00:00:00 --- log: started forth/06.01.23
00:08:20 --- join: mjb1 (n=mjb1@88.202.134.119) joined #forth
00:10:26 --- part: mjb1 left #forth
00:11:07 --- join: mjb1 (n=mjb1@88.202.134.119) joined #forth
00:11:26 --- part: mjb1 left #forth
00:36:25 --- join: swalters_ (n=swalters@6532183hfc82.tampabay.res.rr.com) joined #forth
00:36:32 --- quit: swalters_ (Remote closed the connection)
02:11:54 --- join: Cheery (i=Henri@a81-197-18-99.elisa-laajakaista.fi) joined #forth
02:21:33 --- quit: Cheery (Read error: 104 (Connection reset by peer))
02:24:44 --- join: Cheery (i=Henri@a81-197-18-99.elisa-laajakaista.fi) joined #forth
03:46:56 --- quit: aum ()
04:22:04 --- quit: Cheery ("Leaving")
04:33:10 --- join: tathi (n=josh@pdpc/supporter/bronze/tathi) joined #forth
05:05:32 --- quit: tathi ("leaving")
05:09:49 --- join: PoppaVic (n=pete@0-1pool46-124.nas30.chicago4.il.us.da.qwest.net) joined #forth
05:52:06 --- join: madwork (n=foo@derby.metrics.com) joined #forth
05:58:09 --- join: rehges (n=segher@blueice2n1.de.ibm.com) joined #forth
06:10:03 --- join: swalters_ (n=swalters@6532183hfc82.tampabay.res.rr.com) joined #forth
06:10:29 --- quit: swalters_ (Client Quit)
07:25:42 --- join: tathi (n=josh@pdpc/supporter/bronze/tathi) joined #forth
07:51:23 --- quit: madwork (Read error: 110 (Connection timed out))
07:53:15 --- join: madwork (n=foo@derby.metrics.com) joined #forth
08:04:32 --- quit: PoppaVic ("recycles")
08:06:11 --- join: PoppaVic (n=pete@0-1pool46-72.nas30.chicago4.il.us.da.qwest.net) joined #forth
08:17:30 --- join: sproingie (n=chuck@64-121-2-59.c3-0.sfrn-ubr8.sfrn.ca.cable.rcn.com) joined #forth
08:42:30 --- join: tattrdkat (n=virsys@or-65-40-181-13.dyn.sprint-hsd.net) joined #forth
08:47:19 --- quit: virsys (Connection timed out)
09:11:32 --- quit: PoppaVic ("Pulls the pin...")
09:14:46 --- join: Charnel (n=chatzill@cm-216-185.kabelnet.hu) joined #forth
09:15:08 <Charnel> hello everyone
09:15:15 <tathi> hi
09:15:24 <Charnel> im new to forth
09:15:39 <tathi> ah.  what forth are you using?
09:15:43 <Charnel> gforth
09:16:04 <Charnel> on win32
09:16:18 --- join: Cheery (i=Henri@a81-197-18-99.elisa-laajakaista.fi) joined #forth
09:16:37 --- quit: Charnel (Client Quit)
09:17:03 <tathi> huh.  that was quick.
09:43:02 <Quartus> Just a passing interest. :)
09:46:20 <JasonWoof> heh
10:23:31 <I440r> lol
10:23:48 <I440r> could be connection problems lol
10:23:52 <I440r> or maybe he got scared 
10:23:55 <I440r> ya scared him off!
10:23:59 <I440r> you bastards!
10:24:46 <tathi> :)
10:29:38 <Quartus> tathi is pretty terrifying! :)
10:30:01 * tathi levels a terrifying scowl at Quartus ;)
10:31:57 <docl> I'm s-s-scared! these forthers look meeeean :P
10:35:03 <I440r> stop teasing him, he might brows the logs and never come back
10:35:05 <I440r> :/
10:35:15 <docl> sorry
10:35:19 <I440r> heh
10:35:19 <I440r> lol
10:35:27 <I440r> it wasnt just you, i started it :P
10:35:29 <docl> it probably was connection problems :P
10:36:08 <rehges> but everyone here _is_ mean and scary, no?
10:36:12 <docl> we are nice people! we like new people to learn forth too!
10:36:44 <rehges> that too -- but we _are_ scary!
10:36:50 <rehges> well i am, anyway :-)
10:37:00 <docl> ok I'll give you that :)
10:40:05 <Quartus> I'm personally quite threatening, I can see why he left. :)
10:40:59 * JasonWoof squints at Quartus suspiciously
10:41:26 * JasonWoof mumbles "I don't know about that guy... something to do with the number four..."
10:44:16 <rehges> jasonwoof: do you know how to make a cat do "woof" btw?
10:44:45 <rehges> gasoline.  matchstick.  woof.
10:45:24 <JasonWoof> I love cats
10:45:37 <rehges> how to make a dog go "meow" then?
10:45:38 <JasonWoof> I'll send you my recipies sometime ;)
10:45:42 <rehges> hehehe
10:45:56 <rehges> freezer.  circular saw.  meoooooow!
10:46:31 <JasonWoof> hmm... you're right. you scary :)
10:53:12 <rehges> i'm not the one suggesting eating cats!
10:53:20 <rehges> [well not after burning them, anyway]
10:57:09 --- join: Lars_G (n=lars@unaffiliated/lars-g/x-000001) joined #forth
10:57:24 * Lars_G bows
10:58:24 <Lars_G> I am sorry if this channel is formal, I don't know it since I am new, please dissregard my stumbles as those a cub would do.
10:58:55 <rehges> cub?  circular saw?
10:59:00 <rehges> :-)
10:59:04 <Lars_G> Little dog? hehe.
10:59:08 <Lars_G> Nice to meet you all.
10:59:11 <rehges> formal here?  no way :-)
10:59:24 <Lars_G> Anyhow, Hi all, name's Lars and I am learning Forth.
10:59:40 <Lars_G> I looked at it out of curiosity to be able to do more with the OpenFirmware on my new Mac, and fell in love
10:59:43 <Quartus> H Lars.
10:59:51 <Lars_G> Quartus: Uh oh. Hello.
11:00:10 <Quartus> well, "Hi" is what I meant.  Why the 'uh oh'?
11:00:13 <Lars_G> Quartus: I better shut down now. I was going to ask if there was a cheaper option to Quartus Forth for the palm for someone who is learning.
11:00:20 <Lars_G> I don't think it's such a good idea now ;)
11:00:22 <tathi> heh
11:00:24 <Quartus> There sure is -- the evaulation version is free.
11:00:36 <Lars_G> for a limited time?
11:00:38 <Quartus> No.
11:00:40 <Quartus> Unlimited.
11:00:48 --- quit: rehges ()
11:00:48 <Lars_G> Good it'll do perfectly then.
11:00:59 <Lars_G> Since as long as I am not making commercial software, I dont need self compiled prcs.
11:01:03 <Quartus> Should you happen to be a student or educator, I offer a discount.
11:01:04 <Lars_G> thanks :)
11:01:20 <Lars_G> I always study, but I'm not a student. Thanks anyhow.
11:02:01 <Quartus> Let me know if you have any questions.
11:02:09 <Lars_G> Well I have only one for the future.
11:02:32 <Lars_G> are there/will there be, native compiling for 5.0 arm apps in Q.F. now/in the future?
11:03:05 <Quartus> There's no way to make a pure ARM app under PalmOS5 now, so future considerations will come with future OS releases.
11:03:13 <Lars_G> ok
11:03:38 <Lars_G> Thanks.
11:03:54 <Quartus> Right now there's an ARM assembler that ships with Quartus Forth, so if you have something CPU-intensive to do you can do it that way.
11:04:04 <Quartus> Though why you'd be doing high-horsepower work on a Palm, I'm not sure. :)
11:04:21 <Quartus> There's a full demo of writing an ARM subroutine.
11:04:24 <Lars_G> So I guess my first question on the long list is, could you guys give me some hot examples of Forth being used today? Other than OpenFIrmware? I want to slap examples back at people who look odd at me when I mention forth
11:04:32 <I440r> quartus, why do mountaineers climb mountains?
11:04:36 <I440r> "because they are there"
11:04:40 <Lars_G> Quartus: Well that's true.
11:04:47 <Quartus> Lars, if they're giving you odd looks, they won't be impressed with anything you offer.
11:04:48 <Lars_G> I440r: True as well :P
11:04:55 <Lars_G> Quartus: It's possibly so.
11:04:56 <I440r> heh
11:04:58 <Lars_G> saddly.
11:05:16 <I440r> lars check out my forth compiler at htto://isforth.clss.net - theres alot of example forth in there
11:05:22 <Lars_G> Altough I'll personally try to limit forth so far to the "embedded" side I develope as hobby.
11:05:27 <Quartus> The only way to impress any such person is to express great interest in their favourite language.
11:05:33 <I440r> also, nasa uses forth. the company i just finished a contract with uses forth to do sattelite tracking systems
11:05:38 <Lars_G> Is the creator of AVRForth a regular here too?
11:05:50 <I440r> no but he should be :P
11:06:19 <Lars_G> Quartus: Nod.
11:06:29 <Lars_G> Anyhow I am just learning
11:06:40 <I440r> im always impressed when someone takes up forth
11:06:45 <Lars_G> I only know about 3 ans words so far and how to construct new words.
11:06:48 <I440r> even if its NOT my forth!!! hheh
11:07:18 <Quartus> Lars, which 3? :)
11:07:32 <Lars_G> Quartus: . .s ."
11:07:54 <Quartus> That's all you ever need, the rest is just for show. :)
11:07:55 <I440r> if you can construct new words then you know : and ; too
11:08:00 <Lars_G> hahaha, yeah
11:08:13 <I440r> : new-word 10 0 do stuff here loop ;
11:08:18 <Lars_G> I440r: Yes I know how to construct words, I just didn't consider : a word there, my bad
11:08:34 <I440r> if its not a word its either a number or a typp
11:08:36 <I440r> typo
11:08:40 <Lars_G> hahahaha true
11:09:03 <I440r> there are two things in forth. words and numbers. everything else is a fubar :)
11:09:29 <Quartus> There's also forward-parsed strings.
11:09:42 <I440r> ." is a word
11:09:51 <I440r> (.") is the word ."  compiles
11:09:53 <Quartus> But the text following it up to the " delimiter is not.
11:10:06 <Quartus> Neither is it a number.
11:10:07 <I440r> ok well then theres words and numbers and "data"
11:10:27 <I440r> my bad :)
11:10:50 <Lars_G> Numbers are data too.
11:10:55 <Lars_G> so, there are words, and data
11:10:57 <Quartus> "There are THREE things in Forth.  words, numbers, data, and whitespace.  Ok, FOUR things. There are FOUR things in Forth. :)"
11:11:15 <I440r> well numbers are special, they go on the stack
11:11:18 <I440r> data usually doesnt
11:11:20 <Lars_G> true
11:11:29 <I440r> tho... some peiole seem to think a string stack is a good idea
11:11:33 <I440r> i call it featureitis
11:11:40 <Lars_G> I don't like the idea
11:11:57 <Quartus> words, numbers, data, and delimiters.
11:12:27 <Lars_G> I440r: Aahhh I'm sorry I don't use your forth :P
11:12:31 <I440r> ok, were splitting hairs now :)
11:12:34 <Lars_G> I use GForth and pForth
11:12:41 <I440r> Lars_G, you should!!
!!!
11:12:41 <Lars_G> so far
11:12:48 <Lars_G> why?
11:12:54 <I440r> actually, if you have linux/x867 boxes isforth can help you learn
11:13:03 <I440r> my sources are very well commented
11:13:12 <Lars_G> Quartus: Now that I know you're a human, maybe SOMEDAY I'll buy Q.F. :)
11:13:20 <Quartus> Lars, it'd be much appreciated!
11:13:27 * Lars_G bows to Quartus 
11:13:44 <Lars_G> Please understand I do not now, since forth will not be a source of income. And my budget is tight.
11:13:46 <I440r> i was also working on fully documenting it too at the clss.net url
11:13:56 <Quartus> Not to worry, Lars. I understand.
11:13:59 <Lars_G> I440r: Ok, I'll look to it.
11:14:12 <Lars_G> I440r: Is it pure C, or a C core with the rest in Forth like GForth?
11:14:13 <I440r> your not obliged to but it might help :)
11:14:22 <I440r> and if you have a question you can email me :)
11:14:28 <I440r> no
11:14:35 <I440r> its a 100% pure assembler core
11:14:42 <I440r> and uses ZERO external libraries
11:14:58 <Lars_G> ok
11:15:04 <Lars_G> and the rest in forth?
11:15:13 <tathi> yes
11:15:22 <I440r> it doesnt matter if you dont code x86, there are ppl in here who dont coe x86 who have no problems reading the kernel sources :)
11:15:25 <Lars_G> So linux
11:15:27 <I440r> all the extensions are in 100% pure forth
11:15:37 <Lars_G> I440r: I hate x86, I love RISC archs
11:15:46 <Lars_G> I440r: But I've dibbled in x86 assembly.
11:15:47 <I440r> im trying to learn arm atm
11:15:55 <Lars_G> arm is a decent arch.
11:16:03 <I440r> i actually like the x86 arch, warts and all 
11:16:03 <I440r> lol
11:16:07 <Lars_G> Ewww
11:16:07 <JasonWoof> hmmm... interesting to see you discussing the different things you see in forth source
11:16:09 <Lars_G> nyet.
11:16:25 <JasonWoof> in my system, the compiler only sees two types: words and numbers
11:16:27 <Lars_G> I can't stand beign so tight in GPRegs
11:16:50 <I440r> heh consider it a challenge!
11:16:54 <I440r> i like a good challenge :)
11:17:04 <I440r> register starvation makes you think about your code more :)
11:17:06 <Lars_G> Altough it makes sense. In a form the stack dependence of x86 makes it slightly similar to forth, in that sense
11:17:08 <JasonWoof> but the editor knows the types of the numbers which can be: literals, constants, variables or data
11:17:38 <Lars_G> JasonWoof: So the compiler only needs words and numbers. And whitespace
11:17:56 <I440r> ok, i gotta go do laundry
11:17:57 <I440r> grrr
11:18:03 <I440r> i HATE doing laundry
11:18:09 <JasonWoof> Lars_G: I'm doing funny things... I don't store source in text form
11:18:19 <Lars_G> JasonWoof: How do you store it?
11:18:22 <JasonWoof> Lars_G: the only whitespace is in the editor display
11:18:34 <Lars_G> are you cooping words into IDs?
11:18:45 <JasonWoof> Lars_G: I store a "token" for each word that contains a dictionary index, and a few bits to say what to do with it
11:18:52 <Lars_G> nod
11:18:56 <Lars_G> Not a bad idea at all
11:19:25 <JasonWoof> the bits would indicate: 0) execute it  1) compile it  2) postpone it  3) define it etc
11:19:32 <Lars_G> hmm
11:19:45 <JasonWoof> makes the compiled really simple because it doesn't have to parse
11:19:46 <Lars_G> How about extracting definitions?
11:19:50 <JasonWoof> or maintain any state info
11:20:00 <JasonWoof> I store each definition seperately
11:20:15 <Quartus> He also doesn't pay taxes or wear clothes. 
11:20:16 <JasonWoof> the dictionary entry (which is persistant) links to the definition
11:20:21 <Lars_G> Yes you'd need to parse for that, but I'd extract definitions to a separate table, possibly part of the dictionary, and thus save these bits
11:20:28 <JasonWoof> Quartus: hey, I wear clothes sometimes!
11:20:35 <JasonWoof> hehe
11:20:56 <JasonWoof> Lars_G: I don't follow
11:21:05 <Lars_G> Well I pay taxes, when I buy stuff
11:21:07 <JasonWoof> I really don't do any parsing
11:21:12 <Lars_G> ok
11:21:15 <JasonWoof> except as you type in the editor, if that counts
11:22:22 <JasonWoof> I like the "color" bits because they eliminate state
11:22:45 <JasonWoof> and any sort of forward-parsing
11:22:49 <Lars_G> ok
11:23:10 <JasonWoof> I like how everything is left to right in my forth
11:23:22 <JasonWoof> no creating words
11:23:24 <JasonWoof> etc
11:23:31 <Lars_G> So every forther here has his own forth?
11:23:47 <JasonWoof> not all :)
11:23:53 <JasonWoof> I440r, Quartus, and I do though
11:24:21 <JasonWoof> slava does too. I wouldn't be suprised if others did too
11:24:43 * Lars_G gaps at slava 
11:24:46 <JasonWoof> forth seems to attract tinkerers
11:24:51 <Lars_G> slava: I never tought I'd find you there. How do you do?
11:24:58 <Lars_G> JasonWoof: Well I'm a tinkerer myself, yes
11:26:14 <Lars_G> Does there exist any non deprecated forth hardware in the wild yet?
11:26:27 <JasonWoof> don't think so
11:26:50 <JasonWoof> you can buy a dev board with something based off Chuck Moores design
11:26:59 <JasonWoof> Ignite
11:27:18 <JasonWoof> afaik there really isn't much out there
11:27:35 <JasonWoof> kc5tja is building something
11:28:26 <Lars_G> Interesting
11:28:48 <Lars_G> It should be possible and not too extremely hard to make a Forth cpu off a fpga
11:28:53 <JasonWoof> I'm focusing on software for now
11:29:04 <JasonWoof> sometime later in life I'll delve into hardware
11:29:06 <Lars_G> !
11:29:08 * Lars_G bows to I440r 
11:29:10 <JasonWoof> I don't have the money for it now
11:29:23 <JasonWoof> yeah
11:29:28 <JasonWoof> there have been fpga designes
11:29:43 <JasonWoof> I think the tough part is generating video and communicating with other computers
11:29:52 <JasonWoof> periferals, etc
11:30:11 <Lars_G> Nod.
11:30:22 <Lars_G> Altough code for that is already avaliable.
11:30:24 <Lars_G> www.opencores.org
11:30:38 <JasonWoof> yeah, I understand there are fpga designs available
11:30:43 <Lars_G> the tough part would be integrating it into Forth seamessly and in Forth manner.
11:30:48 <JasonWoof> the trouble is making the rest of the computer
11:31:55 <Lars_G> it's bothersome but not real trouble. the knowledge is scattered all around.
11:32:06 <Lars_G> all you require is patience. specially if you want to use as little money as possible
11:32:58 <Lars_G> What's the channel's rule on swearing?
11:35:02 <JasonWoof> don't think anybody minds particularly.
11:35:06 <JasonWoof> this is logged though
11:35:59 <Lars_G> No trouble.
11:36:13 <Lars_G> My exclamation was, Holy Shit, Moore even visited the channel! incredible.
11:36:21 <JasonWoof> :)
11:36:49 <JasonWoof> yeah, we had a nice interview with him
11:36:59 <Lars_G> Was that one logged as well=
11:37:00 <Lars_G> ?
11:37:04 <JasonWoof> yes
11:37:18 <Lars_G> Do you have an url please?
11:37:25 <Lars_G> So let me get this well. You're the hippie.
11:37:37 <Lars_G> Quartus is the commerciant.
11:37:43 <Lars_G> I440r is the engy......
11:39:03 <Lars_G> ... sorry?
11:39:17 <Lars_G> I sometimes, say stuff I shouldn't.
11:39:18 <JasonWoof> engy?
11:39:29 <Lars_G> engineer, IDK don't know him well enough yet.
11:43:10 <Lars_G> I'll go silent until I have something nice to say
11:44:02 <JasonWoof> huh... forth log of chuck interview seems to have fallen off... found it in a zip archive though
11:45:06 <Lars_G> I accept dccs if you're not inconvenienced
11:45:13 <Lars_G> or email, or im.
11:45:57 <JasonWoof> ok, now I'm hosting a copy: http://jasonwoof.com/downloads/forth_channel_chuck_moore_interview_02-05-04.txt
11:46:42 <Lars_G> Ok thanks
11:47:30 <Lars_G> I440r: Tell me when you're back from laundry
11:48:30 <JasonWoof> there's a few forthers around here that seem to get enough done, or stick to the same thing long enough to have working forth distrobutions and actual users
11:48:46 <JasonWoof> namely Quartus, crc, slava and I440r 
11:49:37 <Lars_G> nod
11:49:59 <JasonWoof> I hope to join them in those qualifications within the next year
11:54:29 <JasonWoof> I created a very usable forth system with an editor and everything (herkforth)
11:54:48 <JasonWoof> but I started over in August
11:56:11 <JasonWoof> I think the biggest barrier to getting a comunity around development in my system will be merging changes
11:56:23 <JasonWoof> I don't have any way to merge changes
11:56:34 <JasonWoof> or know how that will work
11:57:10 <JasonWoof> guess I'll have to write some sort of version control system
12:02:49 --- join: binaryguy (n=Administ@host86-131-188-193.range86-131.btcentralplus.com) joined #forth
12:03:13 --- part: binaryguy left #forth
12:03:28 <Lars_G> JasonWoof: Why not use one that exists?
12:08:31 <JasonWoof> I'm not happy with the development cycle
12:08:37 <Lars_G> How so.
12:08:40 <Lars_G> I440r: back?
12:09:05 <JasonWoof> edit, save, compile/run, quit edit save compile, run, test, quit, edit.
12:09:17 <JasonWoof> etc
12:09:40 <JasonWoof> basically I feel like I can increase programmer efficiency dramatically
12:10:36 <JasonWoof> also I can't stand wasting time searching around for the definitions of things
12:11:54 --- join: snoopy_1711 (i=snoopy_1@dslb-084-058-110-240.pools.arcor-ip.net) joined #forth
12:13:06 --- quit: Snoopy42 (Nick collision from services.)
12:13:19 <Lars_G> I agree and dissagree at the same time
12:13:24 --- nick: snoopy_1711 -> Snoopy42
12:15:29 <JasonWoof> I'm very picky
12:15:40 <JasonWoof> also I want a system that I can edit completely
12:16:52 <JasonWoof> I've integrated the editor and "interpreter" so you can just as easily execute something as enter it into the source
12:19:55 <JasonWoof> this speeds things up for me
12:20:07 <JasonWoof> word-not-found errors happen at edit time
12:20:21 <JasonWoof> looking up definitions of things is a total breeze
12:20:38 <JasonWoof> and I can test things I'm writing immediately without leaving the editor
12:20:52 <JasonWoof> I just hit a couple keys to save and compile the current page
12:21:27 <JasonWoof> in my rewrite (Mist) I'm making compiling automatic
12:21:37 <JasonWoof> but you'll still have to hit a key to save frequently
12:24:10 <Lars_G> Yes, my trouble is the idea of mixing that with the versioning system
12:24:51 <Lars_G> because
12:25:09 <Lars_G> a) People do not code 24/7, that makes realtime versioning less interesting.
12:25:31 <slava> hi Lars_G 
12:25:37 <slava> i'm here because i develop a forth-like language
12:25:41 <Lars_G> b) Versions are generally applied to blocks of change, a realtime versioning will apply to atomic changes making evrsion traveling more difficult, lots mor eentries, less well defined.
12:25:47 <slava> and also because i follow several projects like retro and herkforth
12:25:50 <Lars_G> slava: And I'm here because I fell in love with forth
12:25:57 <slava> heh
12:26:03 <slava> try retroforth, its very clean
12:26:26 <JasonWoof> Lars_G: yeah, I haven't worked out the versioning stuff yet
12:26:32 <Lars_G> c) The iteraction of two programmers working on the same file using a realtime versioning is difficult... can you say toe stepping ^10?
12:26:38 <slava> JasonWoof: i like having word not found errors be edit time too
12:26:41 <slava> JasonWoof: not that i've implemented that yet
12:27:02 <slava> Lars_G: in factor, i just use source files. it works with cvs fine :)
12:27:06 <JasonWoof> Lars_G: I don't plan on doing realtime versioning
12:27:09 <Lars_G> slava: DO you maintain a forth mode for jedit too? ;)
12:27:18 <slava> Lars_G: there's retroforth syntax highlighting in jedit
12:27:28 <slava> and a factor plugin with real time syntax checking and code completion
12:27:46 <JasonWoof> hehe :) I coded tab-completion in herkforth
12:27:55 <slava> its coolest feature is you can select some code, do 'extract word', type a name for a word, and have occurrences of that pattern be replaced with a new colon definition that it adds
12:27:59 <Lars_G> tab completion is confortabl
12:28:00 <Lars_G> e
12:28:16 <Lars_G> slava: WHOA
12:28:17 <slava> JasonWoof: does your forth have a help system?
12:28:23 <Lars_G> slava: Now THAT is a nice feature!
12:28:40 <JasonWoof> slava: cool
12:28:53 <JasonWoof> slava: no. just a comment for each word
12:29:04 <JasonWoof> well, some words don't have comments yet
12:29:16 <JasonWoof> s/some/most/
12:29:19 <Lars_G> SO many forths, so little time
12:29:53 <slava> JasonWoof: i moved my doc comments into separate .facts files, with a styled text markup language. its still factor, a DSL
12:30:10 <slava> there's a gui to browse it and i'm working on an http server responder
12:30:10 <JasonWoof> slava: I saw a screenshot of your styled help stuff
12:30:20 <slava> you'll be able to try a web version in a few hours :)
12:30:26 <JasonWoof> heh
12:30:41 <slava> it already works, i need to fix some formatting issues in the html translation
12:30:51 <JasonWoof> I'm hoping to somehow integrate a documentation wiki with mist
12:30:56 <slava> its much more manageable than xslt, i have to say
12:31:00 <slava> ever used xslt?
12:31:05 <slava> terrible stuff
12:31:14 <JasonWoof> yeah
12:31:18 <JasonWoof> I didn't really see the point
12:31:37 <slava> i have a stylesheet where you change fonts and colors and stuff
12:31:38 <slava> http://cvs.sourceforge.net/viewcvs.py/factor/Factor/Factor/library/help/stylesheet.factor?rev=1.12&view=auto
12:31:49 <slava> its just colon definitions that put hashtables on the stack
12:32:11 <slava> what i like about forth is you can use words for so much stuff you'd need complex designs in other languages
12:33:24 <slava> JasonWoof: this file drives an 'interpreter' for the help markup, and it calls generic words. the ui implements opengl rendering, the http server has an html generator, but its driven by the same piece of code here
12:33:28 <slava> http://cvs.sourceforge.net/viewcvs.py/factor/Factor/Factor/library/help/markup.factor?rev=1.31&view=auto
12:34:07 <JasonWoof> the puzzle for me is figuring out how to integrate documentation into the editor
12:34:19 <slava> http://cvs.sourceforge.net/viewcvs.py/factor/Factor/Factor/doc/handbook/streams.facts?rev=1.3&view=auto
12:34:22 <slava> this is the markup
12:34:35 <slava> the beauty of it is that it is just nested arrays with strings so it doesn't need a parser
12:34:44 <slava> i plan on having the same editor edit code and help markup structures
12:35:14 <JasonWoof> I'm not worried about editing it
12:35:20 <JasonWoof> I'm worried about where/when to display it
12:35:37 <JasonWoof> no use having documentation if people don't know it's there or can't find it
12:35:40 <slava> i have a sort of 'shadow block' system
12:35:43 <slava> http://cvs.sourceforge.net/viewcvs.py/factor/Factor/Factor/library/collections/
12:35:47 <slava> there's pairs of .factor & .facts files
12:37:57 <JasonWoof> I think the comments I have are good for the details of the docs
12:38:02 <JasonWoof> how a particular word  works, etc
12:38:11 <JasonWoof> but I also want to have overviews
12:38:15 <slava> yeah
12:38:18 <slava> examples are good too
12:38:26 <JasonWoof> man I wish the heat would come on
12:38:30 <JasonWoof> yeah
12:38:37 <slava> i have examples in the GUI that show source, and you can click one to run it
12:38:48 <JasonWoof> every once and a while it'd be good to have a paragraph or two that explains how some process works generally
12:38:53 <slava> i'm working on a multi-line edit widget for the GUI
12:39:40 <JasonWoof> like something to explain how I store the source
12:40:10 * docl decides to try jedit again
12:40:11 <slava> do you still plan on storing it using a binray format?
12:40:12 <JasonWoof> I haven't figured out how far I'll go with it
12:40:17 <JasonWoof> yes
12:40:30 <slava> is it an image?
12:40:34 <JasonWoof> yes
12:40:39 <slava> in the sense of smalltalk/lisp
12:40:47 <JasonWoof> like squeak
12:40:48 <slava> with import/export of text?
12:41:01 <JasonWoof> no
12:41:10 <slava> how will you share programs?
12:41:11 <JasonWoof> but I might do that later
12:41:15 <JasonWoof> I don't know
12:41:27 <JasonWoof> haven't worked out multi-person programming yet
12:42:04 <JasonWoof> I think I'll have a special server that you can talk to to get software and commit changes
12:42:23 <slava> the major stumbling block to implementing my editor is interaction with source files
12:42:45 <JasonWoof> yeah
12:42:49 <slava> until i come up with a design that allows an in-image structure editor *and* a module system that's based around conventional source files, i can't implement it
12:42:50 <JasonWoof> that was a pita for me
12:42:59 <slava> so far my image is bootstrapped from source, and i load changed source files in manually
12:43:06 <JasonWoof> I decided that this time around I wouldn't mess with files at all until/unless I was ready
12:43:40 <JasonWoof> the trouble with files is references
12:43:43 <slava> my GUI is essentially read-only; it doesn't have /editing/ other than typing colon definitions one at a time in the listener
12:43:49 <JasonWoof> my data has a lot of references to other data
12:43:56 <slava> but you can browse code
12:44:16 <slava> JasonWoof: you can find a way to serialize all your types of objects
12:44:32 <JasonWoof> all my new "editor" does so far is display 5 definitions
12:44:58 <slava> factor cannot export all types of objects from the image
12:44:59 <JasonWoof> it's easy enough to write a single object to disk. it's tricky to preserve the references between objects
12:45:16 <slava> i can convert any object *tree* into a source literal representation
12:45:23 <JasonWoof> sure
12:45:28 <JasonWoof> but my stuff is not trees
12:45:35 <slava> but if an object graph has cycles or shared structure, i cannot output it as source
12:45:57 <slava> it doesn't hang or anything, it just prints 'meta-syntax' to note sharing, but that does not parse
12:46:16 <slava> if you make an array of 3 elements, then set the first element to the array itself, and print it, you get { & 0 0 }
12:46:28 <slava> but the & doesn't mean anything
12:46:55 <slava> how is ot not trees?
12:47:02 <slava> dictionary -> wordlist -> word -> docs
12:47:04 <slava> dictionary -> wordlist -> word -> definition
12:47:21 <slava> or block -> word -> definition etc
12:48:02 <JasonWoof> hmm
12:48:34 <JasonWoof> I guess my structure isn't a problem
12:49:48 <JasonWoof> I only have references in specific places
12:50:37 <slava> one problem i have to solve soon is garbage collection of compiled code blocks
12:50:53 <slava> i'll need to keep track of which blocks call each other so i can trace references
12:50:59 <JasonWoof> :)
12:51:19 <slava> right now if you reload some compiled code and compile it again the code heap usage grows
12:51:27 <JasonWoof> right
12:51:36 <slava> only data is GC'd
12:51:37 <JasonWoof> I thought a bit about using dynamic memory allocation for compiled code
12:52:22 <JasonWoof> but I don't have dynamic memory allocation or any intention of implementing it soon... so now I think about other things :)
12:52:43 <JasonWoof> my heap is reset when you save
12:53:05 <JasonWoof> that's a form of GC
12:53:12 <slava> oh how does that work?
12:53:18 <slava> it builds a new dictionary?
12:53:31 <JasonWoof> keeps the old dictionary
12:53:45 <JasonWoof> will move bits of it around to make room at some point (haven't implemented that just yet)
12:54:05 <JasonWoof> builds a new kernel which has no references to previously-compiled code
12:54:32 <JasonWoof> makes a copy of all data referenced by source/comments
12:54:59 <JasonWoof> then moves these (the kernel, data, dictionary arrays) to the begining of the memory space, saves them to disk and then runs it
12:56:10 <JasonWoof> took me quite a while to get the compiler and source set up so I could compile this kernel at a different address than it would execute at, and so it had no references to code in the old system
12:56:28 <JasonWoof> also making sure that everything it needed to reference was compiled in it already
12:56:51 <JasonWoof> it ended up being a bit more complex than I hoped
12:57:01 <JasonWoof> but I think I got it right
12:57:28 <JasonWoof> nothing from the new image is used during this process
12:58:06 <JasonWoof> at some point I plan to make it so this bootstrapping process can switch endian
13:00:00 <JasonWoof> the GC part is when it coppies all the data objects. It just goes through the dictionary arrays that reference data objects (name, definition, comment and sometimes CFA) and coppies each into the new image
13:04:35 <JasonWoof> the kernel compiler knows what word should be executed after bootstrapping (you can change what word this is easily)
13:04:56 <JasonWoof> the kernel compiler then compiles all the dependancies of this word, then the word it's self, and sets the entry point of the new image to that word
13:10:25 <JasonWoof> oh, wow. if I made the dictionary out of data objects... then if I bootstrap to a word that does not include the editor, then the new image would not contain the dictionary
13:10:48 <JasonWoof> you should be able to then make pretty small binaries to do specific things
13:15:55 --- join: snowrichard (n=richard@adsl-69-155-177-154.dsl.lgvwtx.swbell.net) joined #forth
13:16:42 --- quit: snowrichard (Client Quit)
13:18:05 * docl fails yet again to get java working well enough to run jEdit
13:18:16 <JasonWoof> right now the dictionary sits at the begining of the memory space (after some header info)
13:18:38 * JasonWoof spreads his toes menacingly at java
13:18:45 <slava> docl: you need sun's jvm then it will work
13:18:58 <docl> darn
13:19:08 <slava> gcj or whatever won't run it
13:19:09 <docl> I have something *called* javac
13:19:25 <docl> and kaffe, and jikes
13:19:26 <slava> do java -version
13:19:46 <JasonWoof> docl: the ibm one works
13:19:47 <docl> Kaffe Virtual Machine
13:19:54 <slava> docl: jedit doesn't run with kaffe
13:19:58 <docl> ok I will try that
13:20:05 <slava> what CPU are you using?
13:20:09 <slava> java's only usable on x86
13:20:13 <slava> to be honest
13:20:19 <slava> the ppc implementations are slow
13:20:21 <JasonWoof> maybe that's why I hate java so much
13:20:21 <docl> yeah it's x86
13:20:24 <slava> i guess the sparc one is ok
13:20:32 <slava> but sparc cpus are slow in general
13:20:33 <JasonWoof> I've always been on ppc
13:20:42 <slava> JasonWoof: it was slower on m68k
13:20:47 <slava> JasonWoof: java 1.0.2 didn't have a jit on m68k
13:20:52 <JasonWoof> think I'll try to stay off java until stuff works on gcj
13:21:12 <JasonWoof> does it have jit on ppc?
13:21:20 <slava> gcj uses gcc's code generation.
13:21:26 <slava> it doesn't have a jit, you compile your java code ahead of time
13:21:37 <slava> but this means that dynamically loaded code runs in an interpreter (every gcj program links against a bytecode vm)
13:21:48 <slava> so if you run jedit, then the editor will be compiled and plugins will be interpreted
13:21:54 <slava> which can be a problem
13:21:56 <JasonWoof> oh
13:22:06 <Lars_G> gcj can compile to .class too
13:22:19 <slava> on x86, gcj compiled code is slower than sun's java jit
13:22:22 <docl> is kaffe the same way?
13:22:29 <slava> it only beats it in load time
13:22:33 <slava> no, kaffe has a jit on x86
13:22:37 <JasonWoof> heh
13:22:46 <JasonWoof> slava: is it much slower?
13:22:49 <slava> kaffe?
13:22:54 <slava> gcj?
13:22:56 <docl> "Unable to locate tools.jar. Expected to find it in /usr/lib/kaffe/pthreads/lib/tools.jar"
13:22:57 <JasonWoof> slava: gcj vs sun
13:23:21 <JasonWoof> I'd think the load time difference would make it worth using gcj
13:23:23 <docl> it's a pain to download sun java
13:23:26 <slava> JasonWoof: on integer and float code that doesn't do much OOP, they're about the same. gcj might have a slight edge, since gcc auto-vectorizes. but on OOP code, gcj is way slower
13:24:06 <slava> JasonWoof: sun's java profiles code at runtime to find methods to inline. you can'd do this with ahead of time compilation, so gcj generates polymorphic code for every method invocation, which can lead to a lot of overhead
13:24:18 <docl> there's several "java" programs from sun's site. do I want jre? jdk?
13:24:20 --- join: snowrichard (n=richard@adsl-69-155-177-154.dsl.lgvwtx.swbell.net) joined #forth
13:24:33 <slava> docl: jre runs compiled java bytecode, jdk compiles source and ships with some other tools
13:24:45 <slava> the jdk has a fucked up license
13:24:48 <slava> you can't redistribute it at all.
13:24:54 <slava> the jre can be shipped with apps with severe restrictions
13:25:15 <docl> either works?
13:25:19 <slava> yes
13:25:20 <slava> jdk is huge
13:25:23 <slava> >30mb
13:25:28 <docl> ick
13:25:36 <slava> jre is ~15mb iirc
13:25:40 --- quit: Cheery (Read error: 104 (Connection reset by peer))
13:25:41 <docl> 16.0
13:25:42 <slava> it gets larger with each release
13:25:49 <slava> is this 1.5/5.0?
13:26:06 <docl> 5.0 update 6
13:26:17 <slava> i remember the _jdk_ used to be 5mb
13:26:17 <docl> linux self-extracting file
13:26:27 <slava> they split out the 'jre' after it got to 11mb or so
13:26:33 <slava> it just kept on getting more bloated
13:26:44 <slava> now, the jre windows installer downloads additional components
13:27:33 <slava> java 6 will have a built-in javascript interpreter and http server, and javac is now part of the core library rather than the jdk
13:27:50 <docl> at least I have broadband
13:27:58 <slava> JasonWoof: you could hide a forth implementation or two in there
13:28:14 <slava> JasonWoof: the hotspot source ships with a scheme interpreter, its used to generate some boilerplate code in an obscure CORBA package
13:28:25 <slava> JasonWoof: they also use huge files of awk and sed macros
13:28:28 <slava> the build process is arcane
13:44:12 <docl> BUILD FAILED
13:44:12 <docl> /home/main/dl/jEdit/build.xml:26: Unable to find a javac compiler;
13:44:12 <docl> com.sun.tools.javac.Main is not on the classpath.
13:44:12 <docl> Perhaps JAVA_HOME does not point to the JDK
13:44:23 <slava> heh
13:44:37 <slava> there's a precompiled package you can use
13:45:21 <docl> java jedit43pre3install.jar
13:45:21 <docl> Exception in thread "main" java.lang.NoClassDefFoundError: jedit43pre3install/jar
13:45:59 <slava> java -jar
13:46:03 <slava> then the name of the .jar
13:46:10 <slava> without .jar, it expects the name of a class
13:46:16 <slava> without -jar
13:46:34 <docl> cool, it's running!
13:48:31 <docl> installed even :)
14:00:59 --- quit: segher (Read error: 104 (Connection reset by peer))
14:06:31 --- join: segher (n=segher@dslb-084-056-131-214.pools.arcor-ip.net) joined #forth
14:21:23 <Lars_G> I440r: you're still not back???
14:25:21 <I440r> im back now
14:27:44 --- quit: snowrichard ("Leaving")
14:58:19 --- nick: tattrdkat -> virsys
15:28:01 --- nick: Raystm2 -> tiff
15:58:52 --- quit: slava (Remote closed the connection)
17:17:59 <crc> jo everyone
17:19:13 <crc> s/jo/hi/
17:36:17 --- join: jungledog (n=dschauer@adsl-64-219-100-33.dsl.lgvwtx.swbell.net) joined #forth
17:54:21 --- log: started forth/06.01.23
17:54:21 --- join: clog (i=nef@bespin.org) joined #forth
17:54:21 --- topic: 'Welcome to #forth. We discuss the Forth programming language, simplicity, and a variety of technical subjects. The rules are simple: don't spam, don't troll, and please speak English.'
17:54:21 --- topic: set by Quartus on [Sat Jan 21 17:32:55 2006]
17:54:21 --- names: list (clog jungledog segher Snoopy42 Lars_G virsys sproingie tathi rsyncx skylan OrngeTide ccfg warpzero docl @crc tiff Teratogen Ray_work saon @JasonWoof)
17:54:27 --- join: I440r (n=I440r___@rrcs-24-242-160-169.sw.biz.rr.com) joined #forth
17:55:14 --- join: Quartus (n=trailer@CPE0001023f6e4f-CM013349902843.cpe.net.cable.rogers.com) joined #forth
17:56:39 --- join: interferon (n=user@vaporeon-22.dynamic.rpi.edu) joined #forth
17:57:49 --- quit: interferon (Client Quit)
17:59:51 --- join: Jim7J1AJH (n=jim@221x115x224x2.ap221.ftth.ucom.ne.jp) joined #forth
18:05:18 --- join: Infight6 (i=WINNT@12-208-98-237.client.insightBB.com) joined #forth
18:19:12 --- nick: tiff -> nanstm
18:20:52 --- quit: tathi ("leaving")
19:36:24 --- join: madgarden (n=madgarde@Toronto-HSE-ppp3713077.sympatico.ca) joined #forth
19:57:17 --- quit: saon (Nick collision from services.)
19:57:39 --- join: saon (i=1000@unaffiliated/saon) joined #forth
20:11:34 --- quit: saon (Read error: 104 (Connection reset by peer))
20:11:52 --- join: saon (i=1000@unaffiliated/saon) joined #forth
20:43:39 --- quit: sproingie (Remote closed the connection)
20:50:25 --- join: amca (n=plump@as-bri-4-1-32.ozonline.com.au) joined #forth
20:50:40 --- part: amca left #forth
22:46:49 --- quit: virsys ("bah")
22:53:55 --- join: virsys (n=virsys@or-65-40-181-13.dyn.sprint-hsd.net) joined #forth
23:29:58 --- join: aum (n=aum@60-234-156-82.bitstream.orcon.net.nz) joined #forth
23:55:10 --- quit: OrngeTide ("bye.")
23:59:59 --- log: ended forth/06.01.23