Conversation with #inferno at Fri Apr 10 16:23:53 2009 on powerman-asdf@irc.freenode.net (irc)
(16:42:25) npe [n=npe@108.192-78-194.adsl-fix.skynet.be] entered the room.
(16:43:09) npe left the room (quit: Connection reset by peer).
(16:47:46) npe [n=npe@108.192-78-194.adsl-fix.skynet.be] entered the room.
(17:11:06) npe left the room (quit: ).
(17:12:08) eno__ [n=eno@adsl-70-137-148-74.dsl.snfc21.sbcglobal.net] entered the room.
(17:12:41) eno left the room (quit: Read error: 104 (Connection reset by peer)).
(17:14:57) eno__ is now known as eno
(17:23:22) npe [n=npe@108.192-78-194.adsl-fix.skynet.be] entered the room.
(17:52:31) mennis [n=mennis@adsl-068-016-104-079.sip.asm.bellsouth.net] entered the room.
(18:00:26) ***mennis is firmly in the mailinglists-can-be-slow camp.
(18:19:17) npe_ [n=npe@ip-81-11-228-147.dsl.scarlet.be] entered the room.
(18:28:06) npe_ left the room (quit: Operation timed out).
(18:30:23) tombohannon left the room (quit: Remote closed the connection).
(18:30:38) tombohannon [n=tombohan@adsl-068-016-104-079.sip.asm.bellsouth.net] entered the room.
(18:35:03) npe_ [n=npe@ip-81-11-228-147.dsl.scarlet.be] entered the room.
(18:35:31) npe left the room (quit: Read error: 110 (Connection timed out)).
(18:37:31) maht-: I've not got the knowledge of how to compile Inferno from the google code, any helpers around ?   where to even start, is it "only works on Plan9" job ? cos on Linux using p9p it fails miserably
(18:38:47) olegfink: hrm, checkout from svn, sh makemk.sh; export PATH=`pwd`/Linux/386/bin:$PATH; mk install
(18:38:53) uriel: using p9p?
(18:39:02) Fish [n=SPARCman@AVelizy-152-1-54-104.w82-120.abo.wanadoo.fr] entered the room.
(18:39:05) uriel: and isn't building documented somewhere? *sigh*
(18:39:20) olegfink: I believe it is, in INSTALL
(18:39:23) maht-: I don;t know, I've searched around and cant' find anything
(18:39:28) uriel: (I seem to remember building it, and I'm clueless, so I must have got documentation somewhere)
(18:39:49) maht-: I'll give olegfink's suggestiona try and report back
(18:39:54) ***uriel sighs at the usual pathetic user-friendliness of inferno
(18:40:30) uriel: yea, /INSTALL seems to have the required info
(18:40:43) uriel: (Surprisingly enough)
(18:40:55) maht-: I'm particularly frustrated because the cdinstall.iso has bitrotted and crashes on both Windows & Linux (in different  ways
(18:41:15) olegfink: i think there was something on inferno-list on that one.
(18:41:50) maht-: yeah, Xorg killed it on Linux no-one's even ackowledged the Widnows problem
(18:41:58) uriel: bitrotting? nah, that never happens, or so I hear from the apologists
(18:42:06) uriel: the .iso is a joke
(18:42:24) maht-: lol @ typing, I get "file not found" sometime when I type a tab / backspace in Acme !
(18:42:47) maht-: and it locks up
(18:43:23) uriel: just give up inferno from VN, use acme-sac like the other  half a dozen people that uses inferno...
(18:43:43) uriel: (at least on windows)
(18:45:34) mjl-: in my experience, compiling from svn is simple on linux, openbsd & plan 9
(18:45:50) mjl-: but haven't used x11 emu for some time, perhaps still crashes with svn inferno-os
(18:45:57) mjl-: but compiling shouldn't be the problem
(18:47:18) gdiaz_ios: compiling it in windows is as easy as is in linux or mac
(18:47:41) mjl-: really? :)
(18:47:50) mjl-: i didn't have success with that
(18:47:54) gdiaz_ios: you just need visual studio express (downloable for free)
(18:48:10) mjl-: first had to download huge compiler stuff.  but i didn't know which version... the names did not ring a bell
(18:48:22) mjl-: that's how it must feel when people first use plan 9
(18:48:25) mjl-: or inferno for that matter
(18:48:37) mjl-: having no idea what the tools are called, and where to get them
(18:48:41) gdiaz_ios: visual studio c++ version
(18:48:48) maht-: charles told me in private email svn version works and he "nedds to ge round to updating the isos"
(18:49:03) maht-: dont worry I wont be doing it on windows
(18:49:12) mjl-: if i were him, i would just delete the iso's
(18:49:17) mjl-: as if anyone wants to install from those...
(18:49:40) maht-: I am using Acme-Sac for editing my source files but I've done all that now, I want to deploy
(18:49:48) mjl-: if you run native, you do something smart yourself.  i you run hosted, just give a binary (or some zip file)
(18:50:07) mjl-: deploy on external machines?
(18:50:23) mjl-: then just make a .tgz or .zip with some .dis files, and the emu binary
(18:51:06) maht-: well tbh I probably want to eventually make a whole ISO but for now just a hosted on Debian
(18:51:32) maht-: I tried the online instructions for native in Qemu but that failed too :(
(18:51:33) mjl-: what do you plan to do with the iso?
(18:51:42) maht-: but that was my Qemu setup
(18:51:51) gdiaz_ios: i did installers for ubuntu, mac and windows
(18:51:52) mjl-: hmm, i've booted inferno on qemu not too long ago
(18:51:53) maht-: oh just have it around (iso) and maybe host it onle
(18:51:59) maht-: online
(18:52:02) mjl-: :D
(18:52:08) mjl-: solid goals :D
(18:52:31) mjl-: i remember i added a different ethernet driver to the inferno kernel, to let it work with the qemu i had.
(18:52:34) mjl-: or so i think
(18:53:07) gdiaz_ios: hum is there a process to build an inferno bootable ISO?
(18:53:10) mjl-: last week i tftp booted inferno, using 9pxeload.  worked perfectly.  luckily inferno native has intel network drivers that worked for me.  otherwise it is less useful
(18:53:13) maht-: no so bothered about the Qemu version, it probably runs slower than hosted
(18:53:31) gdiaz_ios: i could run shuch a process on 9grid.es and build a ISO every week or whatever
(18:53:49) maht-: I was going to use a Hypervisor but found my Xeons don't have VT so that's off
(18:53:50) mjl-: why the obsession with iso's?
(18:54:03) uriel: 15:49 < mjl-> if i were him, i would just delete the iso's
(18:54:04) maht-: just for other people from my end
(18:54:09) uriel: and the idiotic *installers*
(18:54:18) uriel: what a total waste of everyones time
(18:54:27) mjl-: there is only an installer for windows, right?
(18:54:36) uriel: no, also for lunix
(18:54:42) mjl-: oh?  haven't seen that ever
(18:54:44) uriel: it is a total fucking pile of worthless broken shit
(18:55:02) mjl-: do they handle upgrades or something?
(18:55:08) uriel: which fails in the most annoying and incomprensible ways, and it is totally useless
(18:55:12) uriel: no
(18:55:20) mjl-: afaik, just un-tgz'ing a file with newer version is the perfect upgrade strategy here
(18:55:29) uriel: it just 'untars', the 'tars' that are not really tars
(18:55:38) mjl-: ah yes, those :)
(18:55:49) mjl-: the wrap/arch/whatever-they-are-called things
(18:55:58) uriel: so idiotic words fail to describe it
(18:56:04) maht-: oh right, ok, no worries then, I just always found them convenient
(18:56:43) mjl-: binaries are really only useful for windows systems (perhaps macos x too, i don't know how people generally use mac os x)
(18:56:44) uriel: and if you got the paths wrong, or called it from a different dir than it expected, the whole thing went to hell
(18:56:56) uriel: truly awful, stupid, and *useless* shit
(18:57:00) mjl-: but on any linux people either: 1. use package management software  2. compile themselves
(18:57:24) uriel: even on windows, you can distribute the way acme-sac is distributed, just a god damned .zip file
(18:57:32) mjl-: almost never run some external binaries.  unless they are closed-source flash players or adobe readers perhaps...
(18:57:41) uriel: but no, we need the insane idiotic broken installer
(18:57:48) mjl-: true, it doesn't make sense.
(18:58:10) gdiaz_ios: http://9grid.es/store/
(18:58:30) uriel: hey, this is inferno, if it made sense, we shouldn't do it, we might get some new users, or even worse, developers, and we don't want that to happen
(18:58:31) mjl-: hmm, what's the .pkg?
(18:58:34) mjl-: for which platform?
(18:58:37) gdiaz_ios: mac
(18:58:38) mjl-: i should try the msi
(18:58:39) mjl-: ah
(18:58:47) mennis: I won't be updating this site anymore (too busy) but if some one wants to post it elsewhere they are welcome to do so.  http://inferno.corvus.net/category/inferno/ 
(18:59:06) gdiaz_ios: i will update them in a couple of days. . .i have newer version of the included scripts
(18:59:07) mjl-: gdiaz_ios: are those .msi's & .pkg's generated automatically too? :)
(18:59:09) uriel: what is ios?
(18:59:21) mjl-: i suppose i == inferno, os == os
(18:59:26) gdiaz_ios: those files needed a name, and i just named them inferno os
(18:59:28) gdiaz_ios: yes
(18:59:35) mjl-: i didn't recognize it at first either :)
(18:59:49) mennis: ios is Cisco's OS might want to avoid that one.
(18:59:49) gdiaz_ios: should i name them inferno ?
(19:00:11) mjl-: yes, my first thought was with cisco too
(19:00:15) uriel: mennis: hey, you could have told me, I would have added your blog to http://planet9.cat-v.org ;)
(19:00:23) gdiaz_ios: i know about the cisco name, but i didn't cared enougj i guess
(19:00:24) mjl-: i usually use shorthands like "inf" and sometimes "infos"
(19:00:28) mjl-: thought infos kind of sucks
(19:00:59) uriel: gdiaz_ios: anyway, what is in those packages?
(19:01:15) mjl-: gdiaz_ios: have you thought about separating the emu binary from the source+.dis files?
(19:01:34) gdiaz_ios: svn version of march compiled plus a script ios.bat or .sh to run inferno
(19:01:37) mjl-: source+.dis can be updated almost automatically, and changes more often than the emu binaries (i think)
(19:01:44) mennis: uriel: THat information up there is free for anyone to promulgate.  I did it while on vacation thinking it might be a new hobby but I haven't the time in normal day to day life.
(19:01:55) gdiaz_ios: i changed the scripts and now them looks more like the ones from acme-sac
(19:01:56) olegfink: I would say just tarballs of svn snapshots would be terribly useful
(19:02:04) olegfink: especially on plan9
(19:02:10) uriel: mennis: is it ok if I add your blog to planet9? so if you post again it will get agregated..
(19:02:37) mjl-: olegfink: yeah, i agree too.  i can make some of inferno
(19:02:59) mjl-: i already do a daily svn update automatically, and get mail when something is updated.  regenerating a .tgz afterwards is easy
(19:03:02) gdiaz_ios: olegfink there is gsvn script to download it to plan9 easily
(19:03:37) mennis: uriel: Yeah, f there is a place on your site that can host it, I'll migrate it in.
(19:03:56) uriel: gdiaz_ios: what scripts did you change
(19:04:02) olegfink: gabi, easy, but slow and, uhm, not the best way in fact.
(19:04:18) gdiaz_ios: slow the first tome
(19:04:20) gdiaz_ios: time
(19:04:37) mjl-: it does a http request per file i suppose?
(19:04:42) gdiaz_ios: anyway, i just notice itﾫﾫs there :) it is useful to me :)
(19:04:54) olegfink: I agree gsvn is much better than nothing, but I'd still prefer tarballs.
(19:05:20) uriel: mennis: oh, I can do that, http://planet9.cat-v.org is just a 'Blog Planet' that agregates posts from all plan9/inferno related blogs, if you want me to host the concent permanently in http://doc.cat-v.org/inferno/ or such let me know
(19:06:06) olegfink: the more non-chronological stuff available outside the blog paradigm, the better.
(19:06:53) uriel: olegfink: well, I sort of agree, but chronological blogs have their place
(19:07:05) uriel: (and it is just editing one line to add one more to planet9
(19:08:25) olegfink: sure, but most tutorials/writeups have nothing to do with timing, so they're much better when placed in a usual navigation structure like that on doc.cat-v.
(19:08:58) olegfink: but I guess having important documentation in a web forum is worse.
(19:10:06) uriel: nah, the plan9/inferno way of having it nowhere is the best
(19:10:33) uriel: btw, anyone using http://man.cat-v.org ? is it useful?
(19:11:16) mennis: uriel: OK which ever, macht nicht.
(19:11:16) maht-: yes
(19:12:05) maht-: it's safer to link to than plan9 because you get "hoho you must be hosting it on plan9 it's down"
(19:12:30) maht-: bah wtf is iyacc
(19:13:06) eno: inferno version of yacc
(19:13:07) gdiaz_ios: uriel i use it, and it is useful
(19:13:22) uriel: maht-: you didn't follow the instructions carefully enough it seems
(19:13:41) uriel: (set path as instructed in INSTALL)
(19:14:10) uriel: gdiaz_ios: ok, cool, it was an experiment, so I was wondering how well it worked out in practice
(19:14:52) ***uriel is still trying to get hold of copies of Plan 9 1st Ed man pages, and man pages for Unix 8th, 9th and 10th Eds)
(19:15:16) maht-: yeah I didm iyacc fails to get built
(19:15:25) maht-: you have to do it manually
(19:15:33) uriel: really? *SIGH*
(19:16:00) uriel: guess that nobody tried to build inferno from scratch in the last few months ince iyacc was renamed, haha
(19:16:03) maht-: I think so, I have the PATH set properly. I didit manually and it work, though no I'm battling limbo not found
(19:16:08) npe [n=npe@ip-81-11-228-147.dsl.scarlet.be] entered the room.
(19:16:11) uriel: (fill a bug, it might get fixed in a few more months)
(19:16:59) maht-: iyacc is in the mkfile, hmm 
(19:18:56) maht-: oh i was mk all, shoudl have been mk install :)
(19:18:57) mjl-: i had an earlier problem with iyacc too
(19:19:45) ***maht- doesn't like to see compiler warnings scroll past :(
(19:20:02) maht-: ok looks like it worked, emu runs
(19:20:54) maht-: ahaha the fking fonts now
(19:21:01) maht-: I remember that one from the list
(19:21:09) maht-: what a battle
(19:21:24) npe_ left the room (quit: Connection timed out).
(19:23:26) maht-: I was dreaming that google could put inferno on their app engine
(19:25:09) uriel: good that inferno fans got *plenty* of time to waste with worthless shit
(19:26:04) uriel: I have dreams like those too... then I wake up and bang my head against the wall to aliviate the pain and missery I feel when I realize it was all a dream
(19:27:07) Fish left the room (quit: Remote closed the connection).
(19:30:29) Fish [n=SPARCman@AVelizy-152-1-54-104.w82-120.abo.wanadoo.fr] entered the room.
(19:32:11) npe left the room (quit: Connection reset by peer).
(19:33:07) npe [n=npe@ip-81-11-228-147.dsl.scarlet.be] entered the room.
(20:02:26) npe left the room (quit: Read error: 104 (Connection reset by peer)).
(20:02:33) npe [n=npe@ip-81-11-228-147.dsl.scarlet.be] entered the room.
(20:24:12) mjl-: http://www.ueber.net/inferno/tgz/
(20:24:25) mjl-: those are straight inferno-os checkouts
(20:24:49) mjl-: the name is inferno-os-<svnrev>-<yyyymmdd>.tgz
(20:24:49) olegfink: aha, cool.
(20:24:59) mjl-: the -nosvn one doesn't have all the .svn subdirs
(20:25:27) mjl-: pretty sick how those double the file size... i wonder if i didn't mess something up...
(20:25:42) olegfink: no, it should be ~2 times
(20:25:45) mjl-: those files are updated twice a day (only when changed).  only one version wil be there
(20:25:55) anothy: isn't it all the revision history?
(20:26:20) mjl-: perhaps some.  at least not all.
(20:26:26) mjl-: svn needs a server to do a diff afaik...
(20:26:37) olegfink: mjl-: then it'd be nice to have inferno-os-latest appropriately linked?
(20:26:41) eno: git or hg does better
(20:26:54) mjl-: eno: yeah, svn is ancient tech ;)
(20:26:59) mjl-: olegfink: i can do that
(20:27:01) olegfink: iirc in .svn there's mostly a 'base copy'
(20:27:08) eno: git-svn is very good
(20:27:09) mjl-: though i never use the -latest links
(20:27:16) mjl-: actually, i'm not going to put in the latest link
(20:27:24) mjl-: hmm, perhaps i should
(20:27:39) mjl-: if you just want some latest version.  otherwise i'ld say:  open the listing, click on the url
(20:27:48) mjl-: ok, i'll make latest-links...
(20:27:48) olegfink: it has just one use: I don't have to fire up a browser to get the latest inferno.
(20:28:19) gdiaz_ios: just put one and only one inferno-latest.tgz, because does it make sense to have the old ones?
(20:28:25) olegfink: believe it or not, but the browser isn't my main workplace on plan9...
(20:28:59) mjl-: gdiaz_ios: i want the file name to contain the revision number
(20:29:07) mjl-: the date is actually uselses, the timestamp in the listing does that
(20:29:18) olegfink: i.e. my windows life got better after just making a simple http redirect from a short url to the last version of putty.exe :-)
(20:29:47) mjl-: :D
(20:30:03) olegfink: s/last/latest/
(20:30:04) mjl-: yeah, funny how those small things feel like they have a real impact
(20:31:05) eno: inferno-os$ du -s .git
(20:31:05) eno: 27668   .git
(20:31:11) olegfink: I believe it's exactly how you feel when something is 'well-designed'
(20:31:17) eno: this contains all the history
(20:31:47) eno: only slightly larger than nosvn.tgz
(20:32:05) mjl-: yeah, pretty good
(20:32:35) mjl-: it's smart to make delta's against parent versions
(20:32:45) eno: actually it's the top level, including trunk/branch/wiki
(20:33:03) olegfink: wait, just move inferno to hg and export hgfs. :-)
(20:33:28) olegfink: (you can just create a free account at freehg.org and keep it in sync with gcode)
(20:33:34) mjl-: hmms, the symlinks don't work for some reason.  that's bad
(20:33:44) olegfink: (or use any other hg repo)
(20:33:50) olegfink: mjl-: use http redirects
(20:34:12) olegfink: it's 'the right thing as far as I understand'
(20:34:22) olegfink: er, misplaced quote mark
(20:34:31) mjl-: hmm, perhaps i should do that
(20:34:37) mjl-: will do after dinner
(20:34:42) mjl-: which i'll have to cook now
(20:34:46) gdiaz_ios: how difficult is puclish a tar in a website ﾄ)
(20:34:50) eno: i don't think hg-svn integration is as good as git-svn
(20:35:08) eno: so you'll probably ask the upstream to change to hg
(20:35:08) mjl-: no, hg doesn't have very nice svn tools i think.  i looked yesterday or day before that
(20:35:10) olegfink: eno: we'll only have gitfs after this years gsoc.
(20:35:13) olegfink: :-)
(20:35:34) mjl-: i just slightly prefer hg, given existing investments in it
(20:35:42) mjl-: gdiaz_ios: too hard for me :)
(20:36:24) olegfink: mjl-: wait, you said 'dinner'?
(20:36:31) olegfink: what an excellent idea! thanks.
(20:36:56) gdiaz_ios: it's more the web than you i think ☺. . . 
(20:37:19) mjl-: well, i botched the symlinks it seems :P
(20:37:33) mjl-: and otherwise i was too blame too, it's my httpd too :P
(20:38:13) gdiaz_ios: you should put a plan9 or inferno server online. . 
(20:38:35) olegfink: ueber.net is lhttpd iirc.
(20:38:45) eno: http://www.selenic.com/mercurial/wiki/index.cgi/HgSubversion has some hope
(20:39:13) olegfink: mjl-: by the way, does hgfs do tarballs?
(20:39:16) eno: but it's not ready for production use
(20:39:34) mjl-: olegfink: "do"?  as in generate?
(20:40:17) olegfink: more as in serve, but that would probably require generating at least once for every revision.
(20:40:22) mjl-: http://www.ueber.net/hg/hgfs/tgz/
(20:40:39) mjl-: the tgz is generated as you read it
(20:40:49) mjl-: you can open any revision
(20:40:54) mjl-: <= the latest ;)
(20:41:11) mjl-: (it doesn't know the size before hand)
(20:41:18) mjl-: (that's why it is 0)
(20:41:52) olegfink: oh sweet
(20:41:59) olegfink: what's the syntax? http://www.ueber.net/hg/hgfs/tgz/30/ is 404
(20:42:09) mjl-: its $reponame-$rev.tgz
(20:42:24) olegfink: ah, just in tgz/?
(20:42:26) mjl-: i just list the latest revision
(20:42:35) mjl-: yeah, tgz/ "holds" the .tgz files
(20:42:40) mjl-: but you can go up in hiearchy
(20:43:09) olegfink: okay, got it.
(20:43:27) mjl-: instead of the web, you can just do:  mount -A -P net!knaagkever.ueber.net!6867 /n/hg && cd /n/hg/hgfs; du -a
(20:43:28) olegfink: just thought that analogous to files/$rev/ there's tgz/$rev/
(20:44:00) mjl-: but then you'ld have the file $repo-$rev.tgz in that dir.  because you need to make it a file, to let the browser download it
(20:44:08) olegfink: excellent, now just get inferno hosted there. :-P
(20:46:54) mjl-: i think i'm going to try eno's link
(20:46:59) mjl-: if i get around to doing it...
(20:47:46) mjl-: i'm going to forget about the redirect.  it's too annoying and a bit hacky if i still want to make the -latest file show up in the dir listing
(20:47:58) mjl-: really dinner time now
(21:13:04) eno__ [n=eno@adsl-70-137-148-74.dsl.snfc21.sbcglobal.net] entered the room.
(21:14:44) eno left the room (quit: Read error: 104 (Connection reset by peer)).
(21:17:36) eno__ is now known as eno
(22:24:54) eno___ [n=eno@adsl-70-137-173-224.dsl.snfc21.sbcglobal.net] entered the room.
(22:35:32) eno left the room (quit: Read error: 110 (Connection timed out)).