00:03:08 <elliott> loadlin can't load initramfses, right? :p
00:03:17 <elliott> Actually, does the bootloader have to be able to, if you embed it into the kernel?
00:04:05 -!- augur has joined.
00:05:50 -!- nooga has quit (Ping timeout: 255 seconds).
00:07:37 -!- nooga has joined.
00:12:21 -!- FireFly has quit (Quit: swatted to death).
00:17:56 <pikhq> The initramfs loading is identical to initrd loading from a bootloader point of view.
00:18:16 <elliott> pikhq: Right. But you can embed a .cpio directly into the kernel.
00:18:17 <pikhq> Also, if it's embedded in the kernel, the bootloader can be completely ignorant of it.
00:18:43 <pikhq> The .cpio gets transformed into a .o with a symbol pointing to the .cpio.
00:19:01 <elliott> pikhq: How small can you get an ancient DOS floppy if you don't care whether it's usable at the command-line, just to run a program in autoexec?
00:19:09 <elliott> Because I'm thinking that http://busybox.net/~vda/linld/README.txt + DOS might be smaller than lilo.
00:20:07 <elliott> [ ] LINLD.COM 13-Sep-2004 13:11 5.7K
00:20:17 <elliott> So depending on how small I can get DOS...
00:20:19 <pikhq> Uh, IO.SYS + MSDOS.SYS + CONFIG.SYS (can be empty) + COMMAND.COM + AUTOEXEC.BAT (can be empty) is the minimal MS-DOS system.
00:20:43 <elliott> pikhq: AUTOEXEC.BAT would be "linld cl=blah" :P
00:21:00 <pikhq> IO.SYS is the DOS BIOS, MSDOS.SYS is the DOS kernel, and CONFIG.SYS is, of course, a config file.
00:21:01 <elliott> How big are IO.SYS + MSDOS.SYS + COMMAND.COM, roughly? Assume an old, smaller DOS.
00:21:58 <elliott> pikhq: Or, more concretely: Minimal DOS + 5.7K <=> minimal lilo installation?
00:22:24 <elliott> LILO appears to require a /boot partition.
00:22:32 <elliott> Whereas linld could just run from a DOS floppy.
00:23:01 <elliott> pikhq: (BTW, this is for a single-floppy Linux like you were trying to do.)
00:23:03 <pikhq> That might be true of modern DOS, even.
00:23:18 <elliott> "System is 480 kB" --Linux make
00:23:22 <elliott> It lacks some things though.
00:25:28 <elliott> pikhq: Another question: If I have kernel compression enabled, disable "Support initial ramdisks compressed using [blah]", and embed an initramfs... Will that initramfs get compressed?
00:25:33 <pikhq> Uh, FreeDOS, done minimally, is 112K...
00:25:53 <pikhq> You could probably get that down way smaller by removing some useless features.
00:25:59 <pikhq> (FAT32, LFN, tab completion)
00:26:10 <elliott> <elliott> pikhq: Another question: If I have kernel compression enabled, disable "Support initial ramdisks compressed using [blah]", and embed an initramfs... Will that initramfs get compressed?
00:26:34 <pikhq> elliott: The initramfs is just another object file in the kernel if it's embedded.
00:28:05 <elliott> ^?ELF^A^A^A^C^@^@^@^@^@^@^@^@^B^@^C^@^A^@^@^@L<80>^D^H,^@^@^@^@^@^@^@^@^@^@^@4^@ ^@^A^@^@^@^@^@^@^@^@<80>^D^H^@<80>^D^H/^A^@^@3^A^@^@^G^@^@^@^@^P^@^@ZY»^M<81>^D^H<89>Ï1À<8d>Hÿò®<89>Ð<8b>Oü¿/<81>^D^H<81>ùccd^@^O<95>^G<89>Þ<8d>K^D<8d>S^OHt#[<81>;-ccdt^Eö^Gÿt Æ^G^@H<89>ót^M[Ht YHt^EZHt^A_QSö^Gÿu^Ej^VXÍ<80>1ɵ^Hj^EXÍ<80>P<85>Àx9¹ S^@^@<89>Ãj6XÍ<80>ö^Gÿu8SR1Ò²^K¹$<81>^D^Hj^A[j^DXÍ<80>²^A1Ûj^CXÍ<80>Z[¹
00:28:06 <elliott> ^YS^@^@j6XÍ<80>Y[<87>^L$¾^A^@íÀj^UXÍ<80>X<85>Àx^B1À<89>Ãj^AXÍ<80>/dev/cdrom^@^@^@^@^@iso9660^@<enter> ..^@
00:28:09 <elliott> THAT DOES NOT COUNT AS A PROGRAM
00:28:26 <pikhq> (a minimal FreeDOS, BTW, is: command.com, kernel.sys)
00:29:33 <pikhq> AHAHAH. Easy way to get that FreeDOS system down in size.
00:29:44 <pikhq> upx supports DOS executables.
00:30:02 <pikhq> Oh, BTW: you may want to leave that kernel uncompressed, and upx the kernel.
00:30:41 * Sgeo ogles the Nexus S
00:30:59 <elliott> pikhq: I tried UPX'ing the vmlinux and it was bigger than the LZMA'd bzImage.
00:31:17 <elliott> pikhq: I would just like to say that I've just selectively installed pieces of asmutils...
00:31:18 <elliott> arch basename cat chmod chown chroot cmp cp cut date dc dd deflate df dirname dmesg du echo env extname factor false fdisk finger free fromdos gi grep halt head hexdump host hostname id idea ifconfig inflate init kill killall killall5 less ln ls md5 md5sum mount mv nc netstat nice nm nohup od paste pidof ping poweroff ps pwd readelf readlink reboot renice rm rmdir rot13 route scan sh sha1sum size sleep sln strings tail tar tee telnet test todos t
00:31:18 <elliott> ouch tr true tty umount uname update uptime users usleep uuencode watch wc wget which whoami write yes
00:31:25 <elliott> pikhq: Notice how this includes WGET and shit.
00:31:28 <elliott> pikhq: Guess how big all these are?
00:31:39 <pikhq> elliott: Yeah, asmutils is awesome.
00:32:08 <pikhq> No, *motherfucking awesome*.
00:32:13 <elliott> wget doesn't seem to work here :P But whatever!
00:32:31 <elliott> pikhq: Wait... if I have an initramfs, you won't be able to see the kernel, will you? X-D
00:33:38 <elliott> pikhq: Because the kernel will be on the floppy.
00:33:41 <elliott> And there's no block device support.
00:34:27 <elliott> -rw-r--r-- 1 elliott elliott 70K Dec 7 00:33 ../root.cpio
00:34:32 <elliott> Welp, might as well compile it in.
00:35:00 <elliott> pikhq: If I can find a damn smallX tarball, this could actually work.
00:35:08 <elliott> With asmutils and /etc/rc.
00:36:06 <elliott> Oops, I forgot to put init in /sbin. Ehh... I can just say init=/bin/init.
00:36:30 <elliott> pikhq: I have all kernel printing disabled, so time to wait for a minute and hope I get a shell :P
00:37:09 <elliott> DIN'T WORK LOL... or I'm too impatient.
00:37:13 <elliott> I'll rebuild with printk so I can see shit.
00:38:19 <elliott> pikhq: Hmm... does init= actually affect the initramfs?
00:38:30 <elliott> Isn't that for the real root's init path?
00:40:47 <pikhq> With initramfs, init= gets passed to /init.
00:41:53 <elliott> pikhq: ...Maybe I should omit /bin, and just put everything in /. :D
00:42:21 * pikhq sees watercooled cases with fans; cries
00:42:50 -!- nooga has quit (Ping timeout: 260 seconds).
00:43:12 <elliott> pikhq: you do realise just about every watercooled setup uses at least one fan?
00:43:50 <pikhq> elliott: "Fans". As in many. As in several. Tiny. Noisy. Fans.
00:44:01 <elliott> pikhq: "qemu -kernel path/to/bzImage" should be enough to test this, right?
00:44:03 * Sgeo holds his breath and opens IE
00:44:15 <elliott> I'm not sure it's actually doing its initramfs thang.
00:44:35 -!- nooga has joined.
00:44:44 <elliott> pikhq: If it doesn't move in 30 seconds, I'm enabling all this printy stuff. :P
00:45:37 <pikhq> elliott: Also: watercooled computers should have ginormous radiators.
00:46:03 <pikhq> Actually, screw the water.
00:46:11 <pikhq> Computers should have ginormous radiators.
00:46:14 <pikhq> With blinkenlights.
00:46:59 <elliott> │ │ [ ] Use 4Kb for kernel stacks instead of 8Kb │ │
00:49:18 <elliott> pikhq: Hmm... uClinux can be configured to be super-small, right?
00:50:29 -!- madbr has left (?).
00:50:54 <elliott> pikhq: http://www.openinventionnetwork.com/ Hacking the patent system. "Join us, we'll patent everything, and then license it to everyone who agrees not to use their patents against Linux".
00:51:41 <Sgeo> I should not have to manually allow Windows Firewall to make changes to my computer
00:52:02 <elliott> pikhq: Hmm. If a kernel manages to decompress, print out "Booting the kernel.", but then hang... what's up?
00:53:12 <elliott> I have printk and everything.
00:54:37 <pikhq> elliott: Uh, fuck that's not good?
00:54:47 <elliott> pikhq: Having PCI access set to "direct": smart idea?
00:55:50 <elliott> pikhq: Come to think of it, it probably means that it's failing during very early bootup, right?
00:56:04 <pikhq> Exceptionally early.
00:56:21 <pikhq> I cannot be sure it had paging working yet.
00:56:38 <pikhq> If I knew more about the kernel's structure, I'd suggest you kgdb that.
00:57:20 <elliott> pikhq: With qemu? Sounds like a whole new world of fun :P
00:58:40 -!- Mathnerd314 has joined.
00:59:10 <elliott> pikhq: Hypothesis: Plan 9 is the most elegant OS ever designed, in both design and implementation. (Genera is *nicer*, but the code is... heh.)
01:00:51 <elliott> pikhq: Okay, it's STILL not printing anything...
01:01:01 <elliott> Are you sure using -kernel isn't to blame here?
01:01:17 <pikhq> All that does is make the qemu BIOS act as a bootloader.
01:02:22 <elliott> pikhq: Even without namespaces support, IPC, the block layer, anything, I should still get printk right?
01:02:29 <elliott> Even with no drivers enabled?
01:02:45 <elliott> │ │ [ ] Display panel/monitor support │ │
01:02:49 <elliott> I shouldn't need that, surely?
01:02:57 <pikhq> Speaking as someone who has actually had this come up, you should still get printk.
01:03:10 <pikhq> If you have VGA console support and printk, you're golden.
01:03:24 <elliott> pikhq: Not sure I have VGA console support!
01:03:30 <elliott> Is that actually configurable?
01:03:42 <elliott> │ ymbol: VGA_CONSOLE [=n] │
01:03:42 <elliott> │ rompt: VGA text console │
01:03:47 <elliott> HURF DURF IM TARDED CUZ THAT'S NOT ENABLED
01:03:52 <pikhq> There's your problem!
01:04:29 <elliott> │ -> Console display driver support │
01:04:34 <elliott> I don't *have* that last submenu.
01:05:03 <elliott> pikhq: Wait, what? VGA_CONSOLE depends on "VT [=n]". Or is that [=n] just what I have?
01:05:54 <elliott> pikhq: Weird that the console would depend on VTs.
01:06:53 <elliott> "Kernel panic - not syncing: junk in compressed archive"
01:06:59 <elliott> Do I have to construct it specially?
01:07:50 <pikhq> "compressed archive"
01:07:57 <elliott> pikhq: Yes, that one puzzles me.
01:08:10 <pikhq> elliott: At the kernel.
01:08:18 <pikhq> elliott: It should just take a straight cpio.
01:08:27 <elliott> │ │ [*] Initial RAM filesystem and RAM disk (initramfs/initrd) support │ │
01:08:27 <elliott> │ │ (/home/elliott/flinix/root.cpio) Initramfs source file(s) │ │
01:08:27 <elliott> │ │ (0) User ID to map to 0 (user root) │ │
01:08:27 <elliott> │ │ (0) Group ID to map to 0 (group root) │ │
01:08:30 <elliott> │ │ [ ] Support initial ramdisks compressed using gzip │ │
01:08:33 <elliott> │ │ [ ] Support initial ramdisks compressed using bzip2 │ │
01:08:36 <elliott> │ │ [ ] Support initial ramdisks compressed using LZMA │ │
01:08:39 <elliott> │ │ [ ] Support initial ramdisks compressed using LZO │ │
01:08:42 <elliott> │ │ Built-in initramfs compression mode (None) ---> │ │
01:08:47 <elliott> │ CONFIG_INITRAMFS_COMPRESSION_NONE: │
01:08:53 <elliott> │ Do not compress the built-in initramfs at all.
01:09:14 <elliott> pikhq: I figure I'm doin' it wrong as far as cpio goes.
01:10:04 <elliott> pikhq: Ah. Is "-H newc" default?
01:10:45 <elliott> pikhq: Wait. You can just feed the kernel a directory and it'll make a cpio for you.
01:11:29 <elliott> pikhq: "Warning: unable to open an initial console."
01:11:33 <elliott> Now what the fuck does THAT mean?
01:12:26 <elliott> pikhq: You need /dev/{console,null,tty1,tty2,...}
01:13:02 <elliott> pikhq: Er. How does one cp without making it copy the contents?
01:13:07 <elliott> You know, just to cp /dev/console .
01:15:45 <pikhq> rsync --devices /dev/console .
01:16:10 <elliott> pikhq: I have a Linux kernel that boots to a system with 100 programs (99 if you don't count init) in 567K.
01:16:16 <elliott> It has a shell. With tab completion. And line editing.
01:16:26 <elliott> And four devices :P (console, null, tty1, tty2)
01:17:02 <elliott> pikhq: Aaaand ctrl+alt+fN works.
01:17:18 <elliott> TODO: Put chvt and friends into the image, since I have VTs now. Create more ttys. And then, X11!
01:17:38 <elliott> Also: Disable all that nasty kernel printing, and other bloated stuff I enabled, if it helps. (.config.good has all of that removed, so I can just selectively try and migrate back.)
01:18:21 <elliott> Ah, and keep ~/flinix/root owned by root:root...
01:18:37 <elliott> pikhq: This thing boots in about 1 second. 0.2-0.4s if you ignore decompression.
01:18:47 <elliott> Literally. It says "Decompressing Linux...", pauses for a second, floods output and instantly drops you at a shell.
01:19:10 <elliott> Oh, I should probably enable a framebuffer at some point... although smallX doesn't need it, so maybe not.
01:19:31 <nooga> who pastes so much
01:20:19 <elliott> Hmm, poweroff doesn't work, neither does reboot; TODO: remove them,.
01:20:23 <elliott> Halt doesn't do much either.
01:20:47 <elliott> TODO: Add some Ethernet drivers.
01:21:52 <elliott> hmm, whoami segfaults with a simple /etc/passwd
01:25:44 <elliott> pikhq: Can I just say that modern software is stupidly bloated?
01:25:53 <elliott> This qemu window feels so *refreshing*.
01:27:43 <elliott> pikhq: So all I have to do is cram an X server into 700K or so and get slightly over 200K left to put whatever I want in. :P
01:29:07 <elliott> pikhq: The awesome thing is: This has like 99% of what's needed to do networking. Not sure it'll do DHCP.
01:29:28 <elliott> pikhq: But you can *totally* put this on a 386 (well, a 386 working enough to not need any of the machine-specific hacks Linux does which I disabled...) and IRC from it.
01:29:50 <elliott> Stripped printk out of it.
01:29:56 <elliott> It still works and boots instantly.
01:30:32 <elliott> pikhq: Ha, remind me to add /proc sometime so I can use ps.
01:30:42 <elliott> TODO: Add /proc for ps. See if it's worth it.
01:31:20 <elliott> Also: Find out why wc is printing totals twice, second time on the second EOF.
01:31:29 <elliott> Also: Find out why "foo&" isn't working; job control issues?
01:32:11 <elliott> nooga: I'm assembling a tiny Linux with X11 into one floppy disk.
01:32:22 <nooga> how much kernel weights?
01:32:36 <elliott> nooga: Currently it's 536K, kernel with embedded filesystem (it's 480K or so without the filesystem) with 99 programs.
01:32:46 <elliott> nooga: From the asmutils project, which has craaaazy tiny utilities.
01:32:46 -!- oerjan has quit (Quit: Good night).
01:33:29 <elliott> nooga: It even has a very-tiny wget. :P
01:33:35 <elliott> (But no networking support yet; almost.)
01:33:41 <elliott> I'll continue this insanity tomorrow. Toodles.
01:33:42 -!- elliott has quit (Quit: Leaving).
01:39:36 -!- wareya_ has joined.
01:41:33 -!- wareya has quit (Ping timeout: 240 seconds).
02:21:13 -!- GreaseMonkey has joined.
02:31:24 -!- Wamanuz2 has joined.
02:35:09 -!- Wamanuz has quit (Ping timeout: 240 seconds).
03:02:22 -!- TLUL has joined.
03:56:09 <pikhq> http://www.youtube.com/watch?v=heKK95DAKms
03:57:24 -!- Wamanuz2 has quit (Remote host closed the connection).
03:58:19 -!- whoami has joined.
04:00:24 -!- Wamanuz has joined.
04:05:07 -!- Wamanuz has quit (Remote host closed the connection).
04:05:27 -!- Wamanuz has joined.
04:31:07 -!- kmc has quit (Ping timeout: 245 seconds).
04:42:10 -!- Sgeo has quit (Read error: Connection reset by peer).
04:44:43 -!- Sgeo has joined.
07:04:59 -!- clog has joined.
07:04:59 -!- clog has joined.
07:18:32 -!- whoami has quit (Quit: Nettalk6 - www.ntalk.de).
07:49:05 -!- Sgeo has quit (Read error: Connection reset by peer).
07:55:04 -!- evincar has joined.
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:02:22 -!- evincar has quit (Quit: ChatZilla 0.9.86 [Firefox 3.6.12/20101026210630]).
08:15:24 -!- MigoMipo has joined.
09:09:38 -!- GreaseMonkey has quit (Quit: ilua).
09:10:16 <Vorpal> "Error: session already open. Cause: No login cookie found." <-- wtf, seriously?
09:10:30 <Vorpal> I got that from a university web system just now
09:10:48 <Vorpal> due to opening link to a page on it without being logged in
09:29:30 -!- choochter has joined.
09:58:50 -!- TLUL has quit (Quit: *disappears in a puff of orange smoke*).
10:15:58 -!- FireFly has joined.
10:21:21 <Vorpal> elliott (for log reading): I saw this very very bad pun in the yellow text in MC: "sqrt(-1) love you" XD
10:24:07 -!- choochter has quit (Ping timeout: 255 seconds).
10:26:25 -!- choochter has joined.
10:28:39 -!- MigoMipo has quit (Read error: Connection reset by peer).
10:29:13 -!- MigoMipo has joined.
10:55:20 -!- Wamanuz has quit (Read error: Connection reset by peer).
10:55:44 -!- Wamanuz has joined.
11:16:41 -!- oerjan has joined.
11:17:58 <oerjan> 02:21:21 <Vorpal> elliott (for log reading): I saw this very very bad pun in the yellow text in MC: "sqrt(-1) love you" XD
11:18:02 <oerjan> 02:21:29 <Vorpal> that is just so awful
11:18:06 <oerjan> at least it's imaginary
11:21:19 -!- ais523 has joined.
11:47:25 -!- sebbu has quit (Ping timeout: 265 seconds).
12:11:54 -!- sebbu has joined.
12:17:39 -!- Sasha has quit (Ping timeout: 260 seconds).
12:18:15 -!- Sasha has joined.
12:21:35 -!- teuchter has joined.
12:24:17 -!- choochter has quit (Ping timeout: 255 seconds).
12:40:02 -!- tswett has quit (Remote host closed the connection).
13:27:28 -!- nooga has quit (Ping timeout: 240 seconds).
13:29:29 -!- nooga has joined.
13:30:58 -!- sftp has joined.
13:35:45 -!- ais523 has quit (Remote host closed the connection).
13:36:01 -!- ais523 has joined.
13:38:21 <oerjan> http://www.reddit.com/r/AskReddit/comments/ehf49/can_we_please_have_another_antijoke_thread/
13:38:48 -!- choochter has joined.
13:39:52 -!- teuchter has quit (Ping timeout: 245 seconds).
13:57:36 -!- choochter has quit (Quit: lang may yer lum reek..).
15:31:33 -!- elliott has joined.
15:33:35 <elliott> `run allquotes | tail -n 3
15:33:36 <HackEgo> 268) <elliott> So it's not exactly trivial. [Later about same thing] <elliott> It's a trivial C program :P \ 269) <Gregor> "* There is no scientifically-justifiable reason to exclude pornography, which is a vital part of the web ecosystem. However, bear in mind that we're tracing JavaScript, not MPEG and JPEG decoding." <Gregor>
15:33:46 <elliott> `run allquotes | tail -n 2
15:33:47 <HackEgo> 269) <Gregor> "* There is no scientifically-justifiable reason to exclude pornography, which is a vital part of the web ecosystem. However, bear in mind that we're tracing JavaScript, not MPEG and JPEG decoding." <Gregor> ^^^ This bullet-point is my crowning achievement as as a scientist. \ 270) <tswett> That is the mark of
15:33:47 <elliott> `run allquotes | tail -n 1
15:33:49 <HackEgo> 270) <tswett> That is the mark of Gregor right there. <ais523> tswett: except that Gregor didn't write that <tswett> It's still the mark of Gregor.
15:35:38 <elliott> `run allquotes | tail -n 1
15:35:40 <HackEgo> 270) <tswett> That is the mark of Gregor right there. <ais523> tswett: except that Gregor didn't write that <tswett> It's still the mark of Gregor.
15:35:50 <HackEgo> http://codu.org/projects/hackbot/fshg/index.cgi/raw-file/tip/paste/paste.27030
15:35:59 <HackEgo> http://codu.org/projects/hackbot/fshg/index.cgi/raw-file/tip/paste/paste.2865
15:36:11 <HackEgo> Runs arbitrary code in GNU/Linux. Type "`<command>", or "`run <command>" for full shell commands. "`fetch <URL>" downloads files. Files saved to $PWD are persistent, and $PWD/bin is in $PATH. $PWD is a mercurial repository, "`revert <rev>" can be used to revert to a revision. See http://codu.org/projects/hackbot/fshg/
15:37:10 <elliott> ais523: I think you're right about the qdb, it's terrible
15:37:18 <elliott> ais523: I'm going to go and delete all the shitty quotes.
15:37:30 <elliott> If anyone likes them, just look at an older revision :P
15:39:19 <oerjan> i hope you took into consideration that the numbers change after a deletion
15:39:38 <HackEgo> Runs arbitrary code in GNU/Linux. Type "`<command>", or "`run <command>" for full shell commands. "`fetch <URL>" downloads files. Files saved to $PWD are persistent, and $PWD/bin is in $PATH. $PWD is a mercurial repository, "`revert <rev>" can be used to revert to a revision. See http://codu.org/projects/hackbot/fshg/
15:40:04 <elliott> ok, i'll do it properly this time
15:50:40 <elliott> 124) Note that quote number 124 is not actually true.
15:50:45 <elliott> something went wrong there :D
15:50:50 <elliott> oh, everything got renumbered when i removed one quote
15:51:51 * elliott removes the literal tab from quote 128
16:02:10 -!- nopseudoidea has joined.
16:05:18 -!- ais523 has quit (Remote host closed the connection).
16:06:12 <elliott> Okay, I hereby declare that the average quality of quotes in the database is Pretty Good.
16:06:30 <elliott> Wait, never mind, I just made a MISTAKE.
16:07:01 <elliott> Never mind, it resolved itself.
16:07:03 <HackEgo> 99) <Warrigal> Ah, vulva. <Warrigal> What is that, anyway?
16:07:26 <HackEgo> 76) <oklopol> Warrigal: what do you mean by 21?
16:07:29 <HackEgo> 154) <oklopol> but yeah i'm not exactly comfortable with this stuff, to me it seems like if you can unscrew lightbulbs, why couldn't you see into the future, or through walls as well
16:07:29 <HackEgo> 230) Thanks to nooga for constructive criticism, his ideas and being a constant annoyance. --http://theendisnear.no-ip.info/
16:07:29 <HackEgo> 80) <AnMaster> I'm 100% of what sort of magic was involved in it
16:10:44 -!- Sasha has quit (Read error: Connection reset by peer).
16:12:45 -!- Phantom_Hoover has joined.
16:38:17 -!- oerjan has quit (Remote host closed the connection).
16:38:25 -!- oerjan has joined.
16:42:55 <elliott> WHY ARE THERE NO TINY VI CLONES
16:43:05 <nooga> elliott: how is your very damn small linux?
16:43:16 <elliott> nooga: in need of a vi clone :)
16:43:31 <oerjan> elliott: you will have to install TECO instead *MWAHAHAHA*
16:43:37 <nooga> isn't cat and sed enough? :D
16:43:40 <elliott> Phantom_Hoover: No way! This is The Insane Flinix.
16:44:07 -!- Phantom_Hoover_ has joined.
16:44:08 <elliott> Phantom_Hoover: A compleat (well, for some definition of compleat) Linux distribution, with networking, X11 and an IRC client, in one 1.44 meg floppy.
16:44:24 <elliott> Maybe I'll try Emacs; that's gotta be smaller, right?
16:44:38 <elliott> The anti-NIH gods have cast me away from my own life.
16:44:40 <nooga> elliott: http://texteditors.org/cgi-bin/wiki.pl?ViFamily
16:44:57 <oerjan> http://scientopia.org/blogs/goodmath/2010/11/30/the-glorious-horror-of-teco/
16:44:59 <elliott> http://xvi.sourceforge.net/ seems to be the smallest, from their little table there.
16:45:10 <elliott> oerjan: "scientopia"? they renamed it again?
16:45:16 <elliott> oerjan: what fucking kind of name is scientopia :)
16:45:29 <elliott> no wait scienceblogs is still there
16:46:08 <elliott> lol, scienceblogs gave pepsico a blog
16:46:26 <elliott> anyway the problem with xvi is that I need termcap and termcap is ancient vintage software :)
16:46:38 <elliott> gonna be 13K for it and all the supported files
16:46:41 <elliott> but i guess it'll compress well
16:46:57 -!- ais523 has joined.
16:46:57 -!- nopseudoidea has quit (Quit: Quitte).
16:47:28 <oerjan> elliott: wait you are _anti-NIH_ now?
16:47:47 -!- Phantom_Hoover has quit (Ping timeout: 245 seconds).
16:48:05 <oerjan> maybe it's time for my topic idea
16:48:32 -!- oerjan has set topic: The knights who say NIH | http://tunes.org/~nef/logs/esoteric/?C=M;O=D.
16:48:35 <elliott> hmph libtermcap.a is 40K that's huge.
16:48:40 <elliott> oerjan: WTF I THOUGHT OF THAT YESTERDAY
16:48:46 <elliott> fuck you, synchronicitician
16:49:14 -!- nooga has quit (Read error: Operation timed out).
16:49:15 <oerjan> and i've been thinking of it for a while too :D
16:49:19 <elliott> 10K now, and I'll sstrip the resulting xvi
16:49:46 <ais523> also, do you really need support for more than one terminal?
16:49:53 <elliott> ais523: no, but xvi uses termcap
16:49:56 <elliott> so I have to have the library
16:50:00 <ais523> patch it so it doesn't
16:50:02 <elliott> and I figure the real termcap is smaller than terminfo's emulation of it
16:50:12 <elliott> ais523: i'm allergic to vt100 codes
16:50:18 <ais523> they really aren't too difficult
16:50:21 <elliott> ais523: by the way, I cleaned out most of the crap from the quotes file
16:50:30 <elliott> I know they aren't difficult, I just don't wannaaaa :)
16:50:54 <elliott> note: most of, I make no guarantees as to the quality of the quotes file :)
16:50:56 <elliott> and I refuse to remove any fungot quote
16:51:00 -!- nooga has joined.
16:51:12 <elliott> he deserves a lower threshold for what constitutes a good quote, since he did it without a brain
16:51:30 <HackEgo> http://codu.org/projects/hackbot/fshg/index.cgi/raw-file/tip/paste/paste.30285
16:52:02 <elliott> -rwxr-xr-x 1 elliott elliott 127K Dec 7 16:51 xvi
16:52:08 <elliott> I can /not/ afford to spend 127K on an editor.
16:52:18 <elliott> hmm, levee looks the smallest then
16:52:23 <elliott> even so, it's 37K on Mastodon
16:52:24 -!- bsmntbombdood has quit (Remote host closed the connection).
16:52:24 <nooga> elliott: http://texteditors.org/cgi-bin/wiki.pl?TinyEditors
16:52:37 <nooga> elliott: i just wanted to say that you should try levee
16:53:05 <elliott> nooga: I'd already thought of levee, being the official David Parsons Stalker.
16:53:06 <ais523> elliott: most of the old DOS editors were smaller than 64KiB
16:53:11 <ais523> so they fit in a .COM file
16:53:16 <elliott> ais523: they ran on DOS, though
16:53:18 <elliott> not quite the same thing as linux :)
16:53:25 <elliott> sheesh, if I can't get a really tiny editor I'll write my own... in asm!
16:53:30 <elliott> maybe asmutils will accept it
16:53:50 <ais523> what features would you want?
16:54:17 <ais523> for me, the minimum for a usable editor is inserting letters, cursor movement, deleting letters via delete and backspace, and preserving indentation from one line to the next
16:54:26 <ais523> (I don't consider Notepad a usable editor as it lacks the last feature there)
16:55:11 <elliott> ais523: the latter is useless bloat as far as flinix as concerned; you'll be editing, what, config files at the most
16:55:20 <elliott> ais523: I am not sure Flinix actually has a *use* :)
16:55:30 <elliott> ais523: but, uh, features I'd want:
16:55:38 <elliott> the basic : commands like w, o, q
16:55:49 <ais523> oh, I forgot about saving and quitting
16:55:50 <nooga> http://freshmeat.net/projects/virus/
16:55:57 <ais523> you could quit via SIGINT, but there should be a way to save
16:56:07 <ais523> and I said "inserting letters", overwriting them would be bloat
16:56:07 <elliott> the commands "x", "d", "^", "$", "a", "i"
16:56:15 <elliott> the latter two entering insert mode
16:56:19 <elliott> and finally, the command "c" for overwrite
16:56:25 <elliott> plus the vi-style numerical prefixes of those
16:56:34 <elliott> well, numerical and direction
16:56:45 <ais523> elliott: I was trying to stay general, not assuming vi-like
16:56:47 <elliott> nooga: sure, sure, if you wanna be bloaty
16:56:51 <elliott> ais523: well, I want a vi-like editor :)
16:56:53 <elliott> ais523: posix specifies it!
16:57:31 <ais523> elliott: what about this for an idea: you can move the movement commands to arrow keys and home/end/etc
16:57:38 <ais523> that way, you wouldn't have the overhead of entering and leaving insert mode
16:57:48 <elliott> ais523: congratulations, you just invented emacs
16:57:56 <elliott> ais523: (and how do you do :? oh wait, I know, Alt+X!)
16:58:02 <elliott> genius! and we can put Lisp in it, too
16:58:07 <elliott> (fun fact: original ex had lisp)
16:58:21 <ais523> elliott: I'm aware of the irony :)
16:58:43 <elliott> ais523: I actually think insert mode is less code overhead
16:58:50 <elliott> ais523: because arrow keys etc. come as multiple characters
16:58:53 <elliott> so you have to maintain state /anyway/
16:59:19 <ais523> to be fair, Emacs and vim aren't that different; in vim you press esc and i to switch between modes, in Emacs you hold and release control
16:59:55 <elliott> ugh, you stupid configure script
17:00:26 <ais523> but I'm not a configure script!
17:00:40 <elliott> ais523: i'm referring to levee's
17:01:00 <elliott> which only takes $CC as a path, so I can't say "diet -Os pcc", and first tries CC without CFLAGS, so I can't say CC="diet" CFLAGS="-Os pcc"
17:01:07 <elliott> solution: wrapper script! dietpcc!
17:01:16 <elliott> just like pcc, but with fewer calories!
17:02:47 <Gregor> Why can't you say CC="diet -Os pcc"?
17:03:14 <Gregor> Only takes CC as a path.
17:03:17 <nooga> elliott: bundle hedgehog lisp in flinix
17:03:21 <Gregor> That's pretty bizarre.
17:03:50 <elliott> Gregor: It calls some C program that uses exec* of some kind.
17:03:58 <elliott> (Who knows, this is just a guess.)
17:04:04 <elliott> find.o: In function `omatch':
17:04:04 <elliott> (.text+0x1ad): undefined reference to `toupper'
17:04:04 <elliott> find.o: In function `omatch':
17:04:04 <elliott> (.text+0x1bc): undefined reference to `toupper'
17:08:15 <elliott> -rwxr-xr-x 1 elliott elliott 59531 Dec 7 17:07 lev
17:08:18 <elliott> huge and requires termcap too
17:10:49 <elliott> -rwxr-xr-x 1 elliott elliott 54075 Dec 7 17:10 virus
17:12:10 <elliott> -rwxr-xr-x 1 elliott elliott 37691 Dec 7 17:11 virus
17:12:27 <nooga> they've said it's tiny
17:12:35 <elliott> nooga: yeah and that's with every feature disabled, sstrip'd, compiled with pcc, and linked with dietlibc
17:12:49 <elliott> quick, someone name my vi clone with a word involving "vi"
17:12:54 <elliott> ais523! quick! bring out the puns!
17:14:18 <elliott> Vorpal: that is not a word
17:14:24 <Vorpal> elliott, it is almost whimsy
17:16:05 <elliott> viscous, vicious, vixen, villain, vicissitude, vigil, vine, violator, vitriolic
17:17:47 <Vorpal> ^run echo $(( RANDOM % 10 ))
17:17:53 <elliott> no, it should be a thing-that-does i think
17:18:05 <elliott> violator, i'll go with violator
17:18:18 <Vorpal> elliott, did I mix up the bots?
17:18:31 <Vorpal> `run echo $(( RANDOM % 10 ))
17:18:55 <Vorpal> speaking of which, where is fungot?
17:19:03 <Vorpal> fizzie, bring back fungot
17:19:14 <HackEgo> Runs arbitrary code in GNU/Linux. Type "`<command>", or "`run <command>" for full shell commands. "`fetch <URL>" downloads files. Files saved to $PWD are persistent, and $PWD/bin is in $PATH. $PWD is a mercurial repository, "`revert <rev>" can be used to revert to a revision. See http://codu.org/projects/hackbot/fshg/
17:19:40 <elliott> so what's the stupid set of ioctls you need to get a raw terminal
17:19:50 <Vorpal> elliott, you mean to implement one or?
17:20:16 <elliott> Vorpal: no, to put the terminal into one
17:20:41 <Vorpal> elliott, uh what are you trying to do exactly.
17:21:03 <elliott> Vorpal: it's very obvious what i mean to anyone who has done this before, if you haven't done any terminal programming you won't know
17:21:09 <elliott> nooga: yes but all the furries will use it.
17:21:39 <Vorpal> elliott, I have done some. Written a tool that emulated a terminal to be able to send a password to sudo
17:21:48 <Vorpal> or was it su? Was like 3 years ago
17:22:18 <nooga> http://www.ultravixens.net/ uh, the domain seems to be reserved
17:22:27 -!- nopseudoidea has joined.
17:22:39 <nooga> elliott: oh come on, the SADOL interpreter is called BDSM
17:23:09 <Vorpal> elliott, things that may be of use: console_ioctl(4) and termios(3) (the latter I doubt you mean but who knows)
17:23:23 <elliott> Vorpal: yes, it's doing what termios does
17:23:34 <elliott> so i'm going to do it with ioctls
17:23:38 <elliott> ah look, some nice code I can rip off!
17:24:14 <Vorpal> elliott, hm. termios big?
17:24:24 <elliott> Phantom_Hoover_: THIS IS FLINIX NOT KITTEN I TOLD YOU THAT
17:24:40 <Vorpal> elliott, but why? the code is only in one copy on the disk?
17:25:04 <elliott> Vorpal: because, like every abstraction, it uses more code than doing it directly
17:25:19 <elliott> <elliott> Phantom_Hoover: No way! This is The Insane Flinix.
17:25:20 <elliott> <elliott> Phantom_Hoover: A compleat (well, for some definition of compleat) Linux distribution, with networking, X11 and an IRC client, in one 1.44 meg floppy.
17:25:27 <Vorpal> elliott, but since there is bound to be some app that needs it anyway it will still exist in exactly one copy on the system
17:25:54 <Vorpal> elliott, after all, aren't you using a deduplicating filesystem?
17:26:16 <elliott> <Vorpal> elliott, but since there is bound to be some app that needs it anyway it will still exist in exactly one copy on the system
17:26:19 <elliott> you realise this is flinix?
17:26:22 <elliott> no, there will not be any other app that uses it
17:26:27 <elliott> because there is not enough space on a floppy for that.
17:26:45 <elliott> <elliott> <elliott> Phantom_Hoover: No way! This is The Insane Flinix.
17:26:45 <elliott> <elliott> <elliott> Phantom_Hoover: A compleat (well, for some definition of compleat) Linux distribution, with networking, X11 and an IRC client, in one 1.44 meg floppy.
17:26:45 <elliott> <Vorpal> elliott, but since there is bound to be some app that needs it anyway it will still exist in exactly one copy on the system
17:26:51 <elliott> BUY A CLIENT WITH SCROLLBACK ALREADY
17:27:10 <Vorpal> (as in why make that distro)
17:27:24 <elliott> Vorpal: why did you make cfunge
17:27:56 <Vorpal> elliott, because I wanted to code something and was bored, and it looked like an interesting thing to code.
17:28:25 <elliott> Vorpal: so, in short: "because i wanted to."
17:28:41 <Vorpal> elliott, but I though you were busy with kitten?
17:29:54 <elliott> Vorpal: pending on coreutils
17:31:17 <Vorpal> elliott, ah, you took a pause from writing that?
17:31:40 <Vorpal> elliott, also which X will you fit on that floppy
17:33:09 <elliott> and yes, this is my way of avoiding the infinite tedium of writing a coreutils
17:33:26 <elliott> Vorpal: smallX is this server that ran on 386s with 4 megs of ram as part of Small Linux
17:33:35 <elliott> I just need to find a tarball on an ftp server that still exists :)
17:33:51 <elliott> it has two servers, mono (for Hercules mono cards and the like) and 16-colour VGA
17:34:05 <elliott> should suffice to run a very small terminal program at least
17:36:07 -!- Quadrescence has quit (Ping timeout: 245 seconds).
17:36:41 <elliott> -rwxr-xr-x 1 elliott elliott 1639 Dec 7 17:36 vi
17:36:50 <elliott> it goes into raw mode, reads one character, puts the mode back, and exits :)
17:38:42 <nooga> good luck with writing that
17:39:04 <Vorpal> elliott, can't you reuse code for some tools from heirloom?
17:39:17 <elliott> Vorpal: i looked at heirloom code; it may be Olde Tymes but it sure isn't lean
17:39:27 <elliott> Vorpal: in fact i'd say it's about as ugly as BSD core utils. maybe more
17:39:44 <elliott> also i don't /think/ it has mount, not sure
17:39:52 <nooga> elliott: http://hedgehog.oliotalo.fi/ bundle this
17:39:55 <Vorpal> elliott, saves on binary size to merge tools ;P
17:40:03 <elliott> nooga: what lang is it implemented in? C?
17:40:06 <Vorpal> ie. your goals are somewhat conflicting
17:40:22 <elliott> Vorpal: anyway it's more interface bloat i care about, it's just for fun, the size thing
17:40:29 <elliott> Vorpal: asmutils whoop me thoroughly in the ass as far as size goes
17:40:32 <elliott> (and i'm using them in flinix)
17:40:46 <nooga> elliott: who cares, it compiles to 20kB
17:41:13 <Vorpal> elliott, why not use busybox or asmutils then?
17:41:30 <nooga> elliott: and yes, it's in C
17:41:39 <elliott> nooga: 20K? pah, I can get it less
17:42:01 <elliott> nooga: what is the shared secret for :p
17:42:43 <nooga> but it's featureful
17:43:06 <nooga> and proveides a damn small scripting language for flinix
17:44:06 <elliott> nooga: unfortunately, flinix is too useless to need one
17:44:10 <elliott> nooga: what executable is 20K for you?
17:45:45 <nooga> oh, the interpreter
17:46:07 <elliott> -rwxr-xr-x 1 elliott elliott 30067 Dec 7 17:45 hhi
17:46:32 <elliott> UNFORTUNATELY, I would rather spend the bytes on a C compiler.
17:46:39 <elliott> (anyone know a reaaaaaaaaally tiny libc? :P)
17:48:26 <nooga> http://www.fefe.de/dietlibc/
17:49:13 <Vorpal> elliott, yes the one I used on the rcx. That was libc in the "This isn't hosted so fuck you" meaning though
17:49:54 <elliott> nooga: dude, i'm well aware of dietlibc
17:50:07 <elliott> like one meg libc.a or so with stuff enabled
17:50:11 <elliott> maybe a few hundred K with stuff disabled
17:50:20 <elliott> I'm looking for 40K, 50K or so
17:50:23 <elliott> the one in asmutils is 20K
17:50:28 <Vorpal> elliott, after all, the whole OS and any programs have to fit into a 16-bit address space. Together with a rather large ROM.
17:50:29 <elliott> but i have a feeling it does not implement much :)
17:50:52 <elliott> Vorpal: you think that's bad, my fucking kernel takes up 480K to do barely anything at all for me
17:51:16 * elliott considers not supporting command repeats >9
17:52:06 <nooga> when will be the first release?
17:52:49 <Vorpal> elliott, why not drop printf. It tends to take a lot of space. You can have snprintf that does %s and %x, that is enough for anyone
17:53:11 <elliott> Vorpal: i don't even use printf
17:53:16 <elliott> nooga: as soon as I get X working
17:53:29 <elliott> ...somehow, my loop supporting any int is smaller than a conditional just supporting one digit
17:53:37 <elliott> Vorpal: (even in my real programs)
17:53:46 <elliott> wait, no it isn't, it's quite a bit bigger like that
17:56:28 <nooga> then what do you use instead of printf() ?
17:58:23 <nooga> then how to you print numbers?
17:59:42 <elliott> nooga: i have an ltostr routine i carry around :p if I was less silly, I'd just use libowfat, which has formatted printing using write
17:59:47 <elliott> http://www.fefe.de/libowfat/
18:00:03 <elliott> so i'd probably use http://www.fefe.de/djb/ directly
18:11:15 -!- nopseudoidea has quit (Quit: Quitte).
18:13:15 <elliott> Phantom_Hoover_: err, me? in what?
18:13:57 -!- nooga has quit (Ping timeout: 240 seconds).
18:15:06 -!- nooga has joined.
18:16:07 <nooga> my internet connection is flaky
18:16:25 <elliott> Phantom_Hoover_: I don't plan to write any code.
18:16:32 <elliott> So, Julian Assange arrested in this piece of shit nation, denied bail.
18:16:40 <elliott> On the rape "charges", lawl.
18:17:25 <elliott> nooga: yeah you can make it emulate vi by default
18:18:29 <nooga> http://piumarta.com/software/peg/ how awesome
18:18:30 <elliott> -rw-r--r-- 1 elliott elliott 615K Dec 7 18:15 dietlibc.a
18:18:40 <elliott> nooga: you didn't know about PEGs?
18:18:45 -!- MigoMipo_ has joined.
18:19:01 <nooga> i learnt about it when playing with tinyrb
18:19:22 <elliott> so does anyone know about a REALLY tiny libc :)
18:21:21 -!- MigoMipo has quit (Ping timeout: 240 seconds).
18:28:15 <Ilari> Hah: "If two of the three co-authors on this draft cannot agree on the semantics, then it is maybe premature for them to expect the Internet community to do so."
18:29:24 <nooga> then i don't have an idea
18:34:47 <elliott> .suicide:db__n,"Suicide is painless...", EOL
18:34:47 <elliott> .stop:db__n,"You say STOP and I say go...", EOL
18:34:47 <elliott> .nosuchpiddb"Child is 0xDEAD. I'm sorry", __n, EOL
18:44:52 <elliott> Dude, I bloated my system with ethernet drivers.
18:48:38 -!- Quadrescence has joined.
18:51:46 <elliott> 15:37:38 <AnMaster> and it is imperative that I do not use facebook
18:51:51 <elliott> Vorpal will literally explode if he uses facebook.
18:53:18 <elliott> 15:46:26 <AnMaster> Gregor-W, use your own colour matcher!
18:53:18 <elliott> 15:46:32 <Gregor-W> AnMaster: I do. Inverted.
18:53:26 <elliott> Gregor: Didn't you write it to *avoid* doing that kind of stuff? :p
18:54:45 <elliott> ais523: can you tell esr that the problem with open source is that the linux kernel doesn't let you disable some stuff?
18:54:48 <elliott> ais523: i hear he's influential
18:56:29 -!- wareya_ has quit (Read error: Connection reset by peer).
18:56:49 -!- wareya has joined.
18:57:35 <Vorpal> elliott, 1) when was that quote from?
18:57:43 <Vorpal> 2) <elliott> Vorpal will literally explode if he uses facebook. <-- duh no. Implode
18:57:52 <Vorpal> explode if I use facebook? How silly
18:57:58 <elliott> After trying to comprehend:
18:58:01 <Vorpal> everyone knows it is implosion
18:58:07 <elliott> which was said on that day.
18:58:11 <elliott> (C'mon, HackEgo! You can do it!)
18:58:16 <elliott> (I know running grep is hard. But c'mon!)
18:58:23 <elliott> (I know, I know, I blame GNU bloat too.)
18:58:35 <elliott> (But you're on Debian; there's no escape. (Maybe I should port Plash to Flinix.))
18:58:39 <elliott> Vorpal: don't give it MORE things to think about!
18:58:50 <Vorpal> elliott, uptime shouldn't be complex
18:58:51 <HackEgo> 162) <calamari> anmaster gonna give him a birthday bj? <AnMaster> IF ONLY I COULD FIND MY PHONE
18:58:51 <HackEgo> 18:58:27 up 4 days, 22:16, 0 users, load average: 1.55, 1.09, 0.74
18:59:10 <Vorpal> elliott, that looks rather like "different contexts"
18:59:26 <elliott> There were messages in-between :P
18:59:31 <Vorpal> elliott, then it is cheating
18:59:39 <elliott> Yes it is, but Gregor added it and it's his bot.
19:00:11 <Vorpal> elliott, I mean, it is similar to finding hidden messages in the bible (or any other large book)
19:00:24 <HackEgo> 64) <apollo> So... copyright doesn't really apply to God.
19:00:35 <Vorpal> elliott, I'm sure I could create something amusing by picking lines you said over the years...
19:00:35 <elliott> Vorpal: to be fair, you were also adding the relevant person's birthday to your phone at the time.
19:00:41 <elliott> so I think any mocking is perfectly okay
19:01:25 <Vorpal> elliott, also: I use my phone calendar for everything that a calendar is useful for basically.
19:01:34 <elliott> Vorpal: you never did give him his birthday bj though
19:02:11 <Vorpal> elliott, I don't do that kind of stuff :P
19:02:19 <elliott> Vorpal: But you're fine with sodomy?
19:02:28 <elliott> d was my script to remove a quote from my "q" file.
19:02:35 <elliott> It just did grep -v "^$1)"
19:02:40 <elliott> so I could use the same numbers without refreshing
19:02:46 <elliott> (with the main quotes file numbers rejiggle if you remove quotes)
19:02:49 <Vorpal> elliott, only if it enhances the joke :P
19:02:59 <elliott> Which is, admittedly, not the best idea, but it's more elegant to implement :P
19:03:13 <Vorpal> <elliott> Vorpal: But you're fine with sodomy? <elliott> Which is, admittedly, not the best idea, but it's more elegant to implement :P
19:03:20 <Vorpal> see what context (lack of) can do
19:03:26 <elliott> Vorpal: Not much at all? :P
19:03:37 -!- cheater99 has quit (Ping timeout: 245 seconds).
19:03:38 <Vorpal> elliott, indeed it is rather confusing
19:03:39 <Gregor> Sodomy: Definitely elegant.
19:04:03 <elliott> That sounds like it could be smaller than 2.6.36!
19:04:19 <Vorpal> <elliott> It just did grep -v "^$1)" <-- you should use sed's d command
19:04:35 <elliott> Vorpal: But what if the number had a / in it???
19:04:40 <elliott> linux-1.0.tar.bz2 13-Mar-1994 00:00 1.0M
19:04:50 <Vorpal> elliott, doesn't it use N?
19:04:56 <Vorpal> elliott, rather than R
19:05:19 <elliott> oh, i see what you mean but you're wrong
19:05:22 <Vorpal> (Q because, "Rational numbers" starts with a Q)
19:05:24 <elliott> Vorpal: the whole point is that "60" is always the same line
19:05:41 <elliott> i misinterpreted your misinterpretation
19:05:55 <elliott> WHAT LINUX 1.0 DIDN'T USE MENUCONFIG HOW SILLY EH
19:06:11 <elliott> Oh god it's going to prompt me about EVERY DAMN THING.
19:06:24 <Vorpal> elliott, presumably it only started that some time after it became a PITA to not have it
19:06:58 <elliott> ...Wow. I just measured Flinix's memory usage after a cold boot.
19:07:00 <Vorpal> elliott, does it use plain make config?
19:07:15 <Vorpal> elliott, /proc/meminfo?
19:07:39 <elliott> TODO: Figure out why /etc/rc isn't working.
19:07:50 <Vorpal> elliott, weird output format for free(1)
19:08:03 <elliott> Oh, TODO: Use -t for mount in etc/rc...
19:08:06 <elliott> Vorpal: That was my retyping.
19:08:23 <Vorpal> elliott, you could add /proc to fstab
19:08:28 <Vorpal> but I guess that would waste space
19:08:32 <elliott> Vorpal: I don't do mount -a at boot.
19:08:34 <Vorpal> elliott, are you using a compressed fs?
19:08:49 <elliott> Vorpal: Yes and no. It's an initramfs compiled into the kernel, and the kernel is LZMA'd.
19:08:56 <elliott> So yes on floppy, no in RAM.
19:09:03 <Vorpal> elliott, how does it compare to squashfs?
19:09:11 <Vorpal> or does squashfs has too much overhead?
19:09:15 <elliott> Vorpal: How does LZMA compare to anything --> LZMA beats it.
19:09:18 <elliott> Vorpal: Squashfs probably requires the block layer.
19:09:36 <elliott> Vorpal: And I would have to have floppy support since you can't embed normal filesystems into the kernel.
19:09:47 <elliott> Vorpal: Remember, initramfs is literally just a .cpio that gets unpacked into a ramfs in memory :P
19:10:12 <elliott> Guh, why isn't /proc mounting.
19:10:24 <elliott> (Also TODO: Figure out how the fuck to use asmutils less :P)
19:10:26 <Vorpal> elliott, I suspect squashfs does better at larger images since it is somewhat "smart" with how it represents the FS before it goes to the compression bit.
19:10:42 <elliott> Vorpal: "smart"er than cpio?
19:10:53 <Vorpal> elliott, does cpio do block deduplication?
19:11:05 <elliott> Vorpal: LZMA does all the deduplication you need :P
19:11:07 <Vorpal> sure it isn't useful at small sizes since then the compressing will take care of that
19:11:27 <Vorpal> elliott, afaik lzma has a window like most other compression algorithms
19:11:48 <elliott> Vorpal: Anyway, with squashfs I'd need it to be able to read the floppy to read it.
19:11:48 <Vorpal> elliott, which means it won't help if those files end up far from each other in a large (few hundred MB or so) image
19:11:51 <elliott> And the floppy would need formatting.
19:11:57 <elliott> At which point I could just use any filesystem on the floppy and mount it as /.
19:12:12 <elliott> Vorpal: To have it actually useful, I'd have to have a bloated bootloader which can read filesystems :P
19:12:23 <elliott> Vorpal: My bootloader = The very core of FreeDOS + LINLD in AUTOEXEC.BAT.
19:12:26 <Vorpal> elliott, or why not place it inside the initramfs ;)
19:12:29 <elliott> (Well, my bootloader will-be.)
19:12:34 <elliott> Vorpal: Then I don't need an FS :P
19:12:47 <Vorpal> elliott, what, that is a waste. Why not just a hand written boot sector
19:13:21 <Vorpal> elliott, surely that will be less wasteful
19:13:28 <elliott> Vorpal: You try loading Linux in 512 bytes :P
19:13:33 <elliott> It'll be smaller than lilo, probably.
19:13:56 <Vorpal> elliott, well you could use another sector or two. Still smaller than DOS + linld probably
19:14:25 <Vorpal> elliott, just place the kernel starting in the sector after the bootloader
19:14:36 <Vorpal> (and adjust offsets as needed)
19:14:55 <Vorpal> elliott, well it would not need a patch against linux. Just a hand written boot loader
19:15:15 <Vorpal> elliott, what does linux need to load? the kernel image at a specific memory address. Anything else?
19:15:19 <elliott> Vorpal: Boot sectors welcome.
19:15:34 <Vorpal> elliott, I'm not insane enough to write real mode code
19:15:42 <oerjan> Vorpal: i suspect Q is for "quotient" (or whatever the german equivalent is)
19:15:49 <elliott> cmd_lzma = (cat $(filter-out FORCE,$^) | \
19:15:49 <elliott> lzma -9 && $(call size_append, $(filter-out FORCE,$^))) > $@ || \
19:15:52 <elliott> I THINK I CAN DO BETTER THAN -9
19:15:57 <Vorpal> I have never done so. IIRC you have. Thus you are more likely to succeed at that
19:16:05 <Vorpal> hey didn't you write a boot sector some time ago?
19:16:11 <elliott> WHY DOES LZMA NOT HAVE A --CRAZY
19:16:16 <elliott> Vorpal: Yes, but it wouldn't load Linux ...
19:16:20 <Vorpal> elliott, -9 isn't always best iirc
19:16:28 <Vorpal> elliott, for small files especially
19:16:40 <Gregor> I'll bet elliott mounts a case-insensitive filesystem just so he can scream his commands.
19:16:42 <Vorpal> try 1-9 both with and without -e
19:16:44 <elliott> Vorpal: This is lzma not xz fwiw.
19:16:49 <Vorpal> elliott, see which one wins
19:16:55 <Vorpal> elliott, oh, then that might make a difference
19:17:00 <elliott> lzma: invalid option -- 'e'
19:17:13 <Vorpal> elliott, yeah I don't remember if lzma had the same recommendation
19:17:23 <elliott> I'll try "xz --format=lzma -9 -e".
19:18:11 <Vorpal> elliott, I wonder what the file system structure of an initramfs looks like in memory. Is it just the cpio loaded as is
19:18:19 <Vorpal> or does it translate it somehow?
19:18:19 -!- cheater99 has joined.
19:18:33 <Vorpal> elliott, -9 isn't best with xz for small files iirc
19:18:52 <elliott> Vorpal: It loads it into a shmfs or a ramfs if it doesn't have tha-- oh, look, my kernel grew anyway, it isn't xz's fault.
19:19:32 <Vorpal> elliott, also when it comes to 1.0 I doubt it has initramfs, Maaaaaaybe initrd but I would be somewhat surprised at that too
19:20:08 <elliott> Wow, old Linux config was hateful. It doesn't let you pipe yes to it.
19:20:23 <elliott> # This script is used to configure the linux kernel.
19:20:24 <elliott> # It was inspired by the challenge in the original Configure script
19:20:24 <elliott> # to ``do something better'', combined with the actual need to ``do
19:20:24 <elliott> # something better'' because the old configure script wasn't flexible
19:20:25 <Vorpal> elliott, you might be better off with a 2.4 or 2.2 kernel
19:20:27 <elliott> # Please send comments / questions / bug fixes to raymondc@microsoft.com.
19:20:30 <elliott> # Please send comments / questions / bug fixes to raymondc@microsoft.com.
19:20:32 <elliott> # Please send comments / questions / bug fixes to raymondc@microsoft.com.
19:20:34 <elliott> # Please send comments / questions / bug fixes to raymondc@microsoft.com.
19:20:51 <elliott> Microsoft: IN BED WITH LINUX, CIRCA 1994.
19:21:00 <Vorpal> elliott, you mean it isn't a bad joke?
19:21:06 <elliott> IFS='@' read ans </dev/tty || exit 1
19:21:35 <Vorpal> elliott, presumably they didn't care if employees worked on such small hobby projects back then. I mean, they weren't really any threat back then
19:21:40 <elliott> It's someone called Raymond Chen and it MIGHT JUST BE THAT RAYMOND CHEN.
19:21:49 <elliott> Vorpal: http://blogs.msdn.com/b/oldnewthing/
19:22:03 <elliott> Famous blogger on the awful reasons why Windows is so awful.
19:22:29 <elliott> Vorpal: Well, he doesn't consider it awful :P
19:22:35 <elliott> But it's all the gory details of the gory details.
19:23:13 <elliott> IFS='@' read ans </dev/tty || exit 1
19:23:31 <elliott> "</dev/tty" in a program that asks a lot of questions is a synonym for "I hate you".
19:24:43 <Vorpal> elliott, anyway that thing won't have initrd
19:24:53 <elliott> Vorpal: No, but it'll probably be smaller even with floppy support :P
19:25:08 <Vorpal> elliott, augh at the top post of http://notch.tumblr.com/
19:26:18 <elliott> Vorpal: Will the endless bugs in Minecraft ever be fixed instead of terrible puns being posted to the blog? Notch-ance!
19:27:52 <elliott> as86 -0 -a -o boot/bootsect.o boot/bootsect.s
19:27:53 <elliott> make: as86: Command not found
19:28:01 <Vorpal> elliott, I believe the blog host was down for a few hours yesterday or so. It was down with a database error message when I tried at one point
19:28:07 <elliott> Oh joy, it doesn't look like gas syntax.
19:28:26 <Vorpal> elliott, does it look like intel? If so I'm sad. It should use an exotic third option
19:28:35 <elliott> It looks like AT&T but different :P
19:28:53 <Vorpal> elliott, different how?
19:29:06 * Vorpal invents a screwy asm syntax
19:29:16 <elliott> Description: 16-bit x86 assembler and loader
19:29:16 <elliott> This is the as86 and ld86 distribution written by Bruce Evans. It's a complete 8086 assembler and loader which can make 32-bit code for the 386+ processors.
19:29:23 <elliott> Vorpal: Numbers start with # :P
19:29:27 <Vorpal> elliott, "target mov source" or "source mov target"
19:29:44 <Vorpal> elliott, it should use mov
19:29:55 <Vorpal> elliott, I want infix asm
19:30:14 <elliott> see how the prefix is neatly separated from the base register
19:30:17 <Vorpal> elliott, seriously this is intended to be screwy, obscure and confusing :P
19:30:22 <Vorpal> it is not intended to be sane
19:30:23 <elliott> THAT IS SCREWY, OBSCURE AND CONFUSING
19:31:04 <elliott> as -c -o boot/head.o boot/head.s
19:31:04 <elliott> as: unrecognized option '-c'
19:31:05 <Vorpal> elliott, it should use different register names
19:31:25 <elliott> boot/head.S: Assembler messages:
19:31:25 <elliott> boot/head.S:64: Error: suffix or operands invalid for `push'
19:31:26 <elliott> boot/head.S:65: Error: suffix or operands invalid for `popf'
19:31:26 <elliott> boot/head.S:99: Error: suffix or operands invalid for `pushf'
19:31:49 <elliott> Oh, fuck this shit, I'll try Linux 2.0.
19:32:15 <Vorpal> elliott, what about that
19:32:37 <elliott> Vorpal: Yes, except not ex; have an infix operand size modifier.
19:33:02 <Vorpal> elliott, lets rename the registers
19:33:06 <elliott> Vorpal: And also have blocks of commands, so that instead of "bits 32", you do:
19:33:28 <elliott> OK, Linux 2.0.1, if you're bloated I'll fuck your shit up.
19:33:34 <Vorpal> elliott, since 2^5 = 32
19:33:39 <elliott> Your shit... will be maximally fucked up.
19:33:54 <elliott> Vorpal: But that's not what the operation does! Instead, that should be
19:33:54 <Vorpal> elliott, you aren't persistent enoigh
19:34:02 <elliott> Vorpal: But "mov ax, bx" would be
19:34:12 <elliott> (1,2) mov &1111111111111111
19:34:19 <elliott> Vorpal: Because it modifies only the lower 16 bits.
19:34:35 <elliott> Vorpal: Because it's like bitwise and if you squint.
19:35:09 <elliott> Vorpal: All numbers should be trinary.
19:35:34 <elliott> Vorpal: (1,2) mov &10022220020
19:35:35 <Vorpal> elliott, there is a fine line between "screwy" and "malbolge86"
19:35:39 <Vorpal> I think you just crossed it
19:35:41 <elliott> Vorpal: Yes. Let's cross it!
19:35:44 <oerjan> PLuggable ASsembler MAcros
19:35:59 <Vorpal> elliott, so what about base pointer and indexing?
19:36:14 <elliott> Vorpal: Those are done as virtual instructions.
19:36:18 <Vorpal> elliott, also they should be balanced ternary gray code
19:36:22 <Vorpal> for maximum screwiness
19:36:27 <elliott> Vorpal: For instance, what is, in Intel, "mov [ebx], [eax+2]"
19:36:32 <elliott> would be, in this, let me think...
19:36:58 <olsner> you can't have two memory operands in an instruction like that
19:37:07 <elliott> olsner: Okay, okay, okay, fine.
19:37:27 <Vorpal> elliott, did you just increment eax by two first?
19:37:47 <elliott> Vorpal: "Increment, virtually, register 1, by 2. Deaddress, virtually, register 1. Move register 1 to register 2."
19:37:55 <elliott> It's (2,1) incv because we do everything in the most illogical order.
19:37:59 <Vorpal> elliott, what about: mov eax, [eax+2]
19:38:03 <Vorpal> (not sure that is valid even)
19:38:14 <nooga> elliott: which kernel version do you have now?
19:38:42 <elliott> Vorpal: Let me try that again, all on one line.
19:38:57 <olsner> Vorpal: of course you can load into a register you use for the address
19:39:01 <elliott> (-1,1) movv; (2,-1) incv; (-1) deaddrv; (1,-1) mov
19:39:16 <elliott> Vorpal: Registers not part of the architecture work in virtual instructions, since virtual instructions don't actually translate to instructions!
19:39:18 <Vorpal> I meant the other way around
19:39:27 <oerjan> SOurce Language Independent Disassembler
19:39:31 <elliott> Vorpal: So the virtual mapping ends up being:
19:39:49 <elliott> Vorpal: Well, no, you can't :P
19:39:51 <elliott> <elliott> (-1,1) movv; (2,-1) incv; (-1) deaddrv; (1,-1) mov
19:39:59 <Vorpal> elliott, can you in normal asm?
19:40:00 <elliott> The movv creates the map {-1 => 1}.
19:40:06 <elliott> i.e., virtual register -1 is real register 1.
19:40:13 <elliott> Then the invc turns it into {-1 => 1+2}.
19:40:19 <elliott> Then the deaddrv turns it into {-1 => [1+2]}.
19:40:26 <elliott> Then the mov substitutes [1+2] for -1.
19:40:38 <elliott> but I translated what your snippet means in Intel correctly :P
19:40:53 <elliott> virtual instructions use [] instead of ()
19:41:04 <elliott> Only on the side that's virtual.
19:41:16 <Vorpal> elliott, what about rotating the bits of the register numbers one step for each instruction?
19:41:17 <elliott> [-1,1) mov; (2,-1] inc; [-1] deaddr; (1,-1] mov
19:41:20 <olsner> wtf does "virtual" mean?
19:41:28 <elliott> olsner: simple: the calculation is done by the assembler
19:41:42 <elliott> olsner: so in this case, it has this idea of an imaginary register, which is first the value of 1, and which it then imaginarily increments
19:41:49 <elliott> olsner: and it then imaginarily deaddresses it
19:41:57 <elliott> olsner: so when you use -1, it replaces it with [eax+2]
19:41:57 <oerjan> Language Independent QUerying and Interactive Debugger
19:41:58 <Vorpal> rotate it by the hash for the instruction in big endian
19:42:14 <Vorpal> elliott, hey malbolge86
19:42:18 <Vorpal> elliott, it would be fitting
19:42:37 <elliott> oerjan: Creatively Retarded & Addled Computing Kompiler
19:42:57 <Gregor> As opposed to a non-computed "k"ompiler.
19:43:24 <elliott> YES! LINUX 2.0 HAS MENUCONFIG!
19:43:34 <elliott> IT'S LIKE AN ORGASM EXCEPT FOR LINUX KERNEL CONFIGURATION AFTER SUFFERING THROUGH "MAKE CONFIG"
19:43:35 <oerjan> elliott: hey you're not keeping to the theme.
19:43:51 <elliott> oerjan: crack is the fourth form of matta
19:44:02 <Gregor> elliott: The best way to handle make config is yes '' | make config :P
19:44:04 <Vorpal> elliott, I read that as "crack is the forth form of matta"
19:44:08 <elliott> Gregor: It read from /dev/tty.
19:44:15 <Vorpal> as in the programming language forth
19:44:36 <Gregor> elliott: Sweet! ... rm -f /dev/tty && mkfifo /dev/tty && yes '' > /dev/tty &
19:44:43 <olsner> elliott: make /dev/tty a symlink to /dev/fd/0 or something :)
19:44:47 <elliott> Gregor: I approve! (I just hacked the shell script.)
19:44:48 <Vorpal> elliott, yes and I read "matta" in Swedish
19:44:52 <Vorpal> elliott, since it is a valid Swedish word
19:44:58 <elliott> oerjan: fine: Pleonasmtastic Lavish Assembler, Stricken Maliciously from Autonomy
19:45:01 <Vorpal> at which point I did a retake of the whole thing
19:45:05 <oerjan> Vorpal: i thought it a bit unfair that only gas was a programming acronym
19:45:06 <elliott> I hereby claim rights to the word "pleonasmtastic".
19:45:10 <oerjan> elliott: i did plasma abova
19:45:16 <elliott> oerjan: SHUT UP, PLEONASMTASTIC
19:45:21 <olsner> Vorpal: från matta till crack i tre steg!
19:45:45 <Gregor> `translate <olsner> Vorpal: frn matta till crack i tre steg!
19:45:47 <elliott> Vorpal: Horor matta spricka!
19:46:11 <oerjan> Gregor: since when is `translate working again?
19:46:17 <HackEgo> <olsner> Vorpal: From carpet to crack in three steps!
19:46:19 <Gregor> oerjan: Apparently it isn't :P
19:46:21 <Vorpal> elliott, "spricka" = "crack in the sense crack in the wall"
19:46:29 <Vorpal> elliott, not crack in the sense "drug"
19:46:30 <Gregor> Google wasn't happy with me or something :P
19:46:45 <Gregor> `translateto se If I said you had a beautiful body, would you hold it against me?
19:46:59 <Vorpal> elliott, also other errors in it
19:47:10 <elliott> │ │ [*] Compile kernel as ELF - if your GCC is ELF-GCC │ │
19:47:10 <elliott> │ The gcc version 2.7.0 and newer produces the new ELF binary format │
19:47:11 <elliott> │ as default. If you have such a compiler (try "gcc -v"), say Y │
19:47:35 <Vorpal> elliott, that was wider than my IRC window and totally unreadable
19:47:49 <elliott> Vorpal: You're totally unreadable.
19:47:50 <Vorpal> elliott, isn't initrd just gzip-compressed ramfs?
19:47:51 <olsner> elliott: maybe you should get a pre-2.7 gcc then :)
19:47:59 <Vorpal> elliott, that is what your mum said
19:48:04 * Gregor wonders why translateto-se didn't work ...
19:48:08 <Gregor> `translateto es If I said you had a beautiful body, would you hold it against me?
19:48:09 <Vorpal> elliott, both of those lines
19:49:31 <elliott> cc1: error: unrecognized command line option "-m386"
19:49:48 <elliott> cc1: error: unrecognized command line option "-m486"
19:50:14 <Vorpal> elliott, old name for that
19:51:27 <oerjan> ...right language, but completely wrong translation :D
19:51:40 <oerjan> olsner: i know that :D
19:51:52 <elliott> oerjan: no you don't you're a hopeless... nord?
19:52:28 <Gregor> WTF, I'm using the web-services API now :P
19:52:29 <oerjan> olsner: i was just trying that because i knew the translation should have no special characters
19:53:05 <Gregor> `run curl -e http://codu.org/ http://ajax.googleapis.com/ajax/services/language/translate --data-urlencode v=1.0 --data-urlencode q="What the fuck?" --data-urlencode langpair=en|se
19:53:09 <Vorpal> `translateto sv Swedish
19:53:11 <Gregor> `run curl -e http://codu.org/ http://ajax.googleapis.com/ajax/services/language/translate --data-urlencode v=1.0 --data-urlencode q="What the fuck?" --data-urlencode langpair='en|se'
19:53:13 <HackEgo> {"responseData": null, "responseDetails": "invalid translation language pair", "responseStatus": 400}
19:53:17 <oerjan> `translateto sv one three four
19:53:24 <elliott> Did Gregor change anything there?
19:53:26 <Vorpal> oerjan, huh that worked
19:53:26 <oerjan> `translateto sv one two three four
19:53:30 <elliott> Did Gregor change anything there?
19:53:36 <elliott> `translateto sv unix eunuchs
19:53:43 <oerjan> Vorpal: yes, i carefully skipped two to avoid the å
19:53:58 <elliott> One, three, four, seven, twenty-nine.
19:54:13 <olsner> Gregor: the language is sv even though the country is se
19:54:13 <Vorpal> oerjan, oh you mean it fails on åäö in either input or output?
19:54:14 <oerjan> Gregor: i believer `translateto sv still has problem with æøå output
19:54:39 <Gregor> `run curl -e http://codu.org/ http://ajax.googleapis.com/ajax/services/language/translate --data-urlencode v=1.0 --data-urlencode q="What the fuck?" --data-urlencode langpair='auto|sv'
19:54:40 <HackEgo> {"responseData": null, "responseDetails": "invalid translation language pair", "responseStatus": 400}
19:54:56 <Gregor> `run curl -e http://codu.org/ http://ajax.googleapis.com/ajax/services/language/translate --data-urlencode v=1.0 --data-urlencode q="What the fuck?" --data-urlencode langpair='auto|es'
19:54:56 <Vorpal> oerjan, also bachgfj... is the nautural order to write the letters of the alphabet in :P
19:54:57 <HackEgo> {"responseData": null, "responseDetails": "invalid translation language pair", "responseStatus": 400}
19:56:02 <Vorpal> elliott, he wrote äöå. Listing a series of letters of the alphabet you normally do in alphabetic order. Out of habit and convenience I guess.
19:56:29 <Vorpal> elliott, so åäö would be much more natural
19:56:51 <elliott> abcdefz and that's all the letters
19:57:13 <Vorpal> elliott, no.... dzfecab
19:57:21 <Vorpal> elliott, do try to get it right
19:57:26 <elliott> you can't write "fecal" with those
19:57:35 <Vorpal> elliott, same as you listed :P
19:57:53 <oerjan> Vorpal: the order is æøå in norwegian
19:57:54 <Vorpal> <elliott> abcdefz and that's all the letters <Vorpal> elliott, no.... dzfecab <-- that is just a different permutation
19:58:13 <elliott> Vorpal: no you did yours wrong
19:58:45 <Vorpal> elliott, what letters differ
19:59:32 <Vorpal> elliott, no... you have a-f+z So do I
19:59:45 <Vorpal> elliott, you just wrote your list in a different (and very incorrect) order
19:59:58 <oerjan> Gregor: in any case `translateto sv _is_ working sometimes
20:00:15 <Vorpal> Gregor, it fails at unicode it seems?
20:00:33 <Vorpal> `run curl -e http://codu.org/ http://ajax.googleapis.com/ajax/services/language/translate --data-urlencode v=1.0 --data-urlencode q="one three" --data-urlencode langpair='auto|es'
20:00:35 <HackEgo> {"responseData": null, "responseDetails": "invalid translation language pair", "responseStatus": 400}
20:00:37 <Gregor> `run curl -e http://codu.org/ http://ajax.googleapis.com/ajax/services/language/translate --data-urlencode v=1.0 --data-urlencode q="What the fuck?" --data-urlencode langpair='en|es'
20:00:38 <HackEgo> {"responseData": {"translatedText":"¿Qué carajo?"}, "responseDetails": null, "responseStatus": 200}
20:00:40 <Vorpal> `run curl -e http://codu.org/ http://ajax.googleapis.com/ajax/services/language/translate --data-urlencode v=1.0 --data-urlencode q="one three" --data-urlencode langpair='auto|sv'
20:00:41 <HackEgo> {"responseData": null, "responseDetails": "invalid translation language pair", "responseStatus": 400}
20:00:42 <oerjan> Gregor: and the times it didn't work afaict were precisely the times when the output would contain å ä or ö
20:00:47 <Vorpal> `run curl -e http://codu.org/ http://ajax.googleapis.com/ajax/services/language/translate --data-urlencode v=1.0 --data-urlencode q="one three" --data-urlencode langpair='en|sv'
20:00:48 <HackEgo> {"responseData": {"translatedText":"en tre"}, "responseDetails": null, "responseStatus": 200}
20:00:54 <Vorpal> `run curl -e http://codu.org/ http://ajax.googleapis.com/ajax/services/language/translate --data-urlencode v=1.0 --data-urlencode q="one two three" --data-urlencode langpair='en|sv'
20:00:55 <HackEgo> {"responseData": {"translatedText":"en två tre"}, "responseDetails": null, "responseStatus": 200}
20:01:09 <Vorpal> Gregor, so unicode failure
20:01:14 <oerjan> Phantom_Hoover_: something = något in swedish, so confirms the theory
20:01:25 <Vorpal> `translateto sv en två tree
20:01:33 <Vorpal> `translateto sv one two three
20:01:48 <Vorpal> Gregor, what do you do to fail so badly on unicode?
20:02:00 <oerjan> `translateto no something
20:02:07 <Gregor> Vorpal: Note how it worked fine in the above lines.
20:02:20 <Vorpal> Gregor, only in those that I ran curl with
20:02:29 <elliott> │ │ [*] Support Intel processors │ │
20:02:30 <elliott> │ │ [*] Support AMD processors │ │
20:02:30 <Vorpal> Gregor, the script fails when it encounters unicode
20:02:33 <elliott> do you actually need these?
20:02:42 <Vorpal> elliott, I believe so?
20:02:51 <Vorpal> elliott, it probably checks cpuid
20:02:55 <Gregor> `translateto en Hello, world!
20:02:57 <Vorpal> elliott, check the code for what it does
20:03:09 <Gregor> `run cat bin/json | paste
20:03:10 <HackEgo> http://codu.org/projects/hackbot/fshg/index.cgi/raw-file/tip/paste/paste.178
20:03:17 <Vorpal> Gregor, this just confirms the theory that it fails at anything containing unicode in the string
20:03:37 <Gregor> Vorpal: If it's failing for that reason, then it's Python's json module, which seems extraordinarily unlikely to me.
20:03:40 <elliott> hmm qemu has ISA support right?
20:03:43 <Vorpal> `run python --version | tr $'\n' '|'
20:03:56 <elliott> <Gregor> Vorpal: If it's failing for that reason, then it's Python's json module, which seems extraordinarily unlikely to me.
20:03:59 <Vorpal> `run /usr/bin/env python --version | tr $'\n' '-'
20:04:00 <elliott> Gregor: Did you do .decode('utf8')?
20:04:07 <Vorpal> `run /usr/bin/env python --version 2>&1 | tr $'\n' '-'
20:04:12 <Gregor> elliott: http://codu.org/projects/hackbot/fshg/index.cgi/raw-file/tip/paste/paste.178
20:04:15 <elliott> You don't need to do that tr...
20:04:24 <Vorpal> elliott, well I thought it was multi-line
20:04:32 <Gregor> elliott: I have no idea how to Python.
20:04:43 <Gregor> elliott: These are basically the only lines of Python I have ever written.
20:04:43 <elliott> Gregor: What is the filename?
20:05:07 <elliott> `run grep json.loads bin/json
20:05:09 <HackEgo> data = json.loads(sys.stdin.read())
20:05:22 <Vorpal> Gregor, what is argv[1] in print(eval(sys.argv[1]))
20:05:24 <Gregor> elliott: ... you realize that's the same thing I just pasted at you, right ...
20:05:42 <Gregor> Vorpal: data["responseData"]["translatedText"]
20:05:54 <elliott> `run sed -i "s/sys.stdin.read()/sys.stdin.read().decode('utf-8')/g" bin/json
20:05:57 <Vorpal> Gregor, so... what if it contains unicode? Could it be that it fails at that
20:06:08 <elliott> Vorpal: No, that should be OK.
20:06:08 <Vorpal> `translateto sv en två
20:06:13 <Gregor> `translateto sv If I said you had a beautiful body, would you hold it against me?
20:06:15 <Vorpal> `translateto sv one two
20:06:20 <HackEgo> #!/usr/bin/env python \ import json \ import sys \ data = json.loads(sys.stdin.read().decode('utf-8')) \ print(eval(sys.argv[1]))
20:06:32 <elliott> `run sed -i "s/sys.argv[1]/sys.argv[1].decode('utf-8')/g" bin/json
20:06:41 <elliott> `translateto sv If I said you had a beautiful body, would you hold it against me?
20:06:42 <Vorpal> elliott, wrong direction
20:06:46 <Gregor> elliott: There is no Unicode in the arg.
20:06:48 <elliott> Vorpal: I copied *your* line.
20:06:53 <Vorpal> elliott, yes I typoed :P
20:06:58 <Vorpal> `translateto sv one three
20:07:03 <elliott> Gregor: You are total fail at thinking print is a function to start with :P
20:07:07 <elliott> Gregor: OK: Could the result of eval be unicode?
20:07:15 <Gregor> elliott: I - do - not - know - Python
20:07:17 <elliott> I say "yes, yes it could".
20:07:20 <Gregor> elliott: Quite easily.
20:07:32 <elliott> `run echo '#!/usr/bin/env python' >bin/json
20:07:45 <elliott> `run echo 'import sys' >>bin/json; echo 'import json' >>bin/json
20:07:52 <Gregor> Not a fan of echo -e eh :P
20:08:15 <elliott> `run echo "data = json.loads(sys.stdin.read().decode('utf-8'))" >>bin/json
20:08:22 <Vorpal> you could use $'' instead
20:08:26 <elliott> `run echo "print eval(sys.argv[1]).encode('utf-8')" >>bin/json
20:08:28 <elliott> Vorpal: But I don't want to.
20:08:42 <elliott> There, now where's my birthday bj.
20:09:07 <Gregor> `translateto sv If I said you had a beautiful body, would you hold it against me?
20:09:08 <HackEgo> Om jag sa att du hade en vacker kropp, skulle du hålla det emot mig?
20:09:13 <Gregor> `translateto ch If I said you had a beautiful body, would you hold it against me?
20:09:21 <elliott> Gregor: BTW, I hereby release the code I've written under the Give Elliott All Your Sperm Public License, version 1 or later.
20:09:21 <Gregor> `translateto zh If I said you had a beautiful body, would you hold it against me?
20:09:22 <Vorpal> `translateto en en två tree
20:09:24 <Vorpal> `translateto en en två tre
20:09:35 <elliott> Gregor: Did I mention it's viral? (Kinda like my sperm, but I digress.)
20:09:44 <Vorpal> it is just a non-existent word
20:09:46 <Gregor> `translateto zh If I said you had a beautiful body, would you hold it against me? I am no longer infected.
20:09:47 <HackEgo> 如果我说你有一个美丽的身体，你会嫌弃我吗？我不再感染。
20:09:52 <olsner> `translateto sv If I said you had a beautiful body, would you hold it against me?
20:09:55 <HackEgo> Om jag sa att du hade en vacker kropp, skulle du hålla det emot mig?
20:09:59 <Vorpal> `translateto en att vara eller inte vara, det är frågan
20:10:01 <Gregor> `translate 如果我说你有一个美丽的身体，你会嫌弃我吗？我不再感染。
20:10:06 <elliott> Clearly, the Japanese have no word for rape.
20:10:06 <HackEgo> to be or not be, that is the question
20:10:06 <HackEgo> If I said you had a beautiful body would you hold anything against me? I am no longer infected.
20:10:08 <oerjan> Gregor: heh sadly the swedish translation fails to preserve the ambiguity solely because because of pronoun gender
20:10:27 <elliott> `translateto jp everyone is green
20:10:36 <elliott> Wait, the language code isn't jp, is it.
20:10:38 <ais523> elliott: <Commando125> o
20:10:44 <ais523> see, randomly oing is spreading from channel to channel
20:10:44 <elliott> ais523: okokokokokokokokokoko
20:10:49 <Vorpal> <oerjan> Gregor: heh sadly the swedish translation fails to preserve the ambiguity solely because because of pronoun gender <-- which ambiguity?
20:11:15 <oerjan> Vorpal: of "If I said you had a beautiful body, would you hold it against me?"
20:11:17 <olsner> Vorpal: you completely fail
20:11:24 <elliott> i think Vorpal has actually missed the ambiguity
20:11:31 <elliott> `translateto zh Nethack stole my gender.
20:11:34 <Vorpal> I never realised you could read it *THAT* way
20:11:34 <elliott> `translateto zh Nethack stole my gender in space.
20:11:40 <elliott> `translateto zh Nethack stole my gender in space in space.
20:11:44 <elliott> `translateto zh Nethack stole my gender in space in space in space in space in space in space.
20:11:46 <HackEgo> 在太空中的nethack偷走了我的性别空间在空间空间在空间的空间。
20:11:47 <oerjan> the two meanings turn the "it" into "den" or "det" in swedish respectively
20:11:51 <elliott> `translateto zh Nethack stole my gender in space in space in space in space in space in space. in space.
20:11:52 <HackEgo> 在太空中的nethack偷走了我的性别空间在空间空间在空间的空间。在太空中。
20:12:00 <Gregor> `translateto zh This channel is now under the control of red China. Please continue your esoteric activities unperturbed.
20:12:01 <HackEgo> 该通道目前正在红色中国的控制权。请继续深奥的活动不受干扰。
20:12:11 <Vorpal> elliott, I always thought it was the non-physical meaning. And found that line somewhat strange.
20:12:16 <elliott> `translateto sv bork bork bork
20:12:19 -!- Gregor has set topic: 该通道目前正在红色中国的控制权。请继续深奥的活动不受干扰。 | http://tunes.org/~nef/logs/esoteric/?C=M;O=D.
20:12:35 <elliott> `translatefromto sv en bork bork bork
20:12:49 <Vorpal> elliott, it means nothing in Swedish either :P
20:12:57 <elliott> Vorpal: It means all your words!
20:12:59 <Slereah> `swedish The coyote is a living, breathing allegory of Want. He is always hungry. He is always poor, out of luck, and friendless. The meanest creatures despise him, and even the fleas would desert him for a velocipede. He is so spirtless and cowardly that even while his exposed teeth are pretending a threat, the rest of his face is apologizing for it. And he is so homely! -so scrawny, and
20:12:59 <Slereah> ribby, and coarse-haired, and pitiful.
20:13:00 <HackEgo> Zee cuyute-a is a leefing, breetheeng ellegury ooff Vunt. He-a is elveys hoongry. He-a is elveys puur, oooot ooff loock, und freeundless. Zee meunest creetoores despeese-a heem, und ifee zee flees vuoold desert heem fur a feluceepede-a. He-a is su speertless und cooerdly thet ifee vheele-a hees ixpused teet ere-a pretundeeng
20:13:12 <Gregor> `translate 该通道目前正在红色中国的控制权。请继续深奥的活动不受干扰。 elliott: HALP
20:13:13 <HackEgo> The channel is currently being Red China's control. Please continue to esoteric activities without interference. elliott: HALP
20:13:26 <elliott> Gregor: I will pay you $money if you make "`translateto sv ..." redirect to bin/swedish.
20:13:30 <Vorpal> elliott, jaha, det få du väla tro då
20:13:40 <Vorpal> (writing on dialect is always fun)
20:13:45 <Vorpal> (it confuses google translate)
20:13:58 <Gregor> ... I just accidentally typed www.redchina.com while trying to type www.foxnews.com (<-- to test a crashbug)
20:14:08 <elliott> Gregor: Why would you ever type www.foxnews.com.
20:14:17 <ais523> elliott: to test a crashbug
20:14:29 <elliott> ais523: a crashbug on foxnews.com is known as a feature
20:14:46 <olsner> elliott: *crashfeature
20:14:51 <ais523> not really, no input should crash a browser, whether malicious /or/ incompetent
20:14:59 <Vorpal> Phantom_Hoover_, one left. That is all unless I have some I somewhere I forgot.
20:15:01 <Phantom_Hoover_> I'm afraid I mistook the one in the throne room for dirt and broke it.
20:15:11 <Vorpal> Phantom_Hoover_, you fix that. Somehow
20:15:31 <Vorpal> Phantom_Hoover_, get one from the admin. And even if it was dirt why would you break it in my place
20:15:39 <Vorpal> Phantom_Hoover_, should I go around breaking stuff at your place
20:15:58 <Vorpal> Phantom_Hoover_, I accept nothing but a lightstone as repayment
20:16:08 <ais523> I take it lightstones are rare?
20:16:10 <Vorpal> Phantom_Hoover_, you have to get one from the admin
20:16:27 <Vorpal> ais523, you can't get them in the normal way in multi-player
20:16:40 <Vorpal> Phantom_Hoover_, it is very. I use painterly and they are hugely different
20:16:54 <Vorpal> Phantom_Hoover_, so as soon as you see the server admin. Get him to give you one block and replace it
20:17:05 <elliott> Wow, an ISA kernel is much smaller!
20:17:08 <Vorpal> ais523, basically missing feature in multiplayer (no "nether" dimension)
20:17:23 <Vorpal> Phantom_Hoover_, custom painterly?
20:17:32 <Vorpal> Phantom_Hoover_, not the same custom as me
20:17:43 <Vorpal> Phantom_Hoover_, or maybe you use a monochrome monitor
20:17:55 <Vorpal> Phantom_Hoover_, anyway, why did you break it at all. Seriously
20:18:06 <Vorpal> Phantom_Hoover_, even if it HAD been dirt that would be rather rude
20:18:44 <Vorpal> Phantom_Hoover_, I'll go to your place and remove all dirt then?
20:18:55 <elliott> 1 dirt != 1,000,000,000,000 dirt
20:18:56 <Vorpal> elliott, stone is far more common
20:19:49 <Vorpal> elliott, also what about removing one dirt placed in a clearly highly ornamental place. Like in the middle of a stone wall with 4 torches around in a pattern? Except it was lightstone
20:20:25 <elliott> how does one list the types of NIC qemu will emulate?
20:20:29 <elliott> i don't have all the ones in the manpage
20:20:30 <Vorpal> Phantom_Hoover_, anyway I expect you fix this. The place is ruined without a lightstone there.
20:22:08 <Vorpal> Phantom_Hoover_, also if you insist on go about digging in other people's places I guess I could either do that at your place too: "but in my texture pack the glass looked like dirt", eh?
20:22:23 <Vorpal> Phantom_Hoover_, or I could just brick it all up and make a maze of tunnels
20:22:34 <Vorpal> (that don't actually connect anywhere
20:22:57 <Vorpal> Phantom_Hoover_, right, I realise you can't get a replacement right away. But as soon as you see the server I expect you to get one
20:23:38 <elliott> Just throwin' that out there.
20:24:14 <Vorpal> elliott, I'll decide if that happens
20:24:47 <Vorpal> probably going to request a gold block replacement instead or such then. Which will not look very good. Since it doesn't give off light it wouldn't really work
20:25:05 <Vorpal> elliott, besides he said before he was willing to give lightstone if you asked and had an use for it
20:25:15 <Vorpal> and making a sun-like symbol is definitely a use
20:25:38 <Vorpal> (I would have made a bigger one if i had more lightstone)
20:28:19 <Vorpal> Phantom_Hoover_, just logged in to check. 1) very different colour. 2) dirt isn't an acceptable replacement. A gold block would be until you can get a lightstone (you would get the gold block back)
20:33:16 -!- augur has quit (Remote host closed the connection).
20:33:40 -!- augur has joined.
20:34:36 <Gregor> `translateto cy No one actually speaks Welsh.
20:34:38 <HackEgo> Does neb yn siarad Cymraeg mewn gwirionedd.
20:34:53 <Gregor> Wow, that is some poor compression ratio.
20:35:17 <nooga> sometimes i get a Welsh newsletter
20:35:21 <Gregor> I believe that's Welsh for Welsh :P
20:35:36 <nooga> and it's so amusing that i woudn't dare to sign off
20:35:39 <Gregor> `translatefromto cy en Cymraeg
20:36:04 <nooga> welsh is from outer space
20:36:05 <quintopia> seems like would have been inferrable...
20:36:10 <nooga> like hungarian and finnish
20:36:42 <Gregor> `translateto fi The Moon people of Finland welcome our Human comrades.
20:36:43 <HackEgo> Kuu Suomen kansa tervetulleiksi Human toverit.
20:37:21 <fizzie> "Moon Finnish people welcome 'Human' comrades."
20:38:01 <fizzie> Where "welcome" is nounified, not as a verb.
20:38:12 <nooga> księżycowi ludzie z Polski witają ludzi - naszych przyjaciół
20:38:24 <nooga> that looks alien as well
20:38:30 <nooga> like: jak tak to mi sie nie chce
20:38:45 -!- augur has quit (Ping timeout: 260 seconds).
20:39:59 <elliott> /proc apparently costs 67K. Ouch...
20:41:38 -!- goneriku has joined.
20:41:51 <elliott> http://git-annex.branchable.com/ http://git-annex.branchable.com/walkthrough/
20:43:45 <nooga> elliott: get tinywm
20:43:52 <nooga> it's only 50 LOC in C
20:44:03 <elliott> nooga: I know what tinywm is.
20:44:54 <nooga> will you use xterm or something else?
20:45:44 <elliott> nooga: rxvt, or something else if that's too big. Maybe st.
20:45:49 <elliott> mmconfig WINS in size!! Now to try BIOS.
20:47:27 <elliott> Now to try any for gits and shiggles.
20:51:45 <elliott> nooga: find me a smallX tarball and i will
20:51:47 <elliott> tux.org seems to not have it any more
20:53:45 <pikhq> Gregor: It's Welsh for the Welsh language, yes.
20:54:16 <pikhq> And who did that Chinese in the topic?
20:54:20 <Gregor> pikhq: We already verified that.
20:54:23 <Gregor> pikhq: Google Translate X-P
20:54:31 <pikhq> I hates simplified!
20:54:35 <pikhq> It's so much harder to read!
20:54:51 <elliott> pikhq: Red China uses simplified :P
20:55:07 <pikhq> elliott: Your point? It's hard!
20:55:15 <Gregor> `translate zh-TW The glorious Republic of China shall reclaim this topic some day!
20:55:17 <HackEgo> zh-TW The glorious Republic of China shall reclaim this topic some day!
20:55:25 <Gregor> `translateto zh-TW The glorious Republic of China shall reclaim this topic some day!
20:55:27 <HackEgo> 光榮 Republic of中國應收回這個主題的一些日子！
20:55:56 <Gregor> You'd think that Google Translate would know what "Republic of China" is in traditional Chinese ...
20:56:17 <pikhq> Try "中華民國" instead of "中國".
20:56:45 <pikhq> The rest *looks* correct, but I can't say for sure, not being a speaker of a Chinese language.
20:56:56 <Gregor> `translateto zh-TW The glorious 中華民國 shall reclaim this topic some day!
20:57:00 <HackEgo> The glorious 中華民國 shall reclaim this topic some day!
20:57:09 <Gregor> `translatefromto en zh-TW The glorious 中華民國 shall reclaim this topic some day!
20:57:30 <pikhq> Oh, duh, it did omit the 的.
20:57:30 -!- Gregor has set topic: 该通道目前正在红色中国的控制权。请继续深奥的活动不受干扰。 | 光榮的中華民國應收回這個主題的一些日子！ | http://tunes.org/~nef/logs/esoteric/?C=M;O=D.
20:58:03 * pikhq cannot speak Chinese, but *has* communicated in a Chinese/Japanese pidgin before!
20:59:13 -!- xvedejas has joined.
20:59:16 <elliott> nooga: if you could find a mirror of tux.org ftp that's work
20:59:57 <xvedejas> I wish I could read the topic but my Mandarin isn't good enough yet :\
21:00:23 <xvedejas> also it seems to be a mix of traditional and simplified
21:00:49 <goneriku> apparently it's Google Translate shit
21:01:00 <elliott> Slereah: you brought these people, didn't you
21:01:06 <pikhq> goneriku: He spoke Mandarin, I speak Japanese, he didn't speak English well at all. So I wrote using sentences using my very very minimal knowledge of Chinese grammar and words in Japanese that are from Chinese.
21:01:11 <elliott> Slereah: YOU BROUGHT THESE PEOPLE HERE
21:01:17 <pikhq> goneriku: It certainly worked.
21:01:25 <xvedejas> elliott, I came on my own accord
21:01:34 <elliott> Well I blame Slereah for goneriku.
21:01:35 <goneriku> I don't know really know any Chinese or Japanese.
21:01:48 <elliott> All of these things are Slereah's fault.
21:01:49 <xvedejas> elliott, I'm an actual programmer
21:01:50 <Gregor> `translate 该通道目前正在红色中国的控制权。请继续深奥的活动不受干扰。
21:01:51 <HackEgo> The channel is currently being Red China's control. Please continue to esoteric activities without interference.
21:01:58 <Gregor> `translate 光榮的中華民國應收回這個主題的一些日子！
21:02:00 <HackEgo> Glorious Republic of China should recover some of this theme day!
21:02:07 <Gregor> Wow, bad translation X-D
21:02:12 <goneriku> Sler's pretty racist that's true
21:02:14 <elliott> xvedejas: Sorry, this channel is about esoterica and magick!
21:02:19 <goneriku> you should hear his rants about jews
21:02:30 <xvedejas> my favorite programming language is Smalltalk, some consider it sorta esoteric... :P
21:02:32 <goneriku> aw man I was about to say I'd fit in just fine here!
21:02:37 <goneriku> (I know, that's the lolarious part!)
21:02:38 <Gregor> `translateto iw Happy Hanukkah, gentiles!
21:02:38 <elliott> xvedejas: Protip: Ignore Sgeo.
21:02:50 <olsner> smalltalk is about as esoteric as javascript, isn't it?
21:03:00 <Gregor> olsner: Substantially more so.
21:03:09 -!- Gregor has set topic: 该通道目前正在红色中国的控制权。请继续深奥的活动不受干扰。 | 光榮的中華民國應收回這個主題的一些日子！ | חנוכה שמח, הגויים! | http://tunes.org/~nef/logs/esoteric/?C=M;O=D.
21:03:16 <elliott> olsner: Gregor does JS stuff for an academiliving.
21:03:21 <elliott> (If you can call that living)
21:03:25 -!- goneriku has changed nick to gon|away.
21:03:30 -!- augur has joined.
21:03:31 <elliott> So take his opinion with a grain of badly-scoped salt :P
21:03:37 <xvedejas> olsner, I don't really know anything about javascript
21:03:38 <Gregor> elliott: Dude, I browse porn sites for SCIENCE.
21:04:00 <HackEgo> 33) <ehird> pikhq: A lunar nation is totally pointless. <fungebob> ehird: consider low-gravity porn <ehird> fungebob: OK. Now I'm convinced. \ 77) <coppro> SF.net porn :/ <ehird> Oh yeah, baby, gimme that... bloated download page? \ 134) <DoctorDog> I am an inherently pornographic being. \ 156) <ais523> reading playboy for
21:04:02 <Gregor> `quote scientifically-justifiable
21:04:03 <olsner> elliott: I mean to imply that smalltalk is not the least bit esoteric
21:04:06 <HackEgo> 235) <Gregor> "* There is no scientifically-justifiable reason to exclude pornography, which is a vital part of the web ecosystem. However, bear in mind that we're tracing JavaScript, not MPEG and JPEG decoding." <Gregor> ^^^ This bullet-point is my crowning achievement as as a scientist.
21:04:09 <HackEgo> 235) <Gregor> "* There is no scientifically-justifiable reason to exclude pornography, which is a vital part of the web ecosystem. However, bear in mind that we're tracing JavaScript, not MPEG and JPEG decoding." <Gregor> ^^^ This bullet-point is my crowning achievement as as a scientist.
21:04:11 <elliott> Gregor: But you EXCLUDED it!
21:04:11 <pikhq> Gregor: ゴジラが来てる！死にたくないよ！
21:04:19 <Vorpal> <elliott> (If you can call that living) <-- still MS Research?
21:04:23 <Gregor> elliott: I did NOT exclude it. I EXPLICITLY did not exclude it.
21:04:31 <elliott> Vorpal: Yes. Gregor still works at MS Research.
21:04:47 <elliott> Gregor: Wrong! The answer is: Yes.
21:05:08 <Gregor> xvedejas: We're all confused :P
21:05:43 <Vorpal> xvedejas, only under confusing circumstances
21:05:58 <elliott> nooga: If you find a tux.org mirror I will be happy forever.
21:06:06 <xvedejas> I started writing some smalltalk tutorials on a friend's blog, if anyone is interested: http://hackeryblog.blogspot.com/2010/12/learn-programming-with-smalltalk-part-1.html
21:06:15 <elliott> xvedejas: Also, ignore Vorpal, we just haven't figured out a justification for kicking him out yet.
21:06:25 <elliott> But our crack team of crackheads is working on it as we speak.
21:06:30 <Gregor> `translate ゴジラが来てる！死にたくないよ！
21:06:31 <HackEgo> I'm Godzilla coming! I want to die!
21:06:45 <elliott> I'm -- Godzilla! -- coming! I want to die!
21:06:47 <Gregor> elliott: pikhq told me that :P
21:07:06 <pikhq> Gregor: "Godzilla is coming! I don't want to die!"
21:07:18 <elliott> pikhq: I prefer Google's version.
21:07:24 <elliott> Oh! Godzilla! Don't stop! I'm coming! I want to die!
21:07:27 <pikhq> It... Reversed the negative somehow.
21:07:31 <elliott> [[TINY Linux -- " 'Tis Independence 'N Yet "]]
21:07:43 <pikhq> That's fairly impressive, actually.
21:07:45 <elliott> YESSS I THINK I'VE FOUND SMALLX
21:08:28 -!- olsner has set topic: 该通道目前正在红色中国的控制权。请继续深奥的活动不受干扰。 | 光榮的中華民國應收回這個主題的一些日子！ | חנוכה שמח, הגויים! | sed อยู่เสมอดีกว่า Perl! | http://tunes.org/~nef/logs/esoteric/?C=M;O=D.
21:08:30 <elliott> pikhq: Hey, XFree86 had small servers.
21:08:31 <Gregor> pikhq: You're not sexually attracted to Godzilla.
21:08:34 <elliott> [ ]xvg16.zip05-Dec-1999 17:34 740K
21:08:41 <Gregor> pikhq: You're not sexually attracted to Godzilla at all.
21:08:43 <pikhq> Gregor: Only to Mothra.
21:08:44 <Gregor> pikhq: Whatever you say.
21:08:49 <elliott> Gregor: That's like saying Vorpal's not attracted to children!
21:08:53 <Gregor> elliott: Y'know, kdrive ain't big :P
21:09:07 <elliott> Gregor: How big is a 16-colour VGA KDrive? You know, approximately.
21:09:16 <Gregor> Roughly unsupported size?
21:09:25 <pikhq> elliott: Centibits.
21:09:25 <elliott> Gregor: Fine, what's the smallest KDrive?
21:09:37 <Vorpal> elliott, ... Only if they carry a magnet (I'm magnetic)
21:09:42 <elliott> Gregor: If it doesn't fit, LZMA-compressed, on a floppy minus 614K, I can't use it.
21:09:58 <elliott> Gregor: I could fit that VGA16 X server on. :P
21:10:02 <Gregor> LZMA-compressed ... the server itself would probably fit, not sure about all the client libraries.
21:10:21 <elliott> Gregor: I need libX11 and that's all. (Advantage of smallX: it has its own, tiny Xlib.)
21:10:32 <elliott> Gregor: Also, I need *some* kind of room left to put rxvt or whatever in :P
21:11:21 <Vorpal> elliott, couldn't you format the floppy as higher capacity than it really is?
21:11:32 <Vorpal> I seem to remember this being possible (but really really stupid)
21:11:35 <elliott> Vorpal: (1) Not reliably. (2) I refuse to.
21:12:02 <Vorpal> elliott, I wonder where you could find a 2.whatever MB floppy
21:12:09 <olsner> I formatted a floppy at a lower capacity once, unfortunately my special floppy-massage program didn't support the higher original capacity :/
21:12:20 <Gregor> elliott: Heh, so kdrive has enough requirements that that'd probably not work :P
21:12:27 <Vorpal> olsner, "floppy-massage"?
21:12:30 <olsner> IBM used to ship their operating systems on 2.88MB disks
21:12:32 <elliott> Gregor: Oh, and building KDrive has the distinct disadvantage that what the fuck modularised X's build system it is insane and horrible.
21:12:39 <olsner> Vorpal: *formatting then
21:12:45 <Gregor> <3 modular X build system
21:12:58 <xvedejas> what window managers do you all use?
21:13:16 <elliott> I use metacity because Kitten isn't done yet. :p
21:13:22 <elliott> <Gregor> <3 modular X build system
21:13:31 <olsner> it's autocrap, isn't it?
21:13:36 <Gregor> elliott: I've also used what came before.
21:13:44 <Gregor> olsner: Yes, it's autotools, which is why it's so good.
21:14:00 <Vorpal> Gregor, imake was one of the worst ones I ever seen
21:14:10 <elliott> Gregor: autotools is great if you never want to CONFIGURE anything.
21:14:28 <Gregor> Vorpal: Yeah, imake was ... bad. I've built X11R6 (as in, the real X11R6) on HP-UX. Autotooled X = godsend.
21:14:51 <Vorpal> elliott, autotools actually get the shit done. And works when you need to do something a bit unusual. What is your suggestion instead? cmake? scons? Plain makefile doesn
21:14:57 <Vorpal> doesn't* cut it for more complex stuff
21:15:02 <olsner> is imake the one that preprocesses makefiles using cpp?
21:15:19 <Vorpal> Gregor, yes for such a platform I imagine it would be
21:15:20 <elliott> Vorpal: No, seriously, autotools breaks down and fucks itself the minute you step out of a few rigidly-defined GNU-type systems.
21:15:28 <elliott> It works for you because you use GNU/Linux/Typical.
21:15:46 <Vorpal> elliott, not my experience at all. Only thing it doesn't work well on seems to be cygwin
21:16:01 <Gregor> elliott: You realize that for years I worked for Intel doing builds of F/OSS software on like six architectures?
21:16:13 <Gregor> elliott: With every new package, we PRAYED that it was autotools. Because autotools WORK.
21:16:15 <Vorpal> elliott, it works fine on *BSD, it works fine on some more esoteric platforms such as opensolaris. I haven't tried HP-UX though so can't answer for that
21:16:29 <Vorpal> it even worked fine on an old sunos box
21:16:57 <Vorpal> Phantom_Hoover_, did I claim it was?
21:17:03 <Gregor> Phantom_Hoover_: Whatever you want to believe :P
21:17:09 <elliott> "it works fine on some more esoteric platforms such as opensolaris"
21:17:14 <Vorpal> but yes it is in some aspects
21:17:27 <Vorpal> elliott, oh not in the "on topic sense"
21:18:06 <Gregor> Phantom_Hoover_: Esoteric == only known to a small, select group. An unsupported OS that barely managed to get off the ground and is now being squelched is the definition of esoteric.
21:18:55 <Phantom_Hoover_> It should also do something new, or at least defy common practices.
21:20:53 <Vorpal> Phantom_Hoover_, so why would you expect other people to use it?
21:21:39 <Phantom_Hoover_> Well, Gregor's definition makes things like the Seltzer & Friedberg Appreciation Society esoteric.
21:22:02 <Vorpal> Phantom_Hoover_, I never heard of that so I don't know
21:22:12 <Vorpal> it is however an indication that could be the case
21:22:56 <Gregor> elliott: I notice you haven't actually made any further comments about autotools, since autotools are a cross-platform build maintainer's dream, whereas cmake, scons, plain Makefiles and whatever else you're going to list have exactly the problems you mentioned as being problems of autotools.
21:23:21 <elliott> Gregor: I'm not going to list any because every one sucks, I've just been bitten by autotols constantly.
21:23:24 -!- xiaoy has joined.
21:23:49 <Vorpal> Gregor, why did you build them at intel btw?
21:23:59 <Gregor> Vorpal: ... because that was my job?
21:24:07 <Vorpal> Gregor, yes but for what? their linux distro?
21:24:32 <elliott> Vorpal: To maim the babies.
21:24:33 <Vorpal> Gregor, oh heh. They had their own linux distro internally or what?
21:24:52 <elliott> Gregor: Let's not bother trying to make Vorpal understand things!
21:25:09 <Gregor> No, they just had a distro of commonly-used F/OSS tools that were all at the same versions across some six arch/OS combinations.
21:25:30 <elliott> Gregor: please tell me windows was one of them
21:25:41 <Gregor> elliott: Good LORD no.
21:25:56 <elliott> If Linux doesn't have the solution, you have the wrong problem." -- stupid Linux forum signature
21:26:07 <elliott> medical equipment that needs hard real-time is WRONG!
21:26:18 <Vorpal> Gregor, hm, intel would use x86, x86-64, IA-64 and ARM right
21:26:35 <Vorpal> probably not ARM for this
21:26:38 <Gregor> elliott: <jsBot> Gregor: ... If eval() is the answer, chances are you're asking the wrong question
21:26:45 <Gregor> elliott: That makes me grind my teeth every time :P
21:26:58 <Gregor> Vorpal: And SPARC, PA-RISC :P
21:27:14 <Vorpal> Gregor, they use SPARC and PA-RISC? So weird.
21:27:28 <elliott> Gregor: This philosophy is exemplified in #python, where their motto is "You're here for help? Hahahaha. Tell us your entire application architecture and we will rip it apart without trying to understand it."
21:27:37 <elliott> Gregor: (the "wrong question" philosophy)
21:28:42 <Vorpal> elliott, it DOES happen that people ask the wrong question though. Like in #bash, trying to make echo do something weird when it is trivially straightforward with printf.
21:28:49 <elliott> Seriously, I will PAY someone to find a smallX tarball to me.
21:28:59 <elliott> Vorpal: Yes, but with Python it's not "do it this other way", it's "don't do that" at a very high level.
21:29:12 <elliott> I want to hook up X to Y. Don't do that, use Z instead. But I can't use Z. Well fuck you then.
21:29:31 <Vorpal> elliott, tried http://www.superant.com/smalllinux/tinyX01.html ?
21:29:37 <Phantom_Hoover_> What about when the thing they're trying is genuinely dumb?
21:29:46 <elliott> Vorpal: Go on, try clicking one of those tarball links.
21:29:55 <elliott> Phantom_Hoover_: if that was the case i wouldn't criticise as much.
21:30:03 -!- xiaoy has left (?).
21:30:11 <Vorpal> elliott, try the email there?
21:30:18 <elliott> Vorpal: I've found the HTTP interface to tux.org FTP and their smalllinux/smallX pages are 404'd.
21:30:21 <olsner> "genuinely dumb"? you seem to be implying something not-dumb is even possible in python :)
21:30:30 <elliott> Vorpal: I will if I can't find the .tgz.
21:30:34 <elliott> But I'd like to find it rather than bug him...
21:31:01 <Vorpal> elliott, well hopefully that will mean he will update the link to something that works
21:31:12 <elliott> Vorpal: Who knows if he has it?
21:31:26 <elliott> [["Here's some interesting news: KOffice, as a brand name, no longer exists. And with that change, I can't help but wonder if we are soon going to be looking at the end of the KDE name, as well.
21:31:26 <elliott> "It seems that the KDE community has decided to rename the KOffice project to the Calligra Suite project, as well as all the names of the individual applications within KOffice. So, KWord becomes Words, KSpread becomes Tables, and so forth. There's a table on the Calligra suite announcement that lists all of the changes.]]
21:31:35 <Vorpal> elliott, who knows. Who knows if he died from a freak volcano? (Okay that is less probable, but still!)
21:31:38 <elliott> Words, it's like Microsoft Word but there's more of 'em.
21:31:44 <Phantom_Hoover_> olsner, come now, using a particular language doesn't actually make a program stupid.
21:32:14 <Phantom_Hoover_> But programming languages can themselves be stupid ideas, and they can be designed and pitched in such a way that only idiots use them.
21:32:24 <elliott> To see the newest version of these wiki notes go to [[http://www.superant.com/sadrupal/]] ----> 404
21:33:01 <Phantom_Hoover_> elliott, I am sure that somewhere, at some time, someone will have written a sensible program in PHP.
21:33:08 <olsner> Phantom_Hoover_: I'm saying Python is (also) one of those languages
21:33:18 <Vorpal> elliott, "sadrupal"? I thought drupal was a CMS and not a wiki. So weird name
21:33:38 <pikhq> elliott: I don't think you realise how awful imake is.
21:33:52 <Vorpal> olsner, you have to admit python is better than php at least?
21:33:54 <pikhq> elliott: It's C preprocessor on Makefiles.
21:34:36 <elliott> pikhq: i never said it wasn't awful
21:34:39 <Vorpal> pikhq, and since when did not knowing about something stop elliott having extremely strong opinions about it?
21:34:41 <elliott> Vorpal: drupal is na everything
21:34:47 <elliott> I NEVER SAID ANYTHING ABOUT IMAKE GODDAMMIT
21:35:18 <Vorpal> elliott, autotools is an improvement. In fact autotools is the best option there is currently
21:35:33 <Phantom_Hoover_> I've never learnt it, and it's always just been implicitly true.
21:35:37 <elliott> [Removed chvt, deallocvt because they didn't work.]
21:35:46 <elliott> Phantom_Hoover_: Far too many reasons.
21:35:46 -!- xvedejas has quit (Quit: Leaving).
21:36:24 <nooga> norwegians say "precis" all the time
21:36:34 <Vorpal> nooga, that is a different "precis" I think
21:36:47 <pikhq> I had never seen "précis" before.
21:36:55 <Vorpal> pikhq, I had to google it
21:37:04 <Vorpal> pikhq, also I'm surprised. It is something you would use
21:37:08 <elliott> Phantom_Hoover_: Man who admits he hates programming invents programming language that's like Perl except everything is in one big print statement and to get out of print and have code you need to write <?php ... ?>. Then he puts all functions, with inconsistent argument order and naming, into one gigantic namespace, make ?x=y in the URL set $x = y, and makes ' and \ in strings automatically get backslash-escaped so that you can put them into a M
21:37:09 <elliott> ySQL DB directly (seriously).
21:37:10 <olsner> but "a précis" is basically a summary, right?
21:37:22 <elliott> Phantom_Hoover_: Follow natural evolution path towards amazingly crap.
21:37:33 <pikhq> Vorpal: Yes, yes I would.
21:37:51 <elliott> CC arch/x86/kernel/cpu/vmware.o
21:37:52 <elliott> CC arch/x86/kernel/cpu/hypervisor.o
21:38:02 <Vorpal> elliott, which kernel version
21:38:10 <Vorpal> elliott, if it recent I think you should be able to
21:38:13 <elliott> VMware options are *not* enabled.
21:38:31 <Vorpal> elliott, wait, there. That you probably need to enable embedded to remove cpuid strings
21:38:31 <elliott> And nor are any hypervisor options.
21:38:42 <elliott> Vorpal: I have enabled embedded.
21:38:56 <Vorpal> elliott, okay then check if they just contain #ifdef SOMETHINGTHATENABLESME
21:39:07 <Vorpal> elliott, basically empty object files is what I suspect
21:39:19 <elliott> EXPORT_SYMBOL(x86_hyper_vmware);
21:39:27 <elliott> I don't know if it's actually linked in, mind.
21:39:35 <elliott> -rw-r--r-- 1 elliott elliott 1720 Dec 7 20:57 arch/x86/kernel/cpu/vmware.o
21:39:39 <Vorpal> elliott, since I saw altivec.o compile on x86_64. I got so confused about that, that I opened the C file and found it just ifdefed out
21:39:46 <elliott> hypervisor is slightly smaller
21:39:54 <elliott> and isn't ifdeffed out either
21:40:05 <Vorpal> elliott, just export symbol? Nothing else?
21:40:10 <elliott> Vorpal: no, other things too
21:40:15 <elliott> static functions and the like
21:40:16 <elliott> CC arch/x86/boot/video-vga.o
21:40:16 <elliott> CC arch/x86/boot/video-vesa.o
21:40:21 <elliott> ugh, i hope you can disable the latter
21:40:43 <Vorpal> elliott, why aren't you using an older kernel (assuming it is the floppy one still)
21:41:10 <Vorpal> elliott, 2.4 didn't build?
21:41:18 <Vorpal> elliott, if so: what, it is recent enough
21:41:32 <Vorpal> elliott, make xconfig is the sanest option for 2.4 iirc
21:41:44 <Vorpal> elliott, (have fun with Tk)
21:42:00 <elliott> │ │ (0x1000000) Physical address where the kernel is loaded │ │
21:42:03 <elliott> │ │ (0x1000000) Alignment value to which kernel should be aligned │ │
21:42:07 <elliott> wonder if reducing them will shrink kernel :)
21:42:18 <Vorpal> elliott, no it won't I think
21:42:30 <Vorpal> elliott, also if you change that I *think* you need to change the bootloader too
21:43:35 <elliott> what where did my ethernet go
21:43:53 <Phantom_Hoover_> elliott, so wait, where are you going to get software for this?
21:44:09 <elliott> Phantom_Hoover_: um there's barely enough space for software
21:44:16 <elliott> libc is unlikely, they're big and i can statically-link stuff
21:44:21 <Vorpal> elliott, I'm sure 2.4 will be smaller. 2.0 even smaller
21:44:39 <elliott> all i really want at this point is... X, a tiny window manager, and a tiny terminal
21:44:46 <elliott> maybe one day i can squeeze links2 in.
21:44:51 <Vorpal> elliott, "I tried foo, didn't work, I gave up". Okay I won't hold that against you.
21:45:12 <elliott> ok let's see where my ethernet went
21:45:36 <Vorpal> elliott, because it doesn't change anything. I already knew that you weren't very persistent in any endeavour
21:45:36 <elliott> i wonder if i could disable all the non-vi modes of e3
21:45:49 <elliott> Vorpal: sure i am, just not in pointless endeavours.
21:45:54 <elliott> and this is definitely pointless.
21:46:11 <Vorpal> elliott, so procrastinating?
21:46:28 <elliott> and because pikhq wanted to do it
21:46:32 <elliott> so i'd better do it better first!
21:47:39 <Vorpal> nooga, oh btw in case you didn't know. sv:precis = en:exactly.
21:47:53 <Vorpal> (also en:exact, depends on context)
21:50:23 <Vorpal> elliott, you don't need finger?
21:50:43 <elliott> Vorpal: sure I don't, but why not
21:51:05 <Vorpal> elliott, well sure, if you have space left over
21:51:40 <elliott> Vorpal: it's 1622 bytes :P
21:52:08 <Vorpal> elliott, which makes the space you can spend on X 1622 bytes smaller
21:55:56 <Gregor> "And why should a [video game] character conform to and reinforce gender stereotypes? Birdo might have given some transgender kid hope."
21:56:50 <Vorpal> Gregor, I'm not sure how that connection works
21:56:50 <elliott> "One day I can be just like Birdo."
21:57:08 <Gregor> Vorpal: Birdo is the only semi-major transgender videogame character :P
21:57:21 <Vorpal> Gregor, oh, officially?
21:57:30 <Gregor> Vorpal: Not REALLY officially, but pseudo-officially X-P
21:57:45 <Gregor> In that Birdo was officially a gender-confused male, but in later things was officially female.
21:58:20 <Vorpal> Gregor, how did that happen to begin with
21:58:53 <olsner> when talking about post-op transsexuals in past-tense, do you need to keep track of the time the transition was made to get the right pronoun?
21:58:59 <Gregor> Vorpal: In the SMB2 manual, Birdo was labeled as a "boy who thinks he's a girl". In later things I guess they just decided that she's female. But the implications of that are clear, since both are canonical :P
21:59:12 <Gregor> olsner: Pronouns suck :P
21:59:28 <Vorpal> olsner, I'm not sure anybody has a good answer to that
22:00:04 -!- Mathnerd314 has joined.
22:02:16 <Vorpal> olsner, spivak for everyone?
22:02:20 <elliott> <olsner> when talking about post-op transsexuals in past-tense, do you need to keep track of the time the transition was made to get the right pronoun?
22:02:20 <Gregor> "Yoshi is supposedly a male, but lays eggs like a female. Birdo is supposedly a female, but was originally called a male. And now the two of them are a romantic couple? They were both sexually chaotic as individuals -- this new pairing just made your head hurt thinking about it."
22:02:27 <elliott> olsner: um you're always meant to use the post-transition pronoun...
22:02:31 <Gregor> Yoshi + Birdo = the single greatest video game couple in history :P
22:02:56 <elliott> olsner: you don't think about whether someone has a penis or not before choosing a pronoun :P
22:03:06 <olsner> elliott: right, pre/post-*op* is definitely wrong since the operation is independent of changing gender identity
22:03:10 -!- gon|away has changed nick to goneriku.
22:03:18 <elliott> olsner: "when talking about post-op transsexuals in past-tense" --you
22:03:34 <Vorpal> elliott, "to" != "about"
22:03:45 <elliott> Vorpal: um olsner knows what i mean, you don't
22:03:48 <olsner> elliott: bah, I'll just change that sentence to something else then!
22:03:49 <Phantom_Hoover_> elliott, what about before they had any conception of their gender identity differing from the norm?
22:04:18 <elliott> Phantom_Hoover_: it's very rare that people feel 100% comfortable with their identity and then decide to transition...
22:05:13 <olsner> spivak nouns until sexual maturity!
22:05:14 <elliott> Phantom_Hoover_: I can guarantee you that no transperson wants to be referred to as what they used to think their gender was no matter what tense :P
22:05:17 <elliott> (OK, I won't actually guarantee that.)
22:05:25 <elliott> so that's a silly thing to say
22:05:48 <Gregor> TRANSSEXUALS ARE IMMORAL PEOPLE. GOD CHOSE YOUR SEX FOR A REASON.
22:06:01 <Gregor> ... and shoots a tranny with it OH SEE HOW I MADE THAT AMBIGUOUS
22:06:02 <elliott> Gregor: JESUS WAS A GIRL IN A GIRL'S BODY
22:06:14 <olsner> elliott: make that what phantom hoover said instead
22:06:15 <elliott> Gregor: Are you sure it was a bow you shot that tranny with? :|
22:06:21 <Vorpal> Gregor, so what about those cases of ambiguous sex?
22:06:22 <Phantom_Hoover_> Gregor, but what's God's view of the LINGUISTICS of the matter?
22:06:34 <Gregor> elliott: Yes. I put a bow in another bow, and fired the first bow with the second.
22:06:35 <elliott> olsner: Phantom_Hoover_: How about SPIVAK PRONOUNS FOREVER because gendered pronouns are moronic
22:06:49 <Vorpal> Gregor, or the rare condition of having XY but being insensitive to testosterone
22:06:50 <elliott> Is that an [0010] in your pocket or are you just happy to see me?
22:06:58 <olsner> hmm, spivak nouns are just ... ugly
22:07:00 <elliott> Phantom_Hoover_: You think your solution is getting adopted either? :P
22:07:03 <elliott> olsner: ENGLISH IS UGLY :P
22:07:05 <Gregor> Vorpal: GOD WORKS IN MYSTERIOUS WAYS.
22:07:24 <Gregor> elliott: ENGLISH WORKS IN MYSTERIOUS WAYS.
22:07:26 -!- Sasha has joined.
22:07:27 <elliott> i agree with Gregor a magic man did it
22:07:42 <elliott> Gregor: i shouldn't complain about english. good enough for jesus, good enough for me
22:07:45 <Gregor> `addquote <Phantom_Hoover> For instance, Jesus' Y chromosome was clearly GOD'S.
22:07:53 <HackEgo> 237) <Phantom_Hoover> For instance, Jesus' Y chromosome was clearly GOD'S.
22:08:01 <elliott> this is also why I sleep on a cross every night
22:08:09 <elliott> Gregor: That is not even remotely permanently quotable :P
22:08:34 <Vorpal> elliott, I agree with Gregor on this one.
22:08:55 <Vorpal> it can be read in another way without the context
22:10:15 <Vorpal> elliott, any way to quickly get all permutations of a list in python? Quickly as in "easy to write" not "executes fast"
22:10:32 <Vorpal> I feel it should be trivial in a high level language
22:10:59 <Vorpal> like taking the list times itself or such.
22:11:38 <elliott> Vorpal: there is a way but i forget :D
22:11:54 <elliott> Phantom_Hoover_: Perfectly cromulent.
22:12:06 <elliott> Vorpal: itertools has permutations
22:12:25 <fizzie> >>> from itertools import permutations
22:12:26 <Phantom_Hoover_> I also think that, logically, the first person pronoun should be made universally "we".
22:12:28 <fizzie> >>> list(permutations([1,2,3]))
22:12:31 <fizzie> [(1, 2, 3), (1, 3, 2), (2, 1, 3), (2, 3, 1), (3, 1, 2), (3, 2, 1)]
22:12:52 <Phantom_Hoover_> Thus removing all singular/plural distinction in the English pronoun system.
22:12:53 <elliott> Phantom_Hoover_: We are amused.
22:13:04 <olsner> fizzie: "Too slow am I. The phone I blame."
22:13:15 <elliott> Phantom_Hoover_: Also, there IS singular/plural distinction.
22:14:08 <Vorpal> it gave me a list of tuples. Why...
22:14:10 <fizzie> "Is it true that the plural of 'y'all' is 'all y'all'?", I saw asked somewhere.
22:14:41 <fizzie> Well, they're known-length sequences.
22:14:49 <Vorpal> fizzie, ah right, makes sense
22:15:13 <Vorpal> fizzie, I'll just process the result with sed anyway XD
22:15:17 <fizzie> Since it's from itertools, it probably also gives you a generator, not a list.
22:15:33 <Vorpal> fizzie, yes indeed I passed it to list
22:15:43 <Phantom_Hoover_> [[# (50 - X)% of Yancy's genes are eternal genes with no origin, inherited from Fry, who originally inherited them from his mother.]]
22:16:32 <Gregor> Phantom_Hoover_: They're eternal in the sense that if you trace their history, it is an infinite loop.
22:16:54 <Phantom_Hoover_> They're from Fry's mother, whose genes are entirely normal.
22:17:33 <Gregor> Phantom_Hoover_: Isn't Fry's mother's side the side that Fry is on?
22:17:57 <elliott> What kernel module implements qemu's default network card?
22:18:14 <Gregor> Phantom_Hoover_: Oh, then that's a weird statement.
22:18:38 <Phantom_Hoover_> Gregor, but I can see a justification through the haze of my ape brain.
22:18:47 <Gregor> Phantom_Hoover_: Where did you get that quote from?
22:19:50 <Gregor> Phantom_Hoover_: Ohhh, I was thinking about Yancy JR
22:20:16 <Gregor> Phantom_Hoover_: Which made the whole thing kinda nonsense :P
22:20:32 <Phantom_Hoover_> OK, let's reduce the human genome to 4 chromosomes, in 2 homologous pairs.
22:20:49 <elliott> `addquote <Phantom_Hoover> OK, let's reduce the human genome to 4 chromosomes, in 2 homologous pairs.
22:20:49 <HackEgo> 238) <Phantom_Hoover> OK, let's reduce the human genome to 4 chromosomes, in 2 homologous pairs.
22:21:17 <Phantom_Hoover_> (This ignores a few things, most interestingly that Fry's Y chromosome is completely eternal.)
22:22:04 <Gregor> Phantom_Hoover_: Hm, yeah, those are definitely not eternal, they just happen to go backwards and forwards in time before being discarded.
22:25:43 <elliott> Okay, seriously, THERE IS NO WAY A TARBALL DISAPPEARED FROM THE INTERNET.
22:26:15 <elliott> Phantom_Hoover_: I just call that "Saturday".
22:26:49 <oerjan> Phantom_Hoover_: don't forget chromosomal crossover
22:28:27 <Gregor> Phantom_Hoover_: "Depending on what X is (see above), this is either disgustingly incestuous (for X ≅ 5) or entirely inoffensive and legal (for X = 50)." This statement is kinda silly too since all humans have some 99.99% of their genes in common with each other :P
22:28:58 <elliott> I JUST WANT A SMALLX TARBALL
22:29:10 <Phantom_Hoover_> Gregor, yes, but 45% is effectively identical to incest genetically.
22:29:15 <elliott> <Gregor> Phantom_Hoover_: "Depending on what X is (see above), this is either disgustingly incestuous (for X ≅ 5) or entirely inoffensive and legal (for X = 50)." This statement is kinda silly too since all humans have some 99.99% of their genes in common with each other :P
22:29:19 <elliott> There's a reason they call it wincest!
22:29:41 <elliott> [[In my on-going research for alternatives to the X.org full server, I ran across references to a “TinyX” which led me to SmallX, AKA Kdrive. Aside from one, very obsolete reference written for mere mortal users, the whole thing is buried in code-jockey talk.]]
22:29:44 <elliott> If you're not a code-jockey
22:29:48 <elliott> why are you trying to replace X.Org.
22:30:48 <elliott> Seriously. Anyone. SmallX. Please.
22:31:23 <Gregor> elliott: ftp://ftp.mayn.de/pub/really_old_stuff/unix/x11/tinyx/XVGA16.tar.gz
22:31:42 <elliott> Gregor: I love you. How did you find that?
22:31:52 <Gregor> elliott: I googled for the tarball name. It was REALLY difficult.
22:31:59 <elliott> Gregor: That's what I did, so fuck you :P
22:32:06 <Gregor> elliott: I googled it harder.
22:32:13 <elliott> ...OK, googling for XVGA16.tar.gz works WAY better than "smallX tinyX"
22:32:38 <elliott> Gregor: Oh for fuck's sake, it looks like it's just XFree86 TinyX.
22:32:55 <Phantom_Hoover_> OK, I think I have a vague grasp on the thing about the maternal eternal genes.
22:32:58 <elliott> Xtinylib.tar.gz BETTER have some real code.
22:33:13 <Gregor> elliott: I'm betting ... no.
22:33:42 <elliott> Gregor: lolso, teach me how to use the modular X build system.
22:33:56 <elliott> All I want is a single KDrive and Xlib with EVERYTHING DISABLED :P
22:34:02 <Phantom_Hoover_> Probabilistically, 25% of Yancy's genes are inherited from his wife.
22:35:36 <Gregor> elliott: http://freedesktop.org/wiki/Software/Xserver/InstallGuide <-- although slightly out-of-date, this does show the list of modules you have to build before building xserver kdrive. You'll need to provide the --enable-kdrive option nowadays, and ignore the shitload of stuff it builds AFTER the xserver, but otherwise there ya go.
22:35:53 <elliott> "ignore the shitload of stuff it builds AFTER the xserver"
22:36:08 <Gregor> elliott: That's a list of packages essentially, just don't do them.
22:36:26 <elliott> Gregor: Anyway, last I checked the only way to build was (1) 10000000 tarballs or (2) git repository.
22:36:27 <Gregor> elliott: (And note that it shows CVS since that's hyper-old instructions, and everything is in git now, and you should just get the packages anyway, just follow the package names :P )
22:36:43 <Gregor> elliott: For kdrive, I'd estimate 10 tarballs.
22:36:43 -!- MigoMipo_ has quit (Read error: Connection reset by peer).
22:36:51 <elliott> Gregor: over-modular more like
22:37:03 <Gregor> elliott: Yes, it is over-modular, but I'll still take it over imake any day :P
22:37:12 <Gregor> (The main thing is that splitting the proto and lib packages was a weird decision :P
22:38:16 <elliott> http://www.x.org/releases/X11R7.5/src/xserver/
22:38:20 <elliott> Is this the right thing? :p
22:39:15 <elliott> Gregor: And are you suuure it doesn't have 16-bit VGA?
22:44:26 <elliott> checking for XSERVERCFLAGS... configure: error: Package requirements (randrproto >= 1.2.99.3 renderproto >= 0.11 fixesproto >= 4.1 damageproto >= 1.1 xcmiscproto >= 1.2.0 xextproto >= 7.0.99.3 xproto >= 7.0.13 xtrans >= 1.2.2 bigreqsproto >= 1.1.0 fontsproto inputproto >= 1.9.99.902 kbproto >= 1.0.3 xkbfile xfont xau pixman-1 >= 0.15.20 openssl) were not met:
22:48:38 <Vorpal> hm sshfs has some delays that shouldn't be there I think
22:48:54 <Vorpal> sure you don't see a remote file created straight away (on the client side)
22:49:05 <Vorpal> but you should see the files you just wrote surely
22:51:06 <elliott> Gregor: Sooo, is there a simpler way than fetching those manually?
22:51:18 <Phantom_Hoover_> Vorpal, I know this will sound suspicious, but where's the throne room gone?
22:52:01 <Vorpal> Phantom_Hoover_, what throne room?
22:52:29 <elliott> Vorpal: You know, the throne room.
22:52:37 <elliott> It's the room with the throne.
22:52:43 <Vorpal> Phantom_Hoover_, I'm not playing any game. I'm printing a report
22:53:03 <elliott> Phantom_Hoover_: What do you mean where has it gone?
22:53:14 <Vorpal> elliott, oh. That. It's in Buckingham Palace.
22:53:39 -!- Phantom_Hoover_ has quit (Remote host closed the connection).
22:53:56 <Vorpal> well I could tell you. If you were here
22:57:40 <Vorpal> not sure I would. I seriously didn't have time for MC today, yet I had to clean up his mess
22:58:47 <Vorpal> I should just make the pit deep enough to kill on falling. About twice the current depth iirc. Nothing below it. Should be doable. A pain with the obsidian though.
23:00:51 <Vorpal> elliott, I realised while .xz is cool it isn't really usable yet if you want to make sure that everyone can open it.
23:01:02 <Vorpal> (note, windows compat is not an issue in this case)
23:01:07 <Vorpal> (if it was, it would mean zip)
23:01:20 <Vorpal> so tar.bz2 still has it's uses
23:08:14 <pikhq> Vorpal: If you want to make sure everyone can open it, you should use compress.
23:08:31 <pikhq> If you actually want something that *sane distros* can open, use xz.
23:08:40 <Vorpal> pikhq, no, because that would confuse people. Thus they couldn't open it.
23:08:58 <pikhq> Vorpal: Okay, y'know what?
23:09:01 <pikhq> Vorpal: Sharballs.
23:09:04 <pikhq> Vorpal: You may only use shar.
23:09:05 <Vorpal> pikhq, in this case I want a teacher at university to be able to open a lab assignment. I know he use linux. I don't know if he has xz.
23:09:16 <Vorpal> thus best bet: tar.bz2
23:09:32 <Vorpal> but the tar.gz is too large for the email system.
23:09:38 <pikhq> Vorpal: Is his distro more than 3 years old? If not, xz is fine. If it is, I suggest you root that box.
23:09:41 <Vorpal> (lots of data files included)
23:09:56 <Vorpal> ineiros, oh, interesting
23:10:11 <Vorpal> ineiros, any user guide to it?
23:11:02 <Vorpal> fizzie, is it so complex that a single page isn't enough?
23:11:34 <fizzie> http://wiki.hey0.net/index.php/Main_Page
23:11:44 <Vorpal> oh my... what a long commands page
23:12:01 <fizzie> It's got those, and lots of plugins.
23:12:14 <Vorpal> " * Command: /help --- Shows a list of commands (7 per page). "
23:12:39 <Vorpal> ineiros, any kits defined?
23:13:05 <ineiros> Pretty much default configuration, didn't yet check anything else.
23:13:24 <Vorpal> fizzie, also inconsistent. It has /kit without parameters to list kits, but /listwrap instead of /wrap without parameters
23:13:56 <Vorpal> I mean, shouldn't that be a kit :P
23:14:17 <Vorpal> /getpos looks useful. I had a lot of issues trying to work out offsets when building that trap
23:14:41 <elliott> xz sucks because of cpu usage
23:15:07 <Vorpal> elliott, sometimes, such as when preparing a release tarball, that is not an issue
23:15:10 <fizzie> I think /lighter is not a kit because t was ported from some other thing.
23:15:17 <elliott> Vorpal: but unpacking it still uses cpu.
23:15:37 <Vorpal> elliott, technically so does everything unless you have a separate accelerator chip for that algorithm
23:15:58 <elliott> Vorpal: more cpu than gzip though
23:16:07 <Vorpal> elliott, gzip is very cpu-friendly
23:16:17 <Vorpal> and bzip2 uses a lot more cpu for unpacking
23:16:21 <Vorpal> I'm pretty sure about that
23:16:31 <Vorpal> but no one uses that any more
23:16:45 <elliott> i've never cared that much about ultra small compression in common usage
23:16:53 <elliott> networks aren't so slow that wasting minutes compressing a tarball is worthwhile
23:17:22 <Vorpal> elliott, depends on what you plan to do with it. Download it over 3G? EDGE? Pack it on an install cd?
23:17:23 <pikhq> elliott: xz decompression is actually more CPU friendly than bzip2.
23:17:42 <pikhq> Unless you're using xz -9 or something.
23:18:06 <Vorpal> elliott, I used it for testing if it was worthwhile. It wasn't.
23:18:17 <pikhq> It gets you, like, bytes of benefit.
23:18:18 <Vorpal> but no I haven't used -9 beyond some basic testing of xz
23:18:45 <Vorpal> pikhq, note: I only did that as a test. And it saved like 1 MB on an ISO
23:18:53 <Vorpal> took 10-15 minutes to compress
23:18:58 <elliott> I wish lzma(1) had -10 --super-mega-brutish. For Flinix :)
23:19:13 <elliott> Vorpal: Pah! ESO distributed the ICFP ISO COMPLETELY UNCOMPRESSED in... 2009? 2010?
23:19:23 <elliott> And used up our entire bandwidth for the month I might add (100 gigs)
23:19:31 -!- Sasha2 has joined.
23:19:43 -!- Sasha has quit (Ping timeout: 264 seconds).
23:19:47 <elliott> Vorpal: international conference on functional programming. they run a popular programming contest each year
23:19:58 <elliott> they distributed an iso to provide a common testing environment for everyone
23:20:06 <Vorpal> elliott, maybe uncompressed was a bad idea then
23:20:08 <elliott> btw not only functional programmers participate
23:20:13 <Vorpal> if it used the entire bw for that month
23:20:19 <elliott> Vorpal: http://www.boundvariable.org/ you have probably seen this
23:20:22 <elliott> http://www.boundvariable.org/task.shtml
23:20:36 <Vorpal> elliott, ah yes that URL I remember
23:21:37 <elliott> linux is so bloated, what is it spending 480K on
23:22:05 <pikhq> elliott: Herring for penguins.
23:22:37 <elliott> pikhq: *complete copy of Wikipedia to comply with the GFDL
23:22:59 * elliott makes quote search case-insensitively
23:24:02 -!- Sasha2_ has joined.
23:24:08 <HackEgo> 153) * Phantom_Hoover wonders where the size of the compiled Linux kernel comes from. <cpressey> To comply with the GFDL, there's a copy of Wikipedia in there.
23:25:30 <pikhq> elliott: Also herring.
23:25:47 <elliott> pikhq: HOW HARD IS X86 (ANSWER NOT AT ALL LINUX IS BLOAT)
23:26:41 -!- Sasha2 has quit (Ping timeout: 265 seconds).
23:28:11 <elliott> pikhq: Did that kernel of yours ever do anything?
23:28:19 <elliott> OMFG I JUST HAD THE BEST IDEA EVER GREGOR GREGOR GREGOR
23:28:33 <Gregor> GREGOR GREGOR GREGOR Gregor IS the best idea ever.
23:28:42 <elliott> Gregor: You know how much you love SysV?
23:28:56 <Gregor> I lurve it with a spatula.
23:28:59 <elliott> Gregor: #ESOTERIC SHOULD PORT A IT OR A UNIX OF SIMILAR VINTAGE TO I386.
23:29:16 <elliott> How hard can it be! SysV was like the most portable OS ever! And old Research Unixes had stupidly simple kernels!
23:29:19 <Gregor> System V ran on 386 :P
23:29:29 <Gregor> System V ONLY ran on 386.
23:29:40 <Gregor> We should port it to x86_64 though :P
23:29:46 <Gregor> (Except the source to System V is of course not available)
23:29:46 <elliott> System V ran on the DEC VAX and PDP-11 machines. It also added support for inter-process communication using messages, semaphores, and shared memory.
23:29:48 <elliott> System V ran on the DEC VAX and PDP-11 machines. It also added support for inter-process communication using messages, semaphores, and shared memory.
23:30:12 <elliott> The primary platforms for SVR4 were Intel x86 and SPARC; the SPARC version, called Solaris 2 (or, internally, SunOS 5.x), was developed by Sun
23:30:25 <Gregor> Oh yeah, R4 and later >_>
23:30:31 <Gregor> SysVR<=3 sucked anyway :P
23:30:48 <elliott> Gregor: OK then. Unix Nth Edition (for low N. Like say fourth edition because previous versions were written in assembly.)
23:30:57 <elliott> 1st edition code is available at least :p
23:31:26 <elliott> Gregor: "WAIT I KNOW LET'S WRITE OUR OWN UNIX-COMPATIBLE FROM SCRATCH AND DISTRIBUTE IT FREELY OVER THE INTERNET."
23:31:47 <Gregor> elliott: And let's write it in JavaScript!
23:31:48 <Gregor> elliott: And call it JSMIPS!
23:32:52 <elliott> Gregor: Seriously though, research Unix on 386, how cool would that be (answer: mega cool)
23:34:05 <elliott> Gregor: AWESOME AWESOME AWESOME VERSION 1-7 UNIX IS UNDER FOUR-CLAUSE BSD
23:34:21 <elliott> Gregor: (VERSION 7 CAME OUT IN 1979 :P)
23:34:34 <elliott> [[V7 was the first readily portable version of Unix. As this was the era of minicomputers, with their many architectural variations, and also the dawning of the market for 16-bit microprocessors, many ports were completed within the first few years of its release. The first Sun workstations (then based on the Motorola 68010) ran a V7 port by UniSoft; the first version of Xenix for the Intel 8086 was derived from V7]]
23:34:43 <elliott> An x86 port is under active development by Nordier & Associates. The current version is 0.8a. The project has produced a bootable CD image with an installer script.[2]
23:34:49 <elliott> how is anyone as crazy as us
23:34:59 <elliott> Gregor: Dood http://www.nordier.com/v7x86/index.html :P
23:35:12 <Gregor> elliott: How 'bout Unix 32v?
23:35:24 <elliott> There are presently a few supplementary V7/x86-specific documents available:
23:35:25 <elliott> v7x86asref.pdf V7/x86 Assembler Reference Manual
23:35:25 <elliott> v7x86intro.pdf Using V7/x86: A Brief Introduction
23:35:43 <elliott> Gregor: But Unix 32v is Seventh Edition ported to VAX :P
23:35:47 <elliott> Gregor: So it's the same thing.
23:36:09 <Gregor> elliott: I was already typing that before you mentioned v7x86, I was just suggesting it because at least it's already 32-bit :P
23:36:20 <elliott> Gregor: HOWEVER I bet 4th edition would be easy.
23:36:22 <elliott> 4th Edition Nov. 1973 First Unix written in C. It also introduced groups. Number of installations was listed as "above 20". The manual was formatted with troff for the first time.
23:36:29 <elliott> V5 is listed as "Introduced the sticky bit", what a release :P
23:36:44 <elliott> 6th is when people started PORTING it and we want to be the fisrt people to do anything.
23:36:51 <elliott> [[xv6 is a modern reimplementation of Sixth Edition Unix in ANSI C for multiprocessor x86 systems.]]
23:36:55 <elliott> [[It is used for pedagogical purposes in MIT's Operating Systems Engineering (6.828) course.]]
23:37:07 <elliott> Gregor: 4th Edition LET'S DO THIS OKAY
23:37:40 <Gregor> elliott: Instead let's set up a publicly-available free shell service on v7x86
23:37:40 <elliott> [[The fourth edition of Unix was the first version to have a kernel written in a high level language, C, along with some of the commands. A full and complete copy of Fourth Edition no longer exists.]]
23:38:33 -!- ais523 has quit (Remote host closed the connection).
23:38:46 <elliott> "V7 Unix introduced the first version of the modern "Standard I/O" library stdio as part of the system library."
23:39:01 <elliott> [[For many years, MIT had no operating systems course. In the fall of 2002, Frans Kaashoek, Josh Cates, and Emil Sit created a new, experimental course (6.097) to teach operating systems engineering. In the course lectures, the class worked through Sixth Edition Unix (aka V6) using John Lions's famous commentary.]]
23:39:16 <elliott> MIT: Making 2002 feel just like 1975.
23:39:25 <Gregor> <elliott> "V7 Unix introduced the first version of the modern "Standard I/O" library stdio as part of the system library." <-- kinda important X-P
23:39:39 <elliott> Gregor: Whoa, Russ Cox was involved in xv6 X-D
23:39:46 <elliott> Gregor: Also, hey, I swear off stdio and you should too. :p
23:40:22 <elliott> Gregor: But it's terrible!
23:40:51 <elliott> olsner: hmm, you know my microkernel design?
23:42:42 <elliott> I think I just managed to invent an attokernel :P
23:44:15 <elliott> Gregor: OK, I've figured it out.
23:44:27 <elliott> Gregor: We should port the System V kernel as a HURD server.
23:45:26 <Gregor> System V kernel source isn't available.
23:45:39 <Gregor> Except in some enormously derivative form in OpenSolaris.
23:46:34 <elliott> [[ "The network software for Unix was developed on a PDP-11/50, with memory management, two RK05 disk packs, two nine track magtape drives, four dectape drives, 32k words of core, and three terminals. Presently this has been expanded to encompass a DH11 terminal multiplexor, an RP03 moving head disk, a twin platter RF11 fixed head disk, floating point, and 48k of core. User files are stored on the RP03. the RF11 is used as a swap disk and for
23:46:34 <elliott> temporary file storage; one RK05 platter contains the system files, and the second contains login and accounting information. In the near future, the system will be expanded to 128k words of core memory with 10 dial in and 10 hard wired terminal lines"
23:46:34 <elliott> "The base operating system occupies 24.5k words of memory. this system includes a large number of device drivers, and enjoys a generous amount of space for I/O buffers and system tables. A minimal system would require 40k words of hardware memory. It should be noted that Unix also requires the memory management"]]
23:47:03 <elliott> http://tools.ietf.org/html/rfc681
23:47:08 <elliott> FILEDES = OPEN( "/DEV/NET/HARV",2 );
23:47:08 <elliott> PRINTF(" HARVARD IS DEAD");
23:47:08 <elliott> WHILE( (NBYTES=READ(FILEDES,BUF,80)) > 0 )
23:47:09 <elliott> WRITE( 0,BUF,NBYTES ); 4j1
23:47:36 <elliott> WHY IS RFC 681 ALL IN CPITALS
23:47:40 <Gregor> elliott: I can't make v7x86 boot after installing :(
23:48:11 <elliott> Gregor: Probably it requires a bootloader?
23:48:15 <elliott> http://www.nordier.com/v7x86/install.html
23:48:19 <elliott> Note that the install program expects to install to an existing V7/x86 hard disk partition. This can be created with any fdisk compatible utility that allows the partition type to be specified. The V7/x86 partition type is 0x72 (114 decimal).
23:48:43 <Gregor> elliott: Yeah, I did that. I wonder how you think I installed without that.
23:49:03 <elliott> Gregor: Try http://pdos.csail.mit.edu/6.828/xv6/ :P
23:49:23 <elliott> "Understanding exec (exec.c) is left as an exercise."
23:49:37 <Gregor> "Xv6 does boot on real hardware" LAME
23:49:50 <elliott> Gregor: JUST TRY IT MAN IT'S V6 UNIX
23:50:00 <elliott> # Start the first CPU: switch to 32-bit protected mode, jump into C.
23:50:01 <elliott> # The BIOS loads this code from the first sector of the hard disk into
23:50:01 <elliott> # memory at physical address 0x7c00 and starts executing in real mode
23:50:13 <elliott> An introductory OS course?
23:53:50 <elliott> Gregor: xv6 is kinda boring
23:55:47 <Gregor> elliott: http://www.nordier.com/software/c3s.html <-- lawl
23:56:28 <pikhq> elliott: s/is/was/, I'm sure.
23:56:32 <elliott> Gregor: Don't you mean "awesome"?
23:56:35 <Gregor> At present, binaries are available for FreeBSD. Source code should be available at a future date. D'AWW BOO
23:56:42 <elliott> pikhq: xv6 is V6 unix reimpl. for modern hardware
23:57:04 <pikhq> elliott: They no longer use SICP for the intro CS course; I'd imagine the rest of it has been dumbed down similarly.
23:57:11 <elliott> pikhq: No, vx6 is a new thing.
23:57:28 <elliott> And no, all that was changed is 6.001 or whatever it's called :P
23:57:48 <elliott> Gregor: What I would like to see is a ring 0-only OS implementing my syscall inlining idea but it pretty much requires a HLL-based OS...
23:58:01 <elliott> This is a port to V7/x86 of an old BCPL compiler from Martin Richards and the Tripos Research Group at Cambridge University. The port includes both an interpreter and an x86 native code generator. The original software dates from around 1979.
23:58:01 <elliott> The Amsterdam Compiler Kit (ACK) was originally developed at the Vrije Universiteit, Amsterdam, and had its heyday during the 1980s. Several years ago, it was released under a Berkeley-style license. This port adds support for Solaris x86 and for V7/x86.
23:58:34 <pikhq> elliott: SMP in an intro OS course? Fuck MIT is hardcore.
23:58:55 <elliott> I'm not sure it is though.
23:59:11 <elliott> <elliott> Gregor: What I would like to see is a ring 0-only OS implementing my syscall inlining idea but it pretty much requires a HLL-based OS...
23:59:15 <elliott> anyone have any idea how to do this
23:59:18 <elliott> without writing a new compiler?
23:59:32 <elliott> Or, well, without writing a *complicated* new compiler.
23:59:44 <elliott> (for safety, that is; you can easily do this if you let processes do whatever they want)