**** BEGIN LOGGING AT Mon Jan 09 02:59:57 2012
Jan 09 10:55:25 <ant_work> hi bluelightning
Jan 09 10:55:33 <bluelightning> hi ant_work
Jan 09 10:55:49 <bluelightning> how was your weekend?
Jan 09 11:06:57 <ant_work> sorry, phone is always ringing this morinig...
Jan 09 11:07:24 <ant_work> weekend was not bad, windy and sunny here in Italy
Jan 09 11:08:31 <ant_work> just some minor disappointment wrt oe project ;)
Jan 09 11:12:41 <Jay7> hi ppl
Jan 09 11:12:48 * Jay7 is at home finally
Jan 09 11:12:54 <ant_work> hi Jay7
Jan 09 11:26:40 <bluelightning> ant_work: oh? what disappointed you especially?
Jan 09 11:26:45 <bluelightning> hi Jay7
Jan 09 12:08:57 <ant_work> well, I got somehow the feeling that people are satisfied with the status of meta-oe
Jan 09 12:16:03 <bluelightning> I'm not..
Jan 09 12:16:09 <bluelightning> FWIW
Jan 09 12:16:48 <ant_work> I see a strange inertia from the Angstrom and SHR maintainers lately
Jan 09 12:24:20 <ant_work> it looks like someone feels personally injured when you ask to remove smthg from meta-oe
Jan 09 12:25:02 <ant_work> and I'm not talking just about last thread about xserver-common
Jan 09 12:25:27 <ant_work> the assumption is: I'd remove i when the one in oe-core will be better
Jan 09 12:25:36 <ant_work> jeez, it is already better than old cruft!
Jan 09 12:26:43 <ant_work> so, to nail it down, I can't understand why Koen or Martin haven't yet properly unified xserver
Jan 09 12:27:51 <ant_work> and why haven't they proposed systemd for oe-core but just hinted, 'hey, our is bigger
Jan 09 12:28:10 <ant_work> ...anyway...
Jan 09 12:28:47 <Jay7> :)
Jan 09 12:29:19 <Jay7> that's what I've talked some time ago.. layers are good from one side but very bad from other
Jan 09 12:29:47 <Jay7> so instead of fixing things people are reinventing own wheel
Jan 09 12:30:42 <ant_work> yes, but they can't expect us to search their distro layers to find the fixes
Jan 09 12:30:48 <Jay7> so only thing I can propose is create own set of recipes in meta-hh..
Jan 09 12:31:32 <ant_work> in fact we have only klibc recipes in meta-oe. we could even create a meta-klibc layer and host libs + recipes
Jan 09 12:32:04 <ant_work> as for xserver, I'll send a pull msg for oe-core wrt xinput and xinput calibrator
Jan 09 12:32:12 <ant_work> and that should be all
Jan 09 12:32:24 <ant_work> ah, wait, it is worst
Jan 09 12:32:36 <ant_work> it looks like the gcc in meta-oe is 'beter' for armv5
Jan 09 12:32:43 <ant_work> :/
Jan 09 12:32:52 <ant_work> 2 toolchains :/
Jan 09 12:34:30 <Jay7> yeah
Jan 09 12:35:05 <ant_work> on eis even failing with QT iirc :)
Jan 09 12:35:21 <ant_work> I suspect the one in meta-oe :]
Jan 09 12:37:01 <ant_work> you see, it is now clear oe-core as a much wider base of professional testers
Jan 09 12:37:22 <ant_work> it's just silly to duplicate the work in other layers...
Jan 09 12:37:40 <ant_work> *has
Jan 09 12:37:54 <Jay7> so people are creating own layer on top of oe-core for own work
Jan 09 12:39:14 <bluelightning> personally I think meta-oe should be split, it's trying to do too many things
Jan 09 12:40:19 <Jay7> as I understand meta-oe is base for migration from oe-classic to oe-core + layers
Jan 09 12:40:35 <Jay7> but nobody cares about recipes migration
Jan 09 12:40:45 <Jay7> florian was sad about this
Jan 09 12:40:49 <bluelightning> what most people want is the additional recipes, but they get a lot of unexpected baggage with it
Jan 09 12:41:10 <ant_work> Jay7: yes, is the base for that, to be able to build Angstrom images as console-image
Jan 09 12:41:14 <Jay7> bluelightning: you as TSC member may change this :)
Jan 09 12:41:15 <bluelightning> Jay7: it's happening, slowly
Jan 09 12:41:28 <bluelightning> Jay7: I can propose it, as anyone can
Jan 09 12:41:45 <bluelightning> I may have a little more leverage now though :)
Jan 09 12:41:59 <Jay7> sure :)
Jan 09 12:42:02 <ant_work> what I deeply dislike are the copies of *libc and other toolchain bits
Jan 09 12:42:09 <ant_work> this can be in -next
Jan 09 12:42:16 <ant_work> or in -old
Jan 09 12:42:32 <Jay7> people are frustrated imho
Jan 09 12:42:42 <bluelightning> yes, I knwo
Jan 09 12:42:50 <ant_work> only a little bit as for me ;)
Jan 09 12:42:52 <bluelightning> we could use more help in moving things across though...
Jan 09 12:43:21 <Jay7> ant_work: but some zaurus owners behind us are frustrated even more :)
Jan 09 12:43:31 <Jay7> they have no stable builds from 2007..
Jan 09 12:44:27 <ant_work> I don't know what to suggest: packaging feeds are provided by the distro usually
Jan 09 12:44:46 * bluelightning wonders if it's time for a handheld-focused distro again
Jan 09 12:44:47 <ant_work> which one should we consciensciously suggest?
Jan 09 12:45:07 <ant_work> or how you spell it
Jan 09 12:45:30 <bluelightning> I think that's correct spelling at least
Jan 09 12:45:36 <ant_work> :]
Jan 09 12:46:05 <ant_work> do Yocto provide any sort of feeds?
Jan 09 12:46:17 <ant_work> *Yocto people
Jan 09 12:46:18 <bluelightning> yocto isn't a distro
Jan 09 12:46:46 <bluelightning> we don't aim to provide functional images, just ones for testing, and only for the reference machines
Jan 09 12:47:16 <bluelightning> s/functional/fully-functional/
Jan 09 12:48:27 <ant_work> we could easily upload some *images* on ltg but I fear the space for feeds will be short
Jan 09 12:48:39 <Jay7> angstrom is no more focused on hh devices
Jan 09 12:49:07 <Jay7> shr is more like that we need
Jan 09 12:49:13 <Jay7> or sato
Jan 09 12:50:21 <ant_work> I'll continue to test distroless fwiw, this is a pre-requisite imho
Jan 09 12:55:37 <ant_work> btw, are tablet handhelds?
Jan 09 13:09:59 <ant_work> should we start to aim at some tablets potentially running linux? Too many proprietary drivers?
Jan 09 13:12:23 <Jay7> nvidia tegra
Jan 09 13:13:09 <Jay7> mer and meego have support for some models
Jan 09 13:14:23 <Jay7> anyway tablets have large screen and good hardware
Jan 09 13:16:08 <bluelightning> I dunno, I'd just like to have a distro that can accommodate all of the machines in meta-handheld when they are working
Jan 09 13:16:31 <bluelightning> some of those are stuck with older kernels, so cannot support systemd
Jan 09 13:16:44 <Jay7> how we aims to all Z line + HP iPAQ's as I understand
Jan 09 13:17:26 <Jay7> i.e. low memory and qvga/vga screen
Jan 09 13:17:45 <Jay7> and little nand (if any)
Jan 09 13:17:48 <bluelightning> right
Jan 09 13:17:56 <ant_work> makes sense
Jan 09 13:18:36 <Jay7> s/how/now/
Jan 09 13:18:39 <ant_work> I was reading <cs_nbp> florian: morning, my kernel compiled and I have a nBkProOs.img
Jan 09 13:19:00 <Jay7> btw
Jan 09 13:19:18 <Jay7> jlime people aims similar machines
Jan 09 13:19:28 <Jay7> hp jornada's and ben nanonote
Jan 09 13:20:01 <Jay7> may be good idea to work together
Jan 09 13:20:16 <Jay7> they have based on top of classic oe now
Jan 09 13:20:32 <ant_work> I think the basic recipe set exists today, just is badly split around layers
Jan 09 13:21:46 <ant_work> and for existing images, I started with core-image-sato but it's already too big for our purposes
Jan 09 13:22:11 <ant_work> I don't know if the xserver vs. kdrive did inflate soo much..
Jan 09 13:22:17 <ant_work> can't think so
Jan 09 13:22:42 <ant_work> I'd say that maybe the opie-image could be a good start point for users
Jan 09 13:23:04 <ant_work> then an X image, but which? gpe?
Jan 09 13:23:13 <Jay7> do opie have browser with JS now?
Jan 09 13:24:01 <ant_work> I'm not against Angstrom and SHR fwiw, just did not test with oe-core
Jan 09 13:24:50 <Jay7> jama is testing shr on his C3200 afaik
Jan 09 13:26:17 <ant_work> I dislike the idea of 'setup-scripts' playing with my bblayers ;)
Jan 09 13:30:53 <bluelightning> Jay7: old konqueror had rudimentary js support but nothing that would be really useful in a modern context
Jan 09 13:31:14 <Jay7> :(
Jan 09 13:31:42 <bluelightning> ant_work: gpe seems unmaintained unfortunately :/
Jan 09 13:31:57 <bluelightning> I could be wrong but that's the impression I get
Jan 09 13:33:25 <ant_work> heh
Jan 09 13:34:15 <bluelightning> opie is maintained but realistically there are things that will probably never get fixed; I can't imagine writing a new web browser for it for example
Jan 09 13:35:15 <bluelightning> in any case the trouble with these devices is they are really limited and most modern browsers are too memory and cpu-hungry for them
Jan 09 13:35:43 <ant_work> sadly true
Jan 09 13:39:29 <bluelightning> someone suggested NetSurf as a possibility, of course you'd need to build a new UI around it
Jan 09 13:42:18 <bluelightning> ah, no javascript, that's a shame...
Jan 09 14:06:06 <ant_work> I'd be satisfied with a media-player and some sort of mame at the beginning for a small user image
Jan 09 14:06:54 <ant_work> as for pims, I guess people uses mobile phones nowadays
Jan 09 14:11:44 <Jay7> ant_work: fbreader :)
Jan 09 14:12:09 <Jay7> don't forget about it :)
Jan 09 14:12:42 <Jay7> about web-browser, midori is good enough.. but I'm unsure how it will fit into our 64Mb RAM
Jan 09 14:22:26 <ant_work> I've seen someone in oe-core has started some work about 'tiny'
Jan 09 16:39:43 <ant_work> bluelightning: is dvhart on irc now and then? I'd liek to ask him about klibc & initramfs
Jan 09 16:40:08 <bluelightning> ant_work: he should be coming online soon if he's not around now
Jan 09 16:40:32 <ant_work> great, thx
Jan 09 16:41:29 <ant_work> he's also doing linux-yocto-tiny
Jan 09 16:42:18 <ant_work> probably we could make converge his efforts with the ones of Peter Chubb
Jan 09 16:42:39 <ant_work> and park our klibc stuff somewhere ;)
Jan 09 16:44:27 <ant_work> I'm thinking we could decouple linux-kexecboot from zaurus image generation, so that the remaining bits should be very common with other handhelds
Jan 09 16:44:39 <ant_work> just an idea
Jan 09 16:46:37 <ant_work> so one should purposedly bitbake linux-kexecboot and/or zaurus-installer
**** ENDING LOGGING AT Tue Jan 10 02:59:58 2012