| 00:05.26 | *** join/#edev g1powermac_PB (n=g1powerm@unaffiliated/g1powermac) |
| 00:09.38 | *** join/#edev Ian_Daniher_ (n=it@69.61.230.246) |
| 00:18.36 | prpplague | boo |
| 00:18.42 | joe_bleau | hoo |
| 00:18.53 | prpplague | hey joe, whaddaya know? |
| 00:19.49 | joe_bleau | found the debian installer problem on the alix today |
| 00:20.22 | ds2 | Mooo |
| 00:20.22 | prpplague | ds2: anyway, our lib is pretty darn good |
| 00:20.31 | prpplague | ds2: i wanted to try and get more dev going on it |
| 00:20.55 | ds2 | prpplague: Is it entirely opensource or just parts? |
| 00:22.24 | prpplague | ds2: but when opie was all the rage, no one was interested |
| 00:22.28 | prpplague | ds2: yep, gpl |
| 00:24.00 | ds2 | prpplague: is the library usable only from within C or is there a shell script controllable peice like dialog? |
| 00:24.05 | prpplague | ds2: just no one ever did anything with it |
| 00:24.28 | ds2 | prpplague; opie is dead and the other ones have picked up too much dead weight |
| 00:24.37 | prpplague | ds2: it was really designed to be used with C, but there is no reason you couldn't add some hooks for shell usage |
| 00:24.47 | prpplague | ds2: yea, this was 3 years ago |
| 00:25.27 | ds2 | prpplague: Interesting... is it downloadable? |
| 00:25.40 | ds2 | prpplague: I assume you use termlib/ncurses to do it all right? |
| 00:25.51 | prpplague | ds2: no its full fb |
| 00:26.06 | prpplague | ds2: good question i don't think it is, but it will be tomorrow |
| 00:26.53 | ds2 | prpplague: if it is fb, does it cooperate with fbcon or does it replace it? |
| 00:27.45 | *** join/#edev kergoth (n=kergoth@ip68-3-163-239.ph.ph.cox.net) |
| 00:28.09 | prpplague | ds2: it basically replaces it |
| 00:28.25 | prpplague | ds2: think of it as a console done as if it were x |
| 00:30.03 | prpplague | ds2: but these days everyone is into buzz words like android |
| 00:30.09 | *** part/#edev kergoth (n=kergoth@ip68-3-163-239.ph.ph.cox.net) |
| 00:31.52 | prpplague | ds2: i'll post the code tomorrow and you can run with it |
| 00:32.06 | ds2 | prpplague: ah... that's slightly different from what I had in mind |
| 00:32.21 | ds2 | I would like to to cooperate with fbcon so existing console apps just work |
| 00:32.30 | prpplague | ds2: they can |
| 00:32.45 | prpplague | ds2: think of it as a bogus telnet session |
| 00:34.06 | ds2 | prpplague: oh so you use pty's within the lib to support that? |
| 00:34.13 | prpplague | ds2: more like doing: telnet localhost |
| 00:34.19 | prpplague | ds2: yea |
| 00:34.35 | ds2 | as long as you don't really do a telnet localhost ;) |
| 00:34.53 | prpplague | so we normall have 6 different vt's running different windows |
| 00:35.10 | ds2 | if that is the case, adding a OSK on there is pretty close to what I want |
| 00:35.16 | prpplague | ds2: i'll post it tomorrow, either way it will be up on the net |
| 00:35.25 | ds2 | and i really need X, it can be fired up on another VT and dispose of as needed |
| 00:35.37 | prpplague | ds2: yea, you could add the image for the keyboard and a thread to decode the touchscreen |
| 00:35.47 | prpplague | ds2: yep |
| 00:36.04 | ds2 | prpplague: how many libs do you pull in? |
| 00:36.20 | prpplague | iirc just one, ours |
| 00:36.28 | ds2 | I like that |
| 00:36.31 | prpplague | other than the standard libs |
| 00:36.33 | ds2 | X pulls in way too much stuff |
| 00:36.37 | prpplague | indeed |
| 00:36.52 | ds2 | as long you don't yank at GTK, iwhatever, etc |
| 00:37.11 | ds2 | libc, libm, libpthread I don't have an issue with |
| 00:39.05 | prpplague | nope |
| 00:41.00 | ds2 | what about the fonts, does it use plain bitmaps or does it want a fontserver to use TTF stuff? |
| 00:43.40 | prpplague | ds2: hardcoded bitmaps |
| 00:44.53 | ds2 | is there any support for a mouse/cut'n'paste? |
| 00:46.17 | prpplague | nope, as there was no pointing device on our handhelds at the time |
| 00:48.18 | ds2 | output sounds perfect, the inputside seems to lack a few things...but this is more palatable then X |
| 00:49.50 | prpplague | indeed, i once had some grand scheme for it |
| 00:50.18 | prpplague | i had hoped that it would take hold for some midranger(at the time) device to be used with |
| 00:54.39 | ds2 | I don't suppose you have a launcher type functionality on there already? |
| 00:57.06 | prpplague | ds2: indeed |
| 00:57.20 | prpplague | ds2: we need to be able to kick of specific apps the customer might have |
| 00:58.17 | ds2 | prpplague: why not package it up as a ligher alternative to android , etc? |
| 00:58.44 | prpplague | ds2: time and motivations |
| 01:12.32 | *** join/#edev JohnnyL (i=JohnnyL@ool-182ddad4.dyn.optonline.net) |
| 01:13.54 | prpplague | my theme song http://www.youtube.com/watch?v=-xEzGIuY7kw |
| 01:18.19 | JohnnyL | anyone here use scheme? |
| 01:22.58 | *** join/#edev btraynor (n=btraynor@gromit.mixdown.ca) |
| 01:27.41 | *** join/#edev kergoth (n=kergoth@ip68-3-163-239.ph.ph.cox.net) |
| 01:43.10 | joe_bleau | Anyone seen a price on the new Atmel EVK1105 (AVR32 audio reference design)? |
| 02:15.10 | *** join/#edev Ian_Daniher_ (n=it@69.61.230.246) |
| 02:58.36 | *** join/#edev linac (n=lin@125.77.68.62) |
| 03:39.48 | MonMotha | prpplague: around? |
| 03:51.51 | *** join/#edev JoeLlama (n=snork@unaffiliated/joellama) |
| 04:00.29 | *** join/#edev wmat (n=btraynor@gromit.mixdown.ca) |
| 04:06.00 | *** join/#edev JoeLlama (n=snork@unaffiliated/joellama) |
| 04:11.39 | *** join/#edev Ian_Daniher|xo (n=it@69.61.230.246) |
| 04:12.15 | *** join/#edev davidc__ (n=davidc__@S01060002b360aacd.vc.shawcable.net) |
| 04:29.53 | *** join/#edev DrZimmerman (n=theo@77-56-228-246.dclient.hispeed.ch) |
| 04:49.12 | *** join/#edev RobotGuy (n=Dale@pool-96-225-210-142.ptldor.fios.verizon.net) |
| 04:49.13 | *** join/#edev kergoth (n=kergoth@ip68-3-163-239.ph.ph.cox.net) |
| 04:49.44 | *** part/#edev kergoth (n=kergoth@ip68-3-163-239.ph.ph.cox.net) |
| 04:50.05 | *** join/#edev kergoth (n=kergoth@ip68-3-163-239.ph.ph.cox.net) |
| 05:19.24 | *** join/#edev Crofton (n=balister@12.234.232.66) |
| 05:55.55 | *** join/#edev harpal (n=Harpal@122.169.108.195) |
| 06:02.01 | *** join/#edev landley (n=landley@cpe-70-116-30-223.austin.res.rr.com) |
| 06:02.01 | *** mode/#edev [+o landley] by ChanServ |
| 06:03.08 | landley | waves at prpplague |
| 06:29.22 | *** join/#edev sleipnir (n=Sleipnir@p4FDD5C20.dip.t-dialin.net) |
| 06:44.38 | *** join/#edev redblue (i=star@ppp014.108-253-207.mtl.mt.videotron.ca) |
| 08:08.41 | *** join/#edev lyakh (n=lyakh@p57BD28A2.dip0.t-ipconnect.de) |
| 08:11.01 | *** join/#edev hw (n=hw@p578b3905.dip0.t-ipconnect.de) |
| 08:14.35 | *** join/#edev [AD]Turbo (n=turbo@host162-220-dynamic.59-82-r.retail.telecomitalia.it) |
| 08:16.26 | [AD]Turbo | yo |
| 09:49.10 | methril | yo |
| 09:56.24 | *** join/#edev frikker_ (n=blaine@107.135.68.216.DED-DSL.fuse.net) [NETSPLIT VICTIM] |
| 09:56.24 | *** join/#edev Ober (i=jaf@babum.mauthesis.com) [NETSPLIT VICTIM] |
| 09:56.24 | *** join/#edev morrijr (n=john@cpc2-mfld9-0-0-cust297.nott.cable.ntl.com) [NETSPLIT VICTIM] |
| 10:08.07 | *** join/#edev harpal (n=Harpal@122.169.108.195) |
| 10:42.05 | *** join/#edev g1powermac_PB (n=g1powerm@unaffiliated/g1powermac) |
| 11:01.51 | *** join/#edev g1powermac (n=g1powerm@unaffiliated/g1powermac) |
| 11:49.20 | *** join/#edev gizero76 (n=gizero@host7-75-dynamic.9-87-r.retail.telecomitalia.it) |
| 12:17.40 | *** join/#edev linac (n=lin@121.201.33.74) |
| 12:51.04 | *** join/#edev g1powermac_PB (n=g1powerm@unaffiliated/g1powermac) |
| 13:13.37 | *** part/#edev gizero76 (n=gizero@host7-75-dynamic.9-87-r.retail.telecomitalia.it) |
| 13:40.09 | *** join/#edev GPSFan (n=kenm@64.92.145.112) |
| 13:40.09 | *** mode/#edev [+o GPSFan] by ChanServ |
| 14:42.19 | *** join/#edev dmead (n=dan@ip229.sensaphone.com) |
| 15:03.41 | *** join/#edev samrobb (n=samrobb@eng.arriad.com) |
| 15:07.41 | *** join/#edev redblue (i=star@ppp153.108-253-207.mtl.mt.videotron.ca) |
| 15:19.18 | *** join/#edev chouimat|work (n=dieu@209.217.106.98) |
| 15:29.56 | *** join/#edev GPSFan (n=kenm@64.92.145.112) |
| 15:29.57 | *** mode/#edev [+o GPSFan] by ChanServ |
| 15:54.07 | *** join/#edev prpplague (n=dave@mail.americanmicrosystems.com) |
| 15:54.30 | prpplague | MonMotha: ping |
| 16:11.34 | MothaWork | prpplague: pong |
| 16:12.23 | chouimat|work | this match is still 0-0 |
| 16:44.15 | prpplague | MonMotha: you looking for me last night? |
| 16:55.34 | *** join/#edev kergoth (n=kergoth@65.200.49.156) |
| 16:57.00 | *** join/#edev beewoolie (n=beewooli@66.147.194.238) |
| 16:59.03 | kergoth | CosmicPenguin: making progress on that proof of concept git repo generator :) |
| 16:59.56 | CosmicPenguin | awesome |
| 17:00.14 | CosmicPenguin | I'll keep an eye on it - I have every reason to believe that whatever my next job is, distro building will be part of it |
| 17:01.03 | kergoth | the hard part is going to be 1) figuring out how the hell to deal with .inc's, and 2) locating the patch files within the metadata repo.. need to resolve the file:// to the path the fetcher would be copying from |
| 17:01.26 | kergoth | i got the command sequence worked out to create topgit branches for each patch, and it retains the history of patch modifications from the OpenEmbedded metadata repo, including commit info :) |
| 17:01.35 | kergoth | is having fun actually having something hafl way interesting to work on.. |
| 17:03.47 | kergoth | still doesn't even know if this will be something people would want to do, still just a proof of concept, but what the hell, its something to do, and not doing things manually is a good thing |
| 17:05.37 | MothaWork | prpplague: I think I got it figured out |
| 17:12.58 | CosmicPenguin | I have had two callbacks so far and both jobs have indicated a need for distribution building |
| 17:13.15 | kergoth | two already eh, nice to hear, thats quick |
| 17:14.32 | *** join/#edev beewoolie (n=beewooli@66.147.194.238) |
| 17:14.39 | CosmicPenguin | well, call backs |
| 17:14.54 | kergoth | still |
| 17:15.02 | kergoth | good luck with em |
| 17:15.15 | CosmicPenguin | I think my resume hits manager keywords or something |
| 17:15.39 | kergoth | hehe |
| 17:16.53 | prpplague | tries to imagine some keywords for CosmicPenguin |
| 17:18.17 | prpplague | lordy, just not going to get everything i need to complete done today |
| 17:22.06 | *** join/#edev samrobb (n=samrobb@eng.arriad.com) |
| 17:28.09 | *** join/#edev landley (n=landley@cpe-70-116-30-223.austin.res.rr.com) |
| 17:28.09 | *** mode/#edev [+o landley] by ChanServ |
| 17:32.29 | *** join/#edev kergoth (n=kergoth@65.200.49.156) |
| 17:40.35 | *** join/#edev Ei5GTB_ (n=Paul@78.16.178.232) |
| 17:51.25 | agaffney | does anyone know if there's some "feature" in busybox ash that will export foo=bar pairs from /proc/cmdline into the env? |
| 17:51.40 | agaffney | I see that happening here, and I'm pretty sure I don't have any code that does that |
| 17:51.59 | agaffney | every pair except root= and init= from /proc/cmdline are in my initramfs's ash env |
| 17:52.04 | CosmicPenguin | Its implied by the kernel |
| 17:52.17 | CosmicPenguin | anything that that the kernel doesn't consume is passed to the init process |
| 17:53.32 | CosmicPenguin | see linux/init/main.c |
| 17:58.14 | agaffney | CosmicPenguin: ah, so it's the kernel itself that does it? |
| 17:58.26 | CosmicPenguin | yep |
| 17:58.54 | CosmicPenguin | You may have seen this phenonomon in action before - have you ever put a desktop distribution in a different runlevel by passing a number on the kernel command line? |
| 17:58.59 | CosmicPenguin | same mechanism |
| 17:59.21 | agaffney | sure, but I always figured it just looked at /proc/cmdline |
| 17:59.33 | CosmicPenguin | I think that anything of the construct key=value is passed in the environment |
| 17:59.38 | CosmicPenguin | and anything else is passed as a argv |
| 18:00.36 | agaffney | hmm, ok |
| 18:02.42 | agaffney | the odd thing is I was seeing this behavior only with my kernel/initramfs but not when booting without the initramfs |
| 18:03.04 | agaffney | it was causing a problem in gentoo with a grub revision that wasn't putting a \0 at the end of the kernel commandline |
| 18:03.31 | agaffney | so people would get a var in their env like vga=791%(#($*#(*$#($()#*$()$#(#} (continued for a while) |
| 18:03.40 | agaffney | which would cause issues with some commands |
| 18:03.55 | agaffney | but it only happened with a genkernel-made kernel with an initramfs, not with a self-built kernel |
| 18:04.05 | CosmicPenguin | doh - that is unfortunate |
| 18:04.22 | agaffney | so why was that var not there with the custom kernels? |
| 18:04.52 | agaffney | we didn't even know about those vars being there until this issue came up :P |
| 18:05.29 | *** join/#edev kergoth (n=kergoth@65.200.49.156) |
| 18:05.36 | CosmicPenguin | Well, vga= should be consumed by the vesa driver |
| 18:05.43 | CosmicPenguin | even if it was malformed |
| 18:06.21 | agaffney | adding another var after vga= just caused the problem on that one |
| 18:06.25 | agaffney | but vga= was most common |
| 18:06.40 | agaffney | but anyway, it's not getting consumed |
| 18:06.49 | CosmicPenguin | so maybe the self-built kernel had th driver, but the genkernel didn't |
| 18:07.26 | agaffney | I've got a kernel/initramfs booted now that have vga=791 in the env |
| 18:07.42 | agaffney | and it's using a 1024x768 framebuffer, so that *has* to have been consumed |
| 18:07.58 | CosmicPenguin | hmmm - maybe the length of the variable is confusing it |
| 18:08.10 | agaffney | the current kernel doesn't ahve the issue |
| 18:08.11 | landley | CosmicPenguin: actually the kernel filters out options it recognizes. |
| 18:08.15 | CosmicPenguin | like its consuming it, but not omititng it |
| 18:08.19 | landley | console= doesn't get set as an anvironment variable. |
| 18:08.31 | CosmicPenguin | landley: right - which is why vga= making it into the env is a puzzle |
| 18:08.37 | CosmicPenguin | especially if there was something there to consume it |
| 18:08.55 | agaffney | landley: is it only the ones it recognizes *directly* or options recognized by *any* kernel driver? |
| 18:09.01 | landley | Mostly only unrecognized key=value pairs get set as environment variables (it's a fallback behavior), but nothing's stopping recognized ones from setting the relevant env variable on their way out. |
| 18:09.17 | landley | agaffney: it's ones dispatched to a handler functions, so presumably any driver. |
| 18:09.32 | landley | The relevant code's under init/ somewhere, I think... |
| 18:10.19 | CosmicPenguin | I've never seen consumed variables make it to the init env, but I might be wrong |
| 18:11.54 | ds2 | it is easy enough to write an ash parser for those |
| 18:12.31 | agaffney | I just booted my normal kernel with 'vga=normal foo=bar' and those aren't in my env once I boot |
| 18:12.54 | agaffney | if there was something in the gentoo init process that filtered those, it would have been filtered even with the initramfs-based boot |
| 18:13.07 | ds2 | parse() { set - `cat /proc/cmdline`; while [ $# -gt 1 ]; do case $1 in; init=*) ...;; esac; shift; done; } |
| 18:13.45 | agaffney | oh, it's not hard to pull them out |
| 18:13.47 | agaffney | that's not the issue |
| 18:13.54 | agaffney | I just wanted to know why they were there by default |
| 18:14.12 | agaffney | but anyway, I'm not seeing that behavior with my non-initramfs custom kernel |
| 18:14.23 | agaffney | I passed random things on the kernel commandline and they're not in my env |
| 18:14.23 | ds2 | guessing init is stripping them once it enters a runlevel |
| 18:14.39 | agaffney | then that would have happened in the case of the kernel w/initramfs |
| 18:14.50 | CosmicPenguin | try booting with init=/bin/sh and see if they are there |
| 18:15.04 | CosmicPenguin | the initramfs does not use the init binary |
| 18:15.07 | CosmicPenguin | it invokes a shell |
| 18:15.24 | agaffney | yeah, but the switch_root execs /sbin/init |
| 18:15.24 | *** join/#edev lyakh (n=lyakh@p57BD28A2.dip0.t-ipconnect.de) |
| 18:15.27 | landley | Cosmicpenguicon: rdinit=/bin/sh |
| 18:15.34 | landley | Initramfs doesn't use init=, it uses rdinit= |
| 18:16.43 | agaffney | hmm, yep, there they are |
| 18:17.15 | ds2 | looking at the result of ps auxwwwwwe as root, it certainly looks like init cleaned its environment space |
| 18:24.52 | agaffney | I've just modified my /init script in the initramfs to do 'exec env -i switch_root ...' |
| 18:38.33 | *** join/#edev redblue (i=star@ppp068.108-253-207.mtl.mt.videotron.ca) |
| 18:45.11 | *** join/#edev g1powermac_PB (n=g1powerm@unaffiliated/g1powermac) |
| 18:48.15 | *** join/#edev L_Deluxe (n=lou@c-68-53-134-58.hsd1.tn.comcast.net) |
| 18:50.52 | *** join/#edev JohnnyL (i=JohnnyL@ool-182ddad4.dyn.optonline.net) |
| 18:57.40 | solar | agaffney: env -i $(cat /proc/cmdline) switch_root ; # perhaps if you want to keep the exported env vars |
| 18:59.58 | solar | might want a | tr ' ' '\n' | grep = ; in there also. |
| 19:06.54 | ds2 | sigh... I think personal sanity is worth $200... this noisy fan is driving me insane |
| 19:08.33 | ds2 | solar: that sounds dangerous from a security standpoint if people were to ever get access to the bootloader |
| 19:11.58 | *** join/#edev pinglo (n=pinglo@CPE001a70f2faee-CM001371145508.cpe.net.cable.rogers.com) |
| 19:12.55 | *** join/#edev dmead (n=dan@ip226.sensaphone.us) |
| 19:13.29 | landley | ds2: if people get access to the bootloader, don't they have physical access to the machine? |
| 19:13.36 | solar | if people were ever to get access to the bootload enough to export a variable. they could for sure own you in a few hundred thousand other ways |
| 19:14.54 | ds2 | landley: so you beleive in the Cisco argument that physical access == system owned? |
| 19:15.52 | ds2 | solar: I have tried to argue that with people and some folks do not believe is the case |
| 19:16.03 | landley | ds2: I didn't know it was a system argument, I just know I've done it. :) |
| 19:16.30 | landley | That said, some bootloaders can be accessed via a serial connection tunneled over telnet or similar silly things... |
| 19:16.54 | solar | ds2: take the cmdline for example. If an attacker is able to tell the bootloader something. They can surely do things like make another nfsroot or pass init=/bin/sh and the likes. |
| 19:16.56 | landley | s/system/cicso/ |
| 19:17.26 | landley | Take the hard drive, for example. :) |
| 19:18.26 | ds2 | then you decend down the path of let's encrypt the hard drive, put in crypto hashes, etc. I think Tivo has gone down that path |
| 19:19.09 | solar | sure it can be made more difficult to work around |
| 19:20.47 | ds2 | landley: Cisco routers can be rooted by dropping into the monitor, change one value, and it'll boot up to defaults with the config accessible; it has been like this forever and that's their argument for allowing that |
| 19:21.25 | landley | ds2: the xbox people went down that path. |
| 19:22.36 | ds2 | arrg |
| 19:22.54 | ds2 | are there good arguments against using one of those $200 netbooks as a daily dev machine? |
| 19:23.46 | kergoth | ds2: i was thinking about doing that. as long as you have an external monitor and dont mind somewhat long compile times, i dont see much issue with it |
| 19:23.47 | kergoth | shrugs |
| 19:24.11 | ds2 | kergoth: so you their their included UI is tolerable? |
| 19:24.41 | kergoth | who cares? just replace it if need be, i can't see it being that difficult |
| 19:24.49 | kergoth | course i haven't held one in my hands yet, so .. |
| 19:25.12 | ds2 | i don't mind small screens at all.. I've always used the onboard LCD even when it was a mere 10inches or so |
| 19:26.01 | kergoth | i bet a tiled wm could be interesting on one of those |
| 19:28.37 | ds2 | I just need a daily usemachine... cracking keyboard case, a fan that is louder then a 70's land line phone ringer are all wearing my sanity down |
| 19:29.06 | kergoth | understandable |
| 19:29.18 | kergoth | i'm sick of funneling money into machines. i dont need top of the line anymore, i just want shit that works, you know |
| 19:29.19 | kergoth | heh |
| 19:29.39 | ds2 | completely agree... my desktop servers are dual PIII's I got from the dumpster |
| 19:30.48 | *** join/#edev davidc__ (n=davidc__@S01060002b360aacd.vc.shawcable.net) |
| 19:30.54 | MothaWork | kergoth: sadly, if you want to use firefox, "shit that works" seems to be "top of the line" |
| 19:31.02 | kergoth | heh, true |
| 19:31.07 | MothaWork | sears's website has pages that seriously take 20-30 seconds to render on my 2400+ AthlonXP |
| 19:31.19 | kergoth | recently spent too much on a macbook, but mainly cause he wanted to try Mac OS X |
| 19:31.25 | kergoth | MothaWork: that's just sad |
| 19:31.39 | MothaWork | keram: indeed, though I partially blame sears. That website is atrocious |
| 19:31.46 | MothaWork | tons of unnecessary AJAX |
| 19:31.53 | ds2 | MothaWork: I use a ARMv6 machine.... try that for pain |
| 19:32.28 | MothaWork | kergoth: sears.com is also dog slow on IE and Opera |
| 19:34.02 | JohnnyL | Hello Mothra. |
| 19:35.05 | MothaWork | hello |
| 19:41.46 | JohnnyL | what's going on Mothra? |
| 19:42.11 | agaffney | solar: any idea when a directly booted /sbin/init would clean the env of that stuff, but when we use switch_root to run /sbin/init from an initramfs, it doesn't? |
| 19:45.50 | mnemoc | env - ? |
| 19:45.53 | *** join/#edev RobotGuy (n=Dale@pool-96-225-210-142.ptldor.fios.verizon.net) |
| 19:50.14 | MothaWork | JohnnyL: um, work |
| 19:51.39 | JohnnyL | MothaWork your luck you have a job |
| 19:51.42 | JohnnyL | lucky |
| 19:51.46 | agaffney | heh |
| 19:51.52 | JohnnyL | agaffney heh heh |
| 20:02.49 | prpplague | ho ho hum |
| 20:02.52 | prpplague | is it friday yet? |
| 20:03.19 | agaffney | yes! |
| 20:03.35 | prpplague | you lie, it isn't nice to tease people |
| 20:14.12 | agaffney | prpplague: I hope I'm not lying, because my computer says it's friday :P |
| 20:19.49 | prpplague | has pictures of CosmicPenguin panhandling with a sign that reads "will code for beer and altoids" |
| 20:21.27 | CosmicPenguin | One of the guys at AMD was heavily connected to Lefthand |
| 20:21.38 | prpplague | CosmicPenguin: oh yea? |
| 20:21.38 | CosmicPenguin | I told him that if they ever wanted to be tricked out with Linux, I would literally work for beer |
| 20:22.06 | prpplague | hehe |
| 20:22.15 | prpplague | the milk stout is my all time favorite beer |
| 20:28.54 | kergoth | CosmicPenguin: any opinion on how to store foo.bb, foo-native.bb, foo-cross.bb, .inc's, etc in a standalone repo for a package? I'm torn. thinking .bb's with version numbers in the name are unnecessary, since branches can handle any differences there.. could handle any inclusion as a branch.. oe/common, oe/foo, oe/foo-variant, oe/foo-native.. or could jut have foo.bb and foo-native.bb and such in the toplevel of the source tree.. |
| 20:28.55 | kergoth | hmm |
| 20:28.57 | kergoth | scratches head |
| 20:30.06 | CosmicPenguin | Other then with the filename, how are you thinking about handling versioning |
| 20:30.58 | CosmicPenguin | You would have to tag pretty much every git commit for each package, wouldn't you? |
| 20:31.07 | kergoth | upstream sources are on an upstream branch, tagged as upstream/1.2 or whatever for each new version coming in |
| 20:31.30 | kergoth | PR could be automatic via commit / distance from last upstream |
| 20:31.44 | kergoth | PV goes by the tag of most recent upstream merge |
| 20:32.00 | kergoth | OpenEmbedded metadata could be on its own root, or just on a branch of upstream |
| 20:32.07 | kergoth | our patches get one topgit toipc branch per patch |
| 20:32.24 | kergoth | then you just make integration branches that pull all the others together |
| 20:32.40 | kergoth | topgit can automate that. tg update would pull the latest changes from all the dependent topic branches |
| 20:32.54 | kergoth | thats what i'm thinking right now anyway |
| 20:33.04 | kergoth | but variants have me unsure |
| 20:34.59 | kergoth | to look at how debian is doing it now as an example, they have a debian branch for the debian/ changes, an upstream branch for upstream tarball imports, a number of topic branches |
| 20:35.07 | kergoth | and their master is their integration branch, they merge everything into that |
| 20:35.47 | kergoth | there's also a pristine-tar branch, that holds the extra metadata needed to produce a tarball from the upstream branch which is bit identical to the upstream tarball that was imported, in case its needed |
| 20:38.24 | kergoth | CosmicPenguin: http://git.debian.org/?p=collab-maint/mock.git;a=summary note the heads and tags there |
| 20:38.41 | kergoth | that one is simpler, since the debian metadata is in the master branch, not seperated |
| 20:38.46 | kergoth | just upstream & master |
| 20:39.31 | CosmicPenguin | okay - yeah, thats how I would do it |
| 20:40.03 | CosmicPenguin | In that case, I would ignore versioning in the file all together |
| 20:40.11 | kergoth | yeah, thats what i'm thnking too |
| 20:40.17 | kergoth | can do everything with branches |
| 20:40.23 | kergoth | what i'm wondering now is, what about -native and -cross |
| 20:41.00 | kergoth | if you put them in the source tree with the rest, on the same branch, and you add that dir to your OpenEmbedded setup, then -native would have the same ${S} as the regular package, so they couldn't be built simultaneously unless that buildsystem supports an objdir |
| 20:41.04 | kergoth | if you see what i mean |
| 20:41.26 | kergoth | if they're on separate branches, then you could have two local clones of the package, one with the regular and the other with the native changes |
| 20:41.35 | kergoth | i see advantages and disadvantages to each |
| 20:41.48 | kergoth | if the natvie changes are held in a topic branch, you have more branches to keep current when upstream updates |
| 20:42.27 | kergoth | course, unlike debian, we'd probably want an integration branch branched off of each upstream version, so we can pull fixes into all the buildable versions |
| 20:43.15 | CosmicPenguin | no - I like having native in a branch |
| 20:43.31 | CosmicPenguin | it standardizes how natives would be handled |
| 20:43.46 | kergoth | anyway, with topgit holding topic branch interdependencies, and tg update flowing the changes between the branches based on the dependency, it should be painless, and actually conceptually easier than what we have now. just hop in there.. if you need to change the .bb, check out the OpenEmbedded branch and make the change. if you need to change the code, make a topic branch off of upstream, hack away, and perhaps make an integration branch for your distro |
| 20:43.47 | CosmicPenguin | if you need a native for a package, then you know where it is located |
| 20:44.01 | kergoth | yeah, thats true.. |
| 20:44.09 | kergoth | could actually kill the package name being in the bb filename also |
| 20:44.14 | kergoth | just make it '.bb' |
| 20:44.16 | kergoth | or something |
| 20:44.17 | kergoth | heh |
| 20:45.42 | kergoth | one thing i like about this sort of this, is distros could branch in each package they need to modify, easy to track their changes. and if upstream uses git, easy to flow patches upstream, just ask them to pull it |
| 20:47.54 | kergoth | we'd have to write some scripts to simplify some parts, but that's easily done. debian has some scripts for their stuff in unstable already |
| 20:48.04 | kergoth | at any rate.. |
| 21:05.00 | keesj | the Nail kit looks is pretty cool |
| 21:05.04 | keesj | looks |
| 21:05.59 | mnemoc | *is* |
| 21:06.59 | *** join/#edev kergoth (n=kergoth@65.200.49.156) |
| 21:07.21 | keesj | :p |
| 21:07.32 | mnemoc | =) |
| 21:13.44 | keesj | I am searching for a cheap logic annalizer ( something like http://www.saleae.com/logic/features/?q=ii7&gclid=CI3nm97K9ZYCFQ6R1QodA2PUYg ) |
| 21:13.47 | keesj | any tips? |
| 21:14.06 | keesj | preferably working under linux of course |
| 21:44.58 | MothaWork | bitscope? |
| 22:02.41 | *** join/#edev landley (n=landley@m4c0e36d0.tmodns.net) |
| 22:02.42 | *** mode/#edev [+o landley] by ChanServ |
| 22:02.46 | scanlime | I think SparkFun just added a cheap logic analyzer to their inventory... |
| 22:03.15 | scanlime | yea |
| 22:03.17 | scanlime | http://www.sparkfun.com/commerce/product_info.php?products_id=8938 |
| 22:03.39 | scanlime | no idea if it's any good.. but the price is right, and the 24MHz sampling rate isn't awful |
| 22:04.22 | scanlime | Personally I have a BitScope, and it works pretty well as a logic analyzer or mixed signal analyzer- but if you're mostly using it for logic, you'll probably be disappointed. The software for logic is pretty basic, and the sample buffer is relatively small. |
| 22:05.32 | scanlime | I've also used a Cypress FX2 microcontroller as a logic analyzer before- just ship 8 bits of data across USB 2.0 as fast as possible, then analyze it on a PC |
| 22:05.45 | scanlime | that approach is kinda nice, if you want to do detailed analysis on the data |
| 22:06.00 | scanlime | I ended up writing a unix pipe-based set of tools for capturing and analyzing various kinds of signals |
| 22:06.20 | scanlime | You could do the same thing with an FTDI USB FIFO chip |
| 22:17.50 | MonMotha | scanlime: you'll get no argument from me that the bitscope's software sucks but the hardware is at least documented |
| 22:20.18 | *** join/#edev RobotGuy (n=robotguy@pool-96-225-210-142.ptldor.fios.verizon.net) |
| 22:22.20 | scanlime | MonMotha: yeah. Do you know if anyone's started a good open source software replacement for it? |
| 22:24.00 | MonMotha | not that I am aware of |
| 22:39.14 | JohnnyL | Hello Monday Mothra. |
| 22:47.18 | prpplague | scanlime: hmm, with the right stuff, you could use the flyswatter |
| 22:48.06 | landley | Ok, so I grant you that John Glenn could probably perform bus analysis with a fly swatter. |
| 22:48.14 | landley | It's a bit like the chuck norris/bruce schneier thing. |
| 22:48.24 | prpplague | hehe |
| 22:48.25 | landley | (How else do you think Apollo 13 survived?) |
| 22:48.34 | landley | Those nasa guys were _tough_... |
| 22:48.42 | prpplague | landley: indeed they were |
| 22:48.56 | landley | It's too bad we STOPPED DOING THAT back in the 1970's. (Grumble.) |
| 22:49.21 | landley | Tasking a federal bureaucracy with frontier exploration: not very effective, as strategies go... |
| 22:49.29 | prpplague | landley: indeed, not only that, but we stopped working hard to stay ontop |
| 22:49.37 | landley | Armadillo aerospace should have a podcast. |
| 22:50.13 | prpplague | hangs his head low...... grocery shopping after work today :( |
| 22:50.15 | RobotGuy | We are still graduating students from High School who can't read and/or do simple math |
| 22:51.13 | RobotGuy | When I was growing up, you got held back if you couldn't get a handle on the material you should have learned. |
| 22:51.20 | landley | RobotGuy: public school is a joke. |
| 22:51.33 | RobotGuy | landley: Sadly, yes, it is these days. |
| 22:51.49 | prpplague | when my kids moved back to the states from barbados, the teacher was totally amazed at what they knew for their grade level, she said they should have been atleast 2 maybe 3 grades ahead of where they were |
| 22:51.55 | landley | It got a serious jolt due to sputnik, but Regan beat it until it was on life support, and then the "no child allowed ahead act" finished it off. |
| 22:52.05 | landley | An actor and the edumacation president. Sheesh. |
| 22:52.29 | landley | prpplague: If I manage to have kids I'm home schooling 'em. |
| 22:52.36 | RobotGuy | I was pretty much always ahead of my classes. |
| 22:52.46 | landley | Not for religious reasons, but because when I went community college I earned money by tutoring Math 0.0.1. |
| 22:52.52 | prpplague | landley: never heard that statement before, how did Regan damage public education? |
| 22:53.02 | landley | Addition, subtraction, multiplication, division, and ratios. Tought _in_college_ to people who hadn't learned it yet. |
| 22:53.17 | landley | If community colleges do that, what's high school for? (Or most of middle school, for that matter.) |
| 22:53.35 | landley | Give 'em a decent education and then put them in the local community college's remedial classes around age 7. |
| 22:54.09 | prpplague | hehe |
| 22:54.25 | landley | prpplague: the "ketchup is a vegetable" guy? Cutting funding, allowing the religious loonies to start undermining teaching evolution in biology and such, unfunded mandates... The usual. |
| 22:54.59 | landley | Most people didn't notice because the schools were in almost reasonable shape back then, it took a while for the damage to show. |
| 22:55.07 | RobotGuy | There are fanatics in every group. |
| 22:55.27 | prpplague | landley: i generally don't equate funding with quality of education |
| 22:55.30 | landley | Yes, but they shouldn't form your "base". |
| 22:55.52 | landley | prpplague: Once you hit 40 kids to a classroom, what's the teacher there for? |
| 22:56.03 | *** join/#edev linac (n=lin@121.201.102.242) |
| 22:56.03 | landley | Just do a telecourse at that point. |
| 22:56.06 | landley | They had videotape back then. |
| 22:57.12 | landley | Administratively lumping together "gifted and talented" with the retarded students was a nice touch too. |
| 22:57.20 | landley | If you can't be shoehorned into the bell curve, you are a problem. |
| 22:57.22 | landley | Great. |
| 22:57.38 | prpplague | landley: granted that 40 to a class is somewhat extreme, but like the school my kids go to, its about 90% hispanic illegals, they get free meals for breakfast, lunch , and snack right after school, free books, free school supplies, free after school daycare, and free after hours tutoring |
| 22:58.31 | landley | shrugs. |
| 22:58.40 | landley | That's collating problems. |
| 22:59.02 | landley | I also kind of doubt it's _90%_ hispanic illegals. |
| 22:59.18 | landley | They'd probably close that school location if that was the case. |
| 22:59.18 | prpplague | actually the number is 91% |
| 22:59.23 | prpplague | its a title 1 school |
| 22:59.28 | landley | California? |
| 22:59.45 | prpplague | Hurst-Euless-Bedford ISD |
| 23:00.23 | landley | Which is... not in california? |
| 23:00.35 | landley | I'm guessing it's california because they're essentially their own country when it comes to this sort of stuff. |
| 23:01.05 | prpplague | landley: dallas ft.worth area |
| 23:01.15 | landley | Ah, right. I keep forgetting. :) |
| 23:01.34 | landley | shrugs. I dunno your local government. |
| 23:02.07 | prpplague | http://www.ed.gov/policy/elsec/leg/esea02/pg1.html |
| 23:02.14 | landley | It's not an argument _against_ my believe that public school is government funded day care with only a vestigial educational purpose. |
| 23:02.34 | landley | belief. |
| 23:02.55 | prpplague | landley: indeed, problem is they allow illegals to go to school |
| 23:03.10 | prpplague | landley: they give them free stuff for breaking the law |
| 23:04.37 | landley | Technically, they give them free stuff for their parents breaking the law. |
| 23:04.47 | landley | Most school children are minors, with very little say in where they live. |
| 23:05.07 | prpplague | landley: i went totally ballistic the other day, my son got in trouble for talking in class, yes he was wrong and the teacher told himto write "i will not talk in class" but she told him to write it in spanish |
| 23:05.24 | landley | I mentioned this was collating problems. |
| 23:05.35 | prpplague | landley: that is true, but if the parents knew they wouldn't have their kids taken care of for free would they still bring them here? |
| 23:05.46 | landley | Illegal immigration and education of citizens are two different things. They should be handled separately. But they're not. |
| 23:06.15 | prpplague | argh gotta go, they are closing the office |
| 23:06.20 | prpplague | back later |
| 23:17.49 | beewoolie | Looks like some hard-code embedded systems engineering going on here. |
| 23:25.54 | landley | beewoolie: yes, actually. :) |
| 23:26.23 | RobotGuy | Does anyone here understand software defined radio? |
| 23:56.51 | *** join/#edev davidc__ (n=davidc__@S01060002b360aacd.vc.shawcable.net) |