00:00:00 --- log: started forth/19.04.30
00:12:47 --- quit: jedb (Ping timeout: 246 seconds)
00:19:39 --- join: jedb (~jedb@185.128.24.51) joined #forth
00:47:56 --- join: mtsd (~mtsd@94-137-100-130.customers.ownit.se) joined #forth
00:54:27 --- quit: PoppaVic (Ping timeout: 276 seconds)
01:40:16 --- join: dddddd (~dddddd@unaffiliated/dddddd) joined #forth
01:49:35 --- quit: mtsd (Quit: leaving)
02:03:38 --- quit: ashirase (Ping timeout: 258 seconds)
02:07:03 --- join: ashirase (~ashirase@modemcable098.166-22-96.mc.videotron.ca) joined #forth
02:57:33 --- join: dave0 (~dave0@108.060.dsl.syd.iprimus.net.au) joined #forth
03:01:42 <dave0> hi
03:02:47 <proteusguy> howdy dave0 
03:11:16 <dave0> hi proteusguy 
03:23:21 <WilhelmVonWeiner> I have a question about strings in memory I've nevver really thought about.
03:24:49 <WilhelmVonWeiner> When you're interpreting a string instead of compiling, does it go into some kind of scratch buffer or just live in memory forever?
03:55:02 <crc> depends. 11.6.1.2165 calls for the string to go into a temporary buffer (s") when interpreting (6.1.2165 doesn't specify interpret behavior), 6.2.0945 (c") does not specify interpret behavior.
03:55:19 <crc> specifically, for s", 11.6.1.2165 says:
03:55:21 <crc> Parse ccc delimited by " (double quote). Store the resulting string c-addr u at a temporary location. The maximum length of the temporary buffer is implementation-dependent but shall be no less than 80 characters. Subsequent uses of S" may overwrite the temporary buffer. At least one such buffer shall be provided.
03:56:40 <crc> (I use a circular pool of temporary strings in my forth)
03:57:56 <john_cephalopoda> crc: A single buffer would also do though. Since only one buffer is specified, any standard-compliant program must assume that there isn't more than one.
04:02:58 <crc> john_cephalopoda: I don't follow the standard :)
04:04:01 <crc> string handling in standard forth has always been annoying to me
04:07:39 <john_cephalopoda> Yeah.
04:11:50 <crc> http://forth.works/ea3688e8fc6fddaa32ed10079db16370
04:11:50 <crc>  describes my string words
04:24:25 <john_cephalopoda> That's a lot of words :D
04:29:24 <crc> I currently have 487 words in a standard build of my forth on a unix host, (319 in the base image + 168 host-specific)
05:11:36 --- quit: X-Scale (Ping timeout: 244 seconds)
05:45:00 --- join: X-Scale (HydraIRC@a94-133-38-140.cpe.netcabo.pt) joined #forth
06:53:58 <john_cephalopoda> I wrote a fun little morse program
06:54:00 <john_cephalopoda> http://thecutecuttlefish.org/tmp/morse.fth
06:54:21 <john_cephalopoda> If you got OSS emulation enabled in Linux, you can output morse on your speaker.
06:54:38 <john_cephalopoda> Usage: S" MORSE TEST TEXT" MORSE
06:59:24 <john_cephalopoda> WilhelmVonWeiner: Do you see any way to improve the program?
07:24:19 <WilhelmVonWeiner> delete it all and throw your computer out the window, free yourself from these digital chains and escape into the realm of the living, escape the virtual dead
07:24:27 <WilhelmVonWeiner> no lemme have a look
07:25:46 <WilhelmVonWeiner> I would turn it into a jump table
07:26:01 <WilhelmVonWeiner> so that instead of at runtime you're doing all these IF..THENs
07:26:13 <WilhelmVonWeiner> you just execute the word you want
07:26:56 <crc> john_cephalopoda: I did a quick port to retro, can't test the audio, but otherwise works ok: http://forth.works/46ebe0de7fe07ca5acd413f125fce5a3.html
07:27:27 <crc> I'd second the suggestion to use a jump table here
07:28:02 <WilhelmVonWeiner> The - . P syntax is really cool
07:28:59 <john_cephalopoda> Wow, looks cool.
07:29:10 <john_cephalopoda> crc: ^
07:34:15 <WilhelmVonWeiner> I would do something like: `: ;TARGET POSTPONE ; LATESTXT ; IMMEDIATE`
07:34:43 <WilhelmVonWeiner> `: ;target postpone latestxt , ; immediate` even
07:35:01 <WilhelmVonWeiner> then `create sounds` and do all your definitions
07:35:05 --- quit: tabemann_ (Ping timeout: 250 seconds)
07:36:57 <WilhelmVonWeiner> so MORSE can go something like `COUNT SOUNDS + @ EXECUTE` in a loop 
08:05:07 --- join: PoppaVic (~PoppaVic@unaffiliated/poppavic) joined #forth
08:08:57 --- quit: dave0 (Quit: dave's not here)
08:24:11 --- quit: reepca (Ping timeout: 246 seconds)
08:29:59 --- quit: nonlinear (Ping timeout: 245 seconds)
09:25:56 <Zarutian> okay, so you got a Morse output. What about input? And it should be independent on the 'fist' of the operator.
09:26:28 <PoppaVic> heh.. translators
09:27:59 <WilhelmVonWeiner> input appears to be a char array
09:49:49 <john_cephalopoda> What's the difference between COUNT and @ ?
09:50:40 <PoppaVic> ( a - a' c) and ( a - c)
09:50:44 <john_cephalopoda> Zarutian: Morse input is doable but not as trivial.
09:51:02 <PoppaVic> sorry, ( a - a' c) and ( a - w)
09:51:45 <john_cephalopoda> I'd have to check the amplitude and then decide if the pulse is long or short - which is not too much work. But then I would have to dynamically adapt to the 'fist' of the operator.
09:52:59 <PoppaVic> it's two translators - text2morse; morse2text.
09:53:46 <Zarutian> john_cephalopoda: one could use two leaky integrators with hysterisis self-tuning
09:54:10 <john_cephalopoda> A Forth neural network ;þ
09:54:19 <john_cephalopoda> You just train it on signals and it will learn :D
09:58:36 <crc> http://forth.works/83d56beb6ff6ddf0a331afb6bb23093e.html (removing the if's to use a lookup table, making it more idiomatic for retro)
09:59:37 --- quit: Zarutian (Read error: Connection reset by peer)
10:00:10 --- join: Zarutian (~zarutian@173-133-17-89.fiber.hringdu.is) joined #forth
10:02:04 <john_cephalopoda> crc: I saw retro forth being advertised in the German "4th Dimension" forth magazine.
10:02:47 <crc> john_cephalopoda: an old issue?
10:05:17 <crc> hmm, it's in the latest one :)
10:08:36 * crc wasn't aware it was still being listed there. Nice to see it :)
10:12:48 <john_cephalopoda> Apparently somebody is still paying for it
10:15:29 <john_cephalopoda> I should really check out retro for real some time.
10:18:52 <crc> it's been running in the magazine since 2005, so since the retro7/8 days.
10:24:46 <crc> retro is very much non-standard, but it has proven very useful to me over the roughly 18 years I've been working with it.
10:30:56 <john_cephalopoda> Well, Forth isn't really a standard but more a way of thinking.
10:35:44 <john_cephalopoda> The standard has some really inconvenient things that make working with forth a bit of a hassle.
11:03:23 <WilhelmVonWeiner> crc made a really great Forth
11:03:27 <WilhelmVonWeiner> my favourite personally
11:04:03 <john_cephalopoda> I haven't really done anything with retro yet, its paradigm is a bit more complex than standard forth.
11:04:15 <WilhelmVonWeiner> personally it's far simpler
11:04:37 <WilhelmVonWeiner> there's no parser you're potentially controlling, no input stream, no reading ahead
11:05:35 <john_cephalopoda> No reading ahead?
11:05:41 <WilhelmVonWeiner> no reading ahead
11:06:16 <john_cephalopoda> Ah, it does :A? for definitions etc.
11:06:19 <WilhelmVonWeiner> like ' and ['] and postpone
11:06:41 <WilhelmVonWeiner> prefix-based inspired by colorForth
11:11:29 <crc> john_cephalopoda: http://forth.works/book.html#syntax describes the prefixes and interpreter design
11:45:40 --- quit: Zarutian (Read error: Connection reset by peer)
11:45:55 --- join: Zarutian (~zarutian@173-133-17-89.fiber.hringdu.is) joined #forth
11:52:40 <john_cephalopoda> crc: Hmm, "make -f Makefile.linux" fails to compile on my system because the linker can't find strlcat and some others. Looks like those are BSD functions.
11:53:11 <john_cephalopoda> crc: You probably have to link against lbsd.
11:53:40 <crc> I have copies of those functions in a file that should be linked when using Makefile.linux...
11:55:26 <crc> interfaces/io/strl.c and interfaces/io/strl.h
11:56:04 <john_cephalopoda> cd interfaces && cc -O2 -march=x86-64 -pipe  -o ../bin/retro -lm rre.c image-functions.o io/filesystem.o io/floatingpoint.o io/gopher.o io/unix.o
11:56:07 <john_cephalopoda> io/gopher.o: In function `gopher_fetch':
11:56:09 <john_cephalopoda> gopher.c:(.text+0x14f): undefined reference to `strlcat'
11:56:48 <john_cephalopoda> I got a library libbsd.so.0.9.1 on my system.
11:57:27 <john_cephalopoda> But the math library libm seems to be the only thing that gets linked in.
11:59:24 <crc> it looks like your 'make' isn't using the Makefile.linux (the `-lm` comes at the end of the line on linux: "cd interfaces && cc   -o ../bin/retro rre.c image-functions.o io/filesystem.o io/floatingpoint.o io/gopher.o io/unix.o io/strl.o -lm")
12:00:36 * crc rather increasing dislikes dealing with the gnu toolchain...
12:02:24 <crc> what version of make?
12:05:24 <john_cephalopoda> GNU Make 4.2.1
12:05:56 <john_cephalopoda> "man make" says that -f works.
12:06:56 <john_cephalopoda> But when I do "mv Makefile.linux Makefile" and run make, it appears to do it right.
12:07:10 <john_cephalopoda> Only problem is that it doesn't seem to accept my PREFIX... *sigh*
12:07:24 <WilhelmVonWeiner> seems to be a problem with your machine
12:07:57 <crc> http://forth.works/56e043ebd8267d88876ec879a9cd02ae
12:07:57 <crc>  is a shell script to build it on linux, run via `sh filename`
12:09:12 --- quit: gravicappa (Ping timeout: 255 seconds)
12:09:20 <john_cephalopoda> Ok, got it to work now, with qhttps://bpaste.net/raw/54813aedd587
12:09:24 <john_cephalopoda> -q
12:15:05 * crc wonders if it would work if I used GNUmakefile instead of Makefile.linux
12:28:35 --- join: mtsd (~mtsd@94-137-100-130.customers.ownit.se) joined #forth
12:28:40 * PoppaVic chuckles
13:00:12 <crc> I've renamed Makefile.linux and added a troubleshooting subsection to the Linux build instructions
13:00:52 <crc> (in the dev branch, I'll add a bit to the main page relating to this soon)
13:41:13 --- quit: mtsd (Quit: leaving)
13:58:29 <john_cephalopoda> crc: I find it a bit hard to get into retro. It would be nice if basic syntax examples would be given (like "'Hello,_world! s:put" or "#3 #4 + n:put"), also "Source files must be written in Unu format." is mentioned several time with no explanation what Unu format is.
13:58:58 <john_cephalopoda> When I look it up, the first result is "Apply for a Fellowship - biolac.unu.edu"
13:59:08 * PoppaVic chortles
14:00:04 <john_cephalopoda> And for "unu format" (with the quotes), I get tons of romanian sites. Which are probably about a lot of things but not that specific format.
14:01:28 <john_cephalopoda> Looks like there's some handbook appendix that talks about it. Maybe linking it from the top might make sense.
14:01:59 <crc> http://forth.works/book.html#unu-simple-literate-source-files
14:02:23 <crc> (thank you for the feedback on these; I'll address them!)
14:03:41 <john_cephalopoda> Thanks for listening to my feedback :D
14:03:45 <crc> most of the documentation was added in this release, previously there were a handful of text files with some notes and a small-ish set of examples. I'm trying to improve this.
14:04:41 <jn__> does the word Unu have a deeper or alternative meaning?
14:05:50 <crc> unu is a maori word:
14:05:51 <crc> 1. (verb) (-hia) pull out, withdraw, draw out, extract.
14:07:05 <jn__> perhaps that's worth putting into the manual as a side note
14:07:31 <crc> will do!
14:13:56 --- join: dys (~dys@tmo-085-104.customers.d1-online.com) joined #forth
14:15:10 <john_cephalopoda> Aww, history support in retro (-i) would be nice.
14:15:51 <WilhelmVonWeiner> I don't think that kind of triviality should be a focus at the moment
14:15:54 <crc> I use rlwrap to achieve that
14:15:55 <john_cephalopoda> Also going back and forth in the text with arrow keys.
14:15:56 <crc> https://github.com/hanslub42/rlwrap
14:16:08 <WilhelmVonWeiner> iirc crc you were working on rewriting the kernel right?
14:16:45 <crc> WilhelmVonWeiner: mostly the standard library, not the kernel
14:17:07 <WilhelmVonWeiner> do you consider the kernel solved
14:17:11 <crc> WilhelmVonWeiner: the next release will separate the words better, making it easier to leave things out
14:17:47 <crc> WilhelmVonWeiner: the kernel may see small changes, but I don't expect to make any big changes to it in the next few releases
14:17:52 <john_cephalopoda> WilhelmVonWeiner: User interface design is not a triviality. It can make the difference between struggling and being productive.
14:19:20 <crc> WilhelmVonWeiner: the one wildcard is strings. I *might* change strings to be arrays instead of continuing with zero terminated. This would have impacts on the kernel.
14:19:52 * john_cephalopoda . o O ( Thinking Retro )
14:20:01 * john_cephalopoda . o O ( Starting Retro )
14:20:22 <crc> WilhelmVonWeiner: but that's not in the near future. That change would have impacts across the entire source tree, and if I do it, it won't be until I can be sure it'll be better and done right.
14:20:46 <WilhelmVonWeiner> john_cephalopoda: -i isn't much of a user interface though imo, it's for running a couple lines of code while writing a program in another file
14:21:44 --- join: nonlinear (~nonlinear@unaffiliated/discrttm) joined #forth
14:21:48 <crc> I don't use retro interactively very often. Generally I'm running programs written in it, or using an application (e.g., text or block editor) written in it.
15:25:13 --- join: reepca (~user@208.89.170.37) joined #forth
15:46:05 --- quit: dave9 (Quit: dave's not here)
15:46:44 --- join: dave9 (~dave@108.060.dsl.syd.iprimus.net.au) joined #forth
15:57:28 <john_cephalopoda> crc: Hmm, I ran into something confusing. *** 'Just_like_this! s:put ASCII:CR c:put *** (without the ***) is a totally valid retro program and runs in interactive mode. When I put it into a file though (with squiggles ~~~), I get no output.
15:58:17 <john_cephalopoda> It works in a file when I leave out the ASCII:CR c:put part.
15:59:17 <john_cephalopoda> When I replace the c:put with n:put, it outputs the text and 13, just as it should.
15:59:35 <john_cephalopoda> I got no clue what the problem with c:put in the file is.
16:01:53 <WilhelmVonWeiner> there is #retro by the way john_cephalopoda 
16:01:59 --- quit: X-Scale (Ping timeout: 246 seconds)
16:03:48 <WilhelmVonWeiner> well CR isn't NL
16:04:01 <WilhelmVonWeiner> CR is \r which is to the start of the line 
16:04:23 <WilhelmVonWeiner> your shell prompt is probaby overwriting the text
16:04:28 --- join: dave0 (~dave0@108.060.dsl.syd.iprimus.net.au) joined #forth
16:04:35 <WilhelmVonWeiner> try `nl` instead of `ASCII:CR c:put`
16:04:54 <WilhelmVonWeiner> bedtime, peace
16:05:34 <john_cephalopoda> nl works.
16:08:17 <dave0> hi
16:27:16 <crc> WilhelmVonWeiner: that is correct. `nl` does a line feed, not a carriage return (hence nl for new line instead of calling it cr)
16:29:11 <crc> (at the interpreter, the ok prompt calls nl to go to the next line; there is no prompt when evaluating from a file)
16:32:25 --- quit: dave0 (Quit: dave's not here)
16:59:03 --- quit: john_cephalopoda (Ping timeout: 276 seconds)
17:01:28 --- join: X-Scale (~ARM@83.223.225.104) joined #forth
17:12:53 --- join: john_cephalopoda (~john@unaffiliated/john-cephalopoda/x-6407167) joined #forth
17:37:56 --- join: rdrop-exit (~markwilli@112.201.166.63) joined #forth
18:26:02 <rdrop-exit> Good morning Forthwrights :)
18:28:16 <crc> hi rdrop-exit 
18:29:12 <rdrop-exit> Hi crc
19:17:54 --- join: gravicappa (~gravicapp@h109-187-32-71.dyn.bashtel.ru) joined #forth
19:36:54 --- quit: pierpal (Quit: Poof)
19:37:12 --- join: pierpal (~pierpal@host197-221-static.34-79-b.business.telecomitalia.it) joined #forth
20:02:27 --- quit: chunkypuffs (Read error: Connection reset by peer)
20:04:03 --- join: chunkypuffs (~chunkypuf@2a01:4f9:2b:16d5::1) joined #forth
22:23:01 --- quit: dys (Ping timeout: 258 seconds)
23:05:52 --- quit: rdrop-exit (Quit: Lost terminal)
23:59:59 --- log: ended forth/19.04.30