Conversation with #inferno at Mon Sep 28 21:09:29 2009 on powerman-asdf@irc.freenode.net (irc)
(21:59:44) rapidfx left the room (quit: "Leaving.").
(22:03:07) dagle: Mysql makes me wanting to kill my self.
(22:03:49) dagle: Error: entry exists. Show entry: entry does not exist....
(22:13:21) mennis: There is no self.
(22:24:33) uriel: dagle: mysql is responsible for a huge spike in the world wide suicide rates
(01:33:25) mjl0 left the room (quit: Remote closed the connection).
(01:54:35) mennis left the room (quit: ).
(05:11:57) eno__ left the room (quit: Read error: 131 (Connection reset by peer)).
(05:17:15) eno [n=eno@nslu2-linux/eno] entered the room.
(05:29:33) mennis [n=mennis@adsl-065-012-170-146.sip.asm.bellsouth.net] entered the room.
(05:30:42) mennis left the room (quit: Client Quit).
(06:28:41) rapidfx [n=host666@vl-cen-ce1.avtlg.ru] entered the room.
(07:45:10) uriel: anyone checked this thing out: http://www.engadget.com/2009/09/28/38-zipit-wireless-messenger-receives-linux-injection-becomes/
(07:45:14) uriel: this zipit thing seems perfect for inferno
(07:45:57) uriel: for 40$ it is less than half the price of a sheevaplug, and has a screen and kb!
(07:59:58) robot12 [n=robot12@inferno.kgts.ru] entered the room.
(08:00:25) robot12: hi all
(08:06:40) uriel: hi robot12 
(10:55:09) robot12: Microsoft reinvents inferno :) - http://research.microsoft.com/pubs/81154/helios.pdf
(10:55:49) robot12: two concurent projects - Singularity and Helios :) LOL
(10:58:01) mjl-: uriel: quite impressive they got that functionality for that price
(11:05:21) uriel: yea, seems quite awesome
(11:06:18) andguent: there is xml in that pape
(11:06:28) andguent: i don't need to read it to not want *meh*
(11:08:00) uriel: andguent: it is hardware
(11:08:12) uriel: or you mean the helios thing?
(11:08:17) uriel: (about which i know nothing)
(11:15:18) megaboz [i=none@201.80.224.28] entered the room.
(12:07:30) mahtawkmeister: robot12 - the have been going a while, Windows has a few plan9 like features
(12:08:08) robot12: hmm what are they ?
(12:08:17) robot12: I'm using Lunix :) 
(12:46:02) mahtawkmeister: synthetic file systems for a start, introduced witeh nt3.5
(12:46:09) mahtawkmeister: with
(12:55:54) robot12: hmm
(12:56:36) robot12: and lunix have perprocess namespace :) 
(12:56:44) robot12: 9p in kernel :) 
(12:56:59) Fish [n=Fish@AVelizy-152-1-41-12.w82-120.abo.wanadoo.fr] entered the room.
(13:09:05) j123m left the room (quit: Read error: 145 (Connection timed out)).
(13:22:32) j123m [n=j1m@94.190.65.228] entered the room.
(13:41:00) paigeadele left the room (quit: Read error: 145 (Connection timed out)).
(14:05:35) robot12 left the room (quit: "Ухожу я от вас (xchat 2.4.5 или старше)").
(14:41:57) soul9 left the room (quit: verne.freenode.net irc.freenode.net).
(14:41:57) yiyus left the room (quit: verne.freenode.net irc.freenode.net).
(14:50:31) andguent: what is the easiest way in limbo to get a sha1 string from a string?
(14:53:40) mjl-: kr->sha1(buf := array of byte s, len s, digest := array[kr->SHA1dlen] of byte, nil);
(14:53:56) mjl-: oh
(14:54:00) mjl-: s/len s/len buf/
(14:54:52) andguent: ah yes
(14:55:03) andguent: figured out that by myself now. sorry for the noise
(14:55:25) andguent: although the defining the digest in the call is neat. didn't know limbo allowed such syntax
(14:58:05) mjl-: yeah, allows for concise code.  the "len buf" immediately takes the length of the array too
(14:59:43) yiyus [i=12427124@je.je.je] entered the room.
(15:15:18) soul9 [n=none@unaffiliated/johnnybuoy] entered the room.
(16:26:36) idr0 [n=idr@p57A274E1.dip.t-dialin.net] entered the room.
(16:27:41) andguent: my sh script returns: './raise' file does not exist
(16:27:45) andguent: any ideas how to diagnose?
(16:28:01) andguent: it supposed to raise... 
(16:29:22) andguent: also: sh: if: './if' file does not exist
(16:29:25) andguent: strange
(16:29:33) mjl-: load std
(16:29:50) andguent: the thing is...the exact same script works here on my mac
(16:29:54) andguent: but on my server it doesn't
(16:30:01) andguent: same same... i didn't change a thing
(16:32:53) andguent: but thanks it fixes the issue..
(16:32:56) andguent: not sure why though
(16:36:36) andguent: NOTICE: a@qcx.be is now completely inferno served :)
(16:38:37) mjl-: cool
(16:38:48) mjl-: andguent: perhaps you have an autoload directive somewhere?
(16:38:54) mjl-: it's an env var i think
(16:39:07) mjl-: that automatically loads sh modules
(16:39:12) andguent: might be possible afterall i started it on acme/win all the time
(16:39:23) andguent: now i do completely out of the cold
(16:40:23) andguent: mjl-: got your mail :)
(16:40:37) mjl-: good
(16:40:40) mjl-: it's fast too then :D
(16:41:13) andguent: well when i tested it from my usual account. it took like 1 sec or so...
(16:41:28) andguent: but i guess that is due to the latency and stuff
(16:41:51) andguent: the actual smtpd might get in trouble with really big mime parts. but apart from that it should be nearly a noop
(16:53:41) andguent: mjl-: thanks for all your help. it's been very much appreciated :)
(16:54:03) andguent: mahtawkmeiste: thanks for your smtpd.awk...it's been a little inspiration :)
(17:06:09) mjl-: glad to be of help
(17:07:04) andguent: :) ... now for mimefs...to be honest. i am little worried about this one. 
(17:08:39) mjl-: do you want to parse as data comes in?
(17:08:51) mjl-: or read it all, then parse it, then give the styx interface?
(17:09:22) andguent: the second one
(17:09:31) andguent: smtpd doesn't really care about mime
(17:09:40) andguent: from the smtpd point it doesn't matter what's in there
(17:09:48) andguent: the dropper script / mail reader cares
(17:09:48) mjl-: ok, then what's worrying you most?
(17:09:53) mjl-: the mime part or the fs part?
(17:10:01) andguent: a little bit of both i guess
(17:10:12) andguent: more of the fs though
(17:10:29) mjl-: both parts separately seem doable, so combined they should be okay too
(17:10:30) andguent: i can tweak the mime part incrementally. but as of yet i haven't written a file server from scratch
(17:10:48) mjl-: the fs part could be relatively simple btw, if you use the nametree.  if the fs is read-only, that should work
(17:11:00) andguent: nametree?
(17:11:20) mjl-: http://www.vitanuova.com/inferno/man/2/styxservers-nametree.html
(17:11:24) andguent: read only of course...hell at no way whatsoever i would want to write a writable mimefs
(17:11:30) mjl-: the example is a simple but complete limbo program that serves a bunch of files
(17:11:36) mjl-: hehe, smart
(17:11:36) andguent: pain and mutilation would be the ultimate surrender
(17:12:04) mjl-: you can start from the example, and modify the tree.create() calls to put the mime dirs & files in place
(17:12:29) andguent: oh cool. thanks for the pointer. that will save me from some trouble
(17:12:31) mjl-: then the Tmsg.Read can be handled by returning the appropriate part of the buffer from the right mime part
(17:12:49) andguent: but this way or the other i'll have to study 9P in detail ....one day...
(17:13:03) mjl-: normally when i want to write an fs for x, i first write a tool xdump.b, that parses the data and dumps it in plain text form
(17:13:12) mjl-: and then i use that as a starting point to write an fs
(17:13:46) andguent: yeah sure. that is reasonable
(17:14:06) andguent: and the mime part isn't _that_ critical...i can always incrementally update my program
(17:15:08) gualteri [n=speiro@crespins.disca.upv.es] entered the room.
(17:31:32) mjl-: hmm, acme is modifying non-utf-8 files completely
(17:31:35) mjl-: that's not very nice
(17:32:16) mjl-: hmm, brutus is truncating the file, that's even worse
(18:01:40) mennis [n=mennis@adsl-068-016-104-079.sip.asm.bellsouth.net] entered the room.
(18:06:34) Fish left the room (quit: Remote closed the connection).
(18:38:34) paigeadele [i=pthomps@75-146-55-252-Washington.hfc.comcastbusiness.net] entered the room.
(18:38:52) paigeadele left the room (quit: Client Quit).
(18:39:34) paigeadele [i=pthomps@75-146-55-252-Washington.hfc.comcastbusiness.net] entered the room.
(18:43:08) gualteri left the room.
(19:13:32) idr0 left the room (quit: Read error: 104 (Connection reset by peer)).
(20:53:40) rapidfx left the room (quit: "Leaving.").
(21:42:01) paigeadele left the room (quit: Read error: 104 (Connection reset by peer)).
(21:42:23) paigeade1e [i=pthomps@75-146-55-252-Washington.hfc.comcastbusiness.net] entered the room.
(21:46:17) paigeade1e left the room (quit: Read error: 104 (Connection reset by peer)).
(21:49:38) npe [n=npe@94-224-251-223.access.telenet.be] entered the room.
(22:02:30) paigeadele [i=erratic@75-146-55-252-Washington.hfc.comcastbusiness.net] entered the room.
(22:05:05) paigeadele left the room (quit: Client Quit).
(22:05:56) paigeadele [i=pthomps@75-146-55-252-Washington.hfc.comcastbusiness.net] entered the room.
(23:13:12) megaboz left the room (quit: Nick collision from services.).
(23:13:35) megaboz [i=none@201.80.224.28] entered the room.
(00:36:23) mennis left the room (quit: Client Quit).
(01:35:44) mjl-: anyone familiar with sockets & so_linger
(01:35:50) mjl-: and the tcp hangup ctl message
(01:42:12) npe left the room (quit: ).
(02:55:32) sswam [n=sswam@CPE-121-219-182-226.vic.bigpond.net.au] entered the room.
(02:57:35) sswam: hi, I want to build inferno on amd64, seems to need some cflags to pretend its 386, like -m32  but that alone doesn't work, gives errors like /usr/bin/ld: i386:x86-64 architecture of input file `/usr/inferno/Linux/386/lib/libbio.a(bgetc.o)' is incompatible with i386 output
(02:57:42) sswam: after a while.  anyone?
(03:00:01) sswam: I guess I need LDFLAGS too, I'll go look up generic x86 on amd64 instructions somewhere...
(03:00:24) uriel: sswam: did you look to the issue tracker, i think there are some issue(s) about this
(03:00:27) uriel: with ways around
(03:00:31) uriel: but i could be wrong
(03:02:26) sswam: ok
(03:04:15) sswam: maybe I should build it on 386 then copy it across
(03:10:30) uriel: it is a bit annoying that there are no decent pre-built packages
(03:10:49) uriel: but you might want to use just a dis binary
(03:10:50) uriel: (which should be al you need)
(03:11:04) uriel: which you can get form the inferno-bin project
(03:11:31) uriel: http://code.google.com/p/inferno-bin/
(03:13:21) sswam: I don't mind building it, but there should be a note about amd64 in the INSTALL file
(03:13:34) sswam: I guess it's pretty distro specific
(03:14:26) sswam: I think it's working now, I needed GNUTARGET=default CFLAGS=-m32 LDFLAGS="-L/usr/lib32"
(03:14:35) sswam: maybe not the LDFLAGS
(03:16:06) sswam: still getting lots of "cast pointer to int of different size" warnings
(03:16:49) uriel: you might want to post to inferno-list about it
(03:29:40) sswam: ok
(03:36:20) sswam: according to #gcc it should not call ld directly, but use gcc (with the -m32) to do linking too, that's why it needs the extra options
(03:37:59) uriel: fun!
(04:02:20) sswam: it seems to work, anyway
(04:02:57) sswam: now I've got to get plan9 running in virtualbox :)
(04:05:47) sswam: I don't know what web server they use to serve the inferno tgz but I couldn't resume a download on it!  even my homebrew hacky 900 line web server does that
(04:06:17) uriel: tried 9vx?
(04:06:36) uriel: as for the web server, god only knows
(04:06:50) sswam: now I know too: Server: Scylla/1.0
(04:07:00) uriel: but *never* get your inferno stuff from VN, it is a waste of time, pull directly from google.code's mercurial repo
(04:07:08) sswam: oh.  whoops
(04:08:44) sswam: damn these rcs systems that don't let you say "hey I've got all these files here already, try to be efficient eh?"
(04:09:53) uriel: well, you should have used that from the start
(04:11:07) sswam: git doesn't even do that properly - only will refer to an existing git repo not arbitrary files; and it's so easy to do it
(04:12:37) sswam: rm -rf inferno ... try again :)  lol
(04:13:25) sswam: are there some fonts I should snarf from the vn version?
(04:13:56) sswam: 9vx is going to be great, thanks a lot, will beat virtualbox for sure
(04:14:31) sswam: should I get that from hg too?
(04:26:13) uriel: sswam: yes, you shoudld snarf all the fonts
(04:26:31) uriel: and yea, get 9vx from hg, 
(04:31:31) sqweek: http://sqweek.dnsdojo.org/inferno/fonts.tar.gz >_>
(04:33:52) uriel: sqweek: well done!
(04:34:37) uriel: I think I had some instructions on how to fetch the damned fronts somewhre in http://ninetimes.cat-v.org/tips/ but I even forgot about it
(04:34:40) uriel: the whole thing is a pain
(04:37:54) uriel: sqweek: are you cool with this update: http://ninetimes.cat-v.org/tips/2009/02/13/1/ ?
(04:38:16) uriel: (if not it is ok, I will host a copy of the fonts somewhere in cat-v.org)
(04:38:23) sqweek: sure
(04:39:16) sqweek: it's actually just the licensed fonts in there
(04:39:43) sqweek: ie, not the ones you already have in hg
(04:40:10) uriel: yea, that is perfect
(06:24:44) sswam: I already had the vn inferno on my local box, internet is slow, so I downloaded the other to my server, wrote a little program to link each file to a hash-named file in a hash directory, did same at home, and rsync the files, transferred only 6.5Mb instead of 48Mb for the archive, so that's a win.  Probably straight rsync would have worked nearly as well in this case.
(06:25:29) uriel: fun!
(06:26:06) sswam: I have another technique I will try which breaks the files into small chunks first, where the boundaries are chosen based on rolling content checksums (a bit like gzip --rsyncable).  So if a file is mostly the same as another, they will have a lot of identical chunk files.  That can reduce the transfer more too.
(06:26:23) sswam: yeah, fun :)  now I have to wrap it up in a script so it doesn't take all day to do it by hand
(06:26:43) sswam: seems to work anyway
(06:27:25) sswam: I can use that to "clone" repositories for whatever rcs too, because just copying a repo is effectively cloning it anyway I think, more or less maybe some slight config change like where "upstream" is.
(06:29:01) uriel: nice
(06:30:57) sswam: I have an evil plan (and most of program) for syncing binary packages like .debs too, which involves unpacking all compressed files recursively (not lossy ones though!), and then doing something like what I just said on it.
(06:31:24) sswam: might save me a bit of bandwidth, upgrading debian all the time can get annoying
(06:41:54) sqweek: sounds like some sort of venti-sync
(06:46:48) sswam left the room (quit: Remote closed the connection).
(06:46:58) sswam [n=sswam@CPE-121-219-182-226.vic.bigpond.net.au] entered the room.
(06:51:06) uriel: actually, wasn't cnilsen that was working on some kind of bynary sync thing? or was it kuroneko?
(06:51:11) ***uriel can't quite remember
(06:51:22) sswam left the room (quit: Remote closed the connection).
(07:16:26) sswam [n=sswam@CPE-121-219-182-226.vic.bigpond.net.au] entered the room.
(07:25:18) robot12 [n=robot12@inferno.kgts.ru] entered the room.
(07:25:26) robot12: hi all :) 
(07:30:21) jas left the room (quit: ).
(07:42:45) jas [n=jas@69.215.39.41] entered the room.
(08:03:34) j123m left the room (quit: Read error: 145 (Connection timed out)).
(08:36:48) j123m [n=j1m@unas-revda.interra.ru] entered the room.
(09:02:01) rapidfx [n=host666@vl-cen-ce1.avtlg.ru] entered the room.
(09:35:35) mjl-: good day all
(09:36:26) mjl-: sswam: for installing inferno, i always recommend vita nuova's hg checkout snapshot with fonts.  easiest way to get inferno with fonts in a way that is easy to update.
(09:37:07) mjl-: http://www.vitanuova.com/inferno/downloads.html, then the link inferno-20090730.tgz
(09:37:23) mjl-: extracting that, followed by hg pull && hg update gives you the latest & greatest
(09:37:30) mjl-: about installing on amd64
(09:37:55) mjl-: on at least linux you need to do a 'mk nuke' after the 'makemk.sh' (before the 'mk install')
(09:38:38) mjl-: because makemk.sh, which creates mk (needed for the build system), compiles a few libraries without -m32
(09:39:39) mjl-: so when the build system starts (which does use -m32), it tries to link most -m32 code with some of those old non-m32 libs
(09:39:45) mjl-: mk nuke gets rid of those old libraries
(09:40:08) mjl-: but i've only tested that in linux/amd64, perhaps gcc's under bsd's claim different things
(10:28:56) sswam left the room (quit: "leaving").
(10:29:06) sswam [n=sswam@CPE-121-219-182-226.vic.bigpond.net.au] entered the room.
(10:51:42) npe [n=npe@195.207.5.2] entered the room.
(11:44:25) sswam left the room (quit: "leaving").
(12:01:47) j123m left the room (quit: Read error: 110 (Connection timed out)).
(12:21:10) j123m [n=j1m@unas-revda.interra.ru] entered the room.
(12:55:53) sswam [n=sswam@CPE-121-219-182-226.vic.bigpond.net.au] entered the room.
(12:57:59) sswam: trying to build inferno from hg / google code on Linux/386 - it says ../port/devmnt.c:167: error: invalid type argument of '->' (have 'ulong')  and ../port/devmnt.c:1174: error: 'Chan' has no member named 'devno'  ... any ideas?
(12:58:11) sswam: I was able to build the vn release, but this one won't
(12:58:37) mahtawkmeister: on which OS ?
(12:58:42) sswam: Linux/386
(12:59:07) sswam: I'll try to track it down a bit better, apparently something is wrong with "Chan" struct
(12:59:44) mahtawkmeister: here's what I did : http://maht0x0r.blogspot.com/2009/07/updated-compilng-inferno.html
(12:59:51) mahtawkmeister: and as you can see, thats July
(13:00:03) sswam: do you mean, the most recent change to it?
(13:00:55) sswam: I noticed an empty directory problem also, that was easily fixed.
(13:01:07) sswam: should be fixed in hg though
(13:01:24) sswam: (I needed to mkdir /usr/inferno/Linux/386/lib)
(13:01:41) sswam: apparently (and I am surprised) you don't have many Linux/386 users!
(13:02:34) sswam: I'll try following your instructions instead of what is in the package then..
(13:02:41) sswam: the repo I mean
(13:13:57) sswam: someone has been hacking it this monday and apparently broke it
(13:14:20) sswam: or rather didn't finish
(13:14:33) sswam: I might check it out the previous day, he can fix it as he knows what he's doing better than me
(13:14:36) sswam: (forsyth)
(13:14:51) mahtawkmeister: we do have users, just not everyone compiles it from hg every day
(13:15:41) mahtawkmeister: mjl- http://www.vitanuova.com/inferno/downloads.html, then the link inferno-20090730.tgz
(13:15:50) mahtawkmeister: start there really, no much has been added since then !
(13:16:27) mahtawkmeister: Inferno is in active daily development but only a few ppl and we usually use Plan9
(13:16:50) sswam: yeah; I did do that but someone said I should use hg instead - I agree, but it happens to be broken at the moment
(13:17:38) sswam: I could try to fix it but I'll just go back to august 30 instead I think :)
(13:18:06) sswam: should I email forsyth or the list and tell him?
(13:18:46) sswam: I am really impressed with 9vx, it's works a charm on my linux box
(13:19:15) sswam: I can't really run plan 9 as my main os all the time, but on this little eec pc I can run linux, windows, plan 9 and more all at once quite happily!
(13:19:36) sswam: the previous revision seems to be building okay
(13:21:26) sswam: I write a programming language CZ based on C, which aims to make graphics, animation, music etc easy to do in C.  I will try to port it to plan 9, should be fun, plan 9 could do with some graphics demos I think.
(13:25:00) sswam: it compiled ok, now I have a new error: emuinit: cannot load /dis/lib/arg.dis
(13:25:09) sswam: maybe I should run inferno under 9vx instead :(
(13:27:39) sswam: it's working on my server though.  maybe because I symlinked /usr/inferno to my home
(13:28:08) robot12: try emu with -r "root"
(13:29:03) robot12: I start inferno by emu -r $HOME/inferno -c1 -g1200x950 wm/wm
(13:30:45) sswam: hm I think one file didn't compile, don't know why!
(13:30:55) sswam: or more, weird, the compile wasn't finished - sorry
(13:34:27) gualteri [n=speiro@crespins.disca.upv.es] entered the room.
(14:31:48) robot12 left the room (quit: "Ухожу я от вас (xchat 2.4.5 или старше)").
(14:55:54) andguent: do i see it right that -c1 disables bound checking for arrays?
(15:04:00) sswam1 [n=sswam@CPE-121-219-121-250.vic.bigpond.net.au] entered the room.
(15:16:11) sswam left the room (quit: Read error: 110 (Connection timed out)).
(15:37:4(16:52:40) andguent: jas: i just wondered about -b though...
(17:10:09) rapidfx left the room (quit: Remote closed the connection).
(17:18:01) rapidfx [n=host666@vl-cen-ce1.avtlg.ru] entered the room.
(18:21:11) sswam1 left the room (quit: Remote closed the connection).
(19:17:02) npe left the room (quit: ).
(19:34:05) mennis left the room (quit: Read error: 60 (Operation timed out)).
(21:16:56) j123m left the room (quit: Read error: 60 (Operation timed out)).
(21:25:08) mennis [n=mennis@adsl-068-016-104-079.sip.asm.bellsouth.net] entered the room.
(22:19:46) rapidfx left the room (quit: "Leaving.").
(22:49:42) andguent is now known as drunkenturd
(23:58:27) drunkenturd is now known as andguent
(00:30:41) mennis left the room (quit: Client Quit).
(03:11:47) paigeadele left the room (quit: Read error: 110 (Connection timed out)).
(03:41:37) paigeadele [i=pthomps@75-146-55-252-Washington.hfc.comcastbusiness.net] entered the room.