| 00:08.09 | solar | http://tinderbox.dev.gentoo.org/embedded/beagle/All/ (all the bins are here) |
| 00:08.18 | solar | but gimme a sec to make a minimal boot image |
| 00:08.35 | solar | err. You can chroot vs full boot image? |
| 00:08.52 | ds2 | I can chroot, sure |
| 00:18.19 | joe_bleau | I saw a neat Lucent ADSL line card today. Loaded with big FPGAs. |
| 00:18.52 | joe_bleau | I was a bit surprised to see an Altera on the board in the midst of several Xilinx Virtex chips. |
| 00:19.09 | joe_bleau | And an IBM 405 too. |
| 00:19.17 | solar | ds2: quick test if you could plz. linbsd.net/~solar/beagle-20081204.tar.bz2 |
| 00:22.19 | ds2 | okay |
| 00:31.24 | ds2 | BusyBox v1.12.2 (2008-12-03 23:36:02 UTC) multi-call binary |
| 00:31.24 | ds2 | Copyright (C) 1998-2008 Erik Andersen, Rob Landley, Denys Vlasenko |
| 00:31.37 | solar | perfect. |
| 00:31.43 | ds2 | works fine as far as I can tell; entered it with chroot /root/gentoo /bin/sh |
| 00:47.11 | *** join/#edev mjtrangoni (n=mt@host10.190-31-111.telecom.net.ar) |
| 00:47.34 | *** part/#edev mjtrangoni (n=mt@host10.190-31-111.telecom.net.ar) |
| 01:11.01 | *** join/#edev linac (n=lin@222.76.91.240) |
| 01:18.15 | MonMotha | joe_bleau: PPC is very common in telecom |
| 01:18.41 | MonMotha | as for the FPGAs, likely the Altera was used for something different and perhaps they needed something odd that a Xilinx solution wouldn't give them like a different IO standard |
| 01:19.21 | MonMotha | FPGAs are everywhere since the volume usually isn't high enough for an ASIC, but they need to shovel lots of data around and do lots of parallel processing. That's ideal for an FPGA (or ASIC, if you've got the volume, which telecom apps usually don't) |
| 01:24.18 | joe_bleau | Agree to all that, just suprised that they'd have both major FPGA vendors in a single design. |
| 01:24.30 | *** join/#edev JoeLlama (n=snork@unaffiliated/joellama) |
| 01:24.41 | joe_bleau | One side was a big pile of high density press in connectors. |
| 01:24.47 | joe_bleau | Not sure how I feel about those. |
| 01:24.57 | joe_bleau | Just seems a bit *weird*, no solder and all. |
| 01:25.15 | joe_bleau | Obviously, they would not use them if they were not good. |
| 02:10.07 | *** join/#edev Ian_Daniher|XO (n=it@69.61.230.246) |
| 02:25.40 | *** join/#edev it|xo (n=it@69.61.230.246) |
| 02:25.48 | *** join/#edev roxfan (n=dunno@123.56-244-81.adsl-dyn.isp.belgacom.be) |
| 02:26.18 | MonMotha | joe_bleau: I've had mixed results with those press-in connectors |
| 02:26.40 | MonMotha | they tend to work surprisingly well, but when they fail, they are almost impossible to troubleshoot |
| 02:32.58 | *** join/#edev kergoth (n=kergoth@ip68-3-163-239.ph.ph.cox.net) |
| 02:33.51 | *** join/#edev kergoth_ (n=kergoth@ip68-3-163-239.ph.ph.cox.net) |
| 02:45.33 | *** join/#edev it|xo (n=it@69.61.230.246) |
| 02:50.24 | *** join/#edev JohnnyL (i=JohnnyL@ool-182ddad4.dyn.optonline.net) |
| 03:21.17 | JohnnyL | <PROTECTED> |
| 03:45.05 | agaffney | JohnnyL: interesting |
| 03:54.08 | *** join/#edev agaffney (n=agaffney@gentoo/developer/agaffney) |
| 03:54.49 | agaffney | JohnnyL: although, her comment about the only thing changing in computers each year was a tiny piece (referring to the processor, I assume) isn't very accurate :P |
| 04:26.51 | jbevren | how so? ram is still D. its just faster. |
| 04:26.52 | jbevren | lol |
| 04:27.04 | jbevren | we've gone back to serial IO (pci-e, usb, firewire, sata) |
| 04:27.07 | jbevren | ducks |
| 04:27.16 | jbevren | I'm gonna go to bed before I get bumped |
| 04:40.37 | *** join/#edev DrZimmerman (n=theo@77-56-225-121.dclient.hispeed.ch) |
| 04:55.41 | JohnnyL | agaffney i say it's right one. there is a limit to the upgrade based on the board. usually the upgrade will last you another 2 years on average. And they all have varying speeds. When in reality and truth they can all be made the best and fastest possible and they are *not*. |
| 04:56.34 | JohnnyL | it's like going to buy a monitor, the bad ones are cheaper, the expensive ones are quality. While the difference to make them relatively are $0. |
| 04:56.46 | JohnnyL | opps monitor=lcd tv. |
| 04:56.50 | JohnnyL | same difference. |
| 05:21.31 | *** join/#edev viridior1 (n=viridior@ip98-169-228-199.dc.dc.cox.net) |
| 05:29.50 | *** join/#edev JoeLlama (n=snork@unaffiliated/joellama) |
| 05:49.40 | *** join/#edev cbrake_a1ay (n=cbrake@oh-69-34-21-229.sta.embarqhsd.net) |
| 06:32.26 | *** join/#edev harpal (n=Harpal@121.246.75.165) |
| 06:56.57 | *** join/#edev ccjoe (n=joeg@75-169-7-104.slkc.qwest.net) |
| 06:56.57 | *** join/#edev jbevren (n=jbevren@dumt-04-103.dsl.netins.net) |
| 07:05.18 | *** part/#edev JohnnyL (i=JohnnyL@ool-182ddad4.dyn.optonline.net) |
| 07:38.02 | *** join/#edev lyakh (n=lyakh@p57BD22D6.dip0.t-ipconnect.de) |
| 07:57.48 | methril | yo |
| 08:12.23 | *** join/#edev harpal (n=Harpal@121.246.75.165) |
| 08:17.22 | *** join/#edev JoeLlama (n=snork@unaffiliated/joellama) |
| 08:46.29 | *** join/#edev JoeLlama (n=snork@unaffiliated/joellama) |
| 08:54.24 | *** join/#edev hw_ (n=hw@p578b3905.dip0.t-ipconnect.de) |
| 09:27.57 | *** join/#edev boris_OmegA (n=boris@194.51.141.172) |
| 10:20.28 | *** join/#edev mbuf (n=shakthim@94.191.166.51.bredband.tre.se) |
| 10:49.40 | *** join/#edev Ditzi (n=ditzi@190.244.143.167) |
| 10:49.51 | *** part/#edev Ditzi (n=ditzi@190.244.143.167) |
| 12:20.09 | *** join/#edev g1powermac (n=g1powerm@unaffiliated/g1powermac) |
| 12:36.12 | *** join/#edev linac (n=lin@121.201.62.182) |
| 13:07.16 | *** join/#edev g1powermac_PB (n=g1powerm@unaffiliated/g1powermac) |
| 13:46.31 | *** join/#edev hw__ (n=hw@p578b3905.dip0.t-ipconnect.de) |
| 14:09.34 | *** join/#edev mjtrangoni (n=mt@host10.190-31-111.telecom.net.ar) |
| 14:11.21 | *** join/#edev GPSFan (n=kenm@64.92.145.112) |
| 14:11.21 | *** mode/#edev [+o GPSFan] by ChanServ |
| 14:28.09 | *** join/#edev roxfan (n=dunno@234.175-244-81.adsl-dyn.isp.belgacom.be) |
| 14:28.37 | *** join/#edev g1powermac_PB (n=g1powerm@unaffiliated/g1powermac) |
| 14:38.57 | *** join/#edev redblue (i=star@ppp146.108-253-207.mtl.mt.videotron.ca) |
| 14:41.22 | *** join/#edev prpplague (n=dave@mail.americanmicrosystems.com) |
| 14:41.51 | prpplague | ho ho ho |
| 14:41.55 | prpplague | merry freakin thursday |
| 14:50.32 | *** join/#edev samrobb (n=samrobb@eng.arriad.com) |
| 14:51.19 | prpplague | samrobb: greetings |
| 14:51.49 | samrobb | prpplague: Good morning, sir. How goes it? |
| 14:52.06 | prpplague | samrobb: slaying windmills again today |
| 14:52.33 | samrobb | At least you're killing them :-) |
| 14:54.17 | prpplague | you would think so |
| 15:01.24 | *** join/#edev kergoth (n=kergoth@65.200.49.156) |
| 15:01.47 | prpplague | kergoth: hey hey crack daddy |
| 15:01.58 | prpplague | kergoth: did you check the window this morning? |
| 15:01.59 | kergoth | morning |
| 15:03.06 | prpplague | kergoth: just wondering if you saw any pigs flying by |
| 15:06.51 | g1powermac_PB | prpplague: no pigs, but I think I saw a half frozen cow fly by in the heavy winds full of snow, sleet, and rain this morning ;-0 |
| 15:07.55 | prpplague | hehe |
| 15:09.03 | *** join/#edev Chocobo (n=Ask@strongbad.eece.maine.edu) |
| 15:09.09 | g1powermac_PB | oh this is nice, the Fed is thinking of trying to lower mortgage rates, which is perfect timing as we're thinking of buying a house. . . |
| 15:09.36 | prpplague | is soooooo sick of owning a house |
| 15:09.58 | g1powermac_PB | actually, its a house on a 50 acre farm |
| 15:10.11 | prpplague | that would kill me for sure |
| 15:10.15 | Chocobo | Howdy. Is there a way to tell how large integers when using a cross compiler? I don't exactly have a screen to print "sizeof" to. |
| 15:11.06 | g1powermac_PB | prpplague: hehe, thats what equipment is for :-) |
| 15:11.16 | prpplague | Chocobo: seems like there is a kernel module someone wrote that gives that info via a proc entry |
| 15:11.39 | prpplague | g1powermac_PB: equipment doesn't relieve you of the time and stress of doing it |
| 15:11.55 | Chocobo | ooo I am not running linux. it is for a small embedded micro, no linux. |
| 15:12.34 | prpplague | Chocobo: any storage? i.e. flash or sd? |
| 15:12.39 | g1powermac_PB | prpplague: I actually don't find that stuff stressful, I kinda like doing yardwork and construction projects |
| 15:12.59 | Chocobo | prpplague, nope. All I have is the cross compiler. |
| 15:13.00 | prpplague | g1powermac_PB: yea i hear that alot |
| 15:13.28 | Chocobo | it seems like something that should be mentioned in gcc --target-help, but alas... it is not. |
| 15:13.29 | prpplague | Chocobo: i'm sure there is a solution, just never had that problem before |
| 15:13.32 | g1powermac_PB | the only time I find it stressful is when I'm under a crazy deadline |
| 15:13.45 | g1powermac_PB | like when I had to build the store stuff |
| 15:14.03 | g1powermac_PB | hand built all the shelving units on the walls of the store |
| 15:14.33 | g1powermac_PB | also prepped and painted the walls, and setup all the networking and even built the massive register counter |
| 15:22.10 | solar | prpplague: good luck today. |
| 15:22.21 | *** join/#edev chouimat|work (n=dieu@209.217.106.98) |
| 15:22.31 | solar | oh nevermind. thats tmw not today |
| 15:22.31 | prpplague | solar: on the OE stuff? |
| 15:22.46 | prpplague | solar: oh for the meeting? |
| 15:22.59 | prpplague | solar: that's very informal, they just want to see some progress |
| 15:23.00 | solar | yeah. clearly I just woke up |
| 15:27.53 | chouimat|work | morning |
| 15:28.48 | g1powermac_PB | heya chouimat|work |
| 15:30.32 | prpplague | CosmicPenguin: greetings |
| 15:30.36 | CosmicPenguin | hiya |
| 15:30.55 | *** join/#edev Madsy (n=madsy@ti500710a342-0881.bb.online.no) |
| 15:39.19 | CosmicPenguin | If you have more then one PCI bus... |
| 15:39.22 | CosmicPenguin | is it buses or busses? |
| 15:41.12 | *** join/#edev Bluenarf (n=Paul@apollo.paulsnet.org) |
| 15:42.14 | prpplague | CosmicPenguin: http://grammar.ccc.commnet.edu/GRAMMAR/plurals.htm |
| 15:42.27 | prpplague | CosmicPenguin: there is a section that talks about "bus" |
| 15:43.09 | CosmicPenguin | hmm - okay |
| 15:43.21 | CosmicPenguin | I was wondering if we had some sort of geek agreement about that |
| 15:43.24 | CosmicPenguin | probably doesn't matter |
| 15:45.26 | prpplague | CosmicPenguin: it has always been my understand that the word "bus" when refering to electronics was a method of transportation of data, i.e a school bus transport kids |
| 15:49.14 | CosmicPenguin | its a qualifier for a physical thing |
| 15:49.43 | CosmicPenguin | You have to add it on when talking to non geeks |
| 15:52.25 | CosmicPenguin | You know - I know its just advertising |
| 15:52.33 | CosmicPenguin | but Dunkin' Donuts cofee is pretty tasty |
| 15:54.34 | prpplague | CosmicPenguin: oh yea? never tried it or starbucks |
| 15:58.18 | solar | no starbucks? No cellphone? new breed of geek. |
| 15:59.06 | prpplague | didn't know those were requirements |
| 15:59.22 | prpplague | as i was geek long before either existed |
| 16:00.32 | *** join/#edev hw__ (n=hw@p578b3905.dip0.t-ipconnect.de) |
| 16:08.34 | CosmicPenguin | Is there realy somebody in america that hasn't tried starbucks? |
| 16:08.38 | CosmicPenguin | I don't believe it |
| 16:08.40 | CosmicPenguin | I won't believe it |
| 16:09.45 | prpplague | has never been to starbucks |
| 16:10.40 | prpplague | actually that might be 100% correct, seems like i got some coffee in the airport once from starbucks |
| 16:10.59 | CosmicPenguin | goes back to hacking on lar |
| 16:11.55 | kergoth | lar? |
| 16:12.09 | CosmicPenguin | Its the ROM creation tool for coreboot |
| 16:57.23 | CosmicPenguin | Will somebody fly out to Colorado and beat me about the head and shoulders? |
| 16:57.35 | CosmicPenguin | I actually just had the thought - "I should get Pixil up and running again" |
| 16:57.44 | CosmicPenguin | I need a job |
| 16:57.46 | CosmicPenguin | or a hobby |
| 17:02.09 | prpplague | CosmicPenguin: hehe |
| 17:02.19 | prpplague | CosmicPenguin: i'd love to see pixil up and running on the 8050 |
| 17:02.55 | CosmicPenguin | heh |
| 17:03.32 | chouimat|work | is having fun to make a customer happy he as to completely rewrite a module ... |
| 17:04.37 | prpplague | goes an looks at the old pixil pages |
| 17:06.05 | kergoth | damnit, i cant figure out how to set up the branches for this OpenEmbedded script |
| 17:07.27 | prpplague | CosmicPenguin: funny, pixil is almost exactly what we need |
| 17:07.59 | CosmicPenguin | bitrot on line one |
| 17:08.18 | prpplague | CosmicPenguin: indeed |
| 17:08.56 | prpplague | CosmicPenguin: http://www.pixil.org/screenshots.php |
| 17:09.05 | prpplague | CosmicPenguin: those with an orange theme would be just dandy |
| 17:09.28 | prpplague | CosmicPenguin: we you guys using buildroot to do that build? |
| 17:10.12 | CosmicPenguin | not really |
| 17:10.27 | CosmicPenguin | I was just trying to transition people to buildroot at the end of my time there |
| 17:10.43 | CosmicPenguin | I was doing buildroot - but I most of the other stuff was handrolled |
| 17:10.50 | prpplague | ahh |
| 17:11.02 | prpplague | CosmicPenguin: wonder how hard it would be to roll that into OE? |
| 17:11.31 | prpplague | CosmicPenguin: want an 8050 to get pixil running on? hehe |
| 17:11.41 | kergoth | argh, i dunno if i'm gonna be able to retain the history i want to |
| 17:11.42 | kergoth | hmm |
| 17:11.44 | CosmicPenguin | nanox had a terrible build system |
| 17:11.48 | CosmicPenguin | pain in the ass to put in OE |
| 17:12.21 | prpplague | ahh |
| 17:14.38 | prpplague | CosmicPenguin: if you got pixil running on the 8050, you could probably have a steady demand for applications , hehe |
| 17:16.47 | CosmicPenguin | I already tried to start a business hacking on Pixil, remember? |
| 17:18.44 | prpplague | yea i remember, hehe |
| 17:22.04 | prpplague | this oe build is gonna take all day |
| 17:23.03 | kergoth | man i ⥠packaged staging, especially with the pstage dir redirected outside the tmp dir, so i can wipe tmp and do a rebuild and have it pull the staging packages instead of rebuilding everything except what i'm working on |
| 17:23.08 | prpplague | CosmicPenguin: shame so may cool gui's have died |
| 17:25.18 | prpplague | chouimat|work: btw, was a little disappointed in the vellaman kits, the traces on the pcb came off pretty easy if you heated a pad too long |
| 17:25.45 | prpplague | <PROTECTED> |
| 17:26.22 | prpplague | main item we keep getting hammered with is a request for java on our handhelds |
| 17:32.36 | *** join/#edev lyakh (n=lyakh@p57BD22D6.dip0.t-ipconnect.de) |
| 17:44.30 | *** join/#edev Madsy (n=madsy@ti500710a342-0177.bb.online.no) |
| 17:46.03 | *** join/#edev Ian_Daniher|XO (n=it@204.10.219.119) |
| 17:46.06 | solar | pixil looks quite ulgy and dated. |
| 17:46.41 | solar | apps look like proper pda apps however |
| 17:48.47 | joe_bleau | Chocobo: sizeof() is supposed to be a compile time construct |
| 17:49.14 | joe_bleau | I've used it with compile time asserts to verify struct sizes |
| 17:49.48 | joe_bleau | If the size was wrong, it would stop the compilation with an error. |
| 17:53.18 | *** join/#edev Chocobo (n=Ask@strongbad.eece.maine.edu) |
| 17:57.21 | CosmicPenguin | kergoth: you might know - does GNU tar have gzip code built in, or does it fork gzip / bzip2 when the -z and -j flags are used? |
| 17:58.09 | kergoth | don't know, sorry |
| 17:58.42 | prpplague | solar: for a standard pda pixil probably isn't that nice, but for commercial devices, it was lovely |
| 17:59.07 | CosmicPenguin | solar: hey now - I put a bunch of time into Pixil |
| 18:14.46 | *** join/#edev sleipnir (n=Sleipnir@p4FDD41F0.dip.t-dialin.net) |
| 18:19.53 | kergoth | CosmicPenguin: wonder if i should make this proof of concept a two step process. step one, make a git repo for the package's sources + our patches only, no metadata or extra files in the tree, then parse and write back out the recipe with just the git repo and extra files in the SRC_URI, to at least make managing the sources and patches a bit nicer.. then play around with moving away from the monolithic metadata repo as a second step |
| 18:19.54 | kergoth | heh |
| 18:20.05 | *** join/#edev Redhatter (n=vk4msl@2001:388:f000:0:0:0:0:3) |
| 18:21.07 | kergoth | yeah, i think ill do that, makes it easier to wrap my head around too |
| 18:21.08 | kergoth | heh |
| 18:21.12 | kergoth | hmm |
| 18:23.41 | CosmicPenguin | yeah - that sounds good |
| 18:24.09 | CosmicPenguin | My problem is that I keep going back to the debian model - which is somewhat different |
| 18:24.46 | CosmicPenguin | in your case - the OE metadata and the source are fundamentally different, though you are right in that there is no reason why they can't live together |
| 18:27.37 | kergoth | what i want is a capable frontend that can pull down a collection of source repositories and associated recipes into your local working directory for a build. apt like interface to add/remove packages to/from the local tree, and the build (automatically updating the BBFILES for the .conf for this working directory), etc. perhaps a frontend to modifications to the recipes, also. then just run bitbake as usual to actually run the build from there |
| 18:29.56 | kergoth | maybe the operation that sets up the working copy also creates hidden mirror repositories of the upstream ones, so you can push your changes to a central place to affect all your working areas.. |
| 18:29.56 | kergoth | hmm |
| 18:39.15 | T0mW | kergoth: speeding up OE, or just making it more flexible? |
| 18:40.15 | kergoth | long term i want to replace bitbake/oe entirely, but thats just the tools and recipes, i figured i can switch to a less monolithic structure now and leverage it when i replace the tools |
| 18:40.34 | T0mW | nod |
| 18:40.46 | kergoth | lots of people want to be able to more easily change the source and rebuild a package, the active development use case isnt well handled by it today |
| 18:40.53 | solar | CosmicPenguin: don't mean to offend. But it's just very simple looking (UI) wise. But as noted the apps looked ok. |
| 18:41.13 | prpplague | NOTE: Running task 659 of 4789 |
| 18:41.17 | T0mW | yes, it is a bit of a PITA to dev the sources, can be done but carefully. |
| 18:41.17 | prpplague | jeeze |
| 18:41.51 | kergoth | plus being more explicit about what packages are in the working area, you only parse what you'll be wanting to build for the distro & target you're working on |
| 18:42.00 | kergoth | which would reduce the overhead in the cycles |
| 18:42.08 | prpplague | solar: i think CosmicPenguin was joking with you |
| 18:43.04 | kergoth | and having things less monolithic will make it easier to manage access and areas of responsibility. hooks on the repositories can easily enforce distro maintainer access and stuff |
| 18:44.03 | T0mW | ah, got it, current mtn database is all or nothing access. |
| 18:44.20 | prpplague | the quick builds and dev environments are what i want most in an OE like system |
| 18:44.59 | kergoth | T0mW: it uses git now, but its still one big pile of recipes, configuration files, and patches |
| 18:45.05 | kergoth | one big repository |
| 18:45.11 | T0mW | yeah |
| 18:45.56 | prpplague | jokes with kergoth |
| 18:46.07 | T0mW | I'm about to get back into working with OE again. My dev box is healthy again after fixing raid10 array resync prob. That turned out to be a change in udev strategy. |
| 18:46.10 | prpplague | kergoth: it certainly is one big pile of something, hehe |
| 18:46.15 | kergoth | hehe |
| 18:46.56 | kergoth | the metadata is the biggest advantage to using oe, you leverage the work of everyone working on it, 90% of the time, what you want is already available, package wise |
| 18:47.16 | solar | kergoth: btw I like that OE is one big repo. It allows me to look at your tree and find logical fixes I need to import into gentoo |
| 18:47.26 | T0mW | Now, I just have to pull in all the software I've squirreled away on various machines. |
| 18:48.17 | T0mW | kergoth: yes, that I like about OE, the dependency tree. Like I need sshd and OE knows that libssl is also needed. |
| 18:48.19 | kergoth | solar: yeah, there is that.. the way i'm talking would improve collaboration with debian's git setup, and other distros moving to this sort of way of doing things, but we'd have to have a -fast- -easy- interface to extract the patch series for upstream and other perusal, probably a nice web interface |
| 18:48.23 | solar | some of the structure of how things are make no sense however. pkgs living in the subdir of another pkg. |
| 18:48.57 | kergoth | solar: heh, i never wanted packages grouped together, it was supposed to be one package per dir, because decisions about categorization or grouping are too arbitrary and distro specific |
| 18:49.00 | solar | http://gitweb.openembedded.net/?p=openembedded.git;a=tree;f=packages;h=246e44368278b4e0bc06eb9d472fa6cccaa6430f;hb=HEAD (like I know Orbit has to be in there somewhere) |
| 18:49.01 | kergoth | but anyway.. |
| 18:49.27 | prpplague | kergoth: moving away from monotone and to git has made a big improvement in the usage for me |
| 18:49.31 | solar | but darned if I can find it so I can take a quick peek at how you guys are cross-building it. |
| 18:49.47 | kergoth | prpplague: yaeh, it has for just about everybody. so much faster, and easy to publish branches with changes for review and stuff |
| 18:51.34 | prpplague | kergoth: this go around with OE has been alot smoother than in the past |
| 18:56.14 | *** join/#edev g1powermac_PB (n=g1powerm@unaffiliated/g1powermac) |
| 18:57.19 | kergoth | glad to hear it, you've pretty much always struggled withi t |
| 18:58.11 | prpplague | indeed, doing some work with poky has helped my understanding |
| 18:59.18 | kergoth | did their documentation help, or messing with their structure in their svn and stuff? |
| 18:59.34 | prpplague | kergoth: all of the above |
| 19:00.54 | prpplague | kergoth: i still have a problem with configuring alot of the items i want to do, because i still have a buildroot mindset |
| 19:02.31 | prpplague | kergoth: there is a default rootfs that buildroot generates that is about 800k that i use for just about every board bringup, i'd really like to have something similiar for an OE build |
| 19:05.33 | T0mW | prpplague: also, sticking with a 'stable' branch instead of 'dev' helps immensely! |
| 19:06.00 | kergoth | that would be nice. you could use uclibc and leverage our really granular packaging, but even then.. you'd almost want to go down to the file level within the packaging, pick and choose exactly what binaries you want. itd be nice to add that capability to OpenEmbedded |
| 19:06.07 | prpplague | T0mW: yea, problem is right now is hrw has already added the 8050, so i have to run with the dev branch |
| 19:06.24 | kergoth | cherry pick his 8050 changes into stable |
| 19:06.48 | prpplague | kergoth: don't fell confident enough at the momment to do that |
| 19:06.58 | kergoth | git checkout -b mybranch org.openembedded.stable |
| 19:07.08 | kergoth | git log org.openembedded.dev -- find the commits |
| 19:07.15 | kergoth | git cherry-pick .. |
| 19:07.22 | kergoth | :) |
| 19:07.23 | kergoth | ⥠git |
| 19:07.56 | prpplague | hehe |
| 19:07.57 | T0mW | That was a huge problem for my understanding of OE was that I was extracting the dev.openembedded.org branch, where it seems everything is badly broken, then got frustrated at inconsistent builds. Once I abandoned the 'dev' and chose 'stable' all worked really smoothly. Then all I had to do was deal with my mistakes, not buggy recipies, etc. |
| 19:08.22 | prpplague | well so far i've not had any problems |
| 19:08.52 | prpplague | NOTE: Running task 710 of 4789 |
| 19:09.21 | kergoth | T0mW: one of my biggest issues with OpenEmbedded as it stands today is the lack of maintainership, areas of responsibility... we need accountability |
| 19:09.26 | kergoth | machine, distro, and specific recipe maintainers |
| 19:09.28 | kergoth | imo, anyway |
| 19:09.54 | kergoth | i dont even want to know how many of the recipes in the repository today dont even have people really looking after them |
| 19:10.00 | T0mW | kergoth: yes |
| 19:11.17 | prpplague | kergoth: as it is right now, we end up doing two builds, i do one with buildroot to get the system up and running, then when the userland folks start work, they generally start up an OE build |
| 19:11.41 | kergoth | thats a good way to to do it. OpenEmbedded is probably easier to maintain long term for the userland, but its a beast |
| 19:12.13 | T0mW | I just use patch files against OE rather than ask for commit access to the repository as I don't want to run into the possibility that I could do a commit and bork some other area of the repository I didn't want to change. |
| 19:12.44 | prpplague | kergoth: problem is the switch over from the buildroot to OE, we have all kinds of little toolchain issues and basic duplication of work |
| 19:12.52 | kergoth | can always get commit access and just only push oyur changes into a t0mw/foo branch, not the main ones |
| 19:12.59 | kergoth | then get people to look atyour branches and give the go ahead |
| 19:13.03 | T0mW | It would be better if I knew who to contact to negotiate a recipe / source change rather than bother Koen or the others. |
| 19:13.19 | T0mW | more granularity |
| 19:13.22 | kergoth | prpplague: yeah, i know what you mean. theres lots of little variations in the userland between tools, even if you select the same packages and (possibly) same versions |
| 19:13.32 | kergoth | T0mW: agreed |
| 19:13.43 | T0mW | overall, it is a hell of good idea (OE) |
| 19:13.45 | kergoth | T0mW: also, if we knew what was maintained and what wasnt, itd give people that want to contirbute an entry point |
| 19:13.54 | kergoth | they could volunteer to maintain random package Alpha |
| 19:14.02 | T0mW | yeah, I would do that |
| 19:15.12 | kergoth | heh, anyway, didn't mean to steal the channel.. i try not to bother #oe too much with my crazy ideas, since i want to have something to show for it first :P |
| 19:15.35 | T0mW | kergoth: oh heck, it better than listening to Dave cry about his job. |
| 19:15.37 | T0mW | ducks |
| 19:15.40 | kergoth | hehe |
| 19:15.45 | prpplague | T0mW: hehe |
| 19:15.56 | T0mW | ;) |
| 19:16.19 | prpplague | right now, i'm just happy to have one |
| 19:16.23 | prpplague | points at CosmicPenguin |
| 19:16.56 | kergoth | sometimes i worry about my job security, i still have some trouble focusing at times.. but the boss seems happy enough with how i'm doing |
| 19:17.07 | kergoth | besides, theres people here playing with OpenEmbedded now, so i'm supporting them |
| 19:17.27 | chouimat|work | hi kergoth prpplague T0mW |
| 19:17.32 | kergoth | hey chouimat|work |
| 19:18.00 | prpplague | chouimat|work: hey bud |
| 19:18.29 | chouimat|work | tries to understand just some translations just got screwed up |
| 19:19.37 | prpplague | kergoth: i realize i have a disease where in my mind i think "if i know this piece of information, everyone else does as well" |
| 19:20.05 | kergoth | heh, i have that issue too, thats why i never write user documentation, i cant think from their perspective anymore |
| 19:22.21 | chouimat|work | kergoth: currently I have to translate to project manager and senior management ... it's worse than user speak |
| 19:22.30 | kergoth | ew |
| 19:22.58 | chouimat|work | kergoth: and the best answer is "Don't know" ;) |
| 19:23.08 | T0mW | I was at a customer site yesterday, they are doing dev work with the Intel Atom. Pretty impressive: 3000+ BogoMips, 1.5GHz clock and no heatsink on the processor. |
| 19:23.53 | chouimat|work | T0mW: I have a marvell RD-6281 on my desk ... Marvell Kirkwood 1.2GHz in it ... nice beast for an arm based box |
| 19:23.54 | prpplague | hmm, i've seen the same job decription posted a number of times over the last 4 months, i'm wondering if 1) they have more than one opening 2) they can't find the right person for the job 3) the job is a @#$@#$ and no one wants it 4) they aren't willing to pay |
| 19:24.24 | chouimat|work | prpplague: 5) all the above :) |
| 19:24.26 | T0mW | 5) they are keeping the resume pile full |
| 19:24.50 | kergoth | i'd think 2 or 4 are likely |
| 19:24.50 | prpplague | its the same exact description, but posted by about 10 different HR companies |
| 19:25.41 | kergoth | you guys ever seen http://sandbox.movial.com/wiki/index.php/Matrix ? |
| 19:25.47 | kergoth | its an interesting approach |
| 19:26.05 | kergoth | they use git like i was thinking, i like the structure.. but, they use python for the main config and makefiles for the build part |
| 19:26.15 | kergoth | which i find questionable, personally |
| 19:26.21 | kergoth | still, interesting to know about |
| 19:26.46 | chouimat|work | kergoth: I was playing with this last night http://blog.funtoo.org/2008/10/metro-released.html |
| 19:26.59 | kergoth | interesting |
| 19:27.07 | kergoth | will have to look into that |
| 19:28.01 | prpplague | http://seeker.dice.com/jobsearch/servlet/JobSearch?op=302&dockey=xml/4/d/4d9e8e8bffd89c6bb9f824c27ba356ba@endecaindex&source=19&FREE_TEXT=embedded+linux&rating=99 |
| 19:28.05 | kergoth | the thing i dont know about with matrix is, the metadata is in a separate git repo from the source, and that repo gets cloned inside the clones of the sources |
| 19:28.14 | kergoth | i can see advantages and disadvantages to that |
| 19:28.45 | kergoth | disadvantage is itd need its own tags for what version of upstream the metadata is against, etc, or retain all that in separate files, if they were together they could leverage existing tags and branches |
| 19:29.09 | prpplague | compare that listing with this one - http://seeker.dice.com/jobsearch/servlet/JobSearch?op=302&dockey=xml/3/d/3db18a2f57947f8c76a186274377b87a@endecaindex&source=19&FREE_TEXT=embedded+linux&rating=99 |
| 19:29.09 | kergoth | advantage is, you could clone debian's source tree and then pull in the metadata repo for this other setup and build that |
| 19:31.39 | *** join/#edev T0mW (n=Tom@24.115.147.8.res-cmts.tv13.ptd.net) |
| 19:32.36 | prpplague | does that sound like the same job ? |
| 19:33.48 | prpplague | NOTE: Running task 799 of 4789 |
| 19:33.50 | joe_bleau | Didn't the first mention wince? |
| 19:33.52 | prpplague | ho ho hum |
| 19:34.20 | prpplague | yea |
| 19:35.03 | prpplague | joe_bleau: they both have a entry that reads "At least one RTOS (Linux, WinCE, pSOS, Nucleus, REX, VxWorks)" |
| 19:36.13 | joe_bleau | Is one looking for a consultant, the other a w2? |
| 19:37.11 | prpplague | not sure, but basically the same job description has been posted by about 10-15 companies in the last 3-4 months |
| 19:38.03 | *** join/#edev roxfan2 (n=dunno@233.163-243-81.adsl-dyn.isp.belgacom.be) |
| 19:38.19 | joe_bleau | I don't know what the "Tax Term" tag means, but they do look very close. |
| 19:38.33 | joe_bleau | Maybe multiple positions at the same company. |
| 19:38.45 | joe_bleau | Maybe multiple headhunters picking up on a single opening. |
| 19:39.21 | prpplague | thats kind of what i was thinking, but the time period of 3-4 months seems odd to me |
| 19:40.18 | joe_bleau | I know at least one way to find out for sure... |
| 19:42.10 | gandhijee | hey is there a BUILD option that has the path to the linux sources in the buildroot dir? |
| 19:42.36 | gandhijee | and how would i make it so that i could compile a driver for this after the main kernel has been set up? |
| 19:45.25 | prpplague | gandhijee: building an external kernel module via buildroot can be done, but it is generally up to the makefile for the module to handle the paths and such for the build |
| 19:46.07 | gandhijee | yeah, but when i make it i can pass it a KVERS option and KSCR location. |
| 19:46.33 | gandhijee | i was wondering if the buildroot env had the KSCR path somewhere in it, or if i would have to build it by hand |
| 19:46.57 | prpplague | i don't recal it having support for that at the momment, you'd have to add something |
| 19:47.33 | gandhijee | cool |
| 20:36.57 | *** join/#edev Madsy (n=madsy@ti500710a342-1371.bb.online.no) |
| 21:28.16 | *** join/#edev Ian_Daniher|XO (n=it@69.61.230.246) |
| 21:30.20 | *** join/#edev likewise (n=chatzill@82-171-51-231.ip.telfort.nl) |
| 21:34.36 | prpplague | lordy lordy: Running task 1308 of 4789 |
| 21:44.40 | kergoth | hehe |
| 21:44.41 | kergoth | having fun yet |
| 21:54.31 | prpplague | kergoth: loads |
| 22:19.50 | Crofton|work | quad core || make and bitbake threads |
| 22:21.42 | chouimat|work | Crofton|work: not fast enough ... try dual quad core |
| 23:16.01 | *** join/#edev JohnnyL (i=JohnnyL@ool-182ddad4.dyn.optonline.net) |
| 23:16.53 | *** part/#edev JohnnyL (i=JohnnyL@ool-182ddad4.dyn.optonline.net) |
| 23:17.38 | *** join/#edev redblue (i=star@ppp192.108-253-207.mtl.mt.videotron.ca) |