00:00:00 --- log: started retro/06.06.30 00:11:18 --- quit: Cheery ("Leaving") 04:39:51 --- join: nighty (n=nighty@fr-reims-gw.origami-systems.com) joined #retro 06:38:59 --- join: virl (n=virl@chello062178085149.1.12.vie.surfer.at) joined #retro 07:55:31 --- join: dsl749 (n=dsl@66-159-147-101.adsl.snet.net) joined #retro 07:56:49 Hello, can anyone tell me why retroforth 9.1 or 9.2-beta doesn't recognize the "from" command? I'm trying to load the gtk stuff and it fails on that command 08:01:45 yeah, I think the issue is that from currently only works in the generic version 08:02:30 does the generic version work under linux? 08:02:32 to get that running, first build generic, then run the link-generic-whatever script in the extras directory 08:03:08 yeah, but it's a little more complex to set up 08:03:21 how so? 08:03:37 you have to run a script from the extras directory 08:04:03 extras/link-generic-linux 08:05:14 and then run build? 08:05:39 build is first 08:06:12 also, to run the link script you have to be in the bin directory 08:06:51 ./build generic; cd bin; sh ../extras/link-generic-linux 08:06:57 I guess I'll need the source version as well instead of the binary version of the package I have 08:07:05 ./build generic; cd bin; sh ../extras/link-generic-linux 08:07:21 yeah 08:07:30 try the darcs repo if you have darcs 08:07:54 docl, thanks so much for your time, I'll go try this 08:08:08 cool :) 08:08:13 np 08:19:58 --- quit: dsl749 (Remote closed the connection) 08:37:46 who was from DU here ? 08:42:04 du? 08:42:55 OZ 08:42:58 Down Under 08:43:03 Australia 08:43:18 The land where men are men and ship are scared 08:43:22 oh no 08:43:28 that was new zealand :) 08:43:43 sheeps 08:44:31 thin where are you from ? 08:44:37 canada 08:44:41 cool 08:44:44 where from ? 08:44:49 toronto ? 08:44:49 calgary 08:44:56 alberta :) 08:45:04 Maybe I'll see you someday 08:45:08 you? 08:45:09 I'm moving to toronto :) 08:45:12 cool 08:45:13 from where 08:45:16 I am from France 08:45:19 --- join: dsl1169 (n=dsl@69.37.139.52) joined #retro 08:45:20 nice 08:45:25 moving from France to Toronto 08:45:42 whats in toronto? ;) 08:45:46 My company is looking for Aussie staff 08:45:52 thin: my company 08:46:40 over the years there's been like 4-5 forthers from calgary in the #forth chan that i know of 08:46:42 so I'm hanging around the channels to see if anyone would postulate after a harsh filtering from myself :) 08:46:55 but i've never met anyone of them 08:47:00 oh :) 08:47:04 they're not very friendly for some reason ;P 08:47:12 well :) 08:52:03 docl, I'm still having problems getting the generic version to compile. I got fasm, I can run the "build generic" command, I cd to bin, but the sh ../link-generic-linux commands gives me an error of gcc: rf-generic: No such file or directory 08:53:30 when you are in the bin directory, does "ls" show a copy of rf-generic? 08:53:40 dsl1169: cat the script 08:54:00 and then do what the script does except alter it for wherever the rf-generic file is ;P 08:54:05 no, ls doesn't show it 08:54:23 that directory is blank 08:54:27 um thin, that won't work 08:54:57 did ./build generic spit out an error? 08:54:58 docl: i meant manual typing to do whatever that script does ;P 08:55:13 yeah i guess if the file doesn't exist then there's a problem 08:55:17 unles its in another directory 08:55:18 actually I bet that's the problem 08:55:27 build isn't chmodded right by default 08:55:35 chmod it to 755 and try again 08:56:32 I had used chmod 777 alread. ./build from the 901-hosted directory says generic was built without showing any errors 08:56:58 RetroForth 9.1 Build Scriptsrnflat assembler version 1.67.0rnRx :: Release 1rn :: Adding a .bss section to keep file size downrn3 passes, 13010 bytes.rngeneric has been built 08:57:08 hmm. ok this might be apples and oranges since I'm in the 902 darcs repository 08:58:08 no, works in 901 as well 08:58:52 I just ran ./build linux and that gives me a bin/rf-linux! But it still doesn't recognize the "from" token. 08:59:40 right. the link script only works on generic (which works under linux) 09:00:38 ./build generic; cd bin; sh ../extras/link-generic-linux 09:00:52 works for me 09:01:19 let me erase what I've done, unpacka again, and try it again from the top. 09:02:02 tried again by unpacking it again, and the build script wasn't chmodded right so it gave an error 09:03:58 crc needs to update his darcs script so that ./build is chmod'd properly 09:04:44 OK, I got the rf-generic in bin, cd bin, but the sh ../extras/link-generic-linux step gave me this error 09:04:46 sh ../extras/link-generic-linux rnRetroForth 9.1rn----------------------------------------------------------------rnLinking Generic port against libc and libdl.sorn/usr/bin/ld: final link failed: Bad valuerncollect2: ld returned 1 exit statusrnCleaning up...rnmv: cannot stat `rf-generic-bin': No such file or directoryrnDone!rn---------------------------------------------------------------- 09:06:35 have you tried copying the gcc command from the script? 09:06:52 gcc rf-generic -o rf-generic-bin -s -ldl 09:07:04 from the link-generic-linux? no, but I can try it.... 09:07:29 hmm could be a gcc error 09:08:12 if rf-generic exists, the -o part should create rf-generic-bin automatically 09:08:48 assuming you have gcc installed 09:09:25 gcc rf-generic -o rf-generic-bin -s -ldlrn/usr/bin/ld: final link failed: Bad valuerncollect2: ld returned 1 exit status 09:11:20 I wonder if you have libdl installed 09:12:28 yes, it's in /usr/lib/libdl.a and /usr/lib/libdl.so 09:13:31 ls 09:13:52 even without the -s and -l switches at the end of the gcc line I get a "/usr/bin/ld: final link failed: Bad value" error 09:17:27 what gcc and ld versions are you using? 09:18:35 gcc (GCC) 3.3.4 (Debian 1:3.3.4-7) 09:18:40 GNU ld version 2.15 09:18:54 gcc version 3.3.5 (Debian 1:3.3.5-13) 09:19:01 --- join: neceve (n=claudiu@unaffiliated/neceve) joined #retro 09:19:10 shouldn'd be that much difference, you'd think 09:19:44 I'm wondering if it has to do with fasm, since I downloaded the executable not seeing any source 09:20:22 --- join: crc_work (n=cchilder@h-64-105-207-163.phlapafg.covad.net) joined #retro 09:20:25 could be, if it compiled slightly different 09:20:31 hey crc 09:20:53 dsl1169 is having problems linking the generic in 901 09:21:18 try downloading the binary snapshot and running the link script: 09:21:19 http://retroforth.org/902-beta.tar.gz 09:21:25 untar 09:21:27 cd retroforth 09:21:35 ./extras/link-generic-linux 09:24:10 works for me 09:24:16 and then run rf-generic instead of rf-linux. Thanks a lot guys, that did it! 09:24:27 cool :) 09:26:27 glad to help :) 09:35:42 --- quit: crc_work ("(5 hours till work ends)") 09:43:47 --- quit: dsl1169 (Remote closed the connection) 09:48:03 --- quit: nighty ("Disappears in a puff of smoke") 10:06:31 --- quit: neceve (Read error: 113 (No route to host)) 10:10:23 --- join: Cheery (i=Henri@a81-197-12-134.elisa-laajakaista.fi) joined #retro 10:41:59 --- join: nighty (i=nighty@sushi.rural-networks.com) joined #retro 10:42:10 --- join: dsl1667 (n=dsl@69.37.139.52) joined #retro 10:45:41 --- join: crc_work (n=cchilder@h-64-105-207-163.phlapafg.covad.net) joined #retro 10:45:47 Thanks for the help before. I got the generic version to handle the "from" command, but now I'm trying to use the GTK 1.2 functions and I'm having problems. Is the use of system libraires in linux documented anywhere? It looks like "syscall" isn't a function anymore, although I remember it froom the "forthmud.f" code I have from before that I remember working. 10:46:29 the generic port does not have 'syscall', it relies on the ffi words (from, import, etc) 10:46:59 Is there somewhere I can look to see examples of their use? 10:50:48 I tried the example from http://retroforth.net/news/archives/198 and it was OK until the " :: gtk gtk_main_quit bye ; " line where it told me 10:50:51 01:49PM dsl1667 on #retro (R... [Ctrl-N to ChanServ] [IRC 8m] [Lag 0.13s] naimrne gtk_main_quit 10:51:15 :: gtk gtk_main_quit bye ;rn** Unable to find or convert the specified token torn** a number. The token passed was: gtk_main_quit 10:52:01 try this: 10:52:02 from libgtk12.so 10:52:10 0 import gtk_main_quit 10:52:17 does it display an error message? 10:53:03 no but I changed libgtk12.so to libgtk.so 10:53:11 * crc_work has been working with Gtk2 mostly these days; I'm a bit rusty on the 1.x api 10:53:19 ok 10:53:27 I took it from an old news posting. 10:53:30 the library names tend to differ slightly by distro 10:53:32 yeah 10:53:49 I am wondering if it is a problem due to changes in the vocabulary model since 8.3 10:55:02 --- join: swalters (n=swalters@169.139.217.100) joined #retro 10:55:11 The reason I want to do this is dsl linux, a 50 Meg distro, comes stock with GTK 1.2 and I'd love to be able to have them include retro forth as another language because it's so small. 10:55:55 I'll update the 1.2 bindings tonight then 10:56:01 nice idea :) 10:56:21 Thanks a lot, I hate to put you out, but I'd love to see this work. 10:56:30 it's not a problem 10:56:54 It's probably best to keep support for both 1.2 and 2.x availble anyway 10:57:27 hey crc, I have been trying to get the native port working under qemu but it's not quite working 10:57:52 the end.asm file is missing, so I had to comment that out of grub.asm 10:58:35 I'm not sure how to get the resulting file to work, or whether it will. qemu won't load it straight up. 10:59:16 I did get the retro.iso file working by creating a blank hd image, but since there's no partitions on it it's useless 11:00:51 If you don't mind my asking, is it possible to add functions to retroforth and then save the whole thing as a single executable thats starts using a particular function? 11:01:40 I've done that with a shell script, but it's primitive 11:02:11 i.e. save the functions as a text file, then run the executable with -f filename 11:03:58 But there isn't a word that saves the dictionary and sets the starting function as something? I can use the -f switch anyway. Thanks. 11:04:39 no, don't think so. although we had a logging function a while back 11:05:19 hmm. I bet you could make a function that saves a certain range of words to a block and loads them pre-compiled 11:05:42 I remember win32forth had something like that 11:07:15 the dictionary contains a set of pointers that would need updated if saved. but seems like that wouldn't be too hard to change to relative values. 11:12:36 I did a turnkey addition to the linux version; to make it work with the generic and windows ports would require a lot of work though 11:12:59 (since ffi relies on dynamic linking, and addresses can change between launches...) 11:13:20 http://retroforth.org/contrib/ for the turnkey generation for linux 11:13:32 awesume 11:13:34 * crc_work will warn that it's a raw memory image, with an elf header 11:14:04 so the generated binaries will be large at this point. I keep planning to fix that sometime. Maybe I'll try something this weekend :) 11:15:27 yeah.. 11:15:43 ok, now I play again with the X bindings. 11:15:58 that would be great, the code at that link still uses "syscall" 11:16:24 heh, crc, could you fix your display of the program? 11:16:32 virl: no 11:16:45 arg.. 11:17:08 but my 'startx' script is broken due to some of my misconfiguration attempts, so it may not be loading the right settings when I do 'xinit' 11:18:12 dsl1667: I'll take care of that when I update the rest of the code 11:18:54 thanks again, I need to go but I really apprechiate all of the help. 11:19:00 take care 11:19:03 not a problem 11:19:14 --- quit: dsl1667 ("User disconnected") 11:19:15 * crc_work has to go as well.... 11:19:25 ok, see ya later 11:19:31 --- quit: crc_work ("work calls. but only 3 or so hours left") 11:25:38 hmm, do I improve the code so it displays correctly on crcs configuration or do I improve the Xft bindings hmm.. 11:52:19 --- quit: Cheery ("Leaving") 12:23:18 --- quit: swalters ("User disconnected") 12:40:11 --- quit: Ray-work ("User pushed the X - because it's Xtra, baby") 12:41:40 --- join: crc_work (n=cchilder@h-64-105-207-163.phlapafg.covad.net) joined #retro 12:42:32 virl: work on the bindings. I'll try to fix the configuration of X11 on my box 12:43:54 with x11 bindings working, making a cool web browser will be within the realm of possibility :) 12:44:52 that could be done with the GtkHtml library :) 12:45:54 neat! is there a way to access network stuff from generic? 12:46:07 libc should have socket functions 12:46:09 * crc_work will check 12:46:10 --- quit: crc_work ("Less than 2 hours...") 12:52:51 cool, I got the graphical block editor to work 13:06:29 puh, I don't know why xft doesn't display black as black, when I'd like to print a black text through the X bindings it does that. *grr* 13:07:30 is there any other way to print black? 13:10:05 it's so silly. when you want to print in X color text then you need to set the foreground color but when you want to do that in xft then you need fill colors into a struct and then allocating it so that it can be used. *grr!* 13:11:15 I would like it, when it would also use the X mechanism, but it doesn't. I think they smoked some stuff when they developed xft. 13:18:21 and somehow there is a problem with alpha 13:28:51 ah, I fixed. *dancing jahoo* 13:35:02 cool! 13:36:19 ok, now to the crc color problem.. 13:43:13 hmm, if that can be fixed with this 'XColor' stuff? 14:06:31 home :) 14:06:53 early today (it was pretty slow) 14:07:13 good! use the extra time to get yourself a real programming job that pays you 25 bucks an hour! yay! ;P 14:07:44 it won't be slow after the 4th 14:08:00 hiho crc 14:08:06 no work tomorrow; monday should be slow as well. no work on tuesday, then back to normal 14:08:08 hi virl 14:09:52 --- join: fma_steve (n=dsl@69.177.95.155) joined #retro 14:11:33 virl: I am going to change the string implementation to eliminate the need for 'zt' and it's quirks 14:12:49 oh, well, that wasn't needed but ok, when you think you could improve the situation I don't say no ;-) 14:13:58 I tried using a for next construct and a repeat again construct in retroforth 9.2 beta but they failed. Can anyone see why this doesn't print out ten a's? 14:14:01 10 for ." a" next 14:14:23 [[ 10 for ." a" next ]] execute 14:14:36 for/next are "macros" and can only be used in a definition 14:14:46 [[ ]] execute 14:14:57 : test 10 for ." a" next ; test 14:14:58 thanks, that's clearer now. 14:14:58 this form creates an anonymous definition and executes it 14:15:04 np 14:15:07 crc, how does this displays on your box? http://www.rafb.net/paste/results/qVEkyt64.html 14:15:21 crc, what about :: 10 for ." a" next ; execute 14:15:30 thin: that will also work 14:15:41 thats cleaner thant [[ ]] ;P 14:15:42 now I used for drawing XColor functions, except for the line(I was too lazy) 14:16:12 actually why do we have [[ ]] when its identical to :: ; ? 14:16:17 or are there differences 14:16:42 virl: green background, redish text, and a black line 14:16:52 [[ ]] can be used in a definition as well 14:16:56 and can be nested 14:17:01 ah 14:17:13 :: ; is slightly more efficient, but not quite as flexible 14:17:24 lets get rid of :: ; then 14:17:25 both have their places 14:17:35 thin: I am planning to 14:17:40 cool 14:17:50 probably this weekend in fact 14:18:22 so whats the diff between [ ] and [[ ]] 14:18:35 philosophically ;P 14:20:41 [ turns the compiler off, dropping to interpreter 14:20:45 ] turns the compiler on 14:21:02 [[ turns the compiler on, and compiles some support code 14:21:17 WOW! totally different and not very intuitive ;P 14:21:19 ]] ends compilation and patches the support code, also compiling a literal if necessary 14:21:28 one turns it off, the other turns it on 14:21:44 hmm 14:21:45 how about 14:21:50 [: ;] 14:21:57 Not as easy to type 14:21:59 that'd be more understandable but harder to type ; 14:22:00 yeah 14:22:00 heh 14:22:09 ' [[ alias [: 14:22:11 how about :: ;; ? 14:22:14 ' ]] alias ;] 14:22:19 not as visually consistent 14:23:37 crc, green background that's good, very good. 14:23:46 how about using | 14:23:53 instead of [[ ]] just have | | 14:23:59 ;P 14:24:01 nevermind 14:24:09 how to toggle the states cleanly and allow nesting? 14:24:18 just brainstorming is all 14:24:19 yeah 14:24:20 what do you mean with redish text? 14:24:46 the text looks sort of like a very dark red. Of course, that could just be me... 14:25:29 aha.. which one? there should be two.. 14:26:02 one should be the vectored text of xft and one should be a text in a bitmap font(X11 default) 14:28:07 http://retroforth.org/sample2.png 14:29:20 ah, ok. well for me the text is black. ok it matches, but the vector text doesn't appear. have you changed the lib? 14:29:39 /usr/X11R6/lib/libXft.so.1 14:29:50 it's here how is it on your side? 14:31:06 yes 14:31:59 arg. so why doesn't it displays it 14:33:49 I had to change it to /usr/X11R6/lib/libXft.so.2 14:35:09 http://retroforth.org/sample2.png 14:36:15 yeah, that's it. 14:37:08 so it was the wrong path. haha it works! ;-) 14:46:05 updated repo and binary snapshot (not the 'getretro' archive yet) with z" 14:51:56 the next patch will remove :: 14:54:01 so what do you use inside a loc: ... ;loc construct for the return values? without the :: 14:54:13 you can use [[ ... ]] 14:54:18 or say "here ]" 14:54:27 or, if changing an external vector 14:54:32 "here is ... ]" 14:57:08 the "here is" form is pretty readable: 14:57:09 : z" " over + 0 swap c! ; 14:57:12 hmm 14:57:40 here is choose ] erase fname swap move ; 15:12:31 --- quit: fma_steve (Read error: 104 (Connection reset by peer)) 15:13:16 --- join: snoopy_1711 (i=snoopy_1@dslb-084-058-142-245.pools.arcor-ip.net) joined #retro 15:14:50 done 15:14:57 :: is now gone 15:20:45 --- quit: Snoopy42 (Read error: 145 (Connection timed out)) 15:20:57 --- nick: snoopy_1711 -> Snoopy42 15:25:51 does anyone here use >pad 15:29:04 once in a while 15:29:10 ok 15:29:22 * crc is changing reclass: 15:29:27 it will now be: 15:29:35 ( xt "name" -- ) 15:29:46 where xt is the class handler to use for "name" 15:29:58 ahh, cool 15:30:25 I think it'll actually be more useful this way 15:30:43 ' .macro reclass mymacro 15:30:50 yeah 15:30:56 or, to create a new class: 15:31:04 : .classname .... ; 15:31:08 ' .class reclass: .classname 15:31:32 a class is itself of a given class, right? 15:31:51 yes, normally of .class 15:31:57 .class has a class of .class as well 15:33:18 re: >pad, I think a generic move command that keeps the result on the stack would be more useful 15:33:28 I'll work on that 15:35:46 : move: 2dup >r >r move r> r> ; 15:37:44 is there a standard prefix to show that there's a result on the stack? 15:38:28 : and ' seem to denote parsing 15:38:48 nope 15:38:58 I guess ' leaves something on the stack 15:39:27 : usually cleans up 15:44:18 copy, relocate, place 15:44:26 * crc is not sure what to call this word :( 16:01:53 * crc goes with 'copy' for now 16:10:24 several patches, including this are now in the repo. And a new binary snapshot+getretro archive is up 16:29:50 ok *gets the repo* 16:33:18 so :: is gone now. goodbye, dear friend :P 16:34:31 * crc misses it just a little 16:34:57 ah well, it's probably better this way 16:35:17 the less there is to confuse a newbie the better 16:36:04 yeah 16:36:27 : :: here ] ; 16:36:36 I'll define it in the 'compat' module 16:42:11 tinyurl.com/elaj9 16:55:29 --- join: fma_steve (n=dsl@69.182.63.51) joined #retro 17:43:18 docl: I just repaired your subdomain on rx-core.org; I had forgotton to fix that when I switched to boa :( 17:51:54 --- join: jas2o (n=jas2o@WNPP-p-203-54-32-80.prem.tmns.net.au) joined #retro 18:03:19 --- part: jas2o left #retro 18:17:55 bbl, movie 18:23:40 --- quit: fma_steve (Remote closed the connection) 18:25:56 --- quit: nighty (Remote closed the connection) 18:33:24 --- join: nighty (n=nighty@sushi.rural-networks.com) joined #retro 18:36:58 --- quit: nighty (Remote closed the connection) 22:41:37 --- quit: Shain (Read error: 110 (Connection timed out)) 23:05:37 --- join: Shain (i=steve@c-67-161-56-76.hsd1.ca.comcast.net) joined #retro 23:59:59 --- log: ended retro/06.06.30