00:00:00 --- log: started forth/06.01.02 01:05:53 --- join: Raystm2_ (n=Raystm2@adsl-68-93-113-74.dsl.rcsntx.swbell.net) joined #forth 01:06:15 --- quit: Raystm2 (Read error: 104 (Connection reset by peer)) 01:11:45 --- join: Raystm2 (n=Raystm2@adsl-68-93-113-74.dsl.rcsntx.swbell.net) joined #forth 01:11:45 --- quit: Raystm2_ (Read error: 104 (Connection reset by peer)) 01:27:21 --- join: amca (n=plump@as-bri-3-112.ozonline.com.au) joined #forth 02:27:03 --- join: Topaz (n=top@spc1-horn1-6-0-cust128.cosh.broadband.ntl.com) joined #forth 02:45:54 --- join: Cheery (i=Henri@a81-197-18-99.elisa-laajakaista.fi) joined #forth 02:47:45 Hello 02:48:18 hello 02:48:32 How are you? 02:48:36 How is Xell going? 02:48:46 in the psp runs an arm? I thought it's a mobile version of the ps2 chip 02:48:57 ehm, I'm ok. 02:49:07 nope 02:49:24 The PSP runs an R-something chip as far as I know. A R8000? 02:50:32 xell, hmm I'm oscilatting between it will be something good or it's hopeless to work on it. 02:51:38 heck, go for "it's something good" Even if it isnt you will learn from it 02:53:23 how good is that readable? 'initlets', because I'm not a native english speaker I don't know if that sounds moronic or good. 02:54:23 what is an initlet? 02:57:25 only a word to define a xell bytecode segment with an initialisation code sequenz of the xell vm, so it's a bytecode at the beginning of a xell program. I invented it in that second ;-) 02:57:42 hehe 02:57:50 Have you considered "constructor"? 03:05:07 no, because that is so OOPish and any programmer who coded in Java & Co. assoziates with it that it's a _function_, but it's not a function it's _sequenz_ of _bytecode_ and can be saved and routet or compiled to machine language and then it hasn't really much todo with a constructor in OOP sense 03:08:20 isnt "a sequence of bytecode" just another name for function? 03:10:01 no 03:10:34 a word would be function, but a sequence no. 03:11:03 Is a sequnce un named? 03:11:29 (sequence) 03:11:49 yes, it's an unamed stream of bytecode when you so want, where words and other constructs are mixed. 03:12:03 okay 03:12:34 Is it possible to call this bytecode, or does it have to be manually inlined? 03:12:54 or if now call to it and return, can you goto it? 03:12:59 now = not 03:15:27 for example(pseudocode): push: 2 3 4 5 6 7; int: 20; int 21; define blah as: push 23; drop; ; or something like that is a sequence. 03:16:14 * amca nods 03:16:23 Perhaps a initialisation block? 03:16:34 or initblock? 03:17:12 ehm, that would be more a snipplet the opposite of a initlet 03:17:15 an 03:17:37 How is a snippet the opposite? 03:17:51 because it has nothing todo with the initialisation of the vm 03:17:58 Ah! 03:18:17 Sorry, I thought you was talking about code that runs *on* the VM, not the VM code itself 03:18:39 Well myself, I would have the VM initialisation in a seperate function 03:18:48 sry, I don't understand 03:20:04 The init code is used to start up the vm does it? 03:21:11 no, an initlet doesn't start up the vm, it initialisize the vm for the program. for example: 'set cell size to 16 bit', would be an initlet 03:22:36 So it would be like part of the executable image loading block? 03:23:06 but it's bytecode 03:23:22 so? 03:23:33 bytecode are still machine instructions 03:23:38 Just for am imaginary machine 03:25:07 am = an 03:25:16 sorry, my typos are bad today 03:25:29 hell, my are more bad 03:25:41 hehe 03:25:57 But you have an excuse - you dont know English as well as me. ;) 03:26:33 yes, they are machine instructions in that case, but do you know now what I mean? 03:27:02 I think I do. 03:27:48 You have a sequence of instructions in an executable image of bytecode that is run before the bytecode is loaded into memory, or just after? 03:29:11 ehm, yes that is it somehow, but I would substitue image with any form and instructions with bytecode. 03:29:29 * amca nods. 03:29:39 I meant the instructions were instriuctions in bytecode 03:30:06 So it is a consistant executable header that never changes? 03:30:34 no, it's not constant and can change from file to file 03:31:36 I see 03:31:49 but ok, some parts of it are constant, so that you have relative constant header when you so want. 03:31:54 it depends on the problem 03:32:33 because for a gui-program you do some other intilialisations as for a console-program 03:32:34 I see 03:33:12 Why not have the different initialisation possibilities in the VM as opposed to being run on the VM? 03:35:36 Or put another way, why not in the code for the VM have the different possibilities coded and then have a constant value in the executable header saying which initialisation routine to use? 03:41:39 could explain that to me with an example? 03:43:28 well there could be a data byte in the header of the executable image that if it is 0, the VM runs it in console mode, and 1 the VM runs it in gui mode. 03:43:33 xell bytecode files doesn't have any header 03:43:49 when you want they are 'raw' 03:44:40 Well maybe they should :) 03:45:23 See? Now y9ou are learning why executables arent raw code in most OS's. :) 03:45:40 that's a feature of xell, that there isn't a fixed file format, which increases complexity and reduces usability 03:46:07 sry, but I think that's a shit idea, especially for a vm which likes to use streams. 03:47:09 That's okay. 03:47:40 at the beginning xell-bytecode had a header and I removed it completly because I think it makes it complex and it reduces usability for me. 03:47:57 okay 03:49:12 and it doesn't cooperate with xell, because I need to throw away the header and then I need fucking linkers and that is something I don't want. 03:50:03 I see 03:50:41 it's something which I never ever will implement for xell, garbage collecting is something which I would implement before and the hell needs to froze before I implement that. 03:51:39 hehe 03:51:46 I like your description :) 03:52:46 I think I understand what and why you have it as executable bytecode now. It is basically an explicit header. 03:53:27 yes, and it fits nice into xells concept 03:53:47 * amca nods 03:54:13 because any word can be mapped to a bytecode except the bytecode for an api call 03:54:13 Well I dont see any other language or system fitting your needs, so I think that means that Xell is definately worth doing. :D 03:54:39 :-) horray.. 03:55:31 The best criteria for whether it is worth doing something or not is if it has been tried before. If it has, you may as well use other people's try, unless you are doing it for educational reasosn 03:55:33 reasons 03:57:03 I don't know if it has been tried, probably, but I'm also doing it for an educational reason and I'm also doing to have a nice to tool which doesn't fit into todays frameworks really. I need it myself. 03:57:26 well, four good reasons, hey much more than I thought. 03:57:32 :) 03:57:38 Glad you chatted with me? :) 03:57:45 yes... 03:58:03 Good to see I can be useful for something :) 03:59:50 what is good about bytecode for you? 04:00:29 It is machine independant and because your VM doesnt have to worry about practical realities, it can be simpler. 04:01:11 eh, what? 04:01:12 Which is why I am creating a VM myself - so that I can practice with simple asm before I worry about real asm 04:01:24 --- quit: Topaz ("Leaving") 04:01:59 what are you asking "what" about? 04:03:37 ah, sry I didn't understand your answer at the first read. 04:04:30 what are the practical realities today? bloat, comes somewhere to my mind. 04:05:47 The practical realities of implementing a machine in hardware. In all chips, comprimises have to be made between elegance and pragmatics 04:06:18 For example, the x86 chips dont have an orthogonal instruction set 04:07:42 orthogonal instruction set? tell me more ;-) 04:08:35 An orthogonal instruction set is where any registers of the same type can be used with the same instructions. But in zn x86, some instructions can only be used on certain registers, not all of them 04:09:51 zn = am 04:09:56 am = an >.> 04:13:41 --- quit: humulus (Remote closed the connection) 04:15:18 http://en.wikipedia.org/wiki/Orthogonal_instruction_set 04:18:06 --- quit: nballen () 04:24:32 oh, man. that's a really stupid definition. 04:25:02 My definition? Or wikip's? 04:25:18 wikip's 04:25:52 Ah 04:26:04 They use a more rigorous appraoch than my definition 04:26:55 one instruction, an operation on memory or register, but what is exactly a chips main memory? what if when the memory is extern of the chip but it's the main memory from the view of the chip, like x86 04:27:00 --- join: PoppaVic (n=pete@0-1pool73-95.nas24.chicago4.il.us.da.qwest.net) joined #forth 04:27:25 so.. a forth-chip is theoretically a chip with an orthogonal instruction set 04:28:10 As far as I know, the FVM is orthogonal 04:28:20 FVM? 04:28:33 Forth Virtual Machine 04:28:41 ah.. 04:29:04 fvm, hmm do you mean with that the vm which runs under every forth? 04:29:13 in an abstract sense? 04:29:34 The one that has a dictionary and executes the threaded code 04:30:53 so theoretically every forth system 04:31:07 And practicaly, yes :) 04:31:08 ok, I know that all forths out there use threaded code 04:31:42 ok, so xell has also an orthogonal instruction set 04:32:12 Possibly 04:32:55 I'm interested how it could be solved in hw 04:33:25 most of it would be fast onchip memory 04:33:53 --- join: tathi (n=josh@pdpc/supporter/bronze/tathi) joined #forth 04:33:57 Well it can be done in hardware, you just might geta slower and or more expensive chip than is practical 04:34:43 There are forth-chips, and boards 04:46:01 forth chips aren't really slow 04:46:34 virl: I never said they were. I was speaking more about chips in general 04:46:42 ok 04:50:49 someday I could implement it on a FPGA 04:51:18 the problem about a xell chip, it needs to be very flexible 04:51:20 I think others have done that. 04:51:29 Would be a good learning experience 04:52:13 yes, but before that I need to buy a FPGA board and they aren't cheap 04:52:45 * amca nods 04:52:50 Welkl I am off for the moment 04:52:55 Catch you another time 04:55:03 --- quit: amca ("60n3") 04:57:50 --- quit: tathi ("leaving") 05:37:54 --- join: JokeR69 (i=LNIX@12-208-98-237.client.insightBB.com) joined #forth 05:43:45 --- join: snowrichard (n=richard@adsl-69-155-177-154.dsl.lgvwtx.swbell.net) joined #forth 06:07:30 --- quit: PoppaVic ("brb") 06:17:33 --- join: PoppaVic (n=pete@0-1pool73-200.nas24.chicago4.il.us.da.qwest.net) joined #forth 06:33:56 --- quit: snowrichard ("Leaving") 07:19:15 PING 07:59:35 --- join: sproingie (n=chuck@64-121-2-59.c3-0.sfrn-ubr8.sfrn.ca.cable.rcn.com) joined #forth 08:13:48 anybody here? 08:16:30 nope 09:11:48 Hmm.. 09:12:43 Check me, please: between includes/loads: the vocstack/order/defs and base and such all remain as last set? 09:13:21 I'm thinking this is not entirely sensible. 09:15:07 ..as there is no really "neat" way to become "default-state" let alone leave it. 09:25:56 --- quit: PoppaVic ("Pulls the pin...") 10:01:06 --- join: JasonWoof (n=jason@pdpc/supporter/student/Herkamire) joined #forth 10:01:06 --- mode: ChanServ set +o JasonWoof 10:36:06 --- join: Quartus (n=trailer@CPE0001023f6e4f-CM013349902843.cpe.net.cable.rogers.com) joined #forth 10:36:28 Hey all. 10:53:51 hi :) 10:57:45 mornin 11:22:31 http://www.irtc.org/ftp/pub/stills/2005-08-31/winbad.jpg 11:28:47 --- join: nballen (n=nballen@adsl-69-111-248-132.dsl.renocs.pacbell.net) joined #forth 11:33:54 hehe :) and from the category "catastrophy": http://www.irtc.org/ftp/pub/stills/2005-04-30/mineswee.jpg 11:39:06 hi JasonWoof 11:39:21 how is the vm going? 11:39:32 --- join: snoopy_1711 (i=snoopy_1@dslb-084-058-190-071.pools.arcor-ip.net) joined #forth 11:47:28 --- quit: Snoopy42 (Read error: 145 (Connection timed out)) 11:47:41 --- nick: snoopy_1711 -> Snoopy42 11:48:56 --- join: enot (n=panteley@ppp85-140-40-97.pppoe.mtu-net.ru) joined #forth 12:10:19 --- part: enot left #forth 12:12:46 --- join: snowrichard (n=richard@adsl-69-155-177-154.dsl.lgvwtx.swbell.net) joined #forth 12:32:28 --- quit: snowrichard ("Leaving") 12:53:04 --- part: Pragmatic left #forth 13:16:38 virl: my project is going great :) 13:16:45 haven't found any bugs in the vm for quite a while 13:17:46 yesterday I finally found the last bugs in my bootstrapper in mist (the system I'm writing that runs under the vm) 13:18:04 that took forever 13:18:22 now I can start on the editor 13:19:29 what will the editor be like? 13:20:07 slava: somewhere between herkforth and THE 13:20:23 plain text sources or binary? 13:20:29 binary 13:20:32 ok 13:20:41 I've decided mostly what it will look like 13:20:49 I haven't worked out the keyboard interaction fully yet though 13:21:54 I think not going to have input in a separate place anymore 13:22:04 it'll be more like a normal editor where the cursor is right in the middle of the text 13:22:39 although it will not be treated as text I think 13:22:45 individual words will be treated as entities 13:22:54 and maybe a mode where whole definitions are entities 13:23:09 dunno 13:23:18 haven't worked out the whole modes vs no-modes thing 13:23:55 and the flexibility of this new system hasn't fully sunk in yet 13:41:03 eg being able to do graphics, access the keyboard and do things involving timing 13:41:12 eg I could make a cursor that looked like whatever I want 13:41:15 and even make it blink 13:41:27 likewise with highlighting 13:42:01 heh 13:45:08 first it's going to be a list of definitions 13:45:12 with the cursor in the middle one 13:46:02 I'll make it so you can move left and right 13:46:06 and probably up and down 13:46:29 and I'll figure out other keys to go "in" and "back" 13:46:59 "in" would go to whatever is referenced, or "within" the object if that makes sense 13:46:59 --- join: tathi (n=josh@pdpc/supporter/bronze/tathi) joined #forth 13:47:10 on a word it would go to it's definition. 13:47:22 on a string object it would go to the text 13:47:27 i'm going to have a structure editor like that, eventually 13:47:32 there's a lot of details to work out first 13:47:44 yeah, I'm just now starting to think about where to start 13:47:53 I've got code to print out my complete sources in color 13:47:58 that's a start I suppose 13:48:11 next I need to add a cursor and just print one page of sources with the cursor in the middle 13:48:12 i'm stuck on importing/exporting sources in the image. 13:48:23 just started thinking about how to get it in the middle 13:48:24 it needs to be robust, and handle multiple modules with dependencies, and unloading modules cleanly, and so on 13:48:38 ahh 13:48:40 sounds tricky 13:48:45 also i want to be able to edit sources without loading them 13:48:57 for example, windows or unix-specific code on a different OS 13:49:03 ahh 13:49:18 I've simplified things considerably by completely ignoring the outside world 13:49:32 "there is no spoon" 13:49:46 well, i don't want to put all existing factor libraries into the core image 13:49:48 it would be too big 13:49:58 right now i just load source files manually 13:49:59 yeah 13:50:02 I don't have that much code 13:50:05 JasonWoof: did you find the "last" of your bootstrapping bugs? 13:50:11 although I do have almost 500 images 13:50:16 tathi: yeah :) last night 13:50:24 excellent :) 13:50:30 yeah :) 13:51:22 what a relief! 13:51:32 was starting to drive me nuts 13:51:39 heh 13:51:49 I said 500 images... meant almost 500 definitions 13:52:07 i have 5500 in the core 13:52:28 well, almost 500 entries in the dictionary. That includes things without definitions such as numbers and most constants 13:52:41 well, I just started over :) 13:53:04 my current source format will only handle 8K 13:53:08 1800 of those is SDL, FreeType and OpenGL bindings 13:53:20 heh 13:53:35 bindings.... I've got some of those I suppose 13:53:41 probably about 7 13:54:14 6 :) 13:54:17 will have more though 13:54:35 some graphic ones like fill-rect and blit 13:55:04 might have two blits: one screen to screen and one mem to screen 13:55:37 does your VM link against SDL? 13:55:42 yeah 13:56:02 just SDL 13:56:24 might add something later to access the host's clipboard 13:56:33 but I'm not terribly concerned about that 13:56:57 doesn't seem all that important for mist development 13:57:45 I only remember pasting once into herkforth while developing mist, and I could have done it without pasting 13:58:09 oh, heh, also yesterday I made the shift key work 13:58:53 guess I didn't put that in earlier because I was worried my bootstrapping of objects was broken 14:00:08 trying to decide how I want to handle multi-line defs 14:00:30 they make it a bit harder to get the current def in the middle of the screen 14:01:25 I might clip them short 14:02:15 except the current one of course 14:03:52 there aren't many that are that long, and it should stay that way 14:06:29 I'm going to make a big array that says what order stuff is displayed in the editor 14:07:01 so in the editor you can reorder the definitions 14:07:35 i'm wondering how to organize source in my editor 14:07:37 after I've got it usable, I plan to add support for other things to be in there besides definitions: 14:07:56 1) some kind of seperator bar (like a row of stars, or a white line or something) 14:08:03 2) comment line 14:08:09 3) maybe a blank line 14:08:25 hmm 14:08:28 formatting? 14:08:41 eh? 14:08:50 comments, blank lines, etc 14:08:52 not sure about the blank line... 14:10:39 did i show you my help system? 14:10:47 hmmm... I don't intend to actually use that as a commont 14:10:51 meant more like headlines 14:11:15 like so: 14:11:16 ******************* 14:11:16 ** MATH ROUTINES ** 14:11:16 ******************* 14:11:34 don't think so 14:12:09 look at http://factorcode.org/help-hashtables.png 14:12:12 in my system each word can have a comment 14:12:23 it's in the dictionary entry 14:12:37 it'll be displayed when you are looking at the word in question 14:12:59 --- join: segher (n=segher@dsl77-24-100.fastxdsl.nl) joined #forth 14:13:18 grrr, why can't xview display pngs properly anymore? 14:14:17 use a web browser :) 14:15:26 looks like your help is coming along nicely 14:15:30 the formatting is pretty ugly 14:15:44 i'm trying to get it looking ok without writing tons of fancy layout code 14:15:52 slava: fancy 14:16:05 w3m is a web browser 14:16:10 yeah 14:16:31 slava: it pulls the actual definition from the image? 14:16:39 yes 14:17:35 cool 14:18:25 eventually i'd like to have an editor for source and help text (WYSIWYG) 14:18:33 it occured to me after I thought seriously about putting in those comment/headline lines into the editor, that one could then put a great deal of information into the sources 14:18:39 yeah 14:18:41 kindof like Literate Programming 14:18:52 yeah 14:18:55 at the system level my sources are just organized into flat vocabularies, but the help has this outline structure 14:18:59 wysiwyg is good 14:19:22 my UI is a lot more primitive than it looks. there is no editing or text selection 14:19:55 you can click the little expando-dinglers though right? 14:20:03 yes 14:20:11 and click words in the definition to look at their help/definition 14:20:18 cool 14:20:27 or right-click to get a menu of commands like finding usages 14:20:35 I really like the idea of being able to get the big-picture documentation in there somewhere 14:21:01 for understanding specific bits of source, a stack effect on each word and a short comment on many of them is quite sufficient 14:21:24 my docs are kind of wordy but a lot of that stuff is generated by 'macros' 14:21:28 but for some things I think it'd be nice if there were several paragraphs explaining in general terms what's going on, or how something works in the big picture 14:21:35 like the docs I have on my wiki 14:21:40 --- join: amca (n=plump@as-bri-3-95.ozonline.com.au) joined #forth 14:21:47 right now my docs are in latex -- i'm not done converting them 14:21:50 latex didn't turn out very well 14:21:59 it was a pain keeping it in sync with sources, and there was no cross-referencing at all 14:22:03 speaking of which... I'd actually like to hook the wiki into my interface 14:22:23 Hoe goes iyt today? 14:22:35 JasonWoof: that's interesting 14:22:41 erm...how goes it today 14:23:07 amca: ahh, hello 14:23:15 couldn't translate the first time 14:23:38 slava: yeah, that's generally the biggest problem I see with standard text tools, they don't cross-reference well 14:23:56 well, they often do cross-reference, but now what I'd call "well". 14:24:00 i'm thinking of ways to abstract my system of finding callers of words so that it applies to help articles 14:24:18 a general graph word set... otehrwise i'll have to implement similar algorithms a second time 14:24:27 JasonWoof: Yeah, neither did I :) 14:24:43 slava: How is the doc translation going so far? 14:24:58 i've only done one chapter 14:25:12 slava: something that could display a tree/graph of cross-references? 14:25:19 JasonWoof: yeah 14:25:26 be it mostly docs refering to words, or a tree of definitions or whatever? 14:25:39 or word definitions themselves 14:25:42 I've thought on many occations about wanting to see some sort of tree of definitions 14:26:04 trouble is it would never fit on the screen, and once it gets bigger, generally get's impossible to read over medium-sized parts of 14:26:14 you can use expandable outliners 14:26:17 with the possible exception of the idea of doing it in some sort of zooming interface 14:26:46 slava: That sounds like a good idea. You are using it in your GUI eh? 14:26:48 i guess outliners are a primitive form of zooming 14:26:57 yeah, similar idea 14:27:01 i don't really see the value of having fancy graphical effects with shrinking/growing boxes and such 14:27:05 action required to see what's within 14:27:09 i prefer instant feedback and simple mostly-text displays 14:27:15 the boxes are just cues as to what's happening 14:27:28 zoom would be useful for deep outliner exploring 14:27:30 slava: hehe me too. I turn all the animations off on Win that I can 14:27:31 slava: well, yeah, in herkforth the whole tree is at your fingertips, but not much of it is displayed at once. 14:27:51 just the current page of source, and a bit about the word your cursor is on (it's definition and comment) 14:28:13 I think I'd really need that forth implementation which has good ffi. 14:28:41 factor's ffi works pretty well. only thing that's missing is callbacks 14:28:53 I don't do ffi 14:29:13 i prefer it to bloating my VM with bindings for all external libraries 14:29:19 i hate writing wrapper code (or code in general) in C 14:29:41 slava, me too. It's just unnecessary. And very bad practice. 14:29:56 slava: What do you prefer to code in? 14:29:59 its ok if you bind to one or two libraries. 14:30:03 amca: factor 14:30:06 amca: oh, he just loves java 14:30:08 * sproingie winks 14:30:14 :P 14:30:15 :) 14:30:16 but if you modify your runtime for every external library, you get something like php 14:30:28 * sproingie hisses 14:30:37 slava: So you have enough of a Factor system that you can code the rest of the system in it? 14:30:40 i don't like slate's plugin approach either 14:30:48 they hand-code plugins for external libraries 14:30:53 io uses the plugin approach as well 14:30:53 and the VM loads plugins dynamically 14:31:06 amca: I only have a small loader/runtime in C 14:31:07 slava, how did you made the library loading? 14:31:30 slava: cool :) 14:31:53 ok, I gotta go shopping 14:31:58 bbl 14:32:18 Cheery: well, i look up symbols using libdl, and the compiler generates stubs that convert between C and factor types, and call the symbols 14:32:41 in theory its pretty simple, but its extremely CPU and OS specific and somewhat tricky 14:33:29 Thought it works great. 14:33:44 yeah, it works well 14:33:56 Did you found out what I were doing wrong with factor last time when I tested that alien lib? 14:34:05 you didn't compile your word 14:34:17 when you define new words, they are not compiled, they run in the interpreter 14:34:20 why does it need compiling? 14:34:21 but alien-invoke cannot run in the interpreter 14:34:28 ah. 14:34:31 : foo .... alien-invoke ; 14:34:32 \ foo compile 14:34:41 it will be automatic later, when the interpreter goes away 14:35:43 I wonder, why there's only one image and those libs aren't separated from each other? 14:35:55 what do you mean? 14:36:01 And why there is no pure shell version from interpreter? 14:36:19 i don't understand what you're asking 14:36:39 in 0.79 version. 14:36:54 there's that factor.image, then there's f.exe. 14:37:01 right 14:37:13 on windows, we don't support the command prompt 14:37:20 yet? 14:37:33 never. we use i/o completion ports for i/o, and they don't work with the windows console 14:37:43 Aha. 14:37:55 Are you completely sure about that? 14:37:57 yes 14:38:07 weird... 14:38:50 But I just wonder, since currently I think, using factor is quite complex in here windows. 14:38:51 on unix you can run factor in the tty, in fact its the default 14:38:58 why is it complex? 14:39:23 well, first, the SDL interpreter is slow and bulky to use. 14:39:33 it will get faster 14:39:52 second thing is that I find it hard to run a program, compared to other HL languages like python or ruby. 14:39:53 last i used it it was pretty fast 14:40:01 why is it hard to run a program? 14:40:08 "foo.factor" run-file 14:40:32 maybe there should be a shell script front end that invokes that 14:40:41 #!/usr/bin/factorscript 14:40:42 you can list files to load on the command line 14:40:44 yes, but it is inside the interpreter, and that interpreter interface is not good currently. :( 14:41:00 Cheery: the whole point of factor is interactive development with an immersive environment 14:41:03 slava: what are i/o completion ports? 14:41:10 amca: non-blocking i/o feature on win32 14:41:20 slava, say that to end-users. 14:41:26 Cheery: its not a scripting language 14:41:35 even though you can do f.exe factor.image foo.factor 14:41:51 can I do an executable from my program? 14:42:02