| 00:07.12 | *** join/#edev _lifeless (n=patrick@90.150.114.23) |
| 00:12.34 | RobotGuy | Is anyone awake? |
| 00:20.25 | *** join/#edev ico2_ (n=ico2@modem-163.chameleon.dialup.pol.co.uk) |
| 00:35.35 | *** join/#edev Crofton (n=balister@66.207.66.26) |
| 00:37.29 | *** join/#edev chouimat|busy (n=dieu@r2351064.cidc.net) |
| 00:38.28 | RobotGuy | Crofton: Hello |
| 00:57.34 | *** join/#edev davidc__ (n=davidc__@S01060002b360aacd.vc.shawcable.net) |
| 01:10.27 | *** join/#edev RobotGuy (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 01:17.46 | *** join/#edev chouimat|away (n=dieu@kde/developer/chouinard) |
| 01:18.37 | RobotGuy | Knock knock.. |
| 01:29.24 | RobotGuy | prpplague: Are you listening?? |
| 02:19.14 | *** join/#edev ico2_ (n=ico2@modem-869.chimpanzee.dialup.pol.co.uk) |
| 02:20.21 | prpplague | RobotGuy: that sparkfun module is intended to be used with rs-232 only |
| 02:20.37 | prpplague | RobotGuy: if you are wanting to do usb bluetooth, best to use a standard usb dongle |
| 02:20.41 | RobotGuy | prpplague: OK. |
| 02:20.56 | RobotGuy | prpplague: I am going to use the BT module I have and connect it to the hammer console port. |
| 02:21.13 | RobotGuy | prpplague: I was just wondering about the USB. |
| 02:21.14 | prpplague | yep |
| 02:21.24 | prpplague | RobotGuy: the usb connections aren't used |
| 02:21.46 | RobotGuy | prpplague: They are brought out to the pins on the module, as are the SPI connections. |
| 02:22.18 | prpplague | RobotGuy: yep, you have to have different firmware to use those, the module that SFE only has firmware to be used as a rs-232 module |
| 02:22.35 | RobotGuy | prpplague: Why would the USB and SPI and such be connected to pins on the module if they can not be used? |
| 02:22.47 | RobotGuy | prpplague: Ahhhh, OK, different firmware. Hmmmmm. |
| 02:22.51 | prpplague | yea |
| 02:23.10 | RobotGuy | prpplague: Hmmm, that seems to indicate that such firmware must be available somewhere. |
| 02:23.22 | prpplague | RobotGuy: yea for a cost |
| 02:23.52 | RobotGuy | prpplague: I have just attempted a new buildroot toolchain build that errored out. Do you want a pastebin of it? |
| 02:24.10 | prpplague | i'll look but its late |
| 02:24.50 | RobotGuy | prpplague: http://www.pastebin.ca/875528 |
| 02:25.39 | prpplague | hmm, never seen that before |
| 02:25.50 | prpplague | RobotGuy: i not sure how you keep seeing so many issues |
| 02:25.59 | RobotGuy | prpplague: Me either, and it's frustrating. |
| 02:26.05 | prpplague | RobotGuy: i've been using buildroot for 10 years and never had so many problems |
| 02:26.26 | prpplague | RobotGuy: you are using the lastest buildroot posted on the wiki? |
| 02:26.31 | RobotGuy | prpplague: Maybe I should just give up then. |
| 02:27.15 | RobotGuy | prpplague: No, the latest from svn. |
| 02:27.17 | prpplague | RobotGuy: no, just need to standardize on one build, and one distro |
| 02:27.28 | prpplague | RobotGuy: ahh |
| 02:27.56 | RobotGuy | prpplague: I patched in the Samsung stuff and the Squeak package. You did tell me to do that awhile back. |
| 02:28.34 | RobotGuy | prpplague: The newer buildroots have added software in them. |
| 02:28.49 | *** join/#edev _lifeless (n=patrick@89.20.108.146) |
| 02:30.48 | prpplague | RobotGuy: no i said you could do that, i didn't say it was a good idea |
| 02:31.09 | RobotGuy | prpplague: Geeesh. |
| 02:31.45 | prpplague | RobotGuy: you need to focus on what you need to do, and accomplish some small goals one at a time |
| 02:32.23 | RobotGuy | prpplague: I have been doing that. I just want to get a toolchain built! I didn't know "make clean" would zap my toolchain too. |
| 02:32.44 | prpplague | RobotGuy: you done several toolchains |
| 02:32.53 | prpplague | RobotGuy: why do you keep changing? |
| 02:34.13 | RobotGuy | prpplague: I did not mean to zap my toolchain. |
| 02:34.28 | *** join/#edev dijenerate (n=dijenera@72.51.91.199) |
| 02:34.47 | prpplague | RobotGuy: should be able to rebuild it if all you did was a make clean |
| 02:35.08 | RobotGuy | prpplague: Should yes, can, no. |
| 02:36.00 | prpplague | RobotGuy: unfortuantly i can't help you unless i'm using the same buidlroot with the same patches, and the same build |
| 02:36.51 | RobotGuy | prpplague: OK. Well, it's obvious I can't make this all work so maybe I should just send everything back to you and you can give the kit to somebody smarter than me. |
| 02:37.22 | RobotGuy | prpplague: I'll pay for the carrier board next month because I've soldered to it. |
| 02:37.30 | prpplague | hehe, its your kit, i'd like to see you get it working |
| 02:37.51 | prpplague | maybe you should start with the base buildroot until you get your feet wet |
| 02:37.58 | RobotGuy | prpplague: I would like nothing better. I really like Hammer a LOT, and I know it would be great for my robots. |
| 02:38.19 | prpplague | RobotGuy: why not start with the "official" buildroot for the hammer |
| 02:38.54 | RobotGuy | prpplague: I started out that way. |
| 02:39.04 | prpplague | RobotGuy: why did you change? |
| 02:39.22 | RobotGuy | prpplague: To get more software choices and updates with current patches. |
| 02:39.47 | prpplague | RobotGuy: did you really need them to get started with the board? |
| 02:40.12 | RobotGuy | prpplague: Not right away, but the current buildroot has lua in it. |
| 02:40.55 | prpplague | RobotGuy: the best thing to do, is go back the base buildroot, learn all you can, then when you actually need the extra software, then move up |
| 02:41.14 | prpplague | RobotGuy: more likely by the time you need it, the buildroot will be updated anyway |
| 02:42.30 | RobotGuy | prpplague: So are you saying to stay with the stock hammer buildroot until the Hammer stuff is in the official buildroot? |
| 02:42.55 | RobotGuy | I think Hammer is great and one of the best things to happen for robotics lately. |
| 02:43.32 | prpplague | RobotGuy: i think for you, yes, you need to stick with the stock buildroot for the hammer, atleast that way i can help you |
| 02:43.34 | RobotGuy | It's fun to see people's reactions when I show my hammer off and tell them about it. |
| 02:44.33 | RobotGuy | OK, well I will go back to the hammer buildroot, but I am also going to keep tinkering with the current svn buildroot (separate location, of course). |
| 02:46.48 | RobotGuy | I am used to being able to stay updated with the latest and greatest stuff, but I guess that is not possible with buildroot. |
| 02:47.00 | prpplague | RobotGuy: thats fine, but i think you'll get more done by using the buildroot for hammer |
| 02:47.14 | prpplague | RobotGuy: since thats what we tested the most with |
| 02:47.17 | RobotGuy | prpplague: OK. |
| 02:47.47 | prpplague | RobotGuy: and if you don't mind, i'd like to suggestion that you make a game plan for what items you need to get working |
| 02:47.52 | prpplague | RobotGuy: do one at a time |
| 02:49.13 | RobotGuy | prpplague: I already have my kernel configuration. My initrd is under 6 MB compressed. |
| 02:49.41 | prpplague | RobotGuy: what size is the uncompressed image? |
| 02:50.03 | RobotGuy | prpplague: My mistake, that is the uncompressed image size. |
| 02:50.23 | prpplague | RobotGuy: ahh, good deal |
| 02:50.32 | prpplague | long day tomorrow |
| 02:50.36 | RobotGuy | prpplague: And I have the i2c in. Good night then |
| 02:51.05 | prpplague | RobotGuy: just a reminded, you probably don't need the i2c driver unless you have a driver for the specific device you are working with |
| 02:51.17 | prpplague | RobotGuy: what device are communicatiung with via i2c? |
| 02:51.35 | RobotGuy | prpplague: I will be using i2c to communicate with dsPICs and MCP23017 chips. |
| 02:51.54 | prpplague | are their linux drivers for both of those via i2c? |
| 02:52.03 | RobotGuy | prpplague: Plus various sensors. |
| 02:52.17 | RobotGuy | prpplague: There are no linux drivers in the kernel, but there shouldn't have to be. |
| 02:52.19 | prpplague | RobotGuy: that have linux drivers? |
| 02:52.28 | prpplague | RobotGuy: why not? |
| 02:52.44 | RobotGuy | prpplague: I don't see why there needs to be a driver for each specific device. |
| 02:53.01 | prpplague | RobotGuy: how do expect to communicate with them? |
| 02:53.09 | RobotGuy | prpplague: I should be able to just read/write the i2c devices according to what I have read so far. |
| 02:53.24 | RobotGuy | prpplague: The i2c-0, i2c-1, etc devices. |
| 02:53.26 | prpplague | RobotGuy: from user land? |
| 02:53.28 | prpplague | RobotGuy: ahh |
| 02:53.31 | RobotGuy | prpplague: Yes |
| 02:53.47 | prpplague | RobotGuy: that will increase you latency dramatically |
| 02:54.06 | prpplague | and the complexity |
| 02:54.26 | RobotGuy | prpplague: Maybe, but I don't know anything about writing Linux drivers. |
| 02:54.59 | prpplague | okie dokie, well i'm sure there are plenty of bot people that have the same questions and devices |
| 02:55.03 | prpplague | later |
| 02:55.03 | RobotGuy | prpplague: Everything I will be doing with the new W.A.L.T.E.R. v3.0 revolves around i2c. |
| 02:55.07 | RobotGuy | Gnoght |
| 02:55.10 | RobotGuy | G'Night |
| 03:15.54 | *** join/#edev redb1ue (i=star@ppp086.109-253-207.mtl.mt.videotron.ca) |
| 03:42.35 | *** join/#edev roxfan2 (n=dunno@205.189-136-217.adsl-dyn.isp.belgacom.be) |
| 03:55.56 | *** join/#edev ico2_ (n=ico2@modem-869.chimpanzee.dialup.pol.co.uk) |
| 03:57.20 | *** join/#edev _lifeless (n=patrick@90.150.122.106) |
| 04:11.44 | *** join/#edev davidc__ (n=davidc__@S01060002b360aacd.vc.shawcable.net) |
| 04:41.22 | *** join/#edev DrZimmerman (n=theo@2001:5c0:8fff:fffe:0:0:0:6af1) |
| 04:44.02 | *** part/#edev rom10 (n=raj6@CPE00062575886a-CM00186832000a.cpe.net.cable.rogers.com) |
| 05:04.01 | *** join/#edev chouimat|busy (n=dieu@r2351064.cidc.net) |
| 05:56.48 | *** join/#edev ico2_ (n=ico2@modem-869.chimpanzee.dialup.pol.co.uk) |
| 05:57.45 | *** join/#edev davidc__ (n=davidc__@S01060002b360aacd.vc.shawcable.net) |
| 06:01.47 | Mirell | Woo, might get my laptop tomorrow. |
| 06:42.40 | RobotGuy | Cool |
| 06:42.49 | RobotGuy | Mirell: Hello |
| 06:42.59 | RobotGuy | Finally somebody to talk to. |
| 06:48.19 | RobotGuy | I guess not. Bye |
| 07:37.55 | *** join/#edev steliosk (n=Stelios@athedsl-321178.home.otenet.gr) |
| 07:57.34 | *** join/#edev ico2_ (n=ico2@modem-869.chimpanzee.dialup.pol.co.uk) |
| 08:30.22 | *** join/#edev howtoo (n=howtoo@195.159.62.66) |
| 09:23.48 | *** join/#edev hw_ (n=hw@p578b3905.dip0.t-ipconnect.de) |
| 09:37.53 | *** join/#edev redb1ue (i=star@ppp219.108-253-207.mtl.mt.videotron.ca) |
| 09:38.27 | *** join/#edev lyakh (n=lyakh@p57BD33C6.dip0.t-ipconnect.de) |
| 10:07.46 | steliosk | anyone tried to build a .24 kernel for walnut ? |
| 10:08.12 | steliosk | the ppc arch is going away and they have migrated a lot of the ppc boards to powerpc |
| 10:08.22 | steliosk | but i am not sure about the dht-walnut |
| 10:19.24 | *** join/#edev ico2_ (n=ico2@modem-869.chimpanzee.dialup.pol.co.uk) |
| 11:13.29 | *** join/#edev T0mW (n=Tom@24.238.86.158.res-cmts.sth.ptd.net) |
| 11:55.03 | *** join/#edev howtoo (n=howtoo@195.159.62.66) |
| 11:55.44 | *** join/#edev harpal (n=harpal@124.125.76.8) |
| 11:57.15 | *** join/#edev chouimat|away (n=dieu@r2351064.cidc.net) |
| 11:58.58 | harpal | Hey When I try to do hwclock --hctosys than it shows me error hwclock: can't open '/dev/misc/rtc': No such device |
| 11:59.06 | harpal | I have just created file with mknod /dev/misc/rtc c 10 135 but still have same error |
| 11:59.18 | harpal | any one tell me what should be the problem? |
| 12:09.15 | SpeedEvil | you have the rtc module installed, or compiled into your kenel (stab in the dark) |
| 12:19.58 | *** join/#edev ico2_ (n=ico2@modem-2526.cougar.dialup.pol.co.uk) |
| 12:21.06 | *** join/#edev dijenerate (n=dijenera@66.54.118.16) |
| 12:22.13 | *** join/#edev emil|k (n=emil@f054015164.adsl.alicedsl.de) |
| 12:37.11 | *** join/#edev RobotGuy (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 12:56.15 | *** join/#edev redb1ue (i=star@ppp151.108-253-207.mtl.mt.videotron.ca) |
| 13:07.08 | *** join/#edev RobotGuy (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 13:13.04 | *** join/#edev FireEgl (i=FireEgl@4.0.0.0.1.0.0.0.c.d.4.8.0.c.5.0.1.0.0.2.ip6.arpa) |
| 13:18.31 | T0mW | MonMotha: you around? |
| 13:28.09 | harpal | SpeedEvil: Sorry I was not at my desk, I have searched rtc in make menuconfig and than I found many options so which options are required |
| 13:29.15 | SpeedEvil | If you mean 'enhanced real time clock support' |
| 13:29.56 | SpeedEvil | device-drivers -> char devices -> enhanced real time... |
| 13:30.51 | SpeedEvil | actgually - it may be the one below - generic RTC |
| 13:31.20 | harpal | I have some rtc driver in my kernel which I have compiled I have loaded all three drivers |
| 13:31.29 | *** join/#edev __lifeless (n=patrick@90.150.113.222) |
| 13:31.36 | harpal | rtc_lib rtc_cmos rtc_core |
| 13:32.38 | harpal | SpeedEvil: In my kernel config I havent selected Enhanced Real Time Clock support |
| 13:33.23 | harpal | So should I enable it and recomopile? |
| 13:33.37 | harpal | Is there other modules required? |
| 13:35.29 | SpeedEvil | I don't think so |
| 13:36.43 | harpal | ok |
| 13:56.09 | *** join/#edev prpplague (n=dave@mail.americanmicrosystems.com) |
| 13:58.37 | *** join/#edev dijenerate (n=dijenera@66.54.118.16) |
| 14:03.07 | T0mW | prpplague: hey, I need your help in bricking something |
| 14:04.06 | prpplague | T0mW: ok, well , i'm very qualified |
| 14:04.55 | T0mW | LOL |
| 14:05.25 | T0mW | prpplague: on the Zipit Z2 I need to replace the kernel that is in Flash, but I have no access to blob or jtag, ok? |
| 14:05.35 | T0mW | prpplague: but, I do have the MTDBLOCKs |
| 14:05.51 | prpplague | T0mW: to all of the mtdblocks? |
| 14:05.54 | T0mW | yes |
| 14:06.09 | T0mW | let me reboot it. |
| 14:06.21 | prpplague | T0mW: you sure of that? can you pastebin.ca the section of dmesg related to mtd setup? |
| 14:06.43 | prpplague | T0mW: surely they have jtag brought out |
| 14:07.29 | T0mW | nope, another brain-fart idea of management: "security" |
| 14:08.35 | T0mW | prpplague: drat, I will have to get back to you on this, I kicked off OE in rebuilding the working bootsystem. I have nothing to boot right now. |
| 14:08.49 | T0mW | Oh, I do.. |
| 14:11.18 | prpplague | T0mW: how the hell do they flash it |
| 14:14.10 | prpplague | T0mW: one sec phone |
| 14:14.11 | T0mW | prpplague: I think they preprogram the chips |
| 14:15.28 | RobotGuy | Mornin' |
| 14:15.43 | Mirell | <PROTECTED> |
| 14:18.06 | T0mW | prpplague: http://rafb.net/p/YGzQ1576.html |
| 14:22.19 | *** join/#edev ico2__ (n=ico2@modem-2870.chameleon.dialup.pol.co.uk) |
| 14:25.49 | *** join/#edev chouimat|work (n=mchouina@209.217.106.98) |
| 14:27.23 | harpal | I have compiled kernel , now I want somethink like DESTDIR in kernel, is there any way ? |
| 14:27.26 | T0mW | prpplague: we are thinking using kexec, we found a way to get that into Flash |
| 14:27.45 | harpal | I have done it earlier but I forgot how to do that |
| 14:30.01 | prpplague | ~WIAV |
| 14:30.01 | purl | i guess wiav is a patent troll with information at http://trolltracker.blogspot.com/2008/01/4-interesting-new-cases-from-last-2.html and http://www.setexasrecord.com/news/205753-texas-company-seeks-judgement-to-stop-potential-patent-suit |
| 14:30.48 | prpplague | T0mW: yea looks like you have clean access |
| 14:31.04 | *** join/#edev chouimat|work (n=mchouina@209.217.106.98) |
| 14:31.14 | *** join/#edev GPSFan (n=kenm@12.10.255.246) |
| 14:31.14 | *** mode/#edev [+o GPSFan] by ChanServ |
| 14:31.14 | *** mode/#edev [+o prpplague] by ChanServ |
| 14:31.30 | prpplague | GPSFan: hey hey |
| 14:31.44 | GPSFan | prpplague: mornin |
| 14:31.49 | chouimat|work | morning |
| 14:32.16 | RobotGuy | Is it still moring if I haven't slept yet? |
| 14:32.57 | chouimat|work | RobotGuy: yes :) |
| 14:33.21 | *** join/#edev chouimat|busy (n=dieu@r2351064.cidc.net) |
| 14:38.08 | harpal | Any one know that? |
| 14:39.02 | RobotGuy | harpal: What kind of system are you working on? ARM? Other? |
| 14:40.06 | harpal | RobotGuy: I have just x86 system, so I have just compiled kernel and setup all my system to one folder |
| 14:40.38 | RobotGuy | harpal: What is it you need a DESTDIR for? |
| 14:41.05 | harpal | to put my all modules in a folder where I am making system |
| 14:41.39 | RobotGuy | harpal: You do a 'make modules_install' and modules are put in /lib/modules/<kernel-version> |
| 14:42.07 | harpal | option is that I can just install it and put it to my original system and than move it. but any other? |
| 14:42.25 | RobotGuy | harpal: Why do you need to move the modules? |
| 14:43.28 | RobotGuy | harpal: Do you need to make an image with the modules? |
| 14:43.49 | harpal | ya |
| 14:44.15 | harpal | Hey I need to go I will be back tomo |
| 14:44.24 | harpal | RobotGuy: Thanks for your help |
| 14:44.30 | RobotGuy | harpal: I don't know exactly what you are doing, but when I make an image for my Hammer (ARM9), I have a script that moves the modules where I need them. |
| 14:44.40 | harpal | ok |
| 14:50.13 | RobotGuy | Another toolchain build bombed. :( |
| 14:50.34 | *** join/#edev Mirell (n=mirell@otomi.ling.utexas.edu) |
| 14:50.51 | RobotGuy | I haven't built a successful toolchain build since I switched to Kbuntu. |
| 14:51.45 | *** join/#edev sjhill (n=sjhill@real.realitydiluted.com) |
| 14:59.18 | RobotGuy | prpplague: I have a pastebin if you are interested - using the 'standard' Hammer buildroot. |
| 14:59.30 | prpplague | RobotGuy: ok |
| 15:00.40 | RobotGuy | prpplague: http://www.pastebin.ca/876037 |
| 15:01.31 | RobotGuy | I am wondering if the symlink I had for buildroot to the real buildroot caused a problem. |
| 15:02.03 | prpplague | RobotGuy: thats the same problem as before that GPSFan mentioned |
| 15:02.30 | RobotGuy | prpplague: I don't remember that problem. |
| 15:03.13 | prpplague | RobotGuy: you need to make that directory since the install script isn't making it |
| 15:03.26 | RobotGuy | usr/lib |
| 15:03.27 | prpplague | RobotGuy: mkdir /src/hammer/buildroot-h/project_build_arm/walter/root/usr/lib/ |
| 15:03.36 | prpplague | RobotGuy: then rerun make |
| 15:03.44 | *** join/#edev landley (n=landley@cpe-72-177-114-214.austin.res.rr.com) |
| 15:06.19 | RobotGuy | prpplague: Why isn't that directory getting made? |
| 15:06.53 | prpplague | RobotGuy: the installation script is expecting it be there, just needs a tweak to add it |
| 15:07.09 | prpplague | RobotGuy: you can add it to your skeleton if you want |
| 15:11.39 | RobotGuy | prpplague: I added it to buildroot/target/device/Samsung/Hammer/targ*/usr |
| 15:12.34 | prpplague | RobotGuy: that should work, but it really needs to do a mkdir in the installation |
| 15:12.45 | RobotGuy | prpplague: I know. |
| 15:16.07 | RobotGuy | prpplague: I don't see where it has been fixed. |
| 15:16.31 | RobotGuy | prpplague: I mean, as in browsing the buildroot repo. |
| 15:16.59 | prpplague | RobotGuy: yea it probably doesn't show up on some of the builds due to several facts, first that most people aren't building the c++, the other is that their target skeletons probably have the /usr/lib |
| 15:19.40 | prpplague | RobotGuy: yea just looked at the makefile, it needs a directory check |
| 15:20.36 | RobotGuy | prpplague: I figured it would not be difficult to find for somebody who knows makefiles. |
| 15:22.00 | RobotGuy | To me they are Makemyheadhurtfiles |
| 15:24.57 | RobotGuy | Well, I'm not going out today - snow/ice all over the place around where I live. Mostly ice now. |
| 15:25.08 | prpplague | RobotGuy: this kind of stuff is what companies like codesourcery, montavista and timesys do for their toolchains |
| 15:25.37 | RobotGuy | prpplague: No doubt. I can't afford any of them. |
| 15:25.40 | prpplague | RobotGuy: most of it is the basic gnu toolchain, but they spend alot of time tweaking it so you don't have to deal with these types of issues |
| 15:26.24 | prpplague | RobotGuy: indeed, but this goes back to what we were discussing about codesourcery, when codesorucery finds an issue, they general post a patch to fix it and give back to the project |
| 15:26.31 | prpplague | RobotGuy: alot of other companies do not |
| 15:27.27 | RobotGuy | prpplague: I understand all that. Really, I do. :) They are way out of my price range though. |
| 15:27.59 | prpplague | RobotGuy: right, i'm not saying you should buy it |
| 15:29.15 | RobotGuy | prpplague: I am still considering Crossworks, because they do have a Personal License I can afford and DO provide full support, even though they may not give back much to Open Source. |
| 15:29.21 | prpplague | one sec phone |
| 15:29.26 | *** join/#edev g1powermac (n=g1powerm@unaffiliated/g1powermac) |
| 15:32.10 | *** join/#edev linac (n=lin@121.41.36.85) |
| 16:04.16 | prpplague | RobotGuy: surprised that dobermann didn't continue with his project |
| 16:04.26 | prpplague | RobotGuy: a hammer board would have been perfect for him |
| 16:04.37 | prpplague | landley: greetings |
| 16:05.23 | RobotGuy | Is he another Lynxmotion user by chance? |
| 16:06.19 | prpplague | RobotGuy: yea, you were chatting with him |
| 16:06.57 | RobotGuy | prpplague: I thought I remembered him from Lynxmotion. |
| 16:08.07 | RobotGuy | Got another PARTS meeting this Saturday. :) Taking W.A.L.T.E.R. and my Hammer, but I don't have much more to show of Hammer like I was hoping to have. |
| 16:09.09 | landley | prpplague: morning. |
| 16:09.20 | prpplague | RobotGuy: yea, i saw that the meeting is on saturday |
| 16:09.37 | prpplague | RobotGuy: one of these days i'm gonna get the boss to spring for a plane ticket |
| 16:09.59 | prpplague | landley: whats the word on web wackin work? |
| 16:10.05 | prpplague | (say that 10 times) |
| 16:10.53 | RobotGuy | prpplague: How do I get rid of what seems to be old buildroot names (like buildroot-h) without just starting all over again? |
| 16:11.19 | prpplague | RobotGuy: you mean old buildroot projects? |
| 16:12.43 | RobotGuy | prpplague: No. I had a symlink (ln -s buildroot-h buildroot), but that didn't seem to work and I went back to a real name dir buildroot instead of buildroot-h. |
| 16:13.25 | RobotGuy | Buildroot doesn't seem to like being symlinked. |
| 16:13.54 | prpplague | RobotGuy: ahh, you can't, toolchains are directory sensitive, everything is absolute path'd |
| 16:14.30 | RobotGuy | prpplague: That really sucks. |
| 16:14.39 | prpplague | RobotGuy: just the way it works |
| 16:15.32 | RobotGuy | Why does it have to be absolute paths? |
| 16:15.57 | *** join/#edev samuelr (n=samuelr@dsl093-061-094.pit1.dsl.speakeasy.net) |
| 16:17.15 | landley | prpplague: the job's probably mine if I want it, the question is "do I"? |
| 16:17.22 | landley | They gave me a take home test, which I'm doing now. |
| 16:17.33 | prpplague | landley: hehe |
| 16:17.35 | landley | I have to learn php and mysql to finish it. That adds an extra ~4 hours. |
| 16:19.58 | prpplague | landley: fun fun |
| 16:20.25 | prpplague | landley: got confirmation, the nail boards shipped from the pcb house this morning |
| 16:20.31 | prpplague | landley: should be here wed. |
| 16:20.42 | prpplague | landley: then we gotta assemble and test |
| 16:22.32 | *** join/#edev ico2__ (n=ico2@modem-2870.chameleon.dialup.pol.co.uk) |
| 16:22.42 | prpplague | landley: some of the guys found this - http://www.linuxdevices.com/news/NS4837503255.html |
| 16:22.55 | prpplague | landley: appears to be vaporware as it was announced back a year ago |
| 16:24.08 | prpplague | landley: the nail is the same width, 1.1" , but a little longer as their's is 3.2" and the nail is 5" |
| 16:26.19 | landley | I've seen a couple other vaporware ones. Most recently from a european company. |
| 16:26.19 | RobotGuy | That Tux/Stick doesn't have any more RAM or Flash than Hammer though. |
| 16:27.20 | prpplague | oh interesting lcd - http://www.sparkfun.com/commerce/product_info.php?products_id=8568 |
| 16:27.34 | prpplague | landley: yea |
| 16:27.56 | landley | Somebody was using the name "linux on a stick", but googling now I find 5 distros using that phrase... and no hardware. :P |
| 16:28.18 | landley | prpplague: nice. |
| 16:28.55 | prpplague | landley: yea, those distro's are more for thumbdrives |
| 16:30.46 | RobotGuy | This one http://www.sparkfun.com/commerce/product_info.php?products_id=8358 looks interesting. |
| 16:32.41 | prpplague | RobotGuy: yea, the lcd is nice, the serial interface is a waste thou |
| 16:32.58 | RobotGuy | This http://www.sparkfun.com/commerce/product_info.php?products_id=8490 is still the display I really want. |
| 16:33.50 | cajaka | You've guys seen this one: http://www.calao-systems.com/?lng=en ? |
| 16:34.41 | prpplague | cajaka: yea that was posted on linuxdevices awhile back, but i thought it was vaporware as well |
| 16:35.37 | *** join/#edev bbradley (n=bbradley@82-32-35-42.cable.ubr05.azte.blueyonder.co.uk) |
| 16:37.48 | prpplague | hmm thats about $225USD |
| 16:37.59 | prpplague | alot more ram and flash, but much harder to work with |
| 16:40.00 | prpplague | GPSFan: ping |
| 16:45.27 | landley | prpplague: ah yes, that calao one was the vaporware I was thinking of. |
| 16:46.03 | prpplague | landley: i thought it was vaporware as well, but it looks to be available |
| 16:46.56 | prpplague | http://mobile.slashdot.org/mobile/08/01/28/136204.shtml |
| 16:46.58 | prpplague | interesting |
| 16:47.22 | prpplague | mot and nokia seem to be buying up linux companies |
| 16:50.18 | landley | Trolltech ceased being interesting to me when they relicensed QT under GPLv3. |
| 16:50.25 | landley | I'll be moving off KDE when that change makes it through. |
| 16:52.35 | prpplague | landley: indeed, still QT has a large market share on the handheld and mobile area |
| 16:52.48 | prpplague | landley: we get alot of pressure to use QT from out vendors |
| 16:54.05 | chouimat|work | landley: Qt is under commercial/GPL2/GPL3 |
| 16:54.55 | landley | is it? |
| 16:55.04 | landley | I heard they were going gplv3 only, but it could have been garbled. |
| 16:55.11 | landley | It was already under "gplv2 or later", wasn't it? |
| 16:55.19 | chouimat|work | landley: no GPL2 only ... |
| 16:55.44 | landley | Heck, gplv2 and the puppy-killing license. Or CDDL. |
| 16:55.46 | chouimat|work | i'm checking the latest snapshot , just to be sure |
| 16:55.54 | landley | (Is there a license sillier than CDDL? Maybe some of Microsoft's...) |
| 16:56.03 | chouimat|work | landley: wtfpl |
| 16:56.34 | landley | With dual licenses I can ignore the ones I don't plan to use. |
| 16:56.50 | landley | It means I can't contribute to the project, but that's their problem. |
| 16:58.52 | RobotGuy | Is KDE in trouble? |
| 16:59.00 | chouimat|work | no |
| 16:59.56 | chouimat|work | landley: LICENSE.SNAPSHOT.OPENSOURCE.GPL2 and LICENSE.SNAPSHOT.OPENSOURCE.GPL3 in today snapshot |
| 17:01.36 | landley | chouimat: reassuring. |
| 17:01.37 | GPSFan | prpplague: ping |
| 17:01.38 | landley | Ok. |
| 17:01.44 | landley | False alarm then. |
| 17:02.01 | landley | The blurb I read heavily implied that this was a license _switch_ rather than adding another license. |
| 17:02.01 | RobotGuy | Whew |
| 17:02.04 | prpplague | GPSFan: nm, i had a question about your changes on the wiki, but i cleaned it up a little |
| 17:02.40 | prpplague | GPSFan: how is the lcd working for ya? |
| 17:02.54 | GPSFan | prpplague: thanks, I ain't a great wiki artist, so have at it. I didn't get to do anything on it over the weekend, maybe some more later this week. |
| 17:03.25 | prpplague | GPSFan: okie dokie |
| 17:03.26 | RobotGuy | I have done a little bit on the Wiki. |
| 17:03.32 | prpplague | GPSFan: if you want some more lcd's let me know |
| 17:04.49 | landley | chouimat: surprisingly, wtfpl is a relatively sane one. |
| 17:12.26 | landley | Has anybody else had 2.6.24 just segfault on mips? |
| 17:12.34 | landley | I can't get it through ne2k initialization... |
| 17:13.31 | landley | Panics before a shell prompt even comes up, but 2.6.23 worked fine with the same config... |
| 17:36.34 | RobotGuy | The way Ubuntu and Kubuntu handle root access stinks. |
| 17:47.36 | Crofton | anyone familar with using an external tool chain know what the web page means by HOST_TOOLS path? |
| 17:48.09 | *** join/#edev DrZimmerman (n=theo@2001:5c0:8fff:fffe:0:0:0:6af1) |
| 17:48.57 | prpplague | Crofton: for buildroot? |
| 17:49.02 | Crofton | urg |
| 17:49.08 | Crofton | wrong channel |
| 17:49.14 | prpplague | hehe |
| 17:49.15 | prpplague | np |
| 17:49.22 | Crofton | trying to uyse a cros tols gernested tool chain w oe |
| 17:49.59 | Crofton | any clue what the phrase host tools referes to? |
| 17:50.41 | prpplague | Crofton: i would expect that to be the path to the external toolchains execuatables |
| 17:50.46 | prpplague | Crofton: but thats only a guess |
| 17:50.56 | Crofton | heh |
| 17:51.04 | Crofton | there is another variable for that :) |
| 18:01.28 | sjhill | Crofton: host tools means your build host tools, like native sed, gcc, etc. needed to run configure scripts, compile utility applications, etc. |
| 18:02.05 | Crofton | thansk |
| 18:03.57 | *** join/#edev RobotGuy (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 18:06.49 | *** join/#edev DrZimmerman (n=theo@2001:5c0:8fff:fffe:0:0:0:6af1) |
| 18:08.22 | RobotGuy | It looks like I don't have a couple packages installed that I need to have buildroot work right. |
| 18:15.08 | landley | Robotguy: "sudo /bin/bash" then "passwd". |
| 18:15.18 | landley | You can set a root password at any time, it just doesn't set one by default. |
| 18:15.32 | landley | (Note that setting a root password before trying to use the "recovery mode" entry in grub would be a good idea. :) |
| 18:24.10 | *** join/#edev ico2__ (n=ico2@modem-2870.chameleon.dialup.pol.co.uk) |
| 18:27.04 | RobotGuy | landley: I did set a root password. :) I boo booed and somehow got sudu reset from setuid. I always have a current Knoppox CD handy for such ocassions. :) |
| 18:28.29 | landley | Did you try "su -"? |
| 18:29.58 | RobotGuy | I can "su -" now, but could not before. |
| 18:30.42 | RobotGuy | Kubuntu and Ubuntu force you to use sudo to gain root privledges on a normal install. I never liked limits like that. :) |
| 18:31.16 | RobotGuy | Now I need to figure out why fakeroot won't build. |
| 18:32.32 | RobotGuy | It seems to be having problems with libtool commands, but I don't know how to fix that. |
| 18:32.39 | RobotGuy | I just installed libtool. |
| 18:37.33 | *** join/#edev CosmicPenguin (n=nobody@163.181.251.103) |
| 18:37.33 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 19:21.16 | *** join/#edev samuelr (n=samuelr@dsl093-061-094.pit1.dsl.speakeasy.net) |
| 19:29.33 | *** join/#edev emil|k (n=emil@f054013165.adsl.alicedsl.de) |
| 19:36.03 | *** join/#edev dijenerate (n=dijenera@72.51.91.199) |
| 20:24.50 | *** join/#edev ico2__ (n=ico2@modem-1781.chimpanzee.dialup.pol.co.uk) |
| 21:11.49 | *** join/#edev ginger (n=jerneja@mail.balie.nl) |
| 21:13.31 | *** part/#edev ginger (n=jerneja@mail.balie.nl) |
| 21:59.28 | *** join/#edev jburd (n=jburd@59.184.39.158) |
| 21:59.52 | ds2 | is there any reason why a CD4066 won't work as a USB disconnect/connect switch that I can control by software? (trying to automate unplugging/plugging cables into a USB device |
| 22:00.43 | ds2 | quad CMOS switch |
| 22:00.45 | *** part/#edev jburd (n=jburd@59.184.39.158) |
| 22:01.00 | ds2 | sorry, thought it was a generic enough part to be obvious |
| 22:01.35 | prpplague | ds2: yea, you won't beable to use that for disconnecting usb |
| 22:01.47 | ds2 | why? |
| 22:02.16 | ds2 | too much resistance or lack of current handling on the power line? |
| 22:02.21 | prpplague | ds2: because of the differential nature of usb, switching like that will cause the signal operations not to work |
| 22:03.08 | ds2 | maybe it is monday, but I don't see why it won't.. the switch has been used to switch composite video signal lines |
| 22:03.13 | prpplague | ds2: we've that type of switching before with little success |
| 22:03.28 | ds2 | okay |
| 22:03.30 | prpplague | ds2: yea in theory it should work, but it doesn't |
| 22:03.55 | prpplague | ds2: the generally accepted way to control the connection is two fold |
| 22:04.28 | prpplague | ds2: first if the device is being provided a +5v source, it removes the power |
| 22:05.12 | prpplague | ds2: second if the device is providing a usb device connection it disables the pullup resistor, if its a host it pulls the usb line to ground |
| 22:06.30 | prpplague | ds2: have a look at the hammer carrier board schematic - http://www.tincantools.com/assets/Hammer_Carrier_Rev-C.pdf |
| 22:06.40 | prpplague | ds2: let me know when you got it pulled up |
| 22:10.10 | prpplague | ds2: which side are you controlling the disconnection from? |
| 22:12.41 | ds2 | prpplague: got it (sorry, someone dropped by the cube) |
| 22:12.49 | prpplague | ds2: np |
| 22:13.12 | ds2 | prpplague: I am trying to work around either a buggy gadget or buggy hardware and I been manually doing it. What I would like to do is to be able to do this remotely |
| 22:13.35 | prpplague | ds2: ok, so do you have control on the host side or on the slave side? |
| 22:18.28 | prpplague | ds2: if you have control on the host side, simply cutting the +5v to the usb port will generate a disconnect for most usb devices(those that comply with the usb spec) |
| 22:19.11 | prpplague | ds2: if that is impractical to do in a new design, you can use a transition board |
| 22:19.46 | prpplague | ds2: host usb->cable->transition board with power control->cable->usb device |
| 22:23.54 | ds2 | prpplague: I don't really control the hardware on either side, best I can do is insert an intermediary |
| 22:24.16 | prpplague | ds2: ahh, yea the transition board would be best then |
| 22:24.22 | ds2 | I am really slanting towards a 3PDT relay |
| 22:24.27 | prpplague | ds2: how do you plan to trigger the disconenct? |
| 22:25.14 | ds2 | prpplague: thinking of CTS/RTS off the serial port I am using for console driving a diode + 2N2222 |
| 22:25.30 | prpplague | ds2: ahh, interesting |
| 22:25.37 | *** join/#edev ico2__ (n=ico2@modem-3860.chameleon.dialup.pol.co.uk) |
| 22:25.42 | ds2 | "crude" is the term you are looking for =) |
| 22:25.59 | ds2 | i am tired of plugging and unplugging cables on every reboot :( |
| 22:26.02 | prpplague | ds2: well, you can do a quick test, take a short usb extension cable, cut the +5v line in the middle |
| 22:26.14 | adyer | ds2 how about a cheapo 4 port hub and just switch the wall wart DC through a fet? |
| 22:26.14 | prpplague | ds2: then plug in the host and device |
| 22:26.33 | prpplague | ds2: if the device doesn't enumberate, its controlled via the +5v line |
| 22:27.04 | ds2 | I like adyer's idea |
| 22:27.08 | prpplague | adyer: thats doable too, i was thinking more in a production item, but your idea is nice |
| 22:27.10 | ds2 | X10 + USB hub |
| 22:27.12 | *** join/#edev RobotGuy (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 22:27.22 | ds2 | no boards to build since I am officially a software guy anyways |
| 22:27.36 | ds2 | don't wanna scare HR with a soldering iron ;) |
| 22:27.53 | prpplague | hehe\ |
| 22:28.18 | adyer | i like the x10 idea :-) |
| 22:28.25 | ds2 | hahaha, you competing with SFE? =) |
| 22:28.33 | prpplague | indeed |
| 22:28.41 | RobotGuy | prpplague: What are you wanting to do a board for? |
| 22:29.00 | prpplague | RobotGuy: remote controlled usb disconnect |
| 22:29.32 | adyer | lol just make a robot to plug the cable in and out |
| 22:29.38 | prpplague | adyer: hehe |
| 22:30.03 | prpplague | RobotGuy: ? |
| 22:30.23 | prpplague | RobotGuy: what about the bot board? |
| 22:30.38 | RobotGuy | prpplague: I been thinking about that.. :) |
| 22:31.06 | prpplague | RobotGuy: oh, well, we are going into production on the bot board until we send some beta's out |
| 22:31.59 | prpplague | RobotGuy: yea, well we have 10 boards, and we weren't planning to send them out until march or april |
| 22:32.11 | RobotGuy | It should at least have mounting holes to match the Lynxmotion EC-02 electronics carrier, but does not need to be the same size as a Lynxmotion bot board. |
| 22:32.42 | ds2 | prpplague: have you thought of reselling things like BT dongles with known chipsets? |
| 22:33.01 | ds2 | or SD cards with a known Marvell or Athereos chipsets :) |
| 22:33.07 | RobotGuy | ds2: I think SFE already does that pretty well. |
| 22:33.13 | prpplague | ds2: one sec |
| 22:33.19 | RobotGuy | SD would be another thing. |
| 22:33.47 | ds2 | RobotGuy: I don't think SFE ids the USB BT chipsets... the only source I know of is the computergeeks.com routelete |
| 22:34.41 | RobotGuy | I probably would not use all four USB host ports, but I know some fellows who would. I'd probably just connect a thumb drive and wifi adapter. |
| 22:35.52 | RobotGuy | 3 pin headers on all pins that can be GPIOs. :) |
| 22:36.52 | RobotGuy | signal, +5V (or up to 7.2V for servos), and ground. |
| 22:37.42 | RobotGuy | dual power connectors, for logic and servo power. :) |
| 22:38.00 | ds2 | what about galvanically isolated IO pins? ;) |
| 22:38.35 | ds2 | prephaps even a software servo control loop in the kernel? |
| 22:38.40 | RobotGuy | All I/O +5V tolerant, Acroname style I2C headers. |
| 22:38.51 | ds2 | and hardware encoder inputs? |
| 22:38.59 | RobotGuy | ds2: No, that's not needed. We can write our own servo control code if we want that. |
| 22:39.03 | prpplague | ok back |
| 22:39.14 | RobotGuy | ds2: Handed with other hardware such as PICs or AVRs. |
| 22:39.41 | ds2 | RobotGuy: have you thought about porting something like EMC ? |
| 22:39.54 | RobotGuy | Most of us robot folks use dedicated servo and motor controllers. |
| 22:39.56 | prpplague | ds2: yea sfe does pretty well on the bt dongles, once things pick up with sales, we'll probably stock alot of misc stuff like that |
| 22:40.43 | prpplague | RobotGuy: the board is already done |
| 22:40.54 | RobotGuy | I am fond of the Micrichip dsPIC30F4012 - 6 channel motor control PWM and a hardware quadrature encoder module. |
| 22:41.08 | RobotGuy | prpplague: Wow, do you have links to the features? |
| 22:41.28 | prpplague | RobotGuy: no like i said, we aren't going live with it until march or april |
| 22:41.44 | RobotGuy | prpplague: I'm just curious what features it has, but OK. |
| 22:42.46 | prpplague | RobotGuy: pretty much what you said, 4 port hub, +5v tolerant i/o, seperate power connections, standard servo headers |
| 22:43.53 | prpplague | ds2: schematic done |
| 22:44.03 | prpplague | ds2: get the layout guy to do it tomorrow |
| 22:44.10 | ds2 | for the 3PDT relay control? |
| 22:44.40 | prpplague | ds2: naw, this one is a small transition board, usb host connector on one side, usb slave on the other |
| 22:44.51 | prpplague | ds2: with a atmega48 and a db9 connector |
| 22:45.02 | ds2 | hmm |
| 22:45.14 | prpplague | ds2: control the connection via rs-232 |
| 22:45.33 | ds2 | why not through in a a USB-RS232 chip too? |
| 22:45.52 | ds2 | in my case, I lucked out by insisting on a laptop with a RS-232 port...others may not be as lucky |
| 22:45.54 | prpplague | ds2: you said you wanted to via rs-232? |
| 22:46.15 | ds2 | prpplague: not so much I wanted it as much as I happen to have that around |
| 22:46.22 | prpplague | ds2: well i suspect if you have problems with your usb host, you are going to lose connection with the adatper as wel |
| 22:46.29 | ds2 | parallel port is less headache but machine don't have those anymore :( |
| 22:46.37 | ds2 | good point |
| 22:46.38 | prpplague | ds2: ahh |
| 22:47.21 | ds2 | I have X10 modules around + cheap hubs so that might be my quickest solution |
| 22:47.37 | prpplague | ds2: indeed, this is just a tinker item to add to tincantools |
| 22:47.56 | ds2 | best thing to do is for me to talk my boss into getting a ethernet controlled power strip ;) |
| 22:48.04 | ds2 | got boards I need to power cycle anyways |
| 22:48.21 | prpplague | hehe |
| 22:48.27 | ds2 | prpplague: *nod* |
| 22:49.43 | RobotGuy | prpplague: Cool on the bot board. :) |
| 22:50.14 | prpplague | RobotGuy: i'll probably post an updated buildroot tomorrow or wed. |
| 22:50.25 | RobotGuy | Cool. :) :) |
| 22:52.25 | RobotGuy | I have to be missing some component to make buildroot happey. |
| 22:58.36 | prpplague | http://www.robweir.com/blog/2008/01/what-every-engineer-knows.html |
| 22:58.38 | prpplague | good read |
| 23:01.31 | *** join/#edev ico2_ (n=ico2@modem-1688.cougar.dialup.pol.co.uk) |
| 23:11.29 | RobotGuy | I once saved a company about $50,000.00 when I found a power/ground problem on a circuit board design that was just about to go into full production. I was only a temp. |
| 23:11.44 | prpplague | hehe |
| 23:11.46 | prpplague | fun |
| 23:12.33 | ds2 | bah, just inductively couple everything ;) |
| 23:20.18 | *** join/#edev CosmicPenguin (n=nobody@163.181.251.103) |
| 23:20.18 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 23:24.30 | landley | RobotGuy: Join the club. |
| 23:24.42 | landley | Ok, so learning php turns out to be more an 8 hour thing than a 4 hour thing... |
| 23:25.39 | RobotGuy | landley: I still haven't gotten around to learning PHP. I think I will go with Python instead. |
| 23:26.27 | landley | Oh I vastly prefer python. |
| 23:26.39 | landley | I'm just up for a job that uses php (and claims it can be secured; this I've got to see). |
| 23:26.53 | landley | Considering the job is not just local but literally across the street from my wife's workplace... |
| 23:27.03 | landley | And otherwise seems a nice place to work with a fun group of engineers... |
| 23:27.08 | landley | And management that _isn't_ insane. |
| 23:28.02 | landley | Worth a shot. |
| 23:28.11 | RobotGuy | Something is still looking for fakeroot-1.8.10 even though I have replaced it with fakeroot-1.9 |
| 23:28.55 | landley | Python's a great language for cgi. just use: blah = {"thingy":stuff}; print """<html>...%(thingy)s...</html>""" % blah |
| 23:29.16 | landley | Your entire HTML is one text block enclosed in triple quotes, and all your inserted data is a single dictionary calculated at the front and referenced at the end. |
| 23:29.19 | landley | It's easy. |
| 23:29.44 | landley | RobotGuy: my arm toolchains are working fine in FWL. |
| 23:29.53 | landley | It's _mips_ that's gone all weird on me. |
| 23:29.56 | RobotGuy | landley: FWL? |
| 23:30.01 | landley | (Not the toolchain, the kernel. Something buggy in 2.6.24...) |
| 23:30.03 | landley | Firmware Linux. |
| 23:30.09 | RobotGuy | landley: Ah. |
| 23:30.16 | landley | I should probably rename it, but it's had that name since... 2002, I think. |
| 23:30.39 | landley | Back when it _was_ producing a single system image file for a VPN/NAS/FloorWax product. |
| 23:30.46 | landley | (Oh wow did that project have feature creep...) |
| 23:31.07 | landley | Does anyone here actually _use_ mips? |
| 23:31.15 | prpplague | landley: gotta take my step-daughter to the doctor |
| 23:31.21 | landley | ok. Have fun. |
| 23:31.23 | prpplague | landley: i'll be back online in a few hours |
| 23:31.29 | landley | I'll probably still be here. |
| 23:31.29 | prpplague | landley: oh lots of fun |
| 23:31.31 | landley | Fighting with php. |
| 23:31.35 | prpplague | hehe |
| 23:31.38 | prpplague | later then |
| 23:31.40 | landley | Learning! |
| 23:31.45 | RobotGuy | I'll definitely still be here fighting with buildroot |
| 23:32.10 | landley | It was easier. |
| 23:32.11 | RobotGuy | landley: Probably very smart. |
| 23:32.20 | adyer | landley we use a fair bit of mips |
| 23:32.36 | landley | adyer: can _you_ get the ne2k driver in a 2.4 kernel to initalize without panicing the kernel? |
| 23:32.37 | landley | I can't. |
| 23:32.43 | landley | Sorry, 2.6.24 I mean. |
| 23:32.43 | RobotGuy | I'm also in the process of learning Smalltalk. |
| 23:32.54 | landley | RobotGuy: ... why? |
| 23:33.08 | landley | adyer: 2.6.23 worked fine. |
| 23:33.11 | RobotGuy | landley: Because it seems that it will be very useful for robotics. |
| 23:33.13 | adyer | landley we dont use ne2k, SoC has macs built in |
| 23:33.21 | landley | Ah. |
| 23:33.29 | landley | qemu supplies an ne2k. It _used_ to work. |
| 23:33.38 | landley | I suppose I could try to wave an rtl8139 at it instead and see what it does... |
| 23:33.44 | landley | But I'd like to _fix_ this. It's a regression. |
| 23:33.47 | RobotGuy | I haven't seen an ne2K for a LONG time. |
| 23:34.09 | landley | RobotGuy: but they're dirt simple to emulate. |
| 23:34.24 | RobotGuy | landley: No doubt. |
| 23:34.30 | landley | I think I still have 5 of them in a drawer, but I'm running out of systems with a PCI bus... |
| 23:34.52 | RobotGuy | Right now, all I want is a solid working buildroot toolchain, but wonder if there are too many conflicting terms there. |
| 23:34.58 | adyer | dont use qemu either :-) |
| 23:36.39 | adyer | robotguy I feel your pain. buildroot mostly makes me crazy |
| 23:39.01 | RobotGuy | adyer: You have something better that would be compatible with building stuff for Hammer? |
| 23:39.29 | adyer | landley I think there were some recent patches having to do with qemu on the mips linux mailing list recently, but I just skimmed them |
| 23:40.59 | adyer | robotguy I was going to play with scratchbox when I get a "round tuit". I don't have a hammer - our systems are i.MXL based, but similar |
| 23:42.01 | *** join/#edev Mirell (n=mirell@otomi.ling.utexas.edu) |
| 23:43.07 | RobotGuy | adyer: I like the i.MX chips. I would love to have an i.MX31 board. |
| 23:45.02 | landley | adyer: I don't think this is a qemu-specific issue. I narrowed the start of the panics down to this: |
| 23:45.03 | landley | http://kernel.org/hg/linux-2.6/rev/74258 |
| 23:45.30 | landley | I'm not sure that the current panic is the same issue, but no kernel _after_ that boots all the way to a shell prompt for me, and they all did with null pointer dereferences at various points. |
| 23:45.46 | landley | I haven't figured out how to get mips to cough of a stack trace with function names rather than addresses in it, though. |
| 23:45.53 | landley | cough up |
| 23:48.05 | adyer | landley out of my league :-( |
| 23:51.19 | ds2 | isn't there System.map for hand decoding on MIPS? |
| 23:51.51 | landley | adyer: I'm usually out of my league. |
| 23:51.54 | landley | I just don't let that stop me. :) |
| 23:52.14 | landley | ds2: probably. But on other platforms the actual panic message can give you a stack trace with function names. |
| 23:52.26 | landley | Which is really useful, although not so much here. |
| 23:52.44 | landley | The patch that triggered these panics changed the irq locking rules, which is just deeply reassuring... |
| 23:52.56 | landley | I'm going to just try reverting it and see if that's good enough for now... |
| 23:53.01 | landley | Already pestered the patch's author. No response yet... |
| 23:53.41 | landley | (It is nice being able to rebuild the cross compiler, kernel, root filesystem, and package it all up simply by going "cursor up, enter". |
| 23:54.20 | landley | Add in "hg archive -t tbz2 -r 12345 ~/path/to/where/it/expects/linux-2.6.xx.tar.bz2" and then another window where I cursor up on the qemu invocation... |
| 23:54.40 | landley | Yeah, takes 9 minutes or so. I'll live. |
| 23:55.10 | RobotGuy | landley: Doesn't being out of your league cause learning? |
| 23:56.34 | landley | RobotGuy: in my case, waking up causes learning. |
| 23:56.45 | landley | I'm surrounded by things I don't know, I really don't have to seek them out. :) |
| 23:56.51 | RobotGuy | landley: Me too. |
| 23:57.22 | landley | The kernel is an area where I wander away for 6 months, come back, and EVERYTHING'S CHANGED. |
| 23:57.51 | landley | Long time ago I learned that staying up to date on the kernel is a full time job. You either do kernel work or you do _everything_ else. |
| 23:58.31 | landley | Considering that I break the compiler, the C library, all the command line utilities, the gui, the emulator, and its little dog too... |
| 23:58.41 | landley | I have to wander into and frown at all of those. |
| 23:58.57 | landley | With non-kernel packages you can wander through every 6 months and at least have a frame of reference. |
| 23:59.11 | landley | In the kernel, doing that means you get the pain aimed at people maintaining out of tree modules. |
| 23:59.20 | CosmicPenguin | Try being a Kernel and a X developer |
| 23:59.24 | CosmicPenguin | much fun |
| 23:59.33 | CosmicPenguin | if one isn't hosed, the other one is |
| 23:59.33 | landley | "We've replaced the API, locking rules, and memory allocation regime you used to know with Folger's Crystals!" |