <sellout> brett_h: What is asdf:*central-registry* set to?
20:48:36 <brett_h> ("ccl:tools;asdf-install;" (DIRECTORY-NAMESTRING *DEFAULT-PATHNAME-DEFAULTS*))
20:48:49 <sellout> And where is hunchentoot installed?
20:48:51 <brett_h> which I don't really understand, is that a list with a string and another list?  or is the second thing a type?
20:49:01 <brett_h> (sorry, I'm really new)
20:49:18 <sellout> brett_h: It's a form to be evaluated.
20:49:20 <brett_h> I did local install so, ~/.asdf-install-dir
20:49:59 <brett_h> sellout: oh, ok, when evaluated it's ""
20:50:15 <sellout> brett_h: So try (push "/home/brett_h/.asdf-install-dir/systems/" asdf:*central-registry*)
20:50:38 <sellout> (or whatever the actual correct path is  don't expect "~" to expand)
20:51:24 <brett_h> sellout: thanks so much, that did it
20:51:48 <sellout> brett_h: Great  now you might want to add that line to your ccl-init.lisp, so it always happens at startup.
20:51:54 <brett_h> just did :)
20:52:26 <sellout> brett_h: Actually  I'm not sure about this, but (asdf:load-system :asdf-install) might push that path onto your registry.
20:52:45 <brett_h> let me check
20:52:54 <sellout> It would make sense, since asdf-install is putting stuff there, it should make sure you can load stuff from there as well.
20:53:39 <brett_h> hm, that didn't
20:53:58 <sellout> brett_h: Oh well  at least you got it working :)
20:59:09 <brett_h> oh thank god you can rlwrap the REPL :)
21:05:52 <leo2007> any idea why can't I install lift on ccl?
21:05:59 <leo2007> I am getting this Reference to unknown package "MOP".
21:06:51 <brett_h> I had the same problem on sbcl the other day
21:06:55 <brett_h> fwiw
21:07:08 <brett_h> (I haven't tried it on ccl)
21:11:57 <leo2007> thanks for the note.
21:12:32 <brett_h> I wonder if the package is broken or something in general?
21:18:03 <sellout> leo2007: Are you asdf-installing it?
21:18:12 <leo2007> git clone
21:18:12 <sellout> leo2007: I think lift depends on closer-mop.
21:18:28 <leo2007> i tried asdf-install and it failed
21:18:37 <leo2007> and the git clone and failed too
21:18:51 <sellout> leo2007: Same error on both?
21:19:19 <sellout> leo2007: Try asdf-installing closer-mop independently?
21:21:30 <leo2007> sellout: I've installed closer-mop but getting the same error.
21:23:17 <brett_h> I also installed closer-mop on asdf and had the same error
21:23:22 <brett_h> on sbcl*
21:30:27 <leo2007> I give up on lift.
21:30:54 <billstclair> There appear to be only three mop: symbols in utilities.lisp
21:31:15 <billstclair> Import those in packages.lisp, and it will likely work
21:31:25 <billstclair> Though I haven't tried it
21:31:40 <billstclair> They are defined in the CCL package
21:32:55 <billstclair> Strange that he spends effort to import MOP symbols into the package, from an implementation-specific package, but forgot those three
21:34:54 <leo2007> thanks for that.
21:35:08 <billstclair> And, of course, remove the "MOP:" prefix from those three
21:36:09 <leo2007> is it only broken in ccl?
21:36:29 <billstclair> Probably in all the other implementations that don't name their mop package "MOP"
21:36:50 <billstclair> It's very likely that that fix will fix the others too
21:37:26 <billstclair> So, if it works, send it to the maintainer
21:37:56 <brett_h> what does CCL call its MOP package?
21:38:01 <billstclair> CCL
21:38:14 <billstclair> And CCL has only a partial MOP implementation
21:38:21 <billstclair> You need closer-mop to get more of it
21:38:58 <billstclair> It's sad. Had Gregor's MOP book been available when Gail and I wrote CCL's CLOS implementation, back in 1990, it probably would have had the AMOP
21:39:14 <billstclair> But it wasn't, and nobody has updated the implementation for the MOP since then
21:39:34 <sellout> billstclair: Do I hear you saying you want to? ;)
21:39:38 <billstclair> All we had to go from was PCL, which I found completely opaque, so just wrote code from the spec.
21:39:46 <billstclair> Uh, not really
21:40:03 <billstclair> Though it would be a good idea for somebody to do it
21:42:49 <billstclair> It might be fun, if I can find a copy of the AMOP. Used to have the book somewhere, but haven't seen it in years
21:43:44 <billstclair> http://books.google.com/books?id=3X5Gnudn3k0C
21:44:46 <brett_h> billstclair: I just got a copy myself, but I'm very new to CL :)
21:45:14 <brett_h> kinda odd choice for a newbie I guess
21:45:19 <billstclair> heh
21:45:33 <brett_h> I don't expect to get to use CL a lot, I mostly want to use CLOS and learn what I can from it
21:49:03 <billstclair> CLOS is interesting. Most OOP languages are object-centric, but lisp is generic function centric
21:49:13 <billstclair> Can't say I like one or the other better
21:50:05 <brett_h> yeah, it's just fun to learn things that are "very" different
21:50:20 <brett_h> I do python all day, I'd rather learn CL than, say, Ruby
21:50:40 <billstclair> CCL, back in the eighties when it was first written by Coral Software, had Object Lisp, which was object centric
21:51:10 <billstclair> One of my first jobs at Apple was to implement CLOS and change all of MCL's Object Lisp code into CLOS
21:52:34 <billstclair> At least I think Object Lisp was object-centric. Never actually wrote any code in it. Just converted it to CLOS
21:54:59 <brett_h> Apple used CL?
21:55:22 <billstclair> Apple bought Macintosh Common Lisp from Coral Software in 1989 or thereabouts. They sold it to Digitool around 1995
21:55:33 <brett_h> interesting
21:55:55 <billstclair> Gary Byers got the rights to the non-GUI stuff, converted it to run in Linux, and branded it "OpenMCL"
21:56:13 <brett_h> oh, and that's Clozure?
21:56:21 <billstclair> A couple of years ago we reverted to CCL. Originally, it was "Coral Common Lisp". Now it's "Clozure Common Lisp".
21:56:21 <brett_h> I mean, the father of -
21:56:26 <billstclair> The CCL package has been there all along
21:56:32 <brett_h> huh
21:56:56 <sellout> Coral CL -> MCL -> OpenMCL -> Clozure CL
21:56:59 <billstclair> Clozure is a small partnership, including Gary Byers. I work for them.
21:57:04 <billstclair> Thanks, sellout
21:57:19 <brett_h> so is it a support company?
21:57:25 <sellout> billstclair: Did I miss anything? I only arrived for the last name change ;)
21:57:37 <billstclair> That's it, sellout.
21:57:52 <billstclair> Clozure maintains CCL, and does Lisp programming contracts
21:58:17 <billstclair> We also sell iPhone applications for pre-schoolers, under the name "Learning Touch"
21:58:32 <brett_h> originally written in CCL?
21:58:45 <sellout> brett_h: We wish.
21:58:47 <billstclair> No. The Learning Touch apps are plain-old Objective C
21:58:55 <brett_h> do you like Obj-C?
21:59:06 <billstclair> Not much. It's useable, but a bit odd
21:59:15 <sellout> brett_h: I think it's cool, but I'd rather write Cocoa in CCL.
21:59:25 <billstclair> And though they've added garbage collection to the language, the iPhone version doesn't have it
21:59:33 <brett_h> right
21:59:37 <brett_h> I dabbled with a tiny app, it just made me want to check out smalltalk
21:59:54 <sellout> billstclair: Yeah, that's a killer. Just as I get excited about ObjC 2.0, the iPhone comes out without all the cool bits.
21:59:57 <brett_h> clearly this is my geek in me talking, not the businessman :)
22:00:26 <billstclair> Yep. We make noticeable $$ on $2 iPhone apps for 3-year-olds
22:02:11 <billstclair> And then Apple comes out with their Adobe killing app programming tool policy, making it less obviously financially viable to create CCL for iPhone
22:02:21 <brett_h> yeah
22:02:29 <brett_h> Android is growing, who knows
22:02:57 <billstclair> Yes. Android is a much more open platform. It also doesn't yet have nearly the profitability as iPhone apps
22:03:10 <brett_h> for sure
22:03:24 <brett_h> the culture they're growing (and customers they get) seem very different so far
22:03:32 <brett_h> $1 for an app?! crazy!
22:03:35 *brett_h* sighs
22:04:05 <billstclair> I blame my friend Leor Zolman, who wrote the BDS C compiler back in the eighties and was the first to charge no run-time license for a development system
22:04:30 <billstclair> Not that GNU wouldn't have happened anyway...
22:04:35 <billstclair> And I like GNU
22:04:35 <brett_h> hehe
22:04:58 <brett_h> I believe in open source until the top of the stack, if that makes any sense (and this is of course a generalization)
22:05:08 <billstclair> But even a $1/copy runtime license on apps would make lots of money for a development tool vendor
22:05:20 <brett_h> use your python, your apache, your linux and build a business on it, share the rest (as it isn't worth $ to you) with other devs working on other businesses
22:05:36 <brett_h> I don't get people who buy a $5 coffee and roll their eyes at a $2 phone app, though :)
22:05:53 <billstclair> Instead, we give away the programming tools, and have to contract out application software to stay in business
22:06:21 <billstclair> I don't get that either. I'm happy to pay a couple of bucks, just to keep the developer in business
22:06:30 <billstclair> It's an incredible bargain
22:06:38 <brett_h> yep
22:06:58 <billstclair> I feel the same way about music, though I know that the artists see very little of that revenue stream
22:08:54 <brett_h> is there not a CCL based product you guys could create?
22:09:05 <brett_h> seems like if you're pros at it, and it's powerful, product is the way to go
22:10:21 <billstclair> Gotta go. My son is singing in his college choir tonite
22:10:49 <brett_h> thanks for the CLOS/MOP info :)
22:10:50 <brett_h> have fun
22:10:55 <billstclair> k