Time is in GMT-8.
- ** BEGIN LOGGING AT Sat Jan 22 05:59:17 2005
Jan 22 06:00:18 * sandalle looks up Seth's number
Jan 22 06:00:38 afrayedknot i can pm it to you
Jan 22 06:01:02 sandalle got it. :)
Jan 22 06:01:30 sandalle he's up now. :)
Jan 22 06:01:33 sandalle on his way
Jan 22 06:01:40 sandalle doesn't sound as tired as last time I called
Jan 22 06:02:09 * neuron feels convenient to have 3 PM right now :)
Jan 22 06:02:23 * swoolley (~seth@ns0.positivism.org) has joined #smgl-admin
Jan 22 06:02:23 * ChanServ gives channel operator status to swoolley
Jan 22 06:02:30 * low too
Jan 22 06:02:31 sandalle neuron, lucky you. :)
Jan 22 06:02:32 dmlb2000|mobile there he is ;)
Jan 22 06:02:36 low howdy swoolley
Jan 22 06:02:37 afrayedknot if Oadae isnt coming anymore maybe we could move it a few hours forward
Jan 22 06:02:42 * swoolley yawns
Jan 22 06:02:48 sandalle afrayedknot, true, since we had it at this time for him
Jan 22 06:03:19 sandalle shall we get this started?
Jan 22 06:03:31 * dmlb2000|mobile thinks so
Jan 22 06:03:37 * neuron is ready
Jan 22 06:03:38 afrayedknot i guess will go ahead with Kinetix
Jan 22 06:03:43 afrayedknot erm without
Jan 22 06:04:56 sandalle Then I shall call this meeting to order.
Jan 22 06:06:02 sandalle I'm sure everyone's familiar with the RoadMap: http://wiki.sourcemage.org/index.php?page=SourceMageRoadMap
Jan 22 06:06:44 sandalle Our target 1.0 release date is 2005-04-01, roughly two and a half months from now
Jan 22 06:07:03 sandalle ahh, 2005-04-03, sorry
Jan 22 06:07:12 afrayedknot *phew*
Jan 22 06:07:16 sandalle afrayedknot, ;)
Jan 22 06:07:22 neuron :)
Jan 22 06:07:54 sandalle Our 4 main teams, Sorcery, Grimoire, ISO, and Web/Tome (in no particular order) have several targets to meet, some more than others.
Jan 22 06:08:33 sandalle We'll present in the above order as the first two teams have most of everyone here and will give the other two time to wake up.
Jan 22 06:08:56 sandalle At the end I'll open it up for general comments and additional notes
Jan 22 06:09:04 sandalle The goals for the Sorcery Team are listed as:
Jan 22 06:09:14 sandalle * Sorcery supports full dependency ordering [DONE]
Jan 22 06:09:14 sandalle * Sorcery has support for the new compile/install system [DONE]
Jan 22 06:09:14 sandalle * Sorcery's man/help pages are up-to-date and on-line (Karsten Behrmann)
Jan 22 06:09:14 sandalle * Sorcery's API is fully documented [Sorcery BashDoc] (Karsten Behrmann)
Jan 22 06:09:14 sandalle * rewrite/cleanup of summon and constituent downloading code [DONE]
Jan 22 06:09:14 sandalle * rewrite/expansion of libdepends to support a wider array of features: (Andrew Stitt)
Jan 22 06:09:14 sandalle o grimoire_depends
Jan 22 06:09:14 sandalle o optional_grimoire_depends
Jan 22 06:09:14 sandalle o build_depends
Jan 22 06:09:14 sandalle This may also include some improvements to conflicts and adding more
Jan 22 06:09:14 sandalle features there too. CONFIGURE stuff may need some improvements at the
Jan 22 06:09:14 sandalle same time. Oh and triggers too ;) We may be able to pull this all off
Jan 22 06:09:14 sandalle (full depends ordering too) and get rid of our crutch make, which would
Jan 22 06:09:14 sandalle give us a bit more freedom overall.
Jan 22 06:09:14 sandalle * improved upgrade mechanism, we need this to roll new users forward with changes to local file formats. (Andrew Stitt)
Jan 22 06:09:14 sandalle * changes/extensions to state information. (Andrew Stitt)
Jan 22 06:09:14 sandalle Those two tie together fairly closely, and if we are to release a "1.0"
Jan 22 06:09:14 sandalle I dont really want to make all the new installs go through a conversion.
Jan 22 06:09:14 sandalle * improved binary (resurrect) support. This is needed for the iso. This was worked around by hacking some things into place in the iso but it would be better to have our cache tarballs be a little smarter. (Andrew Stitt)
Jan 22 06:09:14 sandalle * Improved dispel support, sorcery is a bit stupid about removing files and could stand to be a lot smarter about it. Similarly our install process is a bit scary as well. (Andrew Stitt)
Jan 22 06:09:14 sandalle * install_root/cross_install sorted out. We currently have somewhat of an anarchy of coding going on here with no clear destination of specification as to how this is supposed to work. [DONE]
Jan 22 06:09:14 sandalle * [alter] (swoolley's strip/prelink script) (Seth Woolley)
Jan 22 06:09:14 sandalle * delve, run individual spell files, more useful once we complete the build split. (Andrew Stitt)
Jan 22 06:09:14 sandalle * all those other little nagging defects we have [Sorcery Bugs]
Jan 22 06:10:46 sandalle Sorcery is currently at 1.11.3 and has most of what we want done, I believe.
Jan 22 06:10:48 afrayedknot i guess that the end of the list
Jan 22 06:11:21 afrayedknot well not really 1.12 has half of the 1.0 features we've vaguely agreed to, and 1.13 has the rest
Jan 22 06:11:50 afrayedknot originally i had all of them in 1.12 but its proven to be too much stuff in one big release so i split it in half
Jan 22 06:12:12 afrayedknot you can of course view the lists at bugzilla
Jan 22 06:12:48 sandalle (http://bugs.sourcemage.org/buglist.cgi?cmdtype=runnamed&namedcmd=Sorcery%20Bugs for those interested)
Jan 22 06:12:57 afrayedknot i guess i'll hilite some of the things that have changed since we last met
Jan 22 06:13:08 afrayedknot cross grimoire dependencies were completed
Jan 22 06:13:29 afrayedknot theres a specification for it somewhere on the wiki
Jan 22 06:13:37 afrayedknot i dont recall where and my browser is being slow
Jan 22 06:13:51 afrayedknot http://wiki.sourcemage.org/index.php?page=Cross+Grimoire+Depends
Jan 22 06:14:15 afrayedknot ive been working off and on completing the new resurrect
Jan 22 06:14:27 benoit great :-)
Jan 22 06:15:08 afrayedknot ive mostly implemented tablet and soon other things will be learning to take advantage of it and its extra state
Jan 22 06:15:26 afrayedknot delve on that list above is complete
Jan 22 06:15:42 afrayedknot we also added gpg checking of sorcery/grimoire tarballs
Jan 22 06:15:57 afrayedknot the sorcery half of that is complete
Jan 22 06:16:27 sandalle afrayedknot, that's swoolley's proposed GPG system, correct?
Jan 22 06:16:47 afrayedknot yea, using the sorcery-pubkeys spell for the keys
Jan 22 06:17:32 afrayedknot im not sure we'll be able to reach complete all the 1.12 and 1.13 work before april though
Jan 22 06:17:39 sandalle afrayedknot, swoolley, do we have a writeup (on the wiki) of what the Grimoire Team needs to do to implement GPG keysigning once it hits stable?
Jan 22 06:17:55 sandalle afrayedknot, what items are the most likely to hold Sorcery back for April?
Jan 22 06:18:13 afrayedknot well, that list above isnt quite accurate
Jan 22 06:18:16 swoolley there's an email to the mailing list with most everything, I can paste it into the bug I suppose.
Jan 22 06:18:35 afrayedknot should eventually have a wiki for it
Jan 22 06:19:12 sandalle swoolley, a wiki explaining the steps would probably help and/or a link to the sm-discuss archive to the e-mail
Jan 22 06:19:26 sandalle afrayedknot, what's incorrect about the list?
Jan 22 06:20:08 afrayedknot well its missing gpg, we've moved build_depends off the list (its an untargetted feature), triggers being updated isnt on the 1.12 or 1.13 lists
Jan 22 06:20:28 afrayedknot gpg isnt on it
Jan 22 06:20:37 afrayedknot i already said that oops
Jan 22 06:20:46 dmlb2000|mobile it's early
Jan 22 06:21:05 neuron It's important
Jan 22 06:21:12 swoolley dmlb2000|mobile: it is my friend.
Jan 22 06:21:24 dmlb2000|mobile ;)
Jan 22 06:21:30 sandalle afrayedknot, why did we take build_depends and triggers off the list?
Jan 22 06:21:51 afrayedknot we could look at the meeting logs
Jan 22 06:22:24 afrayedknot probably because we dont need build depends really badly right now
Jan 22 06:22:37 afrayedknot and merging triggers and depends is a huge project
Jan 22 06:22:45 sandalle afrayedknot, ok
Jan 22 06:23:00 afrayedknot ie destabalizing and evil
Jan 22 06:23:26 sandalle most of this is in devel Sorcery then, correct?
Jan 22 06:24:29 afrayedknot about half of it yes
Jan 22 06:24:46 dmlb2000|mobile afrayedknot, it's really nice too ;)
Jan 22 06:25:22 sandalle afrayedknot, do you know about when you could have everything we want in Sorcery?
Jan 22 06:25:41 afrayedknot probably a few months more past april
Jan 22 06:26:15 afrayedknot i think we ought to figure out when we'll have our iso ready and work backwards, but most of the cauldron team isnt here so we cant really decide that
Jan 22 06:26:47 sandalle afrayedknot, right.
Jan 22 06:27:03 sandalle So June/July looks more like the target for Sorcery?
Jan 22 06:27:17 afrayedknot yea that sounds do-able the moment
Jan 22 06:27:33 afrayedknot we'll see how many new features come up
Jan 22 06:28:04 sandalle afrayedknot, any other items you'd like to bring up for Sorcery?
Jan 22 06:28:27 * benoit has a question about install_root
Jan 22 06:28:28 afrayedknot i think im done
Jan 22 06:28:30 afrayedknot okay
Jan 22 06:28:31 * p3pilot (~george@pcp0010844817pcs.alxndr01.va.comcast.net) has joined #smgl-admin
Jan 22 06:28:41 * afrayedknot probably has an answer
Jan 22 06:28:51 sandalle Any questions about Sorcery?
Jan 22 06:28:57 afrayedknot benoit had one
Jan 22 06:29:12 benoit i intend to integrate sorcery's install_root features (and maybe cross_install...)
Jan 22 06:29:26 sandalle afrayedknot, yep, just making it "official". :)
Jan 22 06:29:29 benoit so. do you think it is quite stable to be used for ISO generation?
Jan 22 06:29:45 sandalle benoit, I believe they're already in stable Sorcery
Jan 22 06:30:03 sandalle benoit, I've used install_root to setup an SMGL system on my SPARC; it works quite well. :)
Jan 22 06:30:31 swoolley h...sh also used it for an entire iso before
Jan 22 06:30:31 benoit i think it should solve some problem i have with the current ISO generation system
Jan 22 06:30:34 afrayedknot some improvements were made in devel to work with tablet and a few corner cases that you wouldnt use for iso building were covered, but for your purposes it should work fine
Jan 22 06:30:41 benoit mainly aliens files.
Jan 22 06:30:59 benoit ok. how about cross_install btw?
Jan 22 06:31:08 benoit i wonder if i could use it to generate a sparc iso for instance
Jan 22 06:31:11 sandalle benoit, it'd be perfect for ISOs, which I think is what Kevin is planning
Jan 22 06:31:14 afrayedknot cross install just means you add --host to configure options
Jan 22 06:31:26 sandalle benoit, our gcc does not support cross-install environments yet. :(
Jan 22 06:31:42 * Asha`man (~Asha`man@adsl-81-7-19.asm.bellsouth.net) has joined #smgl-admin
Jan 22 06:31:42 * ChanServ gives voice to Asha`man
Jan 22 06:31:48 benoit sandalle: you mean it needs a separate cross-gcc ?
Jan 22 06:32:12 sandalle benoit, if you're building for a different architecture, yes
Jan 22 06:32:29 benoit afrayedknot: you also need to use the proper ./configure --prefix and make prefix= options
Jan 22 06:32:43 * vlaaad reminds benoit that sandalle has mutated into SweetPotato :p
Jan 22 06:32:46 afrayedknot uh, thats what install root does
Jan 22 06:32:54 benoit so. if i have a cross compiler already, sorcery can be used to make a "cross ISO" ?
Jan 22 06:32:56 sandalle benoit, those are already done for most packages
Jan 22 06:33:04 benoit afrayedknot: yep. sorry.
Jan 22 06:33:05 sandalle benoit, it should be able to, yes
Jan 22 06:33:14 benoit ok.
Jan 22 06:33:17 sandalle vlaaad, I'm me. :)
Jan 22 06:33:25 neuron So is it easy to use non-/usr/bin/gcc by sorcery ?
Jan 22 06:33:27 benoit i think i'll bug you later on the details
Jan 22 06:33:42 neuron or relatively easy
Jan 22 06:33:47 afrayedknot neuron: thats up to the spells mostly
Jan 22 06:33:50 sandalle neuron, you can modify /var/lib/sorcery/build to use a different gcc I believe
Jan 22 06:33:53 benoit but i think it will be a nice feature to use for ISO generation, it could increase the number of supported platforms
Jan 22 06:33:58 afrayedknot we pass the ./configure option, its up to the spell to do the right thing
Jan 22 06:34:15 sandalle benoit, if we/you have a gcc cross-compiler then yes. :)
Jan 22 06:34:27 afrayedknot lets work on native architecture first...
Jan 22 06:34:37 sandalle afrayedknot, yep. :)
Jan 22 06:34:40 benoit sandalle: right. I have build several cross-gcc (using crosstools from kegel.com)
Jan 22 06:34:57 afrayedknot anyways i think ive answered my sorcery question
Jan 22 06:34:59 benoit afrayedknot: sure :-).
Jan 22 06:35:03 afrayedknot lets not go too far on a tangent
Jan 22 06:35:09 afrayedknot any other questions
Jan 22 06:35:09 sandalle afrayedknot, are any of the Sorcery features from devel going to make it into a test release soon?
Jan 22 06:35:28 afrayedknot well thats why i split them into a 1.12 and 1.13 release
Jan 22 06:35:51 afrayedknot once i get the currentl 19 1.12 bugs done...
Jan 22 06:36:20 sandalle afrayedknot, is there anything general gurus/others can do to help?
Jan 22 06:36:34 afrayedknot test test sorcery when it comes out
Jan 22 06:36:45 sandalle afrayedknot, any ETA on that?
Jan 22 06:36:50 * dmlb2000|mobile uses devel sorcery all the time. ;)
Jan 22 06:36:53 afrayedknot month or two
Jan 22 06:36:57 dmlb2000|mobile dmlb2000|server Jan 22 06:37:00 sandalle dmlb2000|mobile, I noticed. :)
Jan 22 06:37:13 sandalle afrayedknot, k, thanks. :)
Jan 22 06:37:25 sandalle Any other questions for afrayedknot while we have his undivided attention? ;)
Jan 22 06:37:43 low yep
Jan 22 06:38:07 afrayedknot :-)
Jan 22 06:38:15 sandalle low, what's the question?
Jan 22 06:38:18 low is dual/multi-arch support to be included in a devel/test sorcery release quite soon ?
Jan 22 06:38:42 afrayedknot well, its not a 1.12 or 1.13 feature
Jan 22 06:38:46 low would be usefull for a64, em64t and sparc ifaik
Jan 22 06:39:00 low afaik
Jan 22 06:39:10 sandalle low, yes, but there's not much work being done on it
Jan 22 06:39:21 afrayedknot lots of features would be useful for some segment of people
Jan 22 06:39:26 neuron I have one too, is the 'unit testing' official part of sorcery ? Does it have to be finished for 1.0 ?
Jan 22 06:39:28 afrayedknot and some definition of useful
Jan 22 06:39:33 sandalle low, I've put in most of benoit's work into Sorcery's proj2, but I haven't worked on it in a while
Jan 22 06:40:10 afrayedknot unit testing and test development is ongoing, but not currently required for 1.0
Jan 22 06:40:10 sandalle low, we're also wanting to focus on our current platforms, though I'm not going to say "no" to people working on supporting their machines
Jan 22 06:40:19 neuron I can test things on amd64 and sparc platform if need
Jan 22 06:40:33 sandalle neuron, feel free to work on Sorcery's proj2 then. :)
Jan 22 06:40:41 low neuron, same here, tough my spare time is quite limited
Jan 22 06:41:02 neuron sandalle: Any pointer to proj2 ?
Jan 22 06:41:14 sandalle low, there is a bug open for multi-arch support http://bugs.sourcemage.org/show_bug.cgi?id=7472
Jan 22 06:41:20 afrayedknot //sgl/sorcery/proj/proj2
Jan 22 06:41:22 sandalle neuron, //sgl/sorcery/proj/proj2
Jan 22 06:41:26 sandalle afrayedknot, :)
Jan 22 06:41:27 neuron ah, ok
Jan 22 06:41:38 sandalle Any others for Andrew?
Jan 22 06:41:44 low the thing is, if work is done, we must get sorcery and bi-arch synchronised, otherwise it'll be a pain to maintain
Jan 22 06:41:45 afrayedknot talk to me before doing any major work on proj2 though
Jan 22 06:42:04 afrayedknot devel is integrated into proj2 on a regular basis
Jan 22 06:42:12 afrayedknot they are pretty close together
Jan 22 06:42:36 low ok, great, and i suppose once proj2 will be functionnal, proj2 will be integrated into sorcery ?
Jan 22 06:42:37 neuron afrayedknot: yes sir !
Jan 22 06:42:55 sandalle low, into devel I believe, yes
Jan 22 06:43:02 * sandalle gives voice to p3pilot
Jan 22 06:43:19 low sandalle, great :)
Jan 22 06:43:19 afrayedknot low: yes, however like i said, very little feedback and interest in it, so its not been put on the 1.0 schedule
Jan 22 06:43:38 sandalle Any others?
Jan 22 06:44:01 low afrayedknot, afaik, neuron, erics, cuzndragon and prolly a couple others i forgot own a64, so i think there's interest in it :)
Jan 22 06:44:05 afrayedknot if a bi-arch iso is on the iso teams list of goals for 1.0 then that probably should change, but right now its just another fringe feature
Jan 22 06:44:23 afrayedknot right, but you dont need bi-arch for amd64
Jan 22 06:44:33 afrayedknot you can just run pure 64 bit
Jan 22 06:44:42 sandalle or pure 32-bit. :)
Jan 22 06:44:47 afrayedknot right
Jan 22 06:44:49 sandalle same with SPARC
Jan 22 06:44:59 low afrayedknot, pure 64 could work, but things do not work (yet?) in 64 bits mode. grub for example
Jan 22 06:45:19 low sandalle, running 32 when owning a 64 bits cpu is a shame :)
Jan 22 06:46:08 sandalle low, I didn't say it wasn't. :)
Jan 22 06:46:16 alley_cat grub isn't really problem even if you run pure 64 bit as you can just use a 32 bit grub
Jan 22 06:46:21 low afrayedknot, moreover most binaries only programs don't provide 64 bits version
Jan 22 06:46:24 sandalle No other questions/comments for Sorcery?
Jan 22 06:46:32 afrayedknot okay back to SORCERY topics
Jan 22 06:47:14 * benoit has a last comments about amd64 bi-arch support :-)
Jan 22 06:47:27 swoolley low: me too (owning amd64)
Jan 22 06:47:37 sandalle benoit, is it pertinent to Sorcery?
Jan 22 06:47:45 benoit yes. it is.
Jan 22 06:48:09 benoit gcc standard configuration expect to be run on a bi-arch system when compiled on 64 bits architecture
Jan 22 06:48:23 low afrayedknot, 64 bits support is a sorcery topic :) all others distros providing a a64 version do it bi-arch, so there must have a good reason
Jan 22 06:48:26 benoit which mean at least 2 glibc installed ...
Jan 22 06:48:47 afrayedknot low: its sorcery specific in so far as we have arch specs for it, after that, not my problem
Jan 22 06:49:24 benoit afrayedknot: eh eh :-)
Jan 22 06:49:28 sandalle benoit, and gcc, we know this. :)
Jan 22 06:49:35 sandalle afrayedknot, and the junk in proj2. ;)
Jan 22 06:49:53 low afrayedknot, arch spec is not enough, or we would have a a64 iso for quite a while
Jan 22 06:49:57 afrayedknot well yea, but thats for bi-arch, i was unaware that you HAD to have that for 64bits
Jan 22 06:50:01 afrayedknot which seems stupid to me
Jan 22 06:50:11 * swoolley too
Jan 22 06:50:15 alley_cat we do have an (experimental) a64 iso iirc
Jan 22 06:50:26 alley_cat unet has a completely native 64bit system
Jan 22 06:50:37 benoit yes. i just release highly experimental ISO.
Jan 22 06:50:52 low alley_cat, oh, and where can i get that pure 64 bits system ?
Jan 22 06:51:13 alley_cat and completely native is the prefered way to do it imo, bi-arch is a kludge to support proprietary stuff and broken programs
Jan 22 06:51:45 sandalle alley_cat, 64-bit is a bit overdoese for some apps, and so you might want a mixture of 32-bit/64-bit
Jan 22 06:51:56 sandalle low, build it with Sorcery
Jan 22 06:52:22 swoolley I have a tarball somebody made a while ago of a 64-bits system
Jan 22 06:52:24 alley_cat sandalle: even apps taht don't need 64bit usually are faster if compiled for 64bit on amd64 due to hugely bigger number of registers it can use
Jan 22 06:52:44 sandalle alley_cat, true, I'm more familiar with SPARC, which is where my thinking comes from. :)
Jan 22 06:53:04 alley_cat sparc has a useful number of registers with 32bit too, not like broken x86 ;)
Jan 22 06:53:08 sandalle some of this is more ISO conversation, do we have more Sorcery info?
Jan 22 06:53:12 sandalle alley_cat, :)
Jan 22 06:53:13 afrayedknot so people want bi-arch sooner, okay, thats more of an iso topic
Jan 22 06:53:15 low sandalle, tried to do that a couple months ago, and i kept having problems with gcc
Jan 22 06:53:22 afrayedknot lets get off the iso topics and back to sorcery for now
Jan 22 06:53:25 sandalle low, did you file a bug?
Jan 22 06:53:38 afrayedknot like i said if bi-arch becomes a committed 1.0 iso feature sorcery will support it
Jan 22 06:54:10 sandalle afrayedknot, I'd like to hold off on that until we get some people willing to work on proj2. We have plenty of other targets to shoot at in Sorcery for now, IMO.
Jan 22 06:54:19 afrayedknot exactly
Jan 22 06:54:24 low sandalle, filing a bug for something that wasn't official ? :)
Jan 22 06:54:34 sandalle low, the archspec is in Sorcery...
Jan 22 06:55:23 benoit sandalle: maybe we could post a call for developpers on this proj2 thing, that could help us quite a bit... just a thought...
Jan 22 06:55:27 sandalle Last call for any more Sorcery questions.
Jan 22 06:55:43 * benoit is happy with install_root !
Jan 22 06:55:43 swoolley bug 8060?
Jan 22 06:55:48 sandalle benoit, might not be a bad idea
Jan 22 06:55:55 swoolley we want that in 1.0?
Jan 22 06:56:32 low sandalle, right. my bad.
Jan 22 06:56:32 alley_cat i'd say yes
Jan 22 06:56:33 swoolley http://bugs.sourcemage.org/show_bug.cgi?id=8060
Jan 22 06:56:44 * low votes yes too
Jan 22 06:57:03 sandalle benoit, as am I. :)
Jan 22 06:57:50 afrayedknot swoolley: i think so
Jan 22 06:58:00 sandalle swoolley, we could use other distro's numbering scheme
Jan 22 06:58:03 afrayedknot its not that hard
Jan 22 06:58:06 sandalle VERSION-SMGLVERSION
Jan 22 06:58:14 sandalle e.g. 1.0.0-1
Jan 22 06:58:14 afrayedknot bleh
Jan 22 06:58:17 swoolley sandalle: more effort than it's worth
Jan 22 06:58:19 alley_cat using UPDATED for that is easier
Jan 22 06:58:26 afrayedknot until some project uses -'s in there version
Jan 22 06:58:30 afrayedknot which quite a few do already
Jan 22 06:58:33 sandalle alley_cat, except UPDATED doesn't work for stable grimoire in most cases
Jan 22 06:58:45 afrayedknot nothing like ambiguous version strings to get you out of bed in morning
Jan 22 06:58:46 alley_cat sandalle: that's what that bug wants to fix
Jan 22 06:58:46 sandalle swoolley, probably, but figured I'd comment. :)
Jan 22 06:59:22 swoolley the suggestion of using the table UPDATED value would eliminate the stable grimoire problem.
Jan 22 06:59:53 sandalle swoolley, true
Jan 22 07:00:03 swoolley tablet rather
Jan 22 07:00:23 afrayedknot yep
Jan 22 07:00:53 sandalle afrayedknot, how difficult will fixed UPDATED support be?
Jan 22 07:00:59 afrayedknot not very
Jan 22 07:01:19 sandalle so it won't delay Sorcery too much, good. :)
Jan 22 07:01:21 swoolley most of the effort went into thinking about the problem
Jan 22 07:02:22 sandalle any more comments on Bug #8060?
Jan 22 07:02:51 * benoit will read the bug in more details later...
Jan 22 07:03:17 * vlaaad will do so
Jan 22 07:03:45 afrayedknot any more sorcery questions?
Jan 22 07:04:40 afrayedknot going once
Jan 22 07:05:07 neuron ok one
Jan 22 07:05:11 neuron What should I do ? :)
Jan 22 07:05:27 afrayedknot you have time to work on stuff again?
Jan 22 07:05:31 neuron As I have more time now, should I just pickup random bug and gon ?
Jan 22 07:05:32 afrayedknot talk to me after the meeting
Jan 22 07:05:35 neuron ok
Jan 22 07:05:43 sandalle any others? :)
Jan 22 07:06:37 sandalle Thanks for your time afrayedknot. :) I look forward to breaking in a new test Sorcery in a month or so.
Jan 22 07:06:58 afrayedknot you can always break in devel now :)
Jan 22 07:07:08 sandalle afrayedknot, I just updated my laptop to it, hence those two bugs. ;)
Jan 22 07:07:31 sandalle Next up, our Grimoire Team.
Jan 22 07:08:00 sandalle Our Grimoire Team currently has some broad targets:
Jan 22 07:08:02 sandalle * Grimoire has less than 50 bugs open [Bug Report for Codex]
Jan 22 07:08:02 sandalle * Grimoire has converted the most popular (let's say 80% of the grimoire) to the new compile/install system [API Completeness] [API Search] and [API Statistics]
Jan 22 07:08:02 sandalle * All spells are lower-case and no upper-case spells exist in stable [Spell List] [Helper script]
Jan 22 07:08:02 sandalle * Restructure our perforce tree so we can have independant branches of development for new and scary features rather than pushing it all to devel and hoping for the best. [DONE]
Jan 22 07:08:02 sandalle * Remove all duplicate dependencies [Duplicated Dependencies]
Jan 22 07:08:02 sandalle * All spells use config_query instead of query and/or read.
Jan 22 07:08:02 sandalle * All variables in spells should be defined as local (except for config_query stuff)
Jan 22 07:08:02 sandalle * Updated GrimoireGurusHandbook (Arwed von Merkatz)
Jan 22 07:08:29 sandalle I think <50 bugs may be a bit too ambitious, but I really wanted 0 bugs. ;)
Jan 22 07:08:37 alley_cat ok, i'll start with build_api 2
Jan 22 07:08:45 sandalle alley_cat, if you would be so kind as to update us on your progress?
Jan 22 07:09:19 alley_cat right now about a third of devel is still build_api 1, 14% are api 0, i.e. they just need a BUILD_API=2 to convert
Jan 22 07:09:33 alley_cat z-rejected is 100% converted, games is converting at a fast rate recently
Jan 22 07:09:36 afrayedknot last i looked it was 50/50
Jan 22 07:09:42 afrayedknot (for devel)
Jan 22 07:10:03 afrayedknot i can generate daily statistics if you want
Jan 22 07:10:09 alley_cat afrayedknot: 34% are really api1, the rest that isn't api 2 doesn't need anything done to convert except BUILD_API=2
Jan 22 07:10:14 alley_cat i do that already :)
Jan 22 07:10:41 e-type alley_cat: I'm doing my sections not in my personal grimoire
Jan 22 07:10:41 swoolley http://seth.positivism.org/files/searchedapistattotal
Jan 22 07:10:55 p3pilot I have some time now...is it okay to just convert spells that are easy to move to BUIlD_API=2?
Jan 22 07:10:59 alley_cat so i think the 80% target is doable by april 3rd
Jan 22 07:11:03 alley_cat p3pilot: yes
Jan 22 07:11:11 e-type alley_cat: editors almost finished, wm's after that
Jan 22 07:11:34 * e-type is talking gibberish...
Jan 22 07:11:52 e-type what I meant was that I'm not converting editors and wm's in devel but here in my grimoire
Jan 22 07:12:01 p3pilot alley_cat: okay...I will do as I get a chance...my section's are done
Jan 22 07:12:03 sandalle e-type, why not in devel?
Jan 22 07:12:05 alley_cat why don't you do it in devel?
Jan 22 07:12:06 swoolley I have hourly statistics generated on api conversion
Jan 22 07:12:06 e-type I'll put all changes through in one move
Jan 22 07:12:26 sandalle e-type, it might be better to do it as you go so we don't have to review a big changelog. ;)
Jan 22 07:12:32 sandalle swoolley, thanks. :)
Jan 22 07:12:36 swoolley http://seth.positivism.org/files/searchedapistat
Jan 22 07:12:43 e-type sandalle: I simply split up my sections in api 1 and api 2, and then move all from 1 to 2, then into devel
Jan 22 07:12:53 sandalle alley_cat, should we change that to 100% api_2?
Jan 22 07:13:13 alley_cat i think 80% is reasonable
Jan 22 07:13:34 vlaaad no problem for me, it should be done on 3rd april
Jan 22 07:13:40 vlaaad in my section
Jan 22 07:13:54 alley_cat if we get 100% it's nice, but not necessary
Jan 22 07:14:15 alley_cat we have more important work than build_api 2
Jan 22 07:14:35 alley_cat like bugs, i honestly don't believe we can even get close to 50 bugs open in that time frame
Jan 22 07:14:46 p3pilot alley_cat: are the unaintained sections on the website accurate?
Jan 22 07:14:55 alley_cat p3pilot: they should be
Jan 22 07:15:32 vlaaad alley_cat : I think a big problem is unmaintained sections, bugs are accumulating there
Jan 22 07:16:05 alley_cat yes
Jan 22 07:16:20 alley_cat but we can't do much about that unless some general gurus step up to maintain a section
Jan 22 07:16:38 qnr for what it's worth, page_six has an !unmaintained command, and a !bug search section command for checking on unmaintained sections easily
Jan 22 07:16:49 sandalle qnr, thanks. :)
Jan 22 07:16:56 e-type this far I've only done bughunts in my sections, but I can handle other easy bugs too
Jan 22 07:16:59 vlaaad it's good to have general gurus, but I think section maintainers are more important
Jan 22 07:17:15 p3pilot I could take on another section, if mostly simple spells like chat*
Jan 22 07:17:16 sandalle alley_cat, should we revise the bug count?
Jan 22 07:17:29 alley_cat vlaaad: yes, but i can't force people to maintain sections ;)
Jan 22 07:17:35 vlaaad but we can't force people to work only on a section ;-)
Jan 22 07:17:57 vlaaad alley_cat : :)
Jan 22 07:17:59 sandalle I believe unmaintained sections are open for anyone to fix bugs in and do updates, so you don't need to be a maintainer to help on those
Jan 22 07:18:13 alley_cat sandalle: 50 is definitely not a realistic target right now, 100 might be
Jan 22 07:18:24 vlaaad SweetPotato : ok, I will think about it ;-)
Jan 22 07:18:31 sandalle if you want to maintain a section, that's great. :) But if you want to only help fix bugs in it, you don't need to be maintainer
Jan 22 07:18:31 alley_cat not counting new spell bugs of course
Jan 22 07:18:34 Usage: NICK <nickname>, sets your nick
Jan 22 07:18:50 sandalle vlaaad, SweetPotato isn't in here. ;)
Jan 22 07:19:05 sandalle alley_cat, right...I think most of those could be fixed easily and should be all done. ;)
Jan 22 07:19:05 benoit sandalle: i thought it was made clear that each maintainer has to work only in his section...
Jan 22 07:19:09 sandalle alley_cat, is that reasonable?
Jan 22 07:19:27 sandalle benoit, no. it's only maintained sections that you have to be careful of
Jan 22 07:19:30 vlaaad sandalle : I think I missed something with your alias :-D
Jan 22 07:19:42 alley_cat well, i do reassign new spells to the section maintainers, which works quite well, but the unmaintained sections pile up there too
Jan 22 07:19:44 sandalle benoit, and get permission from the maintainer before modifying his section
Jan 22 07:20:05 sandalle http://bugs.sourcemage.org/buglist.cgi?cmdtype=runnamed&namedcmd=Grimoire%20Bugs - 247 bugs currently
Jan 22 07:20:23 vlaaad O.o
Jan 22 07:20:41 swoolley sandalle: that only works for you.
Jan 22 07:20:42 vlaaad are new spells counted in ??
Jan 22 07:20:48 alley_cat and i don't consider new spell submissions as real bugs
Jan 22 07:20:53 alley_cat vlaaad: no
Jan 22 07:21:00 sandalle swoolley, yeah, I just thought of that too. ;/
Jan 22 07:21:36 afrayedknot you have to list the whole ugly query
Jan 22 07:21:36 sandalle alley_cat, they don't look too good sitting there, though, that's why I'm wondering if we should "fix" most of them?
Jan 22 07:22:08 sandalle http://bugs.sourcemage.org/buglist.cgi?query_format=specific&order=relevance+desc&bug_status=__open__&product=New+Spells&content=
Jan 22 07:22:18 alley_cat sandalle: i agree they don't look good, but real bugs just have a much higher priority
Jan 22 07:22:39 sandalle alley_cat, true, and we have plenty of them to keep us busy
Jan 22 07:22:48 sandalle alley_cat, how shall we target all the bugs?
Jan 22 07:23:01 sandalle bug days? post to sm-discuss asking gurus to look at the bug list if they have extra time?
Jan 22 07:23:02 benoit organize bug squashing parties? :-)
Jan 22 07:23:25 * vlaaad thinks it would be a good idea
Jan 22 07:23:29 alley_cat bug days could help a lot, but personally i just don't have the time for such a thing till middle of march
Jan 22 07:23:45 vlaaad but it may be difficult to reunit a lot of people in the same time
Jan 22 07:23:48 sandalle The problem I see with bug days is timezones, much as we have for meetings. :(
Jan 22 07:23:50 alley_cat so if anyone wants to organize such a thing go ahead
Jan 22 07:23:57 sandalle Any volunteers? ;)
Jan 22 07:24:39 alley_cat i don't think timezones are such a big problem as long as some grimoire gurus are there during all times
Jan 22 07:24:49 sandalle alley_cat, good point
Jan 22 07:24:51 alley_cat since our gurus are spread out over timezones too that should work
Jan 22 07:25:30 sandalle alley_cat, do you think we can get the list down to 100 bugs by April?
Jan 22 07:25:45 afrayedknot or june/july
Jan 22 07:25:49 sandalle alley_cat, or do any of the Grimoire items need more time?
Jan 22 07:26:02 alley_cat i think 100 might be a reasonable target
Jan 22 07:26:04 sandalle afrayedknot, right. :) I'm just seeing which teams/items need more time
Jan 22 07:26:53 alley_cat the biggest thing besides bugs in the grimoire is the spell lowercasing, we need volunteers to tackle perl-cpan and php-pear, otherwise we won't get that done
Jan 22 07:26:57 sandalle Anyone have complaints about upping the bug limit to 100?
Jan 22 07:27:23 sandalle alley_cat, I was planning on doing that one of these weekends, as I'd like it all done several months before the deadline so that the old names can be removed before 1.0
Jan 22 07:27:45 alley_cat sandalle: that would be great
Jan 22 07:27:53 * benoit notices Xdialog as a spell name ...
Jan 22 07:28:01 sandalle benoit, yeah, there's quite a few. :(
Jan 22 07:28:08 alley_cat the ones that conflict with other spell names should just get a perl-/php- prefix
Jan 22 07:28:19 sandalle alley_cat, what about hermes/Hermes?
Jan 22 07:28:41 afrayedknot 405 current (some of which may be the depreciated spell
Jan 22 07:28:53 alley_cat or a -perl/-php suffix, i don't really care which one
Jan 22 07:29:07 sandalle I'd prefer the prefix, myself
Jan 22 07:29:34 qnr prefix would seem to make more sense
Jan 22 07:29:37 sandalle alley_cat, how're config_query and variables doing?
Jan 22 07:29:58 alley_cat i don't really know, is there an easy way to find out?
Jan 22 07:30:22 vlaaad alley_cat : devel and mail sections (unmaintained) would need some help...(respectively 17 and 15 opened bugs)
Jan 22 07:30:28 afrayedknot 343 spells need lowercasing
Jan 22 07:30:30 afrayedknot (btw)
Jan 22 07:30:52 sandalle afrayedknot, probably a grep for query/read and remove config_query
Jan 22 07:31:06 sandalle afrayedknot, as for variables, devel Sorcery warns on a few (OPTS and DEBUG, I think)
Jan 22 07:31:20 vlaaad SweetPotato : yes, that's what I did in my section
Jan 22 07:31:25 sandalle vlaaad, ;)
Jan 22 07:31:41 afrayedknot sandalle: if you see the warning, file a bug
Jan 22 07:31:47 afrayedknot actually that goes for anyone
Jan 22 07:31:55 sandalle afrayedknot, yep. :)
Jan 22 07:31:56 * vlaaad is still confusing between SweetPotato and sandalle
Jan 22 07:32:03 afrayedknot it actually warns on a whole wealth of things :)
Jan 22 07:32:18 sandalle afrayedknot, good. :) Is that only for devel, or is that a feature that will make it into test?
Jan 22 07:32:41 sandalle alley_cat, do you believe we can have our items for Grimoire done by July?
Jan 22 07:32:44 afrayedknot eventually, although i may turn it off
Jan 22 07:32:55 sandalle afrayedknot, a menu option, or just remove it?
Jan 22 07:33:09 neuron I think that spells which will made it to stable grimoire should not do any warning
Jan 22 07:33:23 alley_cat sandalle: can't really tell about the query and local variable stuff as i don't have any numbers on those right now
Jan 22 07:33:29 sandalle neuron, that would be prefered, yes. :)
Jan 22 07:33:30 alley_cat the rest should be fine
Jan 22 07:33:41 alley_cat except duplicate dependencies, i'd like to remove that from the targets
Jan 22 07:33:46 sandalle alley_cat, some of the packages have been updated for those, but not most
Jan 22 07:33:48 alley_cat as it's just not really important as they don't hurt
Jan 22 07:34:00 sandalle alley_cat, those are easy bugs, though. ;)
Jan 22 07:34:03 swoolley 3 collab, 4 devel, 1 doc, 1 gnome1-libs, 1 graphics, 2 graphics-libs, 1 net, 74 perl-cpan, 147 php-pear, 1 security, 1 utils, 1 x11-toolkits (capital spell names)
Jan 22 07:34:21 sandalle swoolley, thanks. :)
Jan 22 07:34:23 afrayedknot without lowercase alternatives?
Jan 22 07:34:34 swoolley didn't check
Jan 22 07:34:34 alley_cat sandalle: i'm not saying they shouldn't be done, just that it's not a feature i consider necessary for 1.0
Jan 22 07:34:36 swoolley but I could
Jan 22 07:34:57 sandalle alley_cat, I'm fine with removing it.
Jan 22 07:35:10 sandalle Anyone else want duplicates kept for 1.0 or shall we remove it?
Jan 22 07:35:20 alley_cat i'll add a link to the duplicate dependencies list to the guru handbook
Jan 22 07:35:44 alley_cat i updated that somewhat today
Jan 22 07:35:45 sandalle alley_cat, should it be somewhere else? I don't see it as being a part of a handbook
Jan 22 07:35:49 sandalle alley_cat, thanks. :)
Jan 22 07:36:14 alley_cat i still plan to do a full rewrite of it as i don't think it's very well structured right now, but that will have to wait till march
Jan 22 07:36:30 sandalle alley_cat, k
Jan 22 07:36:46 sandalle Any other comments/questions for the Grimoire Team?
Jan 22 07:36:54 neuron yes
Jan 22 07:37:20 neuron We do have libaccount in each grimoire, wouldn't be better to make from it spell ?
Jan 22 07:37:37 sandalle neuron, why would we want to do that?
Jan 22 07:38:14 swoolley http://seth.positivism.org/files/searched (duplicate dependencies)
Jan 22 07:38:16 neuron I think it would be easier to synchronize, we could say that the spell will be in stable sorcery and only there
Jan 22 07:38:17 afrayedknot its a grimoire library
Jan 22 07:38:30 afrayedknot it needs to be updated for other spells
Jan 22 07:38:36 sandalle neuron, we can control that via the grimoire
Jan 22 07:38:58 afrayedknot i think the problem he's citing is that theres 4 copies that all have to be in sync
Jan 22 07:39:02 * Kinetix (~ajc@S01060050bf065491.vn.shawcable.net) has joined #smgl-admin
Jan 22 07:39:04 alley_cat i don't want that as a spell, the only disadvantage of having it in the grimoire is that people need to remember to integrate it to z-rejected now
Jan 22 07:39:10 neuron Right now we do have four separate ... yea afrayedknot :)
Jan 22 07:39:22 * sandalle gives voice to Kinetix
Jan 22 07:39:42 neuron It could be made 'vital' spell for sorcery, so it won't be dispelled
Jan 22 07:39:46 afrayedknot but, not everyone will update the spell everytime it needs updating
Jan 22 07:39:55 alley_cat i already added that to the guru handbook today
Jan 22 07:39:56 afrayedknot users i mean
Jan 22 07:40:06 sandalle neuron, making it a spell would just make it harder to keep in sync for users
Jan 22 07:40:07 alley_cat it's slightly more work for grimoire gurus, but much nicer for users than a spell
Jan 22 07:40:49 neuron Ok, you're grimoire guru, you know :)
Jan 22 07:40:55 afrayedknot maybe we could have a script to pull from devel into every other grimoire or something
Jan 22 07:41:53 sandalle afrayedknot, I don't think it's updated often enough to warrant a script
Jan 22 07:42:06 afrayedknot shrug okay
Jan 22 07:42:17 neuron So the devel grimoire libaccount is the master one ? (and accounts and groups files)
Jan 22 07:42:23 alley_cat neuron: yes
Jan 22 07:42:31 neuron Does not p4 support symlinks ?
Jan 22 07:42:43 dmlb2000|mobile neuron, nope
Jan 22 07:42:49 neuron ok, no more questions
Jan 22 07:43:49 sandalle Any others for Grimoire?
Jan 22 07:46:34 sandalle Alright then, thanks Arwed for the Grimoire report. :)
Jan 22 07:46:47 sandalle Next up is our ISO Team
Jan 22 07:46:57 vlaaad hum
Jan 22 07:47:05 vlaaad Iv'e got a question
Jan 22 07:47:06 sandalle I'll post what we have queued up for it along with Kevin's comments, and then benoit has some items as well
Jan 22 07:47:08 sandalle vlaaad, yes?
Jan 22 07:47:12 vlaaad sorry
Jan 22 07:47:26 vlaaad just a specifc question for one spell
Jan 22 07:47:26 swoolley doc/rtf2latex2eUnix utils/Linux-PAM x11-toolkits/wxPython security/NetPipe (capital spells outside of perl-cpan php-pear that need conversion)
Jan 22 07:47:34 vlaaad that is not in my section : openoffice
Jan 22 07:47:54 vlaaad currently everyone use openoffice-bin as far as I know
Jan 22 07:48:29 vlaaad openoffice seems indeed very painful to compile and it is not updated since 1.1.0 (we're at 1.1.4)
Jan 22 07:48:49 vlaaad so will the "compiled" spell be kept ??
Jan 22 07:49:02 sandalle vlaaad, I'd like to keep the compile option
Jan 22 07:49:13 sandalle vlaaad, but it needs some work (which I've tried a bit)
Jan 22 07:49:26 vlaaad I think the annoying problem is currently, people who watch at distrowatch.com will see openoffice as 1.1.0 version :-/
Jan 22 07:49:40 sandalle vlaaad, more incentive for us to fix our openoffice package. :)
Jan 22 07:50:02 vlaaad openoffice is a very important software I think
Jan 22 07:50:12 vlaaad ok :)
Jan 22 07:50:13 Asha`man vlaaad: I'm working on it now
Jan 22 07:50:25 Asha`man though I won't guarantee success
Jan 22 07:50:35 vlaaad I took a look at gentoo ebuild, it seemed rather complicated
Jan 22 07:50:47 Asha`man it is
Jan 22 07:51:04 Asha`man it seems every version bump, I have to re-write the damn spell
Jan 22 07:51:07 * e-type cheers for Asha`man
Jan 22 07:51:29 vlaaad Asha`man : Should the spell work for many languages or only english ??
Jan 22 07:52:01 * neuron thinks that the same goes for mozilla ...
Jan 22 07:52:38 Asha`man vlaaad: I hope to add multiple languages eventually, it could be done using ncurses for a menu (optionally of course) unfortunately, I don't know how to use ncurses
Jan 22 07:53:04 vlaaad you can take a look at kde-18n spell
Jan 22 07:53:10 sandalle Shall we continue this discussion in the proper bugs? ;)
Jan 22 07:53:22 Asha`man sandalle: sure
Jan 22 07:53:24 vlaaad sandalle : yes :)
Jan 22 07:53:36 sandalle :)
Jan 22 07:53:44 sandalle And back to the ISO:
Jan 22 07:53:54 benoit yes!
Jan 22 07:54:01 sandalle Our RoadMap has only a few items:
Jan 22 07:54:03 sandalle * ISO installs on IDE [DONE], SCSI, RAID [DONE], and [LVM] systems
Jan 22 07:54:03 sandalle * ISO has LSB/FHS hierarchy and correct file layouts (/etc/services, /etc/group, etc.) [FHS]
Jan 22 07:54:10 sandalle and the ISO RoadMap lists a few more:
Jan 22 07:54:22 sandalle Principles of organization / Critical features
Jan 22 07:54:22 sandalle * (Development) ISOs are to be realized in a 2 weeks iteration:
Jan 22 07:54:22 sandalle * Formal example:
Jan 22 07:54:22 sandalle Week 1: bug fixes + integration of scripts novelty + (compilation) -> friday release
Jan 22 07:54:22 sandalle Week 2: intensive testing + bug report (fixes)
Jan 22 07:54:22 sandalle * Architecture issues
Jan 22 07:54:22 sandalle With the multiplication of kernels and supported arch, it is necessary to have 1.0 ISOs:
Jan 22 07:54:22 sandalle Architecture Kernel glibc
Jan 22 07:54:22 sandalle x86 2.4 non-NPTL
Jan 22 07:54:22 sandalle x86 2.6 NPTL
Jan 22 07:54:22 sandalle ppc 2.6 non-NPTL
Jan 22 07:54:22 sandalle * Scripts enhancements:
Jan 22 07:54:22 sandalle Several scripts are to be improved, please append your name to the list for task acceptance:
Jan 22 07:54:22 sandalle Task Name General description your name here
Jan 22 07:54:22 sandalle ISO creation scripts An easy and more or less automated way to create ISO since there are multiple concurrent attempts at this, TL and PL must decide
Jan 22 07:54:22 sandalle mkinitrd Providing SMGL a real initrd creation script, for installation and general use your name here
Jan 22 07:54:22 sandalle Installation script General installation script maintenance and development your name here
Jan 22 07:54:22 sandalle LVM Integrated LVM support in installation Unet
Jan 22 07:54:22 sandalle hwd Hardware detection script, installation and system Unet
Jan 22 07:54:22 sandalle This is a rough overview of the features we need to have in 1.0.
Jan 22 07:54:22 sandalle Development and post 1.0 planning
Jan 22 07:54:22 sandalle If time allows, it is necessary to work on additional features that could make it in 1.0 but are not considered critical.
Jan 22 07:54:22 sandalle * x86_64 support : as noted by Benoit, sorcery needs to be hacked for bi-arch support [1].
Jan 22 07:54:22 sandalle * sparc support
Jan 22 07:54:22 sandalle * graphical installer: using framebuffer and gtk+2 we can hack a very lightweight graphical frontend - it needs ruby, perl or python scripting though
Jan 22 07:54:29 sandalle (http://wiki.sourcemage.org/index.php?page=ISO+RoadMap)
Jan 22 07:54:58 sandalle most of what's on our RoadMap is done
Jan 22 07:55:08 sandalle but then we had a very short list. ;)
Jan 22 07:55:23 sandalle for 1.0 the main features are:
Jan 22 07:55:47 sandalle * 3 ISOS: x86 2.4 non-NPTL, x86 2.6 NPTL, ppc 2.6 non-NPTL
Jan 22 07:55:55 sandalle (I believe the x86 2.6 is supposed to be non-NPTL as well)
Jan 22 07:56:04 sandalle * hardware detection
Jan 22 07:56:16 sandalle * easy generation of ISOs from a script
Jan 22 07:57:07 sandalle benoit, has been making our ISOs for a while now and his process works, but it seems to be time-intensive and has been a modification of older scripts for a few years now
Jan 22 07:57:18 sandalle Karsten (BearPerson) is working on a newer model, but that is taking time
Jan 22 07:57:25 benoit right.
Jan 22 07:57:47 afrayedknot benoit has been reluctant to commit his work to perforce
Jan 22 07:57:59 benoit as our Team Leader is not there, i will talk only on my work
Jan 22 07:58:19 benoit so far, i did not get much feedback on the ISO
Jan 22 07:58:29 sandalle afrayedknot, I believe he was only commiting working code, but now knows that in devel he can commit what he has without worry of breaking it
Jan 22 07:58:35 afrayedknot quite a few bugs were filed...
Jan 22 07:58:37 benoit that's why I think it would be a good idea to organize a beta test team
Jan 22 07:58:38 sandalle benoit, there are some bugs filed. :)
Jan 22 07:58:58 benoit sandalle: right.
Jan 22 07:59:09 swoolley I'd like to give feedback on an iso creation process, but I haven't seen one with public instructions.
Jan 22 07:59:09 * vlaaad wants to join the beta test team
Jan 22 07:59:49 benoit swoolley: right. because producing a nice script is a recent item added to our todo list
Jan 22 08:00:05 afrayedknot since Unet and BearPerson are both out on personal business it might be best to just hold off on this...
Jan 22 08:00:08 afrayedknot (IMO)
Jan 22 08:00:46 swoolley benoit: I thought it was an old item
Jan 22 08:01:02 benoit anyway, i'm reconsidering how i've build ISO so far
Jan 22 08:01:27 benoit that's why i think sorcery install_root features would simplify things
Jan 22 08:01:32 swoolley true
Jan 22 08:01:42 benoit since it was not working before, i did not use it
Jan 22 08:01:45 sandalle swoolley, Karsten has his here: http://wiki.sourcemage.org/index.php?page=Yet+Another+ISO+Making+Guide and http://wiki.sourcemage.org/index.php?page=New+ISO+generation+scripts and I don't see one for the current ISO on the wiki (I might be blind)
Jan 22 08:01:49 benoit but i haven't check for at least a year
Jan 22 08:02:12 sandalle afrayedknot, we can still get other's feedback
Jan 22 08:02:37 dmlb2000|mobile yeah like mine ;)
Jan 22 08:02:46 swoolley sandalle: ah, ok
Jan 22 08:02:51 sandalle swoolley, it was an old item, but hasn't had any work done on it until recently
Jan 22 08:03:10 sandalle benoit, can you make sure that you keep your Perforce repo up-to-date so we can see what's going on?
Jan 22 08:03:15 benoit so, while reconsidering this, i think it would be good to have a working that would show something new
Jan 22 08:03:28 sandalle benoit, a working ISO that shows what's new or working scripts?
Jan 22 08:03:30 benoit sandalle: yes. that's what i said.
Jan 22 08:03:42 sandalle benoit, just checking. :)
Jan 22 08:04:09 benoit right now, i published a {x86,ppc,x86_64} iso on http://download.sourcemage.org/
Jan 22 08:04:26 benoit i was waiting for at least a review for each ISO
Jan 22 08:04:29 afrayedknot note, these are test isos right?
Jan 22 08:04:35 benoit but so far, NO ONE has tested the x86_64 iso.
Jan 22 08:04:40 afrayedknot neuron has
Jan 22 08:04:42 neuron As I was able to test benoit's iso, I liked the (for me visible) changes
Jan 22 08:04:45 benoit yes. it's highly experimental ISO.
Jan 22 08:05:00 benoit neuron: thanks for your reports btw :-)
Jan 22 08:05:05 sandalle benoit, not many have x86_64 ;(
Jan 22 08:05:13 benoit sandalle: I know ...
Jan 22 08:05:22 neuron I was testing it on amd64, and it has problems with that part, but I would suspect that it will work on x86/32 well
Jan 22 08:05:26 benoit I sent an email to Unet about this ISO
Jan 22 08:05:34 neuron benoit: Thank you for your work
Jan 22 08:05:47 benoit this is experimental since, for instance
Jan 22 08:05:55 benoit - i use gcc 3.4.3 to build it
Jan 22 08:05:59 sandalle benoit, perhaps we should announce test ISOs?
Jan 22 08:06:03 swoolley I can test x86_64
Jan 22 08:06:04 benoit - i use a NPTL glibc
Jan 22 08:06:06 sandalle benoit, did you use gcc 3.4 on all the ISOs?
Jan 22 08:06:09 low i can test too
Jan 22 08:06:21 benoit - i use a new handling mecanism for optional depends
Jan 22 08:06:27 sandalle I believe people wanted non-NPTL as you can upgrade to NPTL, but not downgrade to non-NPTL
Jan 22 08:06:32 benoit but i think this feature is overlapping sorcery's features
Jan 22 08:06:54 benoit so, that's why i call it "experimental"
Jan 22 08:07:05 benoit this is just to show some features that could be in official ISO
Jan 22 08:07:18 benoit and that's why I'm requesting some feedback (on sm-discuss probably)
Jan 22 08:07:28 sandalle Sorcery should be able to be used for most of the building of the ISO, which is what Karsten is playing with
Jan 22 08:07:29 neuron And it has nice 'Intorduction' message :)
Jan 22 08:07:34 sandalle benoit, what features are making it onto the ISO?
Jan 22 08:07:55 benoit sandalle: i've addded evms to handle RAID/LVM alltogether for instance
Jan 22 08:07:56 sandalle benoit, mostly bugfixes? any addons?
Jan 22 08:08:21 sandalle benoit, how large is evms?
Jan 22 08:08:25 benoit right now, this is no completed, but i think it would be nice to show some work :-)
Jan 22 08:08:39 benoit evms is quite small.
Jan 22 08:08:41 Asha`man question, is there a limit to the number of source_urls we want to list in a spell?
Jan 22 08:08:46 sandalle benoit, testing is good. :)
Jan 22 08:08:49 afrayedknot Asha`man: no
Jan 22 08:09:02 benoit and it has the squashfs thing (to make mini CD);
Jan 22 08:09:21 Asha`man afrayedknot: thanks
Jan 22 08:09:37 sandalle benoit, the ISO has squashfs or it uses squashfs?
Jan 22 08:09:43 benoit yes.
Jan 22 08:10:10 sandalle we're also reviewing which scripts to use and if they can be merged into one
Jan 22 08:10:15 neuron I don't like the squashfs personally too much, since it's hard to mount on other computers :)
Jan 22 08:10:16 benoit it has squashfs and uses squashfs, thus reducing the ISO size
Jan 22 08:10:17 sandalle as both can use some work
Jan 22 08:10:26 Asha`man I'm guessing I have about 50 mirrors listed for OO right now
Jan 22 08:10:57 afrayedknot we could always make that a sorcery thing like for GNU and apache
Jan 22 08:11:08 sandalle with benoit's being known to work and Karsten's using Sorcery as much as it can
Jan 22 08:11:30 sandalle benoit, what kind of feedback are you looking for on your ISOs?
Jan 22 08:11:46 sandalle benoit, bugfix works, doesn't work? like the new feature? evms is great/sucks? etc.?
Jan 22 08:11:59 benoit yep. any feedback in fact
Jan 22 08:12:08 benoit but i'd like to have some detailed thing
Jan 22 08:12:13 sandalle benoit, do you have the process you use to build the ISO written down on our wiki? (a quick look didn't show up to me)
Jan 22 08:12:13 benoit "it works" is not a feedback
Jan 22 08:12:26 benoit "i have installed a bootable system" is
Jan 22 08:12:31 benoit "it does not work" is not a feedback
Jan 22 08:12:52 benoit "it says that i have no hard drive" is a feedback (you'll get that on the x86_64 iso i think°
Jan 22 08:12:53 vlaaad in this way, I suggest to make a form for every volunteer
Jan 22 08:13:28 sandalle vlaaad, what kind of form?
Jan 22 08:13:30 vlaaad like I told you benoit, maybe following the menus/sub-menus scheme
Jan 22 08:13:42 benoit vlaaad: yes. that's a good suggestion.
Jan 22 08:14:18 vlaaad with the possibility to note each suggestion for every step of the installation
Jan 22 08:14:49 sandalle benoit, Kevin is working on LVM, does your evms work well and have menu options?
Jan 22 08:15:04 vlaaad we should then determine which format we want ?
Jan 22 08:15:13 benoit sandalle: in fact, i replace the LVM/RAID menu by evms to see how it could work
Jan 22 08:15:32 benoit i've tested by creating a test configuration (using LVM/RAID0/RAID1/RAID5)
Jan 22 08:15:40 vlaaad I thought about an HTML form that could be put on the smgl website, it could be better to store all the tests
Jan 22 08:15:52 benoit but ... since i created / in an EVMS volume ... i cannot boot the installed system.
Jan 22 08:15:55 sandalle vlaaad, what is wrong with using bugzilla to file bugs?
Jan 22 08:16:15 sandalle benoit, did you setup the LVM initrd? I have a script on my site (which I think was listed in the LVM bug)
Jan 22 08:16:25 vlaaad sandalle : well I think it is not appropriate if you want to make a complete review of the ISO
Jan 22 08:16:33 * youpi_ (~youpi@82.230.129.61) has joined #smgl-admin
Jan 22 08:16:42 vlaaad for individual bug it is good
Jan 22 08:16:48 sandalle vlaaad, reviews should probably be e-mailed, but we want bugs to be filed through bugzilla
Jan 22 08:17:07 sandalle are there any comments/questions about the ISO?
Jan 22 08:17:16 vlaaad yes
Jan 22 08:17:17 dmlb2000|mobile Yup
Jan 22 08:17:29 sandalle go ahead
Jan 22 08:17:39 vlaaad what do you think so about a beta-tester team for the ISO ??
Jan 22 08:17:55 sandalle vlaaad, I don't think we need more teams
Jan 22 08:18:03 sandalle our current ones are stretched too thin for my liking
Jan 22 08:18:18 vlaaad well no really a new team, it could be just added to iso team
Jan 22 08:18:29 benoit right now, we don't have enought beta testers... so, they are needed.
Jan 22 08:18:35 sandalle vlaaad, I believe we have volunteers for ISO testing
Jan 22 08:18:45 vlaaad so ISO-team will have permanent testers
Jan 22 08:18:47 benoit sandalle: we don't have enought ...
Jan 22 08:18:47 sandalle I don't believe a separate team is needed, but should rather be the ISO team
Jan 22 08:19:01 vlaaad yes that's what I said :p
Jan 22 08:19:20 sandalle in that the ISO Team makes sure their ISO works for them, then posts the test ISOs to download.sourcemage.org for others to test and put out a call for testers
Jan 22 08:19:22 vlaaad but my english is not always good ;-)
Jan 22 08:19:23 sandalle then they get testers
Jan 22 08:19:33 benoit but we need to organize a bit i think, to have consistent reports... or it would not be usable
Jan 22 08:19:39 sandalle benoit, we never have enough. :(
Jan 22 08:20:15 sandalle I believe all that's needed for testing is to make sure it boots on your machines, post the test ISO to SMGL, post a message asking for testers, and then wait a bit
Jan 22 08:20:29 sandalle since we don't have many people with machines to install on (SMGL hardly ever needs reinstalling, if ever)
Jan 22 08:20:50 benoit sandalle: but ... problem is that i don't have all the hardware to make those test :-(
Jan 22 08:21:17 sandalle benoit, find someone who does and put them on the ISO team. :)
Jan 22 08:21:18 vlaaad sandalle : so no review forms or anything that could help the ISO devs ?? Even if I do it ?? :p
Jan 22 08:21:24 benoit that's why the ISO team would need some beta testers that would follow some guidelines to test.
Jan 22 08:21:39 sandalle vlaaad, I'm not against the forms, but all bugs should go through bugzilla (and any other medium you want)
Jan 22 08:21:50 vlaaad yes I agree with that
Jan 22 08:22:24 benoit sandalle: problem is that i'm not the cauldron team leader so i cannot decide myself or people here will think that i'm working on my own (maybe they have already...)
Jan 22 08:22:33 Kinetix vlaaad: On that thought, how would personal blogs work for that sort of thing in the next website? You do your review, post it to your blog, contact the section head and the section head could have it posted as a news story or linked into an ISO dev section...
Jan 22 08:22:35 vlaaad but I don't know if a bug is welcome, for "I find this colour ugly" or "You should explain better this step"...
Jan 22 08:22:53 sandalle vlaaad, bugzilla is our medium for all bugs, otherwise they get lost
Jan 22 08:23:10 sandalle benoit, we have a few people with PPC who can test, and others with x86_64
Jan 22 08:23:19 alley_cat Kinetix: we'll have a blog system on the new website?
Jan 22 08:23:20 sandalle benoit, and others who want to help can volunteer
Jan 22 08:23:44 sandalle benoit, decide yourself on what?
Jan 22 08:23:46 Kinetix alley_cat: I think it would really help enhance things...
Jan 22 08:23:54 alley_cat Kinetix: i'd like to see that :)
Jan 22 08:24:07 alley_cat i always though developer blogs could be nice for smgl
Jan 22 08:24:44 benoit sandalle: i cannot decide on adding new people to the Cauldron Team.
Jan 22 08:24:59 sandalle benoit, right, but people can still test without being on the ISO team
Jan 22 08:25:17 vlaaad Kinetix : well I don't really know how blogs works but that's sounds good :-)
Jan 22 08:25:19 sandalle So, we have two test ISO types out: old-style scripts and the new custom scripts for testing
Jan 22 08:25:37 sandalle both should be tested, though the new ones should have a new release before more testing is done (there are bugs filed already)
Jan 22 08:26:00 sandalle the old-style (benoit-test*) ISOs are out for testing as well and probably need an announcement
Jan 22 08:26:33 sandalle vlaaad would like to setup (which Kinetix can get into more) a way for people to converse about development (besides sm-discuss I'm guessing?)
Jan 22 08:27:33 sandalle and the ISO may need more time than April to be ready (auto-generation scripts and all), correct?
Jan 22 08:27:58 vlaaad sandalle : well it could be good for ISO, as it is a quite long process, that cannot being reduced imho to a single filed bug...
Jan 22 08:28:07 benoit sandalle: we would need a meeting before to get a decision on what would be on the 1.0 before answering :-)
Jan 22 08:28:36 sandalle benoit, I'll grab Kevin when he gets back and try to setup a meeting the ISO team can attend
Jan 22 08:28:43 sandalle Any other questions/comments for the ISO?
Jan 22 08:28:48 vlaaad yes :)
Jan 22 08:28:59 sandalle We currently have two branches of development, but will be focusing on one in the next week or two
Jan 22 08:29:01 sandalle vlaaad, ? ;)
Jan 22 08:29:14 vlaaad well, I discussed about that on sm-discuss
Jan 22 08:29:22 dmlb2000|mobile Yeah, I do
Jan 22 08:29:32 vlaaad shall we expect reiser4 support in the next ISO ??
Jan 22 08:29:46 vlaaad I know this is just a feature, but...
Jan 22 08:29:50 afrayedknot why dont you file a bug requesting it?
Jan 22 08:29:53 sandalle vlaaad, have you filed a bug? ;)
Jan 22 08:29:53 dmlb2000|mobile dmlb2000|server Jan 22 08:29:55 dmlb2000|mobile dmlb2000|server Jan 22 08:29:58 sandalle dmlb2000|mobile, what's your Q?
Jan 22 08:29:58 vlaaad :p
Jan 22 08:30:02 vlaaad you're right
Jan 22 08:30:03 dmlb2000|mobile BootX for old world macs on the iso
Jan 22 08:30:14 afrayedknot we will never run out of bug numbers
Jan 22 08:30:18 sandalle vlaaad, reiser4 would require a non-vanilla kernel
Jan 22 08:30:18 vlaaad sorry :/
Jan 22 08:30:23 dmlb2000|mobile dmlb2000|server Jan 22 08:30:29 dmlb2000|mobile benoit, you know anything about bootX
Jan 22 08:30:42 lace just wonder, if there are some plans with support of crypto file system on install iso?
Jan 22 08:30:48 lace http://bugs.sourcemage.org/show_bug.cgi?id=7448
Jan 22 08:30:58 sandalle dmlb2000|mobile, can you file a bug for that too, when you get the chance?
Jan 22 08:31:03 dmlb2000|mobile okay
Jan 22 08:31:06 low vlaaad, hmmm reiser4 isn't even stable nor included in kernel.
Jan 22 08:31:08 dmlb2000|mobile will do
Jan 22 08:31:24 benoit dmlb2000|mobile: yep. i've tested with BootX on a friend's machine once.
Jan 22 08:31:28 vlaaad low : it is
Jan 22 08:31:38 vlaaad included in the grimoire
Jan 22 08:31:46 sandalle vlaaad, it's only in test grimoire, I believe
Jan 22 08:31:53 dmlb2000|mobile dmlb2000|server Jan 22 08:32:03 vlaaad so if people wants a rescue CD, they have to download a gentoo or yoper ISO
Jan 22 08:32:04 sandalle benoit, dmlb2000|mobile got your 0.9.3 PPC ISO on an old-world Mac G3. ;)
Jan 22 08:32:10 vlaaad I don't kile this idea
Jan 22 08:32:12 sandalle vlaaad, why's that?
Jan 22 08:32:16 vlaaad s/kile/like
Jan 22 08:32:24 sandalle vlaaad, I use our ISO for rescue every now and then just fine
Jan 22 08:32:48 afrayedknot i think he means if someone tries to install reiser4 on their own
Jan 22 08:32:54 vlaaad if they're native system is in reiser4 format, they will be unable to mount it with the current ISO
Jan 22 08:33:00 low vlaaad, no it's not. it would have been included in kernel otherwise imho
Jan 22 08:33:02 afrayedknot why someone would do that /now/ i dont know
Jan 22 08:33:28 sandalle afrayedknot, testing? weird? ;)
Jan 22 08:33:31 neuron There are more crazy things in the world I would say :)
Jan 22 08:33:44 sandalle vlaaad, reiser4 is still very much experimental, but you can file a bug and it may get put in, though it'd be low priority
Jan 22 08:34:12 sandalle and wouldn't be on all ISOs
Jan 22 08:34:14 vlaaad sandalle : ok, I'll do so :)
Jan 22 08:34:30 sandalle What do people think of adding BootX to the PPC ISO?
Jan 22 08:34:40 neuron So filling bugs against "benoit's" is is now legal ?
Jan 22 08:34:42 sandalle it'd only be for old-world Macs and is rather small (IIRC)
Jan 22 08:34:50 sandalle neuron, it was never illegal...
Jan 22 08:34:52 benoit sandalle: go for BootX :-). it does require no work :-)
Jan 22 08:34:56 sandalle just make sure you specify which ISO you tested
Jan 22 08:35:01 dmlb2000|mobile dmlb2000|server Jan 22 08:35:07 benoit neuron: good point :-)
Jan 22 08:35:10 sandalle benoit, dmlb2000|mobile, we can talk more on the bug once it's filed then. :)
Jan 22 08:35:37 dmlb2000|mobile sandalle, yeah most of what I have to say is to benoit anyway ;)
Jan 22 08:35:42 neuron I hope that noone before filled that you can install sourcemage completely without kernel :)
Jan 22 08:35:54 sandalle neuron, ?
Jan 22 08:36:02 sandalle Any other questions/comments for the ISO?
Jan 22 08:36:16 lace crypto file system ?
Jan 22 08:36:45 neuron sandalle: You can just skip the option to install kernel, as it happened to me :)
Jan 22 08:36:52 sandalle benoit, hav eyou updated the ISOs on sourcemage.org? I get a 404 for http://download.sourcemage.org/iso/benoit-testing-use-only-at-own-risk/smgl-0.9.4-test2-i486.iso
Jan 22 08:36:58 sandalle lace, I'd be fine with crypto
Jan 22 08:37:10 benoit sandalle: no. i just upload this morning.
Jan 22 08:37:12 sandalle neuron, I thought you had to choose either copy or compile? did you file a bug? ;)
Jan 22 08:37:17 sandalle benoit, still uploading?
Jan 22 08:37:25 Kinetix sandalle: No 404 for me here
Jan 22 08:37:27 sandalle lace, would that need menu entries attached as well?
Jan 22 08:37:28 benoit sandalle: no. done.
Jan 22 08:37:54 neuron sandalle: Not yet, that's why I asked if I can :)
Jan 22 08:38:01 lace probably yes
Jan 22 08:38:01 sandalle I get the 404 with wget, but if I click it works fine
Jan 22 08:38:11 sandalle benoit, did you setup the proper permissions? read by all?
Jan 22 08:38:31 neuron That wouldn't be 404, but rather 5xx I think ...
Jan 22 08:38:31 sandalle neuron, wait a sec, I'll make a target for -test2 ISO
Jan 22 08:38:41 benoit sandalle: hummm ... i dont setup anything special. just copy it with rsync
Jan 22 08:38:41 neuron sandalle: sure
Jan 22 08:38:42 sandalle neuron, true
Jan 22 08:38:57 sandalle lace, how much more space does cryptosetup take up?
Jan 22 08:39:39 Kinetix I can wget the test2 iso
Jan 22 08:40:07 benoit lace: if you can explain what's needed for cryto fs, i'll be happy to add it on the next ISO
Jan 22 08:40:31 sandalle Anyone have problems with crypto?
Jan 22 08:40:36 sandalle Any other questions/comments?
Jan 22 08:41:05 lace cryptosetup in my cache: 24922 B, I add more info to bug info
Jan 22 08:41:21 benoit but once again, we need to have a clear statement on what should/could/will be on the official 1.0 ISO
Jan 22 08:41:49 afrayedknot we also need to decide who's scripts we're using, but like sandalle said we'll do that in another meeting
Jan 22 08:42:07 benoit indeed
Jan 22 08:42:38 sandalle we need Kevin and Karsten there as well
Jan 22 08:42:41 sandalle so we can all talk about it
Jan 22 08:42:58 benoit and all the other team member
Jan 22 08:43:04 benoit de_Moya for instance
Jan 22 08:43:16 afrayedknot your fan club?
Jan 22 08:43:24 sandalle I haven't seen him talk lately, but I've seen him online
Jan 22 08:43:25 benoit no.
Jan 22 08:43:28 sandalle and elshaa (IIRC)
Jan 22 08:43:40 benoit just to remind people that there are other member in the Cauldron Team :-)
Jan 22 08:43:47 benoit and elshaa right
Jan 22 08:43:56 sandalle benoit, :)
Jan 22 08:44:21 sandalle Any other questions?
Jan 22 08:44:21 benoit something, i think i'm the only one .... but it's plain wrong :-)
Jan 22 08:44:43 sandalle benoit, the rest are working on other parts: generation, hardware detection, etc.(Part II)