00:00:16 but with timestamps it's either all of the files or none of them 00:00:32 immibis, um. sure. but you could do that with updating the timestamp file listing as you did the backup 00:00:55 wait, you need two files, still easy 00:01:00 like updating the last modified date? but then it isn't the date it was last modified anymore 00:01:06 maybe some people didn't want to change their timestamps when backing up?? (dunno) 00:01:10 immibis, like keeping a list like: 00:01:23 \foo\bar 118273 00:01:23 there is only one time in DOS: last modified 00:01:27 \foo\quux 11231 00:01:29 and so on 00:01:45 last access and creation are supported in FAT32 (but not by FreeDOS for speed reasons) 00:01:54 and when backing up check the last modification time, see if it is newer than what is your list of files 00:01:57 if it is, back it up 00:02:00 otherwise skip it 00:02:04 easy 00:02:05 much more complicated than a simple bit on each file 00:02:13 immibis, that's true 00:02:22 AnMaster, my only regret is that we're stuck with Bash 2.05b, which doesn't support BashFunge :-( 00:02:38 Rugxulo, bashfunge? why does that sound familiar 00:02:41 (probably not THAT hard to port 3.0, though I'm probably not too handy in that area) 00:02:47 * AnMaster is seeing zombies and spooks right now 00:02:48 why? heh, uh ... 00:03:02 Rugxulo, bashfunge was really broken. And I won't fix it. 00:03:07 as in, dead project 00:03:11 considering I've only know about Befunge for less than 3 months, it can't have been that long ;-) 00:03:11 didn't work out well 00:03:22 *sniff* and it was such a good idea :-/ 00:03:42 (worked in Cygwin although I didn't test that hard) 00:03:49 Rugxulo, sure. But well the unlimited size turned out to be a pain 00:03:59 unlimited? oh, B98 again? 00:04:06 God forbid you do a B93-only version ;-) 00:04:08 plus some of the 98-stuff is missimplemented 00:04:22 Rugxulo, yeah. I'm not about to touch that code 00:04:30 lazy ;-) 00:04:33 Rugxulo, anyway, if I did, I would go bash 4 00:04:38 * Rugxulo thinks he called three people lazy today (in jest), heh 00:04:42 some really useful stuff there 00:04:52 meh, might as well go Perl or Python or Ruby or ... 00:05:03 (flavor of month) 00:05:03 Rugxulo, no. zsh would be one step in between 00:05:27 bash3, bash4, perl/python/ruby 00:05:33 isn't correct 00:05:37 bash3, bash4, zsh, perl/python/ruby 00:05:39 is correct 00:06:19 Rugxulo, anyway, isn't everyone else on bash4 nowdays? 00:06:31 no 00:06:34 I've seen many Linux distros still using 3.2 00:06:38 even Cygwin still uses 3.2 00:06:52 Rugxulo, well sure. I have a few around here. Arch, Ubuntu, Gentoo 00:06:57 seems to be 4.x all of them 00:07:15 Ubuntu? if you say so, it wasn't last I checked (but haven't gotten 9.10 yet) 00:07:24 Gentoo upgraded only somewhat recently, but yeah, it's 4.0 here. 00:07:41 can't check *bsd atm. Probably okay on freebsd. No clue about open or net 00:08:05 those obviously don't use Bash by default 00:08:13 Rugxulo, well yes 00:08:22 oh mac OS X is probably very outdated 00:08:35 probably, but I don't have a Mac 00:08:40 nor do I 00:08:44 only classic one 00:08:48 my Dad's five-year-old Panther (10.3.9) still had 2.05b also 00:09:04 Rugxulo, oldest one I ever used was tiger 00:09:09 forgot what it had 00:09:22 btw I saw someone using twm and tcsh today 00:09:29 probably 2.03 or such 00:09:37 -!- puzzlet_ has joined. 00:09:37 -!- puzzlet has quit (Remote closed the connection). 00:09:39 it was a teacher at university, starting a projector to demonstrate stuff 00:09:41 or maybe that was tcsh, who knows 00:09:53 it ran ubuntu, with console login, then starting X manually 00:09:59 had something called "ctwm" 00:10:12 and yes it was definitely csh, think it was tcsh even 00:10:27 xterm too 00:10:40 and that old black/white checker bg 00:10:50 pikhq, ^ 00:12:16 I have not the foggiest idea why he used ubuntu as the base for that. After all that sounds like a pain compared to using arch that way. 00:12:35 * SimonRC goes to bed 00:12:45 night → 00:13:06 * Rugxulo isn't sure ick input works correctly ... or else isn't doing it correctly 00:13:25 AnMaster: My god. 00:13:41 That's... Eeeeew. 00:15:23 ah, I see ... "ONE OH THREE" is Intercal input 00:25:49 -!- Sgeo has joined. 00:27:30 -!- Slereah has quit (Read error: 60 (Operation timed out)). 00:27:48 -!- Slereah has joined. 00:27:50 anyways 00:27:52 -!- Rugxulo has left (?). 01:08:22 -!- Gracenotes has quit (Remote closed the connection). 01:23:20 -!- oerjan has quit ("Good night"). 01:43:07 -!- fax has quit ("Leaving"). 02:19:26 -!- augur has joined. 02:21:54 does anyone have any ideas about debugging dynamic recompilers? 02:37:51 Cry. 02:37:53 Cry a lot. 02:42:19 didn't work :P 02:42:50 But it's cathartic. 02:46:09 "In medicine, a cathartic is a substance which accelerates defecation." <-- wtf? 02:46:39 Well, catheters can be quite cathartic. 02:49:10 lol it must be really bad code if it makes desmume crahs 02:49:13 crash 02:49:21 (ds emulator, which is what my recompiler is running on) 02:51:08 currently it runs the recompiler until a jump instruction, then runs the interpreter until the same instruction, etc... until something doesn't match or isn't implemented 02:51:18 but comparing all of memory takes ages 02:51:24 at each jump instruction 03:13:31 -!- puzzlet_ has quit (Remote closed the connection). 03:13:35 -!- puzzlet has joined. 03:38:57 -!- coppro has joined. 03:39:47 Are there any esoteric languages based on Conway's Game of Life? 03:42:52 yes http://esolangs.org/wiki/Golang 03:43:08 wait, now i can't delete that 03:43:09 damn 03:43:24 haha 03:47:01 REDGREEN is a superset of the game of life: http://catseye.tc/projects/redgreen/eg/life.red 03:47:44 wow that's pretty neat 03:48:32 -!- GregorR has quit ("Leaving"). 03:54:14 -!- augur has quit (Read error: 104 (Connection reset by peer)). 03:54:40 -!- augur has joined. 04:08:11 -!- augur has quit (Read error: 113 (No route to host)). 04:08:31 -!- augur has joined. 04:09:14 Heh. DVD-ROM disks that are claimed to last 1ka. Even if they last in readable shape for half that (500a), one probably isn't going to get any sane data off them... 04:14:48 -!- Gregor has joined. 04:22:44 a? 04:22:58 what's an a? 04:33:46 -!- Gracenotes has joined. 04:39:03 -!- coppro has quit (Read error: 110 (Connection timed out)). 04:47:49 -!- immibis has quit ("ChatZilla 0.9.85 [Firefox 3.5.5/20091102152451]"). 04:56:55 -!- augur has quit (Read error: 113 (No route to host)). 05:35:55 -!- bsmntbombdood has quit (Read error: 60 (Operation timed out)). 05:52:00 -!- bsmntbombdood has joined. 05:59:07 -!- bsmntbombdood__ has joined. 05:59:15 goddamn my connection sucks 05:59:21 did anyone here my nipple story? 06:06:24 -!- bsmntbombdood has quit (Nick collision from services.). 06:06:27 -!- bsmntbombdood__ has changed nick to bsmntbombdood. 06:24:39 AnMaster: It's not "happened once before"; by my count this was at least the fourth occurrence. I probably should save those ?-generated random numbers somewhere so that I could reproduce it. (Or just run the text-generation code again and again until it crashes; can't take that long.) 07:10:57 -!- jix has quit (Read error: 60 (Operation timed out)). 07:11:01 -!- jix has joined. 07:15:49 -!- augur has joined. 07:59:59 -!- clog has quit (ended). 08:00:00 -!- clog has joined. 08:04:36 -!- Sgeo has quit (Read error: 54 (Connection reset by peer)). 08:09:00 -!- Slereah has quit. 08:13:40 -!- Slereah has joined. 08:18:32 -!- bsmntbombdood has quit (Read error: 110 (Connection timed out)). 08:20:40 -!- bsmntbombdood has joined. 08:30:49 -!- Pthing has quit (Remote closed the connection). 09:08:06 -!- bsmntbombdood has quit (Read error: 113 (No route to host)). 09:09:40 -!- bsmntbombdood has joined. 09:50:16 -!- bsmntbombdood has quit (Read error: 110 (Connection timed out)). 09:53:20 -!- bsmntbombdood has joined. 10:15:33 -!- oerjan has joined. 10:16:20 -!- bsmntbombdood has quit (Read error: 110 (Connection timed out)). 10:18:50 -!- bsmntbombdood has joined. 10:19:18 fizzie, mhm 10:37:14 AnMaster: iwc 10:37:42 oerjan, indeed. 10:43:31 great.... zooming a pdf to 150% in evince and in okular gives completely different results 10:43:54 in evince the text is then about twice as big as in okular 10:58:51 I've always assumed that the "100 %" there is supposed to take into account the screen DPI and show it at "natural size"; come to think of it, I wonder if that's true. 11:01:01 For an A4 .pdf, evince's "100 %" seems to be pretty close to natural A4 size, though the width is about 7 mm off; xpdf's "100 %" is utterly too small. 11:01:46 And Adobe Reader's "100 %" is something like 3 cm too wide. 11:02:34 Ah, that's because Reader's configuration is not "Use system setting: 97 pixels/inch" but instead "Custom resolution: 110 pixels/inch". Don't know where that came from. 11:02:52 It defaults to a custom resolution. 11:02:54 Don't know why. 11:03:00 That's strange. 11:04:44 With the system default 97 ppi, it's the same width that Evince does. (The "correct" value -- leading to the right width -- would seem to be somewhere between 99 and 100.) 11:07:29 What's really strange is the xrandr output: TMDS-1 connected 1600x1200+0+0 (normal left inverted right x axis y axis) 367mm x 275mm 11:07:52 1600 points in 367mm would give exactly that 110 ppi. But it's very much incorrect. 11:08:12 -!- oerjan has quit ("leaving"). 11:09:11 fizzie, hm what about okular then? 11:09:30 I found no custom resolution setting in okular 11:09:47 nor in evince 11:10:45 okular's "100 %" seems to mostly match the xpdf one; I guess they were related, anyway? 11:18:15 fizzie, probably 11:23:45 -!- bsmntbombdood has quit (Read error: 113 (No route to host)). 11:26:53 -!- bsmntbombdood has joined. 12:10:23 -!- bsmntbombdood has quit (Read error: 113 (No route to host)). 12:11:05 -!- bsmntbombdood has joined. 12:18:09 hm ok I got a simple hello world to build as a classic PPC app under MPW now 12:18:16 can't make it work as 68k nor as FAT 12:18:30 so should be ready to take on ick if I pre-generate most files 13:03:27 oh hah it seems including the type "long long" makes the 68k compiler crash 13:44:09 -!- bsmntbombdood has quit (Read error: 113 (No route to host)). 13:44:37 -!- bsmntbombdood has joined. 13:56:17 -!- puzzlet_ has joined. 13:56:58 -!- puzzlet has quit (Read error: 60 (Operation timed out)). 14:53:33 -!- bsmntbombdood_ has joined. 14:53:46 -!- bsmntbombdood has quit (Read error: 113 (No route to host)). 14:53:49 -!- puzzlet_ has quit (Read error: 60 (Operation timed out)). 14:56:16 -!- puzzlet has joined. 14:56:18 -!- kar8nga has joined. 15:37:29 -!- bsmntbombdood_ has quit (Read error: 113 (No route to host)). 15:41:21 -!- bsmntbombdood_ has joined. 16:11:03 -!- ais523 has joined. 16:20:53 -!- oklofok has joined. 16:27:37 -!- BeholdMyGlory has joined. 16:29:21 -!- Gracenotes has quit ("brb"). 16:34:27 ais523, some progress with ick on mac 16:34:32 however some questions 16:34:40 the SIZEOF__BOOL thing in config.h 16:34:45 what do you do if there is no _Bool 16:34:50 set it to 0 16:34:53 or to the null string 16:35:02 I think the code accepts either as meaning no _Bool 16:35:04 ais523, and what is the difference between bconfig.h and config.h? 16:35:12 AnMaster: nothing unless you're cross-compiling 16:35:17 oh and I got bin2c to build. but I can't redirect IO do it 16:35:19 to* 16:35:27 nor give it command line arguments 16:35:36 if you are cross-compiling, then bconfig is for the system you build on, whereas config is for the system you're targeting 16:36:01 the output of bin2c doesn't depend on anything but the input files, so feel free to use a prebuilt version of the resulting .c files 16:36:18 ais523, same goes for oil btw I think 16:36:28 and same fix will work 16:36:30 ais523, for ick I have to figure out some way around it 16:36:35 command line bit I mean 16:36:36 are you sure you can't redirect, though? 16:36:42 write a wrapper that uses system and freopen, if you like 16:36:45 and use that to do the redirection 16:36:55 ais523, there is no shell to do it in. As for system(), no clue what it does 16:37:16 according to the Wikipedia article you linked, the shell-equivalent is pretty good at redirection... 16:37:51 ais523, it works for special MPW tools only. Which are not normal applications 16:38:02 ais523, oh and there are two ways to go: MPW (free download) or CodeWarrior (far from). The latter supports emulating some aspects of a terminal for your app. STDIO 16:38:10 Wikipedia said it worked for everything, by hooking syscalls 16:38:47 ais523, hum. didn't work for me 16:39:30 ais523, will check if I can get bin2c to build under MPW. with codewarrior I got that bit to build. 16:39:32 bbiab 16:48:13 -!- fax has joined. 16:52:07 ais523, what creator code should ick, bin2c and so on have? 16:52:12 codes* 16:52:21 that is 4 printable bytes. 16:52:27 not sure really 16:53:00 exactly 4 printable bytes. With printable bytes I mean not just alphanumeric, but also symbols. I have seen SIT! for example. And "MPS " 16:53:11 the latter is MPW in fact 16:53:13 make them all NUL characters 16:53:20 ais523, printable... 16:53:23 beh 16:54:24 ais523, oh and "long long" isn't supported by the 68k compiler 16:54:27 it is by the PPC one 16:54:42 ais523, strdup is missing. So is snprintf 16:54:48 AnMaster: config.h should catch those 16:54:57 they should be #ifdef-guarded 16:55:02 ais523, yes. But it wants size of long long somewhere in there 16:55:19 ais523, oh and I haven't got IO to work at all for 68k apps 16:55:26 as in, freopen to a file doesn't work 16:55:33 that works for PPC apps 16:55:45 (which is how I found out size of the types for it) 16:55:47 ah 16:56:22 ais523, "doesn't work" here means "compiles and runs fine, but no resulting file" 16:56:37 which is very strange 16:57:08 ais523, anyway, what is the file list for ick, avoiding any *.a (I haven't figured out how they work yet) 16:57:34 AnMaster: as in, the sources or the binaries? 16:58:17 ais523, list of all *.o that ends up in it 16:58:19 would work fine 16:58:36 ais523, I have tried figuring out from the Makefile.am but that was quite messy 16:59:28 AnMaster: it's the contents of ick_SOURCES, nodist_ick_SOURCES, and ick_LDADD 16:59:33 * AnMaster looks at bin2c.c.ppc.o 16:59:36 nice filename 17:00:11 ais523, what about the oil stuff. MPW makefile must do all the work in advance as you knopw 17:00:12 so src/feh2.c src/dekludge.c src/ick_lose.c src/fiddle.c src/perpet.c src/uncommon.c parser.y lexer.l idiotism.oil are the sources 17:00:14 know* 17:00:45 feh2.o dekludge.o ick_lose.o fiddle.o perpet.o uncommon.o parser.o lexer.o libidiot.a are the object files 17:01:03 basically, the files generated by oil don't exist until oil runs 17:01:08 ais523, the make here outputs a script. So it must be able to figure out exactly which files would be generated by oil 17:01:19 AnMaster: it depends on the input, though, as to how many there are 17:01:25 *.y and *.l won't work 17:01:40 AnMaster: see the prebuilt/ directory 17:01:41 ais523, well sure. But you surely see the issue here? I can't work around it in any sane way 17:01:43 -!- BeholdMyGlory has quit (Read error: 110 (Connection timed out)). 17:01:44 I thought of that one already 17:01:47 ais523, I know about that 17:01:57 and work around it in an insane way, then 17:02:02 say by invoking the make recursively 17:02:10 once to figure out how many files are needed, once to actually do the building 17:03:18 ais523, also I can't figure out a make rule for files in a subdirectory 17:03:29 well not a generic one 17:04:11 are you sure that using the make version there at all is the right thing to do, if it's that limited? 17:04:17 is there any shellscript-equivalent? 17:05:16 ais523, well, the make one is better than a plain script, becuase compiling and linking is complex. So I don't want to have to write out lots of full command lines. Unless you can tell me how to do functions in the MPW shell. 17:05:22 or at least variables. 17:06:03 I don't know how the MPW shell works at all 17:06:06 ais523, oh btw to add to the confusion. *.a is usually asm here 17:06:26 there's no actual need to library up libidiot.a 17:06:37 except that make understands one library better than an unknown number of object files 17:07:16 ais523, there are some sort of libraries. I have yet to figure out how to create those 17:17:42 ais523, about redirection 17:17:47 I get a very strange error 17:17:56 "Unable to create finder parameter list for this application" 17:18:18 actually, wait 17:18:24 that is due to the command line argument it wants 17:18:28 not due to redirection 17:25:26 ais523, I figured it out. For normal applications, all arguments on command line must be valid existing files 17:25:31 ais523, that sure is a showblocker 17:25:36 wow, that's weird 17:25:55 ais523, any way to work around it in ick and such? 17:26:22 it'd have to get its command line arguments some other way 17:26:30 ais523, standard input? 17:26:42 maybe 17:26:56 ais523, also even after creating the relevant file, it didn't work. Not sure why yet. 17:27:08 it runs now, but gives an error message 17:27:17 (the normal one for no arguments) 17:28:10 ais523, what about converting it to an MPW tool? 17:28:26 * AnMaster looks at an example MPW tool to figure out how messy that would be 17:28:49 okay answer to that is "quite a bit" 17:29:40 or maybe not 17:31:11 ah yes a special resource script 17:32:21 ais523, if this is to go upstream I will need a special subdirectory for mac specific files 17:32:28 any recommended directory name? 17:32:33 macos9 17:32:34 also where is the current darcs repo 17:32:39 I'm using 0.29 here 17:32:44 and, I actually can't rememebr 17:32:49 ah. 17:32:49 I thought you were hosting it 17:33:04 ais523, as I told you ages ago, that server went down. 17:33:14 must have been middle of summer or so. 17:33:29 ais523, and you said iirc you would move it. 17:34:01 ah 17:34:05 well, it's nowhere atm then 17:38:27 -!- oerjan has joined. 17:41:03 ais523, interestingly enough the MPW shell and MPW make makes use of some non-ASCII letters 17:41:13 like "greater than or equal to" for redirecting stderr 17:41:36 or some weird symbol I can't type on my pc for what \ would do in a normal makefile 17:44:32 ais523, oh this is funny. 17:44:39 Volume:foo 17:44:43 means foo on Volume 17:44:47 Volume:foo:bar 17:44:53 means bar in folder foo on Volume 17:44:58 and now the funny part: 17:45:00 :foo 17:45:04 means foo in current directory 17:45:11 heh 17:45:15 revese of UNIX-style 17:45:36 exactly. lucky thing I didn't have a volume called baudot.c 17:45:57 however, I wonder what that thing actually did, since it didn't give an error either 17:47:37 ais523, oh and I managed to lock up MPW by bin2c, when it didn't quit but read on stdin. heh 17:47:57 heh 17:48:18 ais523, the tool thing is similar to a dynamically loaded module 17:48:43 okay bin2c foo < :src:baudot.bin > :baudot.c 17:48:45 that works 17:48:53 (minus the "okay" first of course) 17:49:28 ais523, however, it seems all source files must be in the same directory, no subdirs allowed. Or I just can't figure it out 17:49:52 just write a script to move them all into the same dir, then 17:50:10 for implicit targets at least. Not sure about explicit argets 17:50:12 targets* 17:51:24 ais523, oh and you know that old style s thingy? 17:51:28 that looks somewhat like an f 17:51:38 instead of: all: foo bar 17:51:40 it uses 17:51:45 all foo bar 17:51:49 in the make file 17:52:04 wow 17:52:07 well, sometimes it uses two of them instead 17:52:09 I'm not sure why 17:54:08 ais523, let me take a screenshot... a few seconds 17:56:39 ais523, look at http://omploader.org/vMnJoag 17:56:58 the cursor there is from linux 17:57:20 which is odd, maybe due to screenshot app. Since it doesn't look like that to me 17:57:21 how strange 17:57:49 ais523, what? the cursor or the makefile? 17:57:57 the makefile 17:58:19 ais523, the window behind is that worksheet 17:58:32 think a cross between a shell and emacs *scratch* buffer 17:58:36 and you got it almost right 18:00:12 ais523, anyway, ick building the result itself won't work. You will have to print some lines for the user to run in the MPW shell 18:00:12 I bet 18:00:22 in fact it might be best to make ick too into an MPW tool 18:00:24 just a hunch 18:00:38 yes, probably 18:01:06 ais523, and yeah mpw tools can't run other ones. which is why make prints the commands for you to run 18:01:11 and why oil will cause a headache 18:02:04 ais523, and yes I have no idea about the separate src dir for prebuilt or generated stuff 18:02:38 ais523, btw the way to make things find the source dir is to add a line like: {ObjDir} {SrcDir} 18:02:40 near the top 18:02:50 you can't have several it seems 18:03:17 probably worth experimenting with that to figure out if you can have more than one src dir 18:03:28 ObjDir and SrcDir must of course be defined 18:13:35 ais523, btw do you know how classic mac os did memory management? It is rather horrible: http://en.wikipedia.org/wiki/Mac_OS_memory_management 18:16:04 ais523, aha found how to do multiple directories. You just list them all after that old style s 18:33:28 ais523, what to call the folder where generated files are placed during the build? 18:33:38 ais523, I can't use the object dir 18:33:41 for stuff like generated *.c 18:33:47 because that makes mpw make crash 18:33:57 you're meant to use an out-of-tree build anyway 18:34:00 try macos9/build 18:34:04 well, macos9:build 18:34:30 ais523, and object files? 18:34:44 me has no clue how to go upwards like ".." under *nix and windows 18:34:45 hm 18:34:52 you could use :buildobj and :buildc if you need separate dirs 18:35:06 ais523, :buildc: you mean 18:35:11 well, yes 18:35:12 you need a : at the end if it is a directory 18:35:19 otherwise it complains 18:40:02 ais523, ah found out some more details from reading python's macpath.py 18:40:09 On the Mac, relative paths begin with a colon, 18:40:09 but as a special case, paths with no colons at all are also relative. 18:40:16 Anything else is absolute 18:40:32 oh and :: means "parent dir" it seems 18:40:37 well that should solve some issues 18:40:53 -!- puzzlet_ has joined. 18:52:45 -!- puzzlet has quit (Read error: 110 (Connection timed out)). 18:54:26 -!- ehird has joined. 18:55:34 I am using a trackball! 18:56:12 ehird, works nicely? 18:56:25 Yep. 18:56:34 Need to increase the sensitivity, I think, but I'll leave it for now. 18:56:45 Vintage Logitech TrackMan Marble, circa like 1995. Excellent condition. 18:56:52 PS/2→USB adapter. 18:57:11 Also, I hope the pun in the topic isn't disparaging Go. That's Ken Thompson's project, you know. 18:57:20 — and it looks great, with some warts 18:57:37 ehird, iirc Gregor said it was horrible yesterday 18:57:53 Then he's an idiot. It has a good subset of the old Unix team working on it. 18:57:56 Ken Thompson, Rob Pike, etc. 18:58:08 And its design is mostly clean and elegant. 18:59:31 13:52:16 BTW, have you heard of "Go" yet? 18:59:31 13:52:20 (since you like D) 18:59:34 ... Go is completely unlike D. 18:59:52 bbl 19:00:20 14:04:14 Since people were talking about Google Go the other day, here is my brief opinion from poking arounda t it a bit: Probably the worst language design in at least a decade. Easily the worst toolchain design in the entire history of toolchains. If I never touch this language again, it will be too soon. 19:00:20 HAHAHA, you fail 19:00:42 Yes, the compiler structure is totally unlike any existing compiler. That's why it's much faster and cleaner — it's the same design as the Plan 9 C compilers. 19:00:49 (So, not unlike any existing compiler, but meh.) 19:01:23 14:05:23 Rugxulo: In terms of "real" languages. It's completely horrible. 19:01:23 You are completely off your rocker. Worst language in a decade? Are you absolutely serious? 19:02:13 14:07:32 Rugxulo: Its type system is insane, its toolchain is horrible, its method of "encapsulation" is capitalization, it has no consistent interoperability story, lesse... 19:02:13 Its type system is very usual with some exceptions, and some nice things. Its toolchain is VERY VERY VERY fast and simple. Its SYNTAX for encapsulation is capitalisation. 19:02:35 14:07:59 Rugxulo: THE BINARIES ARE NAMED PER FUCKING ARCHITECTURE 19:02:36 14:08:05 6g? 8g? WTF?! 19:02:36 Which means that cross-compiling is the same as regular compiling, an ADVANTAGE. 19:02:56 14:08:26 Fast is not a pro unless we're compiling in 1965. 19:02:56 It compiles the 100k line go code in seconds 19:03:06 That IS a pro. 19:03:22 14:09:01 It has fork-join-ish concurrency, which makes it "good" in a "hey if we restrict you to a ridiculous degree you won't make mistakes" sense. 19:03:22 It has coroutines. 19:03:24 Of a sort. 19:03:44 14:10:53 --- topic: set to 'Go is a no-go | http://tunes.org/~nef/logs/esoteric/?C=M;O=D' by GregorR 19:03:44 did ken thompson rape you or something 19:06:32 haha this trackball poops if i spin the wheel really fast 19:06:40 "fuck it, I'm not tracking that" 19:11:05 09:25:26 ais523, I figured it out. For normal applications, all arguments on command line must be valid existing files 19:11:06 09:25:31 ais523, that sure is a showblocker 19:11:06 09:25:36 wow, that's weird 19:11:06 09:25:55 ais523, any way to work around it in ick and such? 19:11:07 09:26:22 it'd have to get its command line arguments some other way 19:11:07 just create the files beforehand, duh 19:11:09 :D 19:22:53 Ken Thompson, Rob Pike, etc. 19:22:53 And its design is mostly clean and elegant. 19:22:53 Clean and elegant?!?!? 19:22:57 Syntactically clean maybe, semantically nonsense. 19:23:07 I am Gregor and I will be outraged and call it the puke's dog because it is not what I am used to! I am not required to provide arguments, only RAGE! 19:23:15 HEAR ME ROAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAR 19:23:26 I provided a bunch of arguments when I was arguing against it yesterday. 19:23:33 You didn't, actually. 19:23:36 -!- kar8nga has quit (Remote closed the connection). 19:23:38 You just called various aspects of it stupid. 19:23:52 1) Their half-assed structural subtyping is no replacement for inheritance. Inheritance is useless. 19:23:59 Erm 19:24:08 Inheritance isn't useless rather X-D 19:24:17 Inheritance is bad, and before you argue against that plz google some bit, especially c2 wiki. 19:24:37 It isn't half-assed, it's minimalist. It looks stupid because minimalism does look stupid. 19:24:41 Unix looked stupid. 19:24:55 It being no replacement is your opinion; clearly they are having no issues. 19:25:31 If anyone wants some real description and criticism of Go, http://scienceblogs.com/goodmath/2009/11/googles_new_language_go.php and http://scienceblogs.com/goodmath/2009/11/the_go_i_forgot_concurrency_an.php are good reads. 19:25:57 It's not worth my time arguing this, particularly not on #esoteric or with you, so I'm not going to. 19:26:26 Your arguments continue to stun. Hey, if I'd said that to AnMaster a wild pack of bears would be mauling me round about now. 19:26:31 Funny that 19:27:05 At least you had the chance to give us your interesting arguments: "It's horrible" "It sucks" "That compiler naming thing is stupid" "The toolchain sucks". 19:27:13 I will ponder them. 19:33:33 Also, http://code.google.com/p/go/source/detail?r=4a3f6bbb5f0c6021279ccb3c23558b3c480d995f is such an epic event in the history of computing that anything it's part of cannot possibly be wrong. 19:36:02 heh 19:36:40 (To idiots who don't get it: It's a change by Ken Thompson. [[Ken Thompson was once asked what he would do differently if he were redesigning the UNIX system. His reply: "I'd spell creat with an e."]]) 19:37:29 http://ads.stackoverflow.com/ads/MostWanted_728x90_Static.png ;; by whom?! 19:37:44 ehird, some of those files would be illegal file names. Oh and there are some other issues with it too 19:37:57 AnMaster: Hmph 19:39:39 * ehird tries to emulate the scroll wheel in os x 19:40:01 -!- Sgeo has joined. 19:40:27 ais523: think I'll pwn at Enigma with this trackball? :P 19:40:38 maybe 19:40:51 * Sgeo recently switched to Chrome 19:40:54 give me a level that requires precision 19:40:57 Sgeo: it's a good browser 19:41:00 much better than firefox 19:41:16 I'd like it better if, when switching to a tab, the tab wasn't momentarily blank 19:41:20 But I found an issue discussing it 19:41:23 It isn't 19:41:27 So I'll look into that 19:41:27 Fix your computer 19:41:39 ais523, why not an implicit rule for those bin2c generated files? 19:41:53 ehird, http://www.google.com/support/forum/p/Chrome/thread?tid=020198aa328835c0&hl=en 19:41:57 because I like being explicit 19:42:06 it makes it easier to track everything that's going on 19:42:08 ais523, code repetition 19:42:13 Sgeo: eh, wfm :P 19:42:27 ...? 19:42:33 works for... 19:42:36 Oh 19:42:49 * Sgeo tries the memory-model=high thing 19:42:51 okay need to reboot for this scrolling thing bye 19:42:57 -!- ehird has quit. 19:43:08 Wow 19:43:16 I just added it, and Chrome seems faster in general 19:47:13 ais523, does oil only depend on the generated file oil-oil.c? 19:47:29 yes 19:48:03 -!- ehird has joined. 19:48:03 ais523, btw / is valid in directory names on classic mac os 19:48:07 and file names 19:48:21 As on OS X, though not :. 19:48:31 / is just turned into : for the BSD subsystem. 19:48:37 ehird, well : isn't here of course 19:48:46 : isn't allowed because it'd become /. 19:48:52 Or rather 19:48:56 Because old volume-style things are still supported 19:48:58 iTunes uses them 19:48:58 / is just turned into : for the BSD subsystem. <-- seems backwards 19:48:59 Anyway 19:49:07 AnMaster: It's so you can name things with / in them 19:49:14 "Foo and/or bar" 19:49:22 hm 19:49:27 Let's talk 19:49:27 so 19:49:27 that 19:49:28 I 19:49:28 can 19:49:29 test 19:49:29 this 19:49:31 scroll 19:49:38 Beautiful. 19:49:42 annoying 19:49:49 Your mom is annoying. 19:49:53 Now to try Enigma. 19:50:25 ais523: A level that requires precision plz 19:51:36 * ais523 looks 19:52:18 VI/#15 19:53:21 or V/#95 I suppose, that one's rather infamous 19:53:28 but it does require loads of precision and hardly any thought otherwise 19:53:34 I think this trackball is too good to make me better at Enigma... picking up all my little tremors and the like. 19:53:35 xD 19:54:25 95 requires speed 19:54:26 to get over ramps 19:55:03 it's one of the hardest dexterity levels I know 19:55:41 ais523, oil fails to build 19:55:46 what's the error? 19:56:11 get "Error: illegal combination of types" in the MPW header SizeTDef,h 19:56:16 this was a good deal for £20.07 shipped 19:56:24 heh, ,h 19:56:27 * Sgeo is now in love with Chrome 19:56:29 err 19:56:30 .h 19:56:31 why didn't they make it use mac os filetypes? :( 19:56:33 AnMaster: oh 19:56:36 Sgeo: good 19:56:38 ehird, I can't copy and paste from sheepshaver 19:56:42 finally, you like a good thing 19:56:43 Hopefully, this won't cause issues with other programs, though 19:56:54 AnMaster: old mac os had no file extensions, so i thought they were emulated with , 19:56:57 but they're being emulated with . 19:57:03 real filetypes would be more fun 19:57:27 ehird, hah no. The creator is 'MPS ' (without the quotes, with the space) and the file type is 'TEXT' (again without the quotes) 19:57:46 creator 'MPS ', file type 'CHED' 19:57:51 source code being 'CSRC' 19:57:55 I can actually close other tabs without watching the browser freeze1 19:57:59 oh great 19:58:01 (CHDR is also ok for headers, whatever) 19:58:04 editor doesn't display line numbers 19:58:05 :-D 19:58:07 eh 19:58:09 neither does mine 19:58:15 well it does at the bottom 19:58:23 ehird, this one doesn't 19:58:31 Sgeo: you can hold down ctrl+t for 10 seconds with chrome without it lagging much 19:58:32 it's great 19:58:33 * AnMaster begins counting the 16 lines from the top the error was on 19:58:44 okay 19:58:45 so it was: 19:58:50 AnMaster: and then look at the most suspicious line 19:58:52 typedef unsigned int size_t; 19:58:54 that it didn't like 19:58:55 fun how cpp fucks up everything innit. 19:58:57 why on earth 19:59:07 change it to 19:59:11 typedef unsigned size_t; 19:59:13 ehird, actually I know why 19:59:14 i bet it'll work 19:59:17 it was due to config.h 20:00:02 /* Define to `unsigned int' if does not define. */ 20:00:03 /* #undef size_t */ 20:00:05 well 20:00:11 there was no sys/types.h 20:00:13 so I did define it 20:00:20 however, it does exist in stdlib.h and such 20:00:22 sigh 20:00:29 just #include 20:00:59 ais523, can you make it check in stddef.h instead? That is the standard place for size_t 20:01:09 sys/types.h is POSIX specific 20:01:14 AnMaster: do you want this trackball? it's absolutely wonderful, and with a shallow learning curve, it's just that i think my hands are too small :P 20:01:21 POSIX specific? Hope you use it in cfunge! 20:01:22 AnMaster: send a bug report to the autoconf people, then 20:01:26 ehird, you pay for postage? 20:01:30 but I'll make that change sometime later if I remember 20:01:33 Um ... yes? 20:01:43 ehird, what is the catch 20:01:52 Oh I thought you meant 20:01:56 do the UK people pay for postage 20:01:56 xD 20:01:59 -!- zzo38 has joined. 20:02:09 No, I'm not paying postage :-P 20:02:13 Do you know anything about writing operating systems? What is the code for loading data from the boot disk into RAM? 20:02:30 slv %r, ($pla) 20:02:32 ttr 20:03:23 ais523, which programs depends on config.h and which on bconfig.h? 20:03:52 stuff that's only used during the build depends on bconfig.h 20:03:54 such as oil and bin2c 20:03:58 I found a page on the wiki called [[Sammich]] but it seems it so far is only a example program, do you think anything should be done with that? 20:04:03 everything else depends on config.h 20:04:18 ais523, right. So only oil and bin2c? You used "such that" 20:04:23 zzo38: hopefully the author would come back later 20:04:29 AnMaster: those are the only two I can think of right now 20:04:41 and yes, claiming that size_t existed in sys/types.h worked. oil compiled 20:04:53 ais523: OK. 20:04:57 AnMaster: It's ergonomic (fits your whole hand!) and has a 3D scrollball! All you have to do is configure your computer to do it. And the clever thing is that the trackball IS the scrollball! No moving your hand! Um, okay, so technically it's just making your system emulate a scroll wheel with the trackball; shut up. 20:04:59 * AnMaster test runs oil with no arguments on nix to see what happens 20:05:26 ah, does not give an error. No good then 20:05:31 because that will lock up MPW 20:05:32 AnMaster: you have to give it an input file, IIRC 20:05:41 and I have not yet figured out how to send EOF 20:06:05 ehird, for left hand? 20:06:17 Right hand. 20:06:30 ehird, meh. I'm using left atm for my mouse. 20:06:31 You say you use a mouse in both, so quit whining. :P 20:06:44 ehird, yes indeed I do 20:06:50 Um, it's vintage! Circa 1995! 20:06:54 It has the oooooooooooooold Logitech logo. 20:07:22 Not that old one; the one before that. 20:07:39 I already started writing a operating system, I might call it 4RM, or maybe it should be called CHAOS instead (for "CHAOS Has An Operating System") 20:07:44 http://www.goodlogo.com/images/logos/small/logitech_logo_2485.gif 20:07:46 This one. 20:08:04 It, um, has screws in it for easy user-servicability? 20:08:09 And an easy way to pop out the trackball! 20:08:15 And a quite-long PS/2 cord! 20:10:03 ais523, where does oil put the files? Current directory won't work. And changing current directory seems to confuse stuff 20:10:13 My thumb sure is getting good exercise. 20:10:17 I think it's current directory 20:10:29 It, um, has screws in it for easy user-servicability? <-- very nice 20:10:31 ais523, meh 20:10:43 AnMaster: Yes, it's refreshingly non-user-hostile. 20:11:03 ehird, it was expensive as new? 20:11:06 AnMaster: you could modify the code so that a particular #definition in bconfig.h put them somewhere else 20:11:13 AnMaster: Nope 20:11:20 ehird, what the hell... 20:11:22 AnMaster: Pretty average price for a niche mouse, I imagine. 20:11:27 Let's say £25? 20:11:30 ais523, hm. 20:11:31 Probably less, even. 20:11:31 -!- zzo38 has left (?). 20:11:38 ehird, mhm 20:11:39 Computer stuff was cheap in 1995. 20:12:33 It's basically new, anyway; it came from some sort of warehouse and the only imperfections are a slight crack in the plastic around some hole that I don't know what it is (barely noticeable) and some sticky stuff on a small part of the end of the PS/2 cord. 20:12:48 Although it might need some lubrication to work perfectly, not sure. 20:13:48 It has dots on the wheel???? I CANNOT SELL THIS ANY MORE 20:13:57 No no wait! Convenient indentations on the three buttons help finger placement. 20:15:38 -!- Asztal has quit (Read error: 60 (Operation timed out)). 20:16:06 -!- Asztal has joined. 20:16:39 ehird, dots? 20:16:58 ehird, you don't like that trackball? 20:16:59 Little black spots on the red ball in an intriguing pattern. It's so that the optomechanical (I think) magic can work. 20:17:12 I love it! I just need to grow my hand to use it properly, I think :P 20:17:22 ehird, just wait a few years then? 20:17:32 most likely it will work perfectly in a few years then 20:17:45 Well, sure, that would work. 20:17:45 ehird: the (Microsoft, and therefore Logitech) trackball I have at home has a red ball whose undersurface is made out of little rings of red 20:17:52 ais523: Funky. 20:17:54 (it has a plastic covering on top of that, though) 20:17:55 Wait... 20:18:00 Microsoft, and thterefore Logitech? 20:18:02 *therefore 20:18:12 What crack-smoking planet are you on? xD 20:18:17 ehird: Microsoft-branded mice and trackballs are made by Logitech, IIRC 20:18:19 at least, they used to be 20:18:22 Really/ 20:18:25 *Really? 20:18:26 just they have the Microsoft logo on 20:18:29 ais523, "used to" 20:18:30 That must be why they're good. 20:18:55 ehird, my first MS mice had the logitech logo on the bottom of it. in small print. My current one doesn't 20:18:55 Wikipedia (LOGO) Forever 20:18:57 Fuck off, Wikipedia. 20:18:58 AnMaster: for all I know they still are 20:19:05 Fuuuuuuuuuuuuuuuuuuuck Offffffffffffff 20:19:16 ehird, what? 20:19:27 On Wikipedia pages. 20:19:44 apparently someone found out when they opened up a mouse and there was a logitech logo on the circuit board 20:19:49 ais523: what type of trackball is it? thumb? 20:19:52 (this is on a more recent mouse that didn't have the logitech logo anywhere visible) 20:19:54 ehird: yes 20:19:54 or fingertips? 20:19:59 thumb trackball 20:20:01 ais523: right or left? 20:20:04 ais523, hm 20:20:05 for a right hand 20:20:15 (because everyone in my family's right-handed) 20:20:21 colour? (basically equivalent to age) 20:20:26 grey 20:20:33 light or dark 20:20:35 dark 20:20:40 recent then 20:20:44 2000s 20:20:48 yep, sounds about right 20:20:50 or slightly before, maybe 1998 at earlier 20:20:52 *earliest 20:21:20 ais523, oil sure takes a long time, or is locked up. 20:21:23 it's new enough that the computer that was bought at the same time has an Intel integrated 3D graphics card 20:21:36 AnMaster: it shouldn't take /that/ long 20:21:36 ais523: oh, then it'll be the latest model, probably 20:21:39 ais523, it certainly doesn't ever leave yeild the cpu. 20:21:42 nobody makes *new* trackballs any more 20:21:44 so hard to find out 20:21:51 it's just a bison/yacc-generated compiler 20:22:02 ais523, mouse can be moved, the spinning waiting thing is frozen though 20:22:17 ais523, 40 seconds so far 20:22:26 ais523: will you help me get c-intercal working on macintosh system 6? 20:22:27 something wrong? 20:22:43 AnMaster: it shouldn't be doing that 20:22:53 ehird: well, I've been helping AnMaster do it for mac os 9 20:23:14 Yes. System 6 is far better, and very different. 20:23:17 ehird, I have it part working. bin2c works perfectly. 20:23:26 Macintosh System 6 is not Mac OS 9. 20:23:33 Take your uncouth modernity elsewhere, heathen. 20:23:33 ehird, I build stuff as MPW tools. Should be portable to 68k 20:23:50 MPW is available for System 6, I think. 20:23:52 ehird, oh except that long long isn't supported, and I have no clue about the type sizes 20:23:53 But seriously. 20:23:58 Everything is different. 20:24:05 * ais523 downloads PDP-11 emulator 20:24:13 http://www.xkcdb.com/?6753 20:24:15 ehird, well sure. 20:24:21 System 6 is from such an era that it can only address 8 MiB of RAM. 20:24:29 ais523, there is oilout00.c. It is empty 20:24:36 AnMaster: seems dubious 20:24:40 ais523, do you accept both \r and \n or only \n? 20:24:41 And its standard UI is entirely black and white. 20:24:44 something's almost certainly gone wrong 20:24:52 http://upload.wikimedia.org/wikipedia/en/9/96/Sys6screenshotbusy.png ;; Anything like Mac OS 9? Didn't think so 20:24:59 ais523, I think the C compiler does some sort of \r \n swapping for IO by default 20:25:05 and the rules for line endings with OIL are that \n is always accepted, \r\n may leave stray \r in the output but is otherwise accepted 20:25:10 ais523, somewhat like windows does \n <-> \r\n 20:25:18 ais523: classic macintosh is just \r 20:25:25 ooh, and I read files in binary mode 20:25:28 that could explain al ot 20:25:29 ehird, yes but MrC swaps \r and \n 20:25:30 *a lot 20:25:34 Mister C! 20:25:36 AnMaster: even for binary-mode files? 20:25:37 ais523, no such thing iirc? 20:25:55 AnMaster: fopen("filename","rb") 20:25:55 ehird, yeah I thought that too. But it actually means Macintosh RISC C compiler 20:26:07 RISC is such a lovely name. 20:26:08 RISK 20:26:10 It's like... 20:26:11 Daring. 20:26:19 ais523, I'm unable to locate docs atm. So I don't know if b affects anything 20:26:22 This architecture? Exciting. MACINTOSH RISK C COMPILER 20:26:30 AnMaster: my guess is it causes the program to read \r as \r 20:26:39 ais523, and that doesn't work? 20:26:42 you'll therefore want to change the newlines to \n in the original idiotism.oil source code 20:26:47 if they're as \r atm 20:27:00 either that, or modify oil.y to be able to handle classic Mac newliens 20:27:01 *newlines 20:27:02 ais523, they should be \n due to having just copied the file 20:27:09 hmm 20:27:16 try changing the newlines to some other format to see what happens 20:27:43 let me open it in BBEdit 20:27:56 BBEdit started on System Software 6! 20:27:59 SIX POWER! 20:28:00 BBEdit lite claims it is unix line endings 20:28:07 ehird, do you happen to have the non-lite version? 20:28:18 No. 20:28:21 Well, maybe for System 6. :P 20:28:21 damn 20:28:31 ehird, I have bbedit lite 4.6 20:28:40 last I have been able to find and extract 20:28:43 Try the google 20:28:49 found a later lite version but the download was corrupt 20:28:52 ehird, I did... 20:29:00 Okay :P 20:29:34 ehird, tell me if you have better luck however 20:31:09 ais523, anyway, debugging is near impossible on this. Since I have to restart the system when it locks up. Can't get the key combo for macsbug to work 20:31:18 it was cmd-alt-power button iirc 20:31:26 which I have no way of pressing 20:31:51 Sure you do, just press keys until it happens 20:32:26 ehird, it doesn't. I checked in the key mapping file for sheepshaver. There is no such mapping 20:32:33 Dun dun DUN 20:35:24 ehird: do you think you'll get C-INTERCAL running on an emulated Mac OS 6 before or after I get it running on an emulated PDP-11? 20:35:38 after, I'm terminally lazy 20:35:54 Hey, the ball seems to have lubricated itself. Maybe my skin oils helped. 20:41:13 ais523, how to make oil print some sort of trace info? 20:41:24 ais523, like processed lines or such 20:41:31 AnMaster: by telling Bison to and regenerating the file 20:41:33 since it is generated code... 20:41:42 eh, i wish this scrollwheel emulator didn't move the pointer 20:41:42 ais523, how do you tell bison that hm 20:41:51 and didn't stop when the pointer reaches a screen edge 20:51:12 that's kind-of a silly scrollweheel emulator 20:51:26 it's meant to emulate the click-and-drag hand tool in e.g. pdf readers 20:51:43 but i inversed the axes so it's basically a hold-middle-button-for-scrollbar thiingy 20:51:44 *thingy 20:51:48 except for, you know, those two flaws 20:53:19 I'm becoming better at using this trackball, I think. 20:54:47 It's weird having full mouse movement when clicking. 20:55:07 gah, why does PDP-11 UNIX v5 use "chdir" not "cd"? 20:55:13 chdir() 20:55:31 ais523, how does convickt use those bin2c generated files? 20:55:42 it doesn't, it uses the originals 20:55:47 the generated files are linked into the standard library 20:55:48 ais523, oh? How? 20:55:54 ld 20:55:56 presumably 20:55:57 AnMaster: by fopening them and looking at the contents 20:56:00 how did you think? 20:56:06 http://imgur.com/5PUVF.jpg 20:56:07 oh, the generated files are .c files 20:56:14 so they're compiled then linked in the normal manner 20:56:17 ais523, why does it do it that way rather than use those generated c files? 20:56:27 self-containedness 20:56:43 basically, convickt looks at the .bin files (so you can add more) 20:56:55 ais523, you could add more anyway but ok 20:56:59 whereas libick.a contains a compiled version so the resulting programs are self-contained 20:57:05 (not on the fly though) 20:57:11 ais523, why is it needed in libick.a? 20:57:19 I/O, I gueess 20:57:21 *guess 20:57:23 AnMaster: so that it's possible for generated programs to do I/O 20:57:57 ais523, true, but isn't that all in ASCII? 20:58:15 well 20:58:26 hmm... how am I going to shut this ancient UNIX down? 20:58:31 it doesn't seem to have any of the usual methods 20:58:31 ais523: kill init 20:58:36 ehird: I was just about to try htat 20:58:37 or use init 1 20:58:39 etc 20:58:41 init 0, maybe 20:58:50 ais523: how is your emulation environment set up? 20:58:55 I am very much intrigued 20:58:58 ehird: I killed init, and it's still running 20:59:03 as in, the simulation 20:59:03 ais523, PDP-11? You didn't try ick on it before? 20:59:06 whereas init itself is dead 20:59:07 AnMaster: no, I didn't 20:59:08 ais523: done init 1 already? 20:59:11 which clearly needs correcting 20:59:14 also, heh 20:59:17 can you use the shell? 20:59:17 ehird: I'm not entirely sure where init is 20:59:21 ehird: yes, sh still works 20:59:27 find / -name init 20:59:30 or is there no find? 20:59:30 f: 1 20:59:39 -name: not found 20:59:57 man: not found 20:59:58 hmm.... 21:00:11 find init 21:00:20 find / 21:00:27 ais523: um, is this gui or command line thing? 21:00:27 ugh, I just control-Ded my login shell 21:00:30 and it seems to have frozen 21:00:33 also, it's command-line 21:00:34 hit me up with ssh and i'll try and get it to shutdown 21:00:44 ((((also play around with it)))) 21:01:19 ehird: I killed the emulator with SIGTERM, that seems to have worked 21:01:30 yeah but I'll do it properly! :( 21:02:14 ais523, issue. You depend on unistd.h in convickt. You do not use the config.h check for it 21:02:22 AnMaster: well caught 21:02:27 ais523: how much do i have to pay you :| 21:02:27 on line 29 in convickt.c 21:02:38 -!- Pthing has joined. 21:02:40 ehird: download and run the emulator yourself if you like 21:02:48 what, it comes with unix? 21:02:56 ehird: yes 21:03:01 well, has a separate download for it 21:03:04 link 21:03:07 http://simh.trailing-edge.com/ 21:03:15 trailing edge, heh 21:03:19 "software kits" contains an image of a system installed with UNIX 21:03:22 methinks pdp-11s are much older than the trailing edge 21:03:25 three, in fact 21:03:36 "If your host system is Alpha/VMS," 21:03:38 ais523, what in unistd.h is it that you use in that file? 21:03:41 Congratulations! You don't need a vintage system emulator. 21:03:44 AnMaster: not sure offhand 21:03:44 ais523, because it is a blocker atm 21:03:53 take it out and see 21:03:53 remove it and see what the error is? 21:04:03 ais523: you're using V5 unix? 21:04:05 ehird: yes 21:04:07 just checking 21:04:10 nice and old 21:04:18 V1 would be nicer of course :P 21:04:20 because if ick runs on that, it probably runs on v6 and v7 21:04:21 Death to SysV! 21:04:28 sysv changed a lot 21:04:36 i think it invented runlevels, for instance 21:04:39 so scratch my `init 1` idea 21:04:40 (also, file correctly identified unix_v5_rk.dsk as a "PDP-11 executable") 21:04:42 ais523, hm only gets a warning then 21:04:44 about line 90 21:04:48 AnMaster: what's the warning? 21:04:51 sec 21:04:56 wait 21:05:00 SysV is 5 21:05:02 ais523, oh just about assignment in if 21:05:06 Unix System V, commonly abbreviated SysV (and usually pronounced — though rarely written — as "System 5"), is one of the versions of the Unix operating system. It was originally developed by American Telephone & Telegraph (AT&T) and first released in 1983. Four major versions of System V were released, termed Releases 1, 2, 3 and 4. System V Release 4, or SVR4, was commercially the most successful version, being the result of an effort, marketed as Unix 21:05:06 Unification, which solicited the collaboration of the major Unix vendors. It was the source of several commercial common Unix features. 21:05:08 ais523, so not related 21:05:10 no idea what V5 is, then 21:05:11 ehird: also make a backup of the .dsk file, it's used to persist data in 21:05:16 there's no V release 5... 21:05:26 maybe they're releases of System I, or whatevef 21:05:28 *whatever 21:05:40 I very much doubt that 21:05:45 if(!(ick_datadir=getenv("ICKDATADIR"))) 21:05:47 ais523, that line 21:05:58 AnMaster: oh, for getenv 21:06:09 ais523, which is stdlib.h says my man page 21:06:12 that looks very ifdef-guardable 21:06:19 or if it's stdlib.h, just use that instead 21:06:20 http://en.wikipedia.org/wiki/Version_7 21:06:28 cool, V7 is the first release to widely escape ma bell 21:06:29 ais523, ... read what I said above 21:06:32 ais523, " ais523, oh just about assignment in if" 21:06:37 on the mentioned line 21:06:40 ais523: V7 looks like the best old unix 21:06:46 V5 isn't even mentioned in Wikipedia... 21:06:49 just V6 21:06:49 well, it's likely to have the most features 21:06:50 For historical relationships, just look at http://www.levenez.com/unix/unix_a4.pdf or something. 21:06:58 V6 is 1975... 21:07:12 V7 != SysV, just to clarify 21:07:13 anyway, I think I'll target V7 21:07:15 um, I think 21:07:17 because it introduced lex and yacc 21:07:19 ais523, the next few files for convickt crashed the emulator 21:07:21 so I'll be able to do a complete install 21:07:25 ais523: how boring :P 21:07:27 ais523, after spitting errors 21:07:37 ais523, too quick to see the errors though 21:07:37 ehird: less boring if you can do the whole thing, rather than relying on prebuilts 21:08:19 where do i put the unix files? 21:08:22 in a separate dir? 21:08:34 ais523, the clc-cset file crashes the emulator when compiling 21:08:35 PDP10 or PDP11, btw? 21:08:38 after some errors 21:08:41 ehird: PDP-11 21:08:45 From unix_a4.pdf, one path of descent goes V5 -> V6 -> USG 1.0 -> CB UNIX 1...3 -> System III -> System IV -> System V. 21:08:47 I'm unable to see what errors due to the speed 21:08:53 just "make" yes? 21:08:54 for simh 21:09:02 ais523, might try this on a real old mac. Should be a bit more stable 21:09:05 ais523: Well V7 isn't SysV. 21:09:14 wait, /bin/exit is an /actual program/? htf does that work? 21:09:25 ehird: ofc not, or Caldera wouldn't have released it under a BSD-style licence 21:09:40 admittedly, this is before they became tSCOg... 21:09:44 ais523: kill(meinproctree()->parent) 21:09:47 is how /bin/exit iwll work 21:09:52 *will 21:10:09 anyway, sysv was developed concurrently with unix up to the tenth edition... 21:10:17 ais523, is there a /bin/cd= 21:10:19 gcc -std=c99 -U__STRICT_ANSI__ -g -lrt -lm -D_GNU_SOURCE -I . PDP1/pdp1_lp.c PDP1/pdp1_cpu.c PDP1/pdp1_stddev.c PDP1/pdp1_sys.c PDP1/pdp1_dt.c PDP1/pdp1_drm.c PDP1/pdp1_clk.c PDP1/pdp1_dcs.c scp.c sim_console.c sim_fio.c sim_timer.c sim_sock.c sim_tmxr.c sim_ether.c sim_tape.c -I PDP1 -o BIN/pdp1 21:10:20 ld: library not found for -lrt 21:10:20 s/=/?/ 21:10:25 /* exit -- end runcom */ 21:10:27 main() { 21:10:28 seek(0, 0, 2); 21:10:30 } 21:10:31 Many new features were introduced in Version 7. 21:10:31 Programming tools: lex, yacc, lint, pcc, and make. 21:10:38 that's an incredibly short program 21:10:41 yeah, methinks targeting pre-V7 is crazy 21:10:45 ais523, and it makes no sense 21:10:55 end runcom = end running command, duh 21:11:04 hahaha 21:11:07 I just got kill to kill itself 21:11:14 suicide(1) 21:11:15 by guessing what process number it'd be allocated 21:11:17 *blink* 21:11:19 ais523: oi, help me compile simh :-P 21:11:30 ehird: mkdir BIN 21:11:32 make 21:11:37 that's all I needed to do 21:11:41 >_< 21:11:46 I have no -lrt 21:11:50 is the problem 21:11:52 what's it supposed to be 21:11:54 ehird, try dropping it 21:11:57 and see if it works 21:12:00 AnMaster: ? 21:12:01 ehird: what OS are you on? 21:12:02 since sometimes that stuff is in libc 21:12:03 ... 21:12:05 ah 21:12:06 dropping 21:12:07 in that sense 21:12:08 ais523: osx 21:12:11 so, bsd 21:12:26 ehird: export OSTYPE 21:12:29 then make 21:12:37 ifneq (,$(findstring darwin,$(OSTYPE))) 21:12:38 OS_CCDEFS = -D_GNU_SOURCE 21:12:38 else 21:12:38 OS_CCDEFS = -lrt -lm -D_GNU_SOURCE 21:12:38 endif 21:12:39 (according to the docs, you need that export command on OSX) 21:12:42 ehird, it is for "real time" stuff. like clock_gettime() and other high res and/or low latency interfaces 21:12:45 too bad it never sets OSTYPE 21:12:49 * ehird sets OSTYPE=darwin 21:13:00 make OSTYPE=darwin did it 21:13:13 congrats 21:13:13 ehird, was it mentioned in the readme? 21:13:14 or such 21:13:25 i didn't bother reading it, this package looks really crufty :P 21:13:29 or rather, compatible, i guess 21:13:44 "0readme_38.txt" isn't exactly hope-inspiring 21:13:51 ehird, it could have used autotools. You wouldn't have needed to check then. 21:14:02 It could also have done it entirely differently. 21:14:07 False dichotomy heyoooooooo 21:14:48 ehird: it's version 3.8 21:14:53 yes, but still 21:14:54 and the 0 is to sort it to the start of the list 21:16:05 * ehird reads the output of "help" and goes "._." 21:16:17 at something unixdiskname.dsk is what I need to do first, I think 21:16:20 and then go 21:16:34 False dichotomy heyoooooooo <-- of course. I was joking... 21:16:47 AnMaster: Oh. 21:16:54 Can you start doing more funny jokes? :P 21:17:04 ais523: what're the commands you need? 21:17:16 ehird, no. That would be breech of contract. 21:17:16 ehird: http://simh.trailing-edge.com/pdf/simh_swre.pdf 21:17:30 ais523: *look of disapproval* 21:17:34 AnMaster: *breach 21:17:52 also, http://imgur.com/5PUVF.jpg 21:18:04 damn this pdf is loading slowly 21:18:13 hey safari, why'd you gotta crash like dat 21:18:14 for UNIX v7, where \ represents newline, it's set cpu u18 \ set rl0 RL02 \ att rl0 unix_v7_rl.dsk \ boot rl0 \ @boot \ : rl(0,0)rl2unix 21:18:29 where the @ and : are prompts 21:18:38 sim> cpu u18 21:18:38 Unknown command 21:18:47 I'm doing BIN/pdp11 21:19:02 I'll just skip the cpu line 21:19:08 I guess calling pdp11 handles that 21:19:27 don't skip it 21:19:29 it's needed 21:19:39 It didn't work. 21:19:50 And besides, isn't that for the main simh binary? to select pdp11? 21:19:53 Also, i t's _rk 21:19:54 *it's 21:20:03 ehird: no, that's for the pdp11 binary to select what sort of pdp11 21:20:10 try cding to BIN and then trying again, as it works for me 21:20:42 Still doesn't work, and cpu isn't listed in help. 21:20:46 What, exactly, are you calling? 21:21:27 ehird: "set cpu u18" 21:21:29 not "cpu u18" 21:21:35 >_< 21:21:35 Oops 21:21:49 * ais523 wonders how to copy files onto the filesystem ther 21:21:51 *there 21:22:13 sim> set cpu u18 21:22:13 Disabling XQ 21:22:14 sim> set rl0 RL02 21:22:14 sim> att rl0 unix_v5_rk.dsk 21:22:14 sim> boot rl0 21:22:14 @boot 21:22:16 @ 21:22:18 -!- Gracenotes has joined. 21:22:18 So much for : being a prompt 21:22:23 ok, that's strange 21:22:33 oh, the instructions are different for v5 21:22:37 type unix at the @ prompt 21:22:39 rather than boot 21:22:58 then give the username as root, no password 21:22:58 Any other instructions different? 21:23:00 Like the cpu line? 21:23:25 the cpu line's the same, but you're supposed to use rk and rk0 not rl and rl0 21:23:35 Would explain the filenames 21:23:46 yes 21:23:49 RK02, as well? 21:23:51 instead of RL02 21:24:03 oh, you don't need that set line at all 21:24:03 nope 21:24:32 set cpu u18 \ att rk0 unix_v5_rk.dsk \ boot rk \ @unix \ login: root 21:24:37 Woot 21:24:41 I'm all V5'd up 21:24:42 you're in? 21:24:46 Fuck is this slow 21:24:59 needs moar backspace 21:25:08 heh, ^H backspaces on screen but not in reality 21:25:19 sweet, the kernel is actually called /unix 21:25:35 ehird: changes to the filesystem persist 21:25:43 i copied to pristine.dsk 21:25:52 I'm keeping the originals in the .zip files 21:26:06 ha! /usr/bin actually has files owned by the user bin 21:26:08 since /usr = /home 21:26:12 also, V5 has yacc 21:26:15 v7 has man 21:26:20 so shut your face with your V7 crap :( 21:26:29 cmoooooooooooon 21:26:36 V5 has the best cc ever, so cryptic 21:26:41 # cc -h 21:26:43 un: _maain 21:26:45 *_main 21:26:56 haha, and i still have an a.out 21:27:12 oh, if you hit backspace enter, it cancels hte command 21:27:55 oh, there isn't a way to kill INIT, it seems 21:28:01 *kill init 21:28:05 # ps 21:28:05 No swap 21:28:05 f: 0 +??^??~??[??J[R??S? ????? R? ? 21:28:06 f: 1 /etc/init 21:28:12 although if you send it SIGHUP, it drops to single-user mode 21:28:13 Interesting use of /etc. 21:28:30 ehird: this is before the FHS was invented, I think 21:28:35 Of course. 21:28:40 The FHS is Linuxcrap. 21:28:44 Interesting is in a good way. 21:28:44 also, that process 0 looks suspicious, but I'll chalk it up to a simulator hiccup 21:29:10 The FHS has one good thing about it: it is *a* standard. 21:29:26 However, not everyone supports said standard, so that doesn't do you much good. :P 21:29:38 Standards are overrated 21:29:42 this ps seems to be broken 21:29:49 it prints "No namelist" regardless of what arguments you give it 21:29:53 # /etc/init 21:29:54 No match 21:29:55 erm 21:29:56 # /etc/init 21:29:57 /etc/mtab: non existent 21:29:57 No match 21:30:10 and it hands 21:30:12 *hangs 21:30:42 I think these systems were designed to be turned off by pressing the off switch 21:30:48 like DOS was 21:30:51 I can do this! 21:30:55 erm 21:30:59 what number is hup again? 21:31:02 1 21:31:15 kill -1 1, catchy 21:31:20 okay, so that's single user mode; nothing happened 21:31:27 and init is still there 21:31:32 * ehird runs sync 21:31:34 good practice, you know 21:31:46 -!- fax has quit ("Leaving"). 21:31:52 * ais523 wonders what filesystem this is 21:32:13 RK? 21:32:18 or RL for you 21:32:30 * ehird notes: /etc/umount 21:32:34 may be useful for shutting down 21:32:44 obviously halting won't happen, but I think I can get rid of the unix 21:33:04 ehird: Playing with UNIX v5? 21:33:09 yup 21:33:11 haha, the v7 one comes with some mail in the mailbox 21:33:19 that says "Secret mail has arrived." 21:33:26 xD 21:33:30 (you can control-D the single user mode to get multiple-user mode) 21:33:45 (and root's password is root; it was the second password I tried) 21:33:56 # ls /dev 21:33:56 mem 21:33:56 null 21:33:57 tty8 21:33:59 ain't no disks therre... 21:34:01 *there 21:34:09 ehird: I must admit, I'm curious: is it possible to get any vaguely modern software working on there? 21:34:14 # cat /mailbox 21:34:15 From root Fri Mar 21 12:15:23 1975 21:34:15 yo root 21:34:15 # 21:34:21 pikhq: we're planning to get C-INTERCAL working on there 21:34:24 pikhq: well, ais523 is working on V7 (not SysV) and C-INTERCAL 21:34:25 ais523: YOU! 21:34:26 ais523: Spiffy. 21:34:26 You. 21:34:28 I am not so insane. 21:34:40 anyone have the source code for uudecode handy? 21:34:44 (something simple, like, oh, modern PCC?) 21:35:02 huh, /usr/c has a bunch of source and binaries 21:35:16 pikhq: Hey now, with V7 you can have ORIGINAL pcc! 21:35:21 And with V5? Original authentic cc, bitch. 21:35:22 ah, http://www.nr.com/utils/uudecode.c.txt 21:35:24 under BSD4 21:35:29 with a copyright date of 1983 21:35:32 this looks suitably old 21:35:53 no return type declarations 21:35:57 think you'll be fine! 21:36:09 ehird: Modern PCC is "original". Just maintained. :P 21:36:16 nah 21:36:21 pcc has been updated for decades 21:36:30 * ais523 uses cat to transfer the source of uudecode 21:36:36 Yeah, I know. 21:36:43 anyone want to guess where the disk device is? 21:36:48 wow, this emulation is slo 21:36:49 *slow 21:36:53 probably about original speed 21:36:58 You may actually be able to build some GNU programs on there. 21:36:59 heh, /usr/sys has headers in it 21:37:02 ais523: faster, i'm sure 21:37:07 pikhq: But I won't. :P 21:37:09 I'm using /usr/root as my home directory 21:37:21 ais523: ~ is / 21:37:24 ehird: cat's catting so slowly I can read the individual characters as they arrive 21:37:30 (at least some of them have shockingly ugly code for the sake of being able to build on K&R C compilers) 21:37:32 i know this because your mailbox goes there 21:37:34 ehird: yes, but that's a really fugly location to use 21:37:35 ais523, hm those were warnings not errors. And a real mac didn't crash 21:37:42 AnMaster: you tested on a real mac? 21:37:54 Though I suspect that's primarily GCC these days. 21:37:56 ais523, yes an old first model ibook 21:37:57 I keep mistaking this for DOS because it's such a pain to use and using dir(1)... 21:38:15 the fact that it's chdir, not cd, doesn't help 21:38:20 ais523, which has a very lound whining harddrive (was less loud when it was new) 21:38:25 * ehird writes some code with ED! 21:38:29 Ed! Ed is the standard editor! 21:38:30 Which I doubt could build *on* a v5 machine -- nowhere near enough RAM. 21:38:40 ais523, flat battery, glitchy power connector 21:38:42 # ed 21:38:42 i 21:38:42 ? 21:38:43 wow, this program seems to predate the # key 21:38:45 well guess why I prefer emulator 21:38:50 well, not quite 21:38:57 what I mean is, cat ignored all #s in my input 21:39:00 some sort of emulator thing? 21:39:06 ignores all special chars 21:39:08 including @ 21:39:08 for me 21:39:12 probably emulator thing 21:39:16 I mean, the prompt IS #... 21:39:23 ais523, "fatal error: Unable to open ick_lose.h"? 21:39:23 # echo 'ab#cd' 21:39:25 acd 21:39:29 # is obviously the backspace character 21:39:29 in uncommon.c 21:39:32 any clue? 21:39:36 you could set it to anything you want back then, after all 21:39:37 XD 21:39:45 AnMaster: where is ick_lose.h? 21:39:50 heh, no puts???? 21:40:01 ehird: I transferred the file by catting 21:40:04 # ls a#b 21:40:04 b not found 21:40:05 clever 21:40:15 ugh, this thing predates vi 21:40:19 duh 21:40:21 vi is a BSD thing 21:40:25 unix ignored it even after it was invented 21:40:25 ais523, ah include issue 21:40:26 thought so, probably 21:40:26 ais523: just use ed 21:40:34 ok, I'll have to remember how ed works now... 21:40:52 "Ed is the standard text editor." 21:40:52 # cc hello.c /lib/crt0.o 21:40:52 Multiply defined: /lib/crt0.o;savr5 21:40:53 un: _puts 21:40:55 scratch that idea 21:40:58 ais523: use a to insert 21:41:00 . to stop inserting 21:41:01 w foo 21:41:03 ^D to quit 21:41:05 . is print 21:41:07 ^ and $ work 21:41:10 // searches 21:41:18 I think you can do +1 and -1 to move around 21:41:20 that should do you 21:41:24 and s/// works too, doesn't it/ 21:41:28 maybe. 21:41:29 who knoows 21:41:31 *knows 21:41:36 still, I wonder how to type a literal # character 21:41:39 maybe I'll use ??- 21:41:41 * ??= 21:41:51 actually having a legit reason to use trigraphs would amaze me, but why not? 21:41:52 # cc hello.c /lib/libca.##.a 21:41:52 un: _puts 21:41:52 I guess it really does predate puts 21:42:00 I love the short names of the libs 21:42:05 -la -lc -lf -ly? fuck yeah! 21:42:35 woot, it has printf 21:42:48 # ed 21:42:48 a 21:42:48 main (##() { 21:42:49 printf("hello world\n"); 21:42:49 } 21:42:49 . 21:42:50 w hello.c 21:42:52 37 21:42:54 # cc hello.c 21:42:56 # ./he##a.out 21:42:58 hello world 21:43:00 # a.out 21:43:02 hello world 21:43:04 I can program, me 21:43:06 (sorry for the flood, I'm just happy) 21:43:08 I especially like . being in PATH 21:43:10 xD 21:43:14 Back then a security flaw was, uh, something that it doesn't mean today 21:43:52 -!- immibis has joined. 21:43:59 hmm, it seems to ignore $ 21:44:00 didn't Stallman specifically not add a wheel group to GNU su on the basis that he thought that people should have a chance to wrest control of a computer away from its current admins by guessing the root password? 21:44:10 ais523: probably 21:44:12 ais523: Yes. 21:44:17 he also had no password on unix for years 21:44:26 did that computer allow remote logins/ 21:44:37 probably once telnet was invented 21:44:40 he openly invited people to use his account 21:44:54 For ages, the easiest way to get online was to dial into the AI lab at MIT and log in as RMS. 21:45:07 oh, of course; telnet is modern 21:45:11 phone lines fuck yeah! 21:45:21 I imagine many more people used the RMS account without knowing what "RMS" meant 21:45:24 uudecode.c:40: External definition syntax 21:45:27 wtf does that mean? 21:45:28 I'd imagine. 21:45:33 ais523: "extern"? 21:45:36 also, use cc, not pcc 21:45:38 it's even more cryptic 21:45:41 I'd bet RMS got backronymmed 21:45:49 Remote Management System 21:45:54 oh, it doesn't like my trigraph 21:45:57 ofc, that was C89 21:46:04 ... You have *trigraphs*? 21:46:16 pikhq: I trigraphed the program I was sending because the # character was being ignored 21:46:16 someone backronym RMS into something resembling "open-access" 21:46:21 ais523: Ah. 21:46:26 remote m'login sir? :P 21:46:34 ais523: just write a c program that prints out # 21:46:37 with putchar 21:46:38 duh 21:47:00 ooh, I could probably write a c program that did s/??=/#/ 21:47:10 Either that or a C99 preprocessor. 21:47:36 * ais523 stty erase '^D' 21:47:40 wait, that's wrong... 21:47:44 * ais523 stty erase '^X' 21:47:53 ais523, I didn't run oil, but now I have something in oilout00.c on mac 21:47:57 WITHOUT running it 21:48:00 that is on the real mac 21:48:07 hmm 21:48:14 ais523, what is the raise(SIGSEGV) stuff in it? 21:48:15 for 21:48:26 backwards compatibility 21:48:50 I think 21:49:02 as in, that always used to segfault 21:49:10 and there was no reason to change the behaviour when we fixed the actual bug 21:49:26 '^X' isn't accepted 21:49:32 ehird: you on v5? 21:49:35 it was by v7 21:49:41 ah 21:49:42 # echo ^ls 21:49:42 a.out 21:49:42 hello.c 21:49:42 num.c 21:49:43 wat 21:50:06 ehird: kill character, probably 21:50:11 that deletes the whole line 21:50:16 ah, indeed 21:50:48 no sed aaaaaaaaaaa 21:51:51 # cat hello.c 21:51:51 num; 21:51:51 main(){ 21:51:52 num2; 21:51:52 num=rand(); 21:51:52 num2=num; 21:51:53 printf("%d\n",num2); 21:51:55 } 21:51:57 # cc hello.c 21:51:59 5: Lvalue required 21:52:01 7: num2 undefined 21:52:05 hmph 21:52:07 you need "auto num2;" in a function, don't you? 21:52:32 Unless you prefer "int" 21:52:37 no sed aaaaaaaaaaa <-- use ed 21:52:51 AnMaster: I am 21:52:53 Deewiant: meh! 21:53:05 Or "register"? 21:53:15 don't, you'll probably run out of registers 21:53:17 ais523: /usr/c is the c compiler sources 21:53:22 C compilers were rather literal back then 21:53:22 Or even "static" would work there 21:53:22 it's... upsettingly moddern, the code 21:53:26 I mean, not olde times at all, really 21:53:32 apart from its simplicity 21:53:42 * ais523 stty erase '^X' <-- what does that mean 21:53:52 uudecode.c:98: void undefined; func. main 21:53:53 set erase to control-x 21:53:54 duh 21:53:59 ehird, oh. 21:53:59 ais523: yep 21:54:09 ais523: ed, s/void//g :-D 21:54:14 I think that does all lines, maybe not 21:54:14 ? 21:54:23 this ed doesn't like s/// 21:54:33 ha 21:54:37 /void/d then 21:54:41 dunno how to do it for every line 21:55:00 ehird: you deleted the whole line 21:55:02 ais523, okay oil seems to require input file to use \r. 21:55:14 ais523: they should have had types on another line! 21:55:21 everyone knows the One True Declaration style looks like 21:55:22 void 21:55:26 main(int argc, char **argv) { 21:55:36 (—recent convert from straight k&r :P) 21:55:38 ehird, no. That is C89 21:55:43 (Though straight k&r has *no* return types.) 21:55:45 try K&R C 21:55:49 AnMaster: it's easy to munge into k&r c, though 21:55:59 remove the type decl line, and delete one word after ( and , 21:55:59 -!- fax has joined. 21:56:03 ehird, unprotoize 21:56:04 (who cares about types) 21:56:12 yay, finally compiled uudecode 21:56:43 ais523, oil works fine of the input file uses \® 21:56:45 \r* 21:56:47 go figure 21:56:48 ... 21:56:54 ais523: does it workk? 21:56:55 *work 21:56:58 if it does, I'll be amazed 21:57:00 ehird: not sure, haven't tried it yet 21:57:03 uudecode uses quite some characters! 21:57:08 I'd stick to alphanumericals 21:57:08 * ais523 finds a small file to test with 21:57:16 easy to write your own converters for that 21:57:35 assign one letter to "read two letters and combine" 21:57:43 (recursive; we can handle one or two levels of recursion) 21:57:52 combine as in, 36-bitwise 21:58:00 erm 21:58:02 not 36-bitwise 21:58:14 5.16-bitwise :P 21:58:18 it almost worked, except it deleted the last character of the filename 21:58:26 and then the converter just stuffs ascii into that place 21:58:28 using the additional thingy 21:58:38 like a 30 line decoder, and a 40 line encoder 21:58:43 oh, there was an @ in there 21:58:57 ais523, conclusion is that mac doesn't support the b binary mode for fopen 21:58:59 right? 21:59:10 which means you always get the \r \n swap with fopen 21:59:18 AnMaster: or that I didn't use it in oil.y 21:59:19 probably you need some other mac specific interface for binary 21:59:25 ais523, or that 22:00:25 ais523, I can't find *any* fopen in oil.y. And freopen only for stdout to a file 22:00:41 yacc does it, no? 22:01:17 oh, if yacc's doing it automatically 22:01:22 then it's probably /it/ opening in text mode 22:01:45 binary mode is the stupidest thing ever 22:01:51 oil takes input on stdin 22:01:52 hmm... in what way? 22:01:54 "Like... *actually* open this file." 22:02:11 "When I say fopen, as in open this file and give me this contents, right, this flag means open this file and ACTUALLY GIVE ME THE CONTENTS." 22:02:20 ehird: oh, you mean the fact it's needed is stupid, rather than using it is stupid 22:02:27 yeah 22:02:33 blame micro$oft 22:02:33 ehird, ais523 oil reads on stdin, and no fopen and generated freopen in oil-oil.c 22:02:41 so no yacc isn't doing anything automatically 22:02:45 but yes it is in text mode 22:02:47 oh, stdin's always in text mode 22:02:52 immibis: Hello, 12-year-old AOLer. 22:02:56 You found the $ key. 22:03:02 Why not use it to insult a money-grabbing corporation? 22:03:06 You know... like all of them. 22:03:25 i'd be really surprised if 203-97-111-43.cable.telstraclear.net was an AOL address... 22:03:26 It's the stupidest angle of attack you could possibly mount on Microsoft stuffed into obnoxious leetspeak. Stunning. 22:04:43 I'm doing pretty well with this trackball, considering how long it takes anyone to be proficient with a mouse. 22:04:52 test.t.Z: compressed with 16 bits, can only handle 12 bits 22:04:54 hmm 22:05:58 yay, this version of compress has an option to make it use legacy 12-bit behaviour 22:06:38 * ais523 vaguely wonders how many years transferring the entire C-INTERCAL distribution like this would take 22:07:05 -rw-r--r-- 1 ais523 ais523 1759163 2009-11-13 22:00 ick-0.-2.0.29.pax.Z 22:07:24 Not too many, then. 22:07:29 I was about to say. 22:08:00 entering files using cat manages a few bytes per second 22:13:15 using pcc or cc? 22:15:48 hmm, I guess Go is why Russ Cox has been so silent lately 22:16:14 -!- ehird has set topic: GO IS A DIVISIVE ISSUE IN THIS CHANNEL, also $pun | http://tunes.org/~nef/logs/esoteric/?C=M;O=D. 22:16:59 ehird: cc 22:17:09 pcc is more likely to compile ick, but cc is far cooler 22:17:12 and far harder to understand 22:18:28 ais523, odd thing: :: is /../ but ::: is /../../ I would have expected that to require four : 22:19:17 I wouldn't 22:19:20 as : is . 22:19:57 hm 22:24:21 ais523, so on a real mac now bin2c, oil and convickt builds. Note that the clc-charset thingy make sheepshaver crash 22:24:44 oh and the bin2c generated files are generated. Same goes for the oil generated files 22:24:49 I wonder why? messing with binary files? 22:24:57 ais523, why what? 22:25:19 ais523, clc-charset file? well it crashes compiling that file 22:25:36 after printing some of the warnings about suspect assignment in if. 22:25:50 turning off warnings doesn't help 22:26:28 why would turning off warnings help? 22:26:29 -!- Slereah_ has joined. 22:26:33 ais523, oh on a real mac those weird symbols are easy to type 22:26:34 looking at the warnings might work better 22:26:40 right-alt plus d and f 22:27:06 ais523, 1) too fast crash to see them, thought maybe the fact that printing the warnings made the window scroll could cause a crash 22:27:36 2) seeing the warnings on a real mac makes it possible to see that they are the same in sheepshaver. On the real mac they are all about doing = instead of == inside if 22:27:37 as I said 22:27:55 ais523, basically: emulator bug 22:29:29 ah 22:36:19 -!- Slereah has quit (Read error: 110 (Connection timed out)). 22:39:41 -!- fax has quit ("Leaving"). 22:43:02 -!- ais523 has quit (Remote closed the connection). 22:48:53 -!- fax has joined. 22:49:19 what is this tiny unix system that you are talking about? 22:49:24 and who is using it? 22:50:29 or maybe it is ancient as well as tiny 22:50:41 ehird is the one using it, right? 22:50:52 and ais523. 22:50:58 V7 and V5, him and me respectively. 22:51:04 it's tiny only because it is ancient. 22:52:16 ehird, how tiny? 22:52:29 Just... PDP-11 tiny. 22:52:37 It's just Unix. 22:52:39 emulated or on an extra computer? 22:52:45 Mainline Unix, straight from ma bell. 22:52:51 SimonRC: neither of us have a pdp-11, no :P 22:53:04 ok 22:53:15 does it do TCP/IP? 22:53:22 Um, no? 22:53:26 This is 1970s stuff. 22:53:29 V7, 1975 iirc. 22:53:31 V5, earlier. 22:53:41 ah 22:54:10 TCP/IP dates from the 70s 22:54:47 Might I just say that Go's make-based build system is wonderfully simple? 22:54:49 TARG=hello 22:54:50 GOFILES=hello.go 22:54:50 22:54:50 include $(GOROOT)/src/Make.$(GOARCH) 22:54:50 include $(GOROOT)/src/Make.cmd 22:55:00 Can substitute pkg for cmd to make a library, iirc. 22:55:01 Recursive make considered harmful 22:55:13 And, of course, $(GOARCH) for amd64/386/arm/whatever to cross-compile. 22:55:15 Deewiant: Um... 22:55:19 That's good, because it's not recursive make. 22:55:24 I'm talking about the system for Go progarms. 22:55:26 *programs 22:55:29 Not the toolchain. 22:55:44 (And really, recursive make is useful if you just want a lazy "okay fucking build everything plz".) 22:56:18 Oh, and it has the nice property that cross-compiling works even with that minimal file... 22:56:23 "make GOARCH=386" etc. 22:57:18 Rather bruteforce approach: 22:57:18 $ make clean 22:57:19 rm -f *.[568vq] hello 23:00:49 -!- fax has quit ("Leaving"). 23:01:03 ehird, it forgot *~ 23:01:07 always useful to clean out 23:01:12 That's your editor's fault. 23:01:45 ehird, that was sarcasm. Because I have seen way too many makefiles do that 23:01:47 * ehird decides that an RPN calculator is the best way to play around with Go 23:01:50 AnMaster: Ah. 23:01:58 AnMaster: Well it's funny, now that I know it's sarcasm. 23:02:04 Which is an improvement in some sense :P 23:02:13 ehird, improvement over what? 23:02:22 Things that aren't funny even when explained. 23:02:35 oh 23:02:40 gorpn: pronounced "gorpn". 23:02:48 Fnord xyzzy gorpn quux. 23:02:58 ehird, what *is* gorpn? 23:03:02 Go RPN. 23:03:07 i.e. an RPN calculator written in Go. 23:03:11 My creativity KNOWS NO BOUNDS. 23:03:18 write a roguelike! 23:03:18 After that it'll be an implementation of an esolang, probably. 23:03:21 ehird, what did Gregor say about Go? I forgot 23:03:27 SimonRC: Don't feel like binding ncurses 23:03:35 AnMaster: "IT'S HORRIBLE AND I HATE HOW THEY NAMED THE COMPILERS AND THE TOOLCHAIN SUCKS" 23:03:40 AnMaster: No justification was given. 23:03:44 ehird, write a text adventure! 23:03:52 Oh, and something about the syntax being acceptable but the semantics being nonsensical? I don't even remember. 23:03:53 ehird, I didn't read it too closely 23:03:58 I think Gregor was PMSing. 23:04:01 you needn't bind ncurses 23:04:01 im playing around with go by working with it in acme on a plan 9 cpu server that im drawterming to, its the ultimate Rob Pike fanboy experience 23:04:08 ehird, PMSing? 23:04:12 Males can in fact do that when presented with something sufficiently repulsive to them! 23:04:15 That's a scientific fact. 23:04:27 AnMaster: jfgi 23:04:35 mycroftiv, rob pike fanboy experience? Do I even want to know what that is 23:04:49 It's like an orgasm, but composed of love for Rob Pike. 23:04:57 -!- Pthing has quit (Read error: 104 (Connection reset by peer)). 23:05:08 ehird, how do you fell about a makefile using ð instead of \ ? 23:05:14 as syntax I mean 23:05:16 (Rob Pike is one of the old guard Unix guys, huge force in Plan 9 development, wrote acme for instance, blah blah) 23:05:24 AnMaster: what is that char supposed to be? 23:05:26 AnMaster: Hurr I am Mac and I don't do charsets 23:05:30 SimonRC: icelandic d :-P 23:05:35 (And also Go) 23:05:36 okaaaaay 23:05:37 -!- Pthing has joined. 23:05:54 ehird, you have broken unicode atm? Or was that a joke about MPW makefiles (because you already seen them?) 23:05:57 yeah seeing Ken Thompson Rob Pike and Russ Cox working together on this is definitely the Revenge of Plan 9 crew, weve all been wondering what they were up to 23:06:05 AnMaster: Latter 23:06:12 SimonRC, look at http://omploader.org/vMnJoag 23:06:16 Rob Pike supports software patents and that makes me sad 23:06:28 ehird, then what is the difference between one of those old-style s and two of them? 23:06:31 But he's so cool that it FADES INTO IRRELEVANCE 23:06:34 I have been unable to find proper docs 23:06:35 :/ 23:06:37 AnMaster: No idea 23:06:42 I think it's : 23:06:43 so I have been looking at examples 23:06:44 and :: 23:06:55 ehird, hm 23:07:07 ehird, now what was the diff in normal make 23:07:11 * AnMaster has been using : mostly 23:07:12 AnMaster: that looks like the hack some languages used to fit their semi-latin chars into 7 bits 23:07:29 SimonRC, charset would be MacRoman 23:07:38 they tended to replace {|}[\] or whatever 23:07:47 SimonRC, {} is used there too. See 23:07:58 well, a lot of the old guard has kind of had to eat crow on their early positions on free software and linux, ken thompson was pretty negative about the linux kernel in interviews in the 90s 23:08:00 and [] and | works 23:08:42 :-S 23:08:52 mycroftiv: well, the linux kernel is pretty shit to be honest. 23:09:11 and free software is just another copyright-centered movement, except with a really crazy guy at the helm 23:09:14 ehird, anyway I'm pretty sure that I can type \ on it *tries* 23:09:15 mycroftiv: so he doesn;t dislike it as much now? 23:09:43 The greatest lie of the free software movement was claiming it used copyright to subvert copyright. It did no such thing; it is exactly in the spirit of copyright law and would fail horribly without it... 23:09:51 ehird, yes I can. AltGr-Shift-7 23:09:55 gives me \ 23:10:02 lol 23:10:11 well, being employed by google and having his work be focused on it means that as a practical matter, i doubt he would make the same comments - a few years ago in an interview he basically took it all back and said that he thought linux was OK because it got the job done 23:10:13 try using the backslash key? 23:10:31 immibis, on a classic mac yes. Using Swedish keyboard layout 23:10:36 immibis, what did you expect? 23:10:36 I thought the point of FS was to stop people having programs they can't fix? I recall RMS telling a story about a printer driver with a bug, and being angry he didn't have thhe sourec to fix it and donate a patch. 23:10:55 at least on linux/pc with Swedish keyboard it is just AltGr-+ 23:11:16 fmt.Printf("%s\n", os.Stdin) 23:11:17 → 23:11:18 %s(*os.File=&{0 /dev/stdin 0}) 23:11:21 Fun and weird! 23:11:33 erm, maybe that %s is being iinterpreted literally 23:11:51 nope 23:12:09 weird stringification 23:12:31 ehird, maybe that is what Gregor was talking about 23:12:37 or maybe not 23:13:30 nah 23:17:37 [23:10] SimonRC: I thought the point of FS was to stop people having programs they can't fix? I recall RMS telling a story about a printer driver with a bug, and being angry he didn't have thhe sourec to fix it and donate a patch. 23:17:38 That was its progenitor. 23:17:55 Stallman's method, however, was nothing short of perverse: enforcing a bewildering array of restrictions to get to the paltry grants. 23:18:31 Taking dominion of the infinite space of some information is exactly what copyright is intended for. The GPL user stands guard, slaying anyone who dares do what he was not told he could do. 23:18:50 That's not subverting copyright, that's being its BFF. 23:19:08 Good lord, is this channel having simultaneous licensing and language wars? 23:19:27 (There *is* a way to subvert copyright by granting all rights, thus meaning that anyone who receives a copy can know that all the author can do is sit there chillin'.) 23:19:36 Gregor: It is possible to talk about an issue without having a flamewar. 23:19:45 And, in fact, we haven't been warring about Go at all. 23:19:51 i dont think anyone is taking the bait to rise up and defend the gpl, though I certainly can mentally simulate what Eben Moglen etc would have to say 23:20:13 Heck, even the BSD and MIT licenses don't achieve it. 23:20:28 if i can talk to an imaginary Richard Stallman in my head, and hear what he has to say, does that mean i have a cognitive subsystem that passes the stallman turing test? 23:20:31 There's still that one restriction (include this notice), and so the legal guardian is as primed as ever in the space-of-concepts. 23:20:40 ehird: You can remove that. 23:20:41 It makes no difference how many; only whether there are any. 23:20:44 ehird: I have on occasion :P 23:20:48 Gregor: Then it isn't the BSD/MIT license, duh. 23:20:59 And? 23:21:07 I was mentioning the BSD/MIT license... 23:21:29 The license text I settled on as the simplest, solid way to grant all rights for any given work (not just software) is, lemme find it... 23:21:50 "No rights reserved." :P 23:21:52 Permission is granted to copy, modify and distribute this work. 23:21:52 23:21:52 THERE IS NO WARRANTY. 23:21:59 Gregor: Public domain is... shaky. 23:22:07 That was a joke. 23:22:10 Yes. :P 23:22:18 I like "THERE IS NO WARRANTY" 23:22:20 Saying "Public domain, or if you prefer see LICENSE" is the best of both worlds. 23:22:26 In proper no-warranty style ALL CAPS. 23:22:39 Gregor: JUST IN CASE IT GOT DROWNED OUT IN THOSE FEW WORDS ABOVE IT 23:22:44 lol 23:23:13 Some software licenses actually grant the permission to use the software. 23:23:20 Yeah, uh, I don't think we need your permission, dude. 23:23:41 licenses are for distribution anyway... 23:23:47 maybe if it was an EULA :P 23:24:30 as a matter of fact, i think the right to use does have to be explicity granted 23:24:38 at least in usa law under current interpretation 23:24:46 USA law is a giant suckfest. 23:24:47 see apple vs. psystar 23:24:50 And not the good kind. 23:24:54 you can't do that with a license, though 23:25:01 licenses simply don't apply to the situation of usage, no? 23:25:02 But anyway, that's just because Apple has redefined using as a kind of copying. 23:25:07 otherwise they'd be EULAs 23:28:30 -!- clog has joined. 23:28:30 -!- clog has joined. 23:28:30 see, i thought that loading software from physical media into computer memory *is* a copyvio unless you are explicitly granted the right to do so as a necessary precondition of using it 23:28:33 Which is to say, it should cover only copying. 23:28:36 Copyright really ought to be /dev/null. 23:28:43 Which is to say, it should die a painful death. 23:28:51 Anyway, I left out "use" because I specifically made it usable for non-software works. 23:30:43 ehird, why is public domain "shaky"? I never understood what the issue would be... 23:31:07 Because most laws of copyright don't let you release the copyright of your work until you've been dead for some years. 23:31:11 It's not well-defined what process is necessary to declare something as being in the public domain. 23:31:18 That too. 23:31:28 i dont think the concept of public domain exists or has very similar ;meaning between different countries, whereas copyright is more harmonized due to treaties 23:31:45 ehird, why is that? 23:31:56 "Dual-copyrighting" as "public domain or if you want it can be copyrighted under license-that-grants-permission-to-everything" is the best solution. 23:32:00 AnMaster: Because they simply don't grant you the right. 23:32:09 They say "If someone makes a work it's copyrighted until 80 years after their death." 23:32:21 And at no point do they say "Oh, and the creator can release the copyright to their work, btw." 23:32:47 by the way, in terms of crazy copyright news, anyone hear a few months ago about egypt attempting to claim copyright on the pyramids and the sphinx? 23:33:19 lol 23:33:41 Have the "authors" been dead for OVER NINE-THOUSAAAAAAAND years? 23:33:56 ok, my time sense is off - it was two years ago, heres a 'reputable source' news link http://news.bbc.co.uk/2/hi/middle_east/7160057.stm 23:34:27 i'm pretty sure mycroftiv exists in a realm outside of time 23:34:30 either that or he's stoned 24/7 23:34:41 (or both) 23:35:10 using Plan 9 alters brain chemistry as much or more than any merely physical molecule 23:35:25 i very much doubt that 23:35:57 * Sgeo actually feels safe opening links directly from XChat now 23:36:12 When I used Firefox, if I tried that, XChat would freeze for some time 23:36:14 So, using a trackball, I think my thumb is now a contortionist. 23:36:17 That's some exercise. 23:36:24 I guess until FIrefox was able to process and start loading it 23:36:59 ehird: I misread that as "cartoonist" 23:37:01 And was confused. 23:37:11 Quite. 23:37:31 mycroftiv: haha, my linux distro will confuse you plan 9 weenies — "/bin/8g foo.go? What kind of language is g?" 23:38:11 Er... Isn't stuff like that supposed to be in /usr/bin? 23:38:25 ehird, and what is a "contortionist" 23:38:27 in plan 9 everything gets union bound to /bin 23:38:33 http://www.justfuckinggoogleit.com/ 23:38:39 Ilari: /usr is the most retarded idea ever 23:38:41 trufax 23:38:50 ehird, I'm on a phone atm. To do so I would have to close the irc client 23:38:53 well not in the day when it was invented (i'm talking about non-/home-style usr) 23:39:09 but nowadays it just adds needless hierarchy cruft. 23:39:16 thus why everything goes in /bin in mine 23:39:19 (including /bin/linux) 23:39:28 ehird, so you still refuse to explain it then? 23:39:35 AnMaster: it's someone who's bendy. 23:39:41 ehird, ah 23:39:42 and it's not refusing, it's not explaining it 23:39:50 you can't cause someone to do an active thing (refuse) merely by prompting them 23:40:15 * Sgeo can read DIlbert from the beginning! 23:40:21 (Or at least, the first published one) 23:40:53 Ok, all of these I read already 23:41:07 What's the one where Dilbert dies? What's the one where Dogbert dies? 23:41:16 eh? 23:41:33 The dates of the dilbert strips 23:41:37 AFAIK, /usr is useful since it contains most of the program executables and data. Idea is that /usr would be put on different filesystem. 23:42:04 But nobody actually does that :) 23:42:22 Ilari: What Gregor said :P 23:42:39 http://www.dilbert.com/fast/1995-10-16/ The start of the Dogbert dies arc 23:43:34 But nobody actually does that :) <-- I do 23:43:40 on servers 23:43:45 tiny / 23:43:54 then separate /tmp /var /usr 23:43:58 oh and /home 23:44:14 /opt being a symlink to /usr/opt 23:44:48 http://www.dilbert.com/fast/1995-10-22/ is one of my all-time favourites IIRC 23:45:35 I'm going to have to get rid of the instinct to simply copy the URL 23:45:42 SimonRC, 999999999... 23:45:56 actually, I suspect I have a book with that time period in it, which is why many of them looked familiar 23:46:07 Two months after I was born, that latter one. 23:46:10 SimonRC, I think I do, to 23:46:14 To the day. 23:46:26 ehird: argh you bloody young people that are almost overtaking me in skills 23:46:42 SimonRC: Thanks for the compliment? 23:46:51 (psst. most kids are fucking dumb) 23:47:11 yeah, but you aren't 23:47:29 except when you come out with jewels like not knowing htere were multiple moon landings 23:47:36 THAT WAS A BRAINFART >_< 23:47:48 Curse your accursed good memory 23:49:08 Ooh, Chrome froze 23:49:14 Hm, not for very long 23:53:46 mycroftiv: the go tutorial has a cat... with a flag :-D 23:53:56 okay, so it's just -n, which was in pre-cat-v, but still 23:55:49 * oerjan imagined a picture of a cat waving a flag 23:56:08 hello kitty style 23:56:42 "cat came back from Berkeley waving flags" —Rob Pike 23:59:19 rob pike did say though "the days of one tool per job are over, and the eulogy was written by perl" - still not sure exactly how he meant that, maybe just being realistic 23:59:47 <3 Pimp Code