Conversation with #inferno at Wed Apr  8 05:28:28 2009 on powerman-asdf@irc.freenode.net (irc)
(05:43:22) tombohannon left the room (quit: ).
(05:54:54) newmanbe [n=btdn@138.74.131.25] entered the room.
(05:55:34) newmanbe left the room (quit: Connection reset by peer).
(05:56:29) newmanbe [n=btdn@138.74.131.25] entered the room.
(06:01:18) newmanbe left the room (quit: Connection reset by peer).
(06:02:09) newmanbe [n=btdn@138.74.131.25] entered the room.
(06:11:22) newmanbe left the room (quit: Read error: 104 (Connection reset by peer)).
(06:11:22) newmanbe [n=btdn@138.74.131.25] entered the room.
(06:32:14) mennis [n=mennis@adsl-219-169-69.asm.bellsouth.net] entered the room.
(06:35:21) mennis left the room (quit: Client Quit).
(06:38:39) newmanbe left the room (quit: "Leaving").
(06:44:23) npe [n=npe@108.192-78-194.adsl-fix.skynet.be] entered the room.
(06:49:57) fortyseventeen left the room (quit: "beep").
(07:33:29) npe left the room (quit: ).
(07:50:33) eno left the room (quit: Read error: 104 (Connection reset by peer)).
(07:55:41) eno [n=eno@nslu2-linux/eno] entered the room.
(08:19:02) npe [n=npe@108.192-78-194.adsl-fix.skynet.be] entered the room.
(08:20:55) npe left the room (quit: Client Quit).
(09:02:23) npe [n=npe@108.192-78-194.adsl-fix.skynet.be] entered the room.
(09:12:14) andguent [n=andre@p4FF67D3E.dip.t-dialin.net] entered the room.
(09:14:10) eno__ [n=eno@adsl-70-137-138-31.dsl.snfc21.sbcglobal.net] entered the room.
(09:19:30) eno___ [n=eno@adsl-70-137-138-31.dsl.snfc21.sbcglobal.net] entered the room.
(09:20:12) npe left the room (quit: ).
(09:24:58) eno left the room (quit: Read error: 110 (Connection timed out)).
(09:30:55) eno__ left the room (quit: Read error: 110 (Connection timed out)).
(10:27:41) npe [n=npe@195.207.5.2] entered the room.
(11:10:36) underspecified [n=eric@isa7-dhcp-116-220.naist.jp] entered the room.
(11:11:25) underspecified left the room (quit: Client Quit).
(11:11:46) underspecified [n=eric@isa7-dhcp-116-220.naist.jp] entered the room.
(12:06:31) gualteri [n=unknown@crespins.disca.upv.es] entered the room.
(12:17:57) underspecified left the room (quit: Read error: 113 (No route to host)).
(12:56:44) olegfink: The 28 byte fields are used for names (user
(12:56:46) olegfink: IDs, file names) and are represented as strings and include a terminating null byte.
(12:57:17) olegfink: so no file name served by 9P/Styx can be longer than 28 bytes (and even less unicode characters)?
(13:03:06) mjl-: i thought that was a restriction of the original 9p, not 9p2000
(13:03:12) mjl-: where did you read that?
(13:03:27) mjl-: brb
(13:03:43) C-Keen: olegfink: where did you get that limit?
(13:04:08) olegfink: ah, that's probably the case
(13:04:28) olegfink: I'm reading ipwl on styx instead of mans as god intended.
(13:05:15) C-Keen: olegfink: aah yeah
(13:08:37) olegfink: ah, no all styx messages have variable length
(13:09:20) olegfink: mjl-: could you explain a bit on the statefulness of Styx? ipwl briefly mentions that when comparing to NFSv2
(13:09:42) olegfink: is that understood in the sense that all fid mapping are stored on the server side?
(13:09:48) mjl-: which page does it mention it?
(13:10:36) uriel: wtf is ipwl?
(13:10:44) olegfink: bottom of page 4 and iirc somewhere else.
(13:10:53) uriel: ah, the book
(13:11:07) uriel: good, that thing is ancient, certainly predates 9p2000
(13:11:18) uriel: s/oo/o/
(13:11:22) mjl-: yes, fids map to at least a qid on the server side.  servers usually keep more state info (e.g. a buffer to perform reads from) for some of its fids (depends on qid usually)
(13:11:41) olegfink: ah, yeah, and p. 186
(13:12:56) olegfink: aha, that certainly makes sense, but renders all my speculations on keeping state in the client inapplicable at best.
(13:16:35) olegfink: does that mean that the state of the fid in respect to revision should also be kept on the server? doesn't make much sense to me
(13:17:48) mjl-: yes
(13:18:04) olegfink: yes to 'it should'?
(13:18:07) mjl-: it's like an fd.  the server has the fd, so it should know what the fd represents (e.g. some revision of the file)
(13:18:32) mjl-: yes, to "state of fid wrt revision should be kept on server"
(13:19:41) olegfink: then how do you do the infamous fork
(13:20:23) mjl-: reopen the file...
(13:20:39) mjl-: that's why i said it was hairy... you cannot always get to the file again
(13:20:44) aguent [n=andre@p4FF66FF7.dip.t-dialin.net] entered the room.
(13:20:45) mjl-: e.g. with clone-based file systems
(13:20:50) mjl-: good day aguent 
(13:22:42) olegfink: hrm, no, I definitely want to keep state in the client (and no I don't want NFS)
(13:22:58) mjl-: perhaps you should have a look at Op]
(13:23:17) mjl-: i haven't read (or remembered) the details, but i think it is less dependent on fids as handles to files
(13:23:39) mjl-: Op is nemo's layer/replacement on/of 9p
(13:32:39) olegfink: yes, I know Op, should have a look, thanks.
(13:36:42) andguent left the room (quit: Read error: 110 (Connection timed out)).
(13:38:49) olegfink: but that all is starting to look very different from my expectation.
(13:54:15) gualteri_ [n=unknown@crespins.disca.upv.es] entered the room.
(14:10:09) gualteri left the room (quit: Read error: 110 (Connection timed out)).
(14:18:17) gdiaz_ios [n=none@217.14.40.23] entered the room.
(14:28:09) aguent: mjl-: oh yes for you too :)
(14:28:40) tombohannon [n=tombohan@h117.255.18.98.static.ip.windstream.net] entered the room.
(14:42:44) tombohannon left the room (quit: ).
(14:54:26) megaboz [i=none@201.80.224.34] entered the room.
(15:24:48) underspecified [n=eric@softbank220043052007.bbtec.net] entered the room.
(15:47:30) tombohannon [n=tombohan@adsl-068-016-104-079.sip.asm.bellsouth.net] entered the room.