| 00:04.47 | rwhitby-away | ByronT: is Notes.txt becoming a separate file, or still at the end of the README? (I'd prefer the latter) |
| 00:15.37 | CIA-3 | 03rwhitby * 10unslung/Makefile: Promoted mtr, adns and fetchmail. |
| 00:15.48 | jacques | neat |
| 00:16.00 | jacques | rwhitby-away, BTW, libcurl builds native and tests at 100% |
| 00:17.19 | CIA-3 | 03rwhitby * 10unslung/Makefile: Promoted libcurl to cross and native. |
| 00:20.56 | rwhitby-away | jacques: thought any more about automatic and reversible /bin/sh -> bash? |
| 00:21.18 | kergoth | i'd recommend against that |
| 00:21.36 | kergoth | all the major distros now -have- to keep /bin/sh as bash because of scripts assuming it is. |
| 00:21.51 | kergoth | i'd leave it as a posix sh (dash, busybox ash), and scripts that use bashisms should have #!/bin/bash |
| 00:21.57 | kergoth | if they use #!/bin/sh and require bash, they're broken. |
| 00:21.59 | kergoth | heh |
| 00:22.42 | kergoth | there are better ways of making the _user_'s default shell different |
| 00:23.01 | kergoth | without imposing such foolish assumptions. your startup scripts dont need, nor should they use, bash |
| 00:23.08 | kergoth | </rant> |
| 00:27.01 | rwhitby-away | jacques is suggesting it, not me |
| 00:27.39 | kergoth | well, whomever. i'm recommending against it. if jacques is the one suggesting it, have him read what i just wrote |
| 00:32.29 | *** join/#nslu2-linux garpinc (~garpinc@pool-141-154-232-220.bos.east.verizon.net) |
| 00:40.56 | garpinc | Anyone knows if cif support is compiled into nslu2 kernel? |
| 00:41.11 | *** join/#nslu2-linux Koops (~weiner@adsl-19-97.swiftdsl.com.au) |
| 00:42.38 | Koops | anyone know how to change the busybox "vi" so that it takes up more available rows on the screen? It doesn't seem to take the normal commands that vim does. |
| 00:43.05 | *** join/#nslu2-linux NSLU2-IRC (~NSLU2-IRC@firewall.reactivity.com) |
| 00:44.16 | danwalkowski | rats, I tried to install crosstool-native, and got "no space on device". it's trying to write to flash, isnt it |
| 00:44.27 | jacques | kergoth, you make good points. I guess I was just looking for the expedient solution to some native build problems (broken configure/install scripts) - I was recommending /bin/bash -> sh for native devel systems. |
| 00:44.58 | jacques | danwalkowski, more likely your conf partition which is not large enough |
| 00:45.56 | rwhitby-away | danwalkowski: you need 117MB free disk space in /opt |
| 00:46.10 | danwalkowski | ah. well that sucks. let me check |
| 00:46.27 | rwhitby-away | you can move /opt onto the data partition .... |
| 00:47.27 | danwalkowski | 100% used. |
| 00:47.58 | danwalkowski | and it's only 144M to start with |
| 00:48.01 | danwalkowski | 114 i mean |
| 00:48.17 | danwalkowski | is it clogged with semi-installed pkgs now? |
| 00:48.29 | danwalkowski | the 7 parts seemed to mostly download |
| 00:49.10 | rwhitby-away | yeah, but you just use -force-reinstall |
| 00:50.44 | danwalkowski | cd / |
| 00:50.44 | danwalkowski | ls |
| 00:50.48 | danwalkowski | oops :) |
| 00:52.12 | danwalkowski | can i just cp -R /share/hdd/conf/opt over to .../data/opt?and then fix the opt link at /? |
| 00:52.38 | danwalkowski | that seems like a recipe to break a lot of paths |
| 01:08.28 | CIA-3 | 03djf * 10unslung/sources/bzflag/control: initial checkin |
| 01:09.25 | CIA-3 | 03djf * 10unslung/make/bzflag.mk: initial checkin - builds cross (not enough ram native) and packages, but package needs tweaking |
| 01:09.29 | danwalkowski | rwhitby? any suggestions on a safe way to move /opt? |
| 01:09.44 | danwalkowski | or at least free up some space? |
| 01:09.51 | [g2] | root@nslu2:/var/tmp# flash_eraseall /dev/mtd2 |
| 01:09.51 | [g2] | Erasing 128 Kibyte @ e0000 -- 87 % complete. |
| 01:10.33 | *** join/#nslu2-linux NSLU2-IRC (~NSLU2-IRC@firewall.reactivity.com) |
| 01:15.06 | danwalkowski | ok, so install crosstool-native ran out of space on /conf. now I am left with no crosstool, and no space. |
| 01:15.17 | danwalkowski | how do I clean out what it downloaded? |
| 01:15.18 | rwhitby-away | danwalkowski: since everything should access it through /opt, then moving it should be fine |
| 01:15.30 | rwhitby-away | rm -rf /opt/armeb |
| 01:15.59 | danwalkowski | thank you! :) |
| 01:16.24 | danwalkowski | I didnt know where it stashed it |
| 01:16.51 | [g2] | rwhitby-away, dyoung-logger I'm working on an update mechanism for OpenSlug where users can reflash the kernel and possibly the jffs2 or switchbox paritions from the commandline. |
| 01:16.52 | rwhitby-away | (the "away" in rwhitby-away means that I'm multi-tasking with Real Life, and responses may be delayed) |
| 01:17.19 | rwhitby-away | [g2]: cool - we can use that for Unslung too |
| 01:17.24 | [g2] | I've got the mtd-utils' flash_eraseall erasing the kernel partition |
| 01:17.44 | [g2] | rwhitby-away, most likely |
| 01:17.51 | rwhitby-away | we probably have room to put the mtd utils in the recovery shell |
| 01:18.14 | [g2] | I don't image that they are super large :) |
| 01:18.40 | [g2] | two keys are loop mount support and mtdram support |
| 01:19.30 | [g2] | From what I've read we'll need mtd ram support to be able to create a mtd ram parition in memory that allows a jffs2 loop mount |
| 01:19.49 | [g2] | I think that's the piece I was missing before for jffs2 loop mounting |
| 01:20.07 | [g2] | and obviously some way to get off the current root :) |
| 01:20.10 | rwhitby-away | [g2]: you've seen how we do maintenance mode in Unslung? |
| 01:20.29 | [g2] | from the web interface and then a reboot ? |
| 01:20.45 | [g2] | is it written up anywhere ? |
| 01:21.04 | rwhitby-away | nope, but it's all in the switcbox linuxrc |
| 01:21.18 | rwhitby-away | touch /.ramdisk to enable it |
| 01:22.02 | [g2] | I saw the code in the /linuxrc |
| 01:22.12 | [g2] | and the ordering |
| 01:22.46 | [g2] | and that requires 2 reboots right |
| 01:23.04 | [g2] | 1 to get into maint mode, 1 to get out |
| 01:23.21 | [g2] | also, it's not all secured |
| 01:24.15 | rwhitby-away | yes, 2 reboots |
| 01:24.18 | rwhitby-away | "secured" ? |
| 01:24.24 | [g2] | with OpenSlug, all the commands will work within dropbear |
| 01:24.54 | rwhitby-away | maintenance without rebooting will be cool |
| 01:25.21 | [g2] | well you'll probably need the reboot for kernel changes to take effect |
| 01:25.50 | [g2] | and you'll probably want to reboot on rootfs changes, but I'd liked it to be 1 reboot |
| 01:26.01 | [g2] | it will be for the kernel changes |
| 01:26.21 | *** join/#nslu2-linux NSLU2-IRC (~NSLU2-IRC@res-24-197-102-116.spa.sc.charter.com) |
| 01:26.37 | [g2] | I left the kernel erased and the box comes up and waits for an image via upload |
| 01:26.37 | Carsten | yo yo |
| 01:26.42 | [g2] | hey hey |
| 01:26.51 | Carsten | ahh its u again :) |
| 01:27.00 | Carsten | remember me from a week ago...test or soemthign like that lol |
| 01:27.24 | test | :) |
| 01:27.25 | test | lol |
| 01:32.08 | test | so quite |
| 01:34.37 | danwalkowski | rwhitby: all the links in /conf/opt/bin that are part of the armeb package have incorrect ard links |
| 01:35.06 | danwalkowski | they do not use /opt path, they hardcode /share/hdd/conf/ |
| 01:36.22 | danwalkowski | which makes my moving conf and then attempting to re-install armeb probably moot, because things will break during install, yes? |
| 01:41.31 | [g2] | rwhitby-away, jacques is the partition table stored in the last block ? |
| 01:46.02 | jacques | yes |
| 01:48.04 | [g2] | jacques, how does the kernel find the partition table and I'm guessing it might be able to be moved pretty easily |
| 01:48.48 | rwhitby-away | danwalkowski: those links should be using /opt/armeb/... |
| 01:49.25 | jacques | [g2], why do you want to move it? |
| 01:49.44 | rwhitby-away | [g2]: redboot expects it there too |
| 01:50.14 | rwhitby-away | you wouldn't be able to push such a change upstream |
| 01:50.22 | [g2] | I'm wondering aloud about creating 1 big partition for the rootfs including the kernel |
| 01:50.53 | rwhitby-away | you can do that, just put the new table in the last block :-) |
| 01:51.29 | [g2] | I'm wondering about sercom compat |
| 01:51.58 | [g2] | thinking 1 big part where the kernel is and 1 one block for the rootfs |
| 01:52.04 | rwhitby-away | danwalkowski: all the symlinks in /opt/bin definitely go through /opt - if yours have /share/... in them, then you must have done that manually |
| 01:52.12 | [g2] | just pushed back 16 bytes for the lenght info |
| 01:52.37 | rwhitby-away | [g2]: what's the driver for breaking compatibility? |
| 01:52.51 | rwhitby-away | shall we just get a beta out first? |
| 01:53.23 | [g2] | it's happening |
| 01:53.41 | [g2] | the driver is just replacing the kernel partition probably isn't very proper |
| 01:54.07 | [g2] | because there's no guarentee that the library modules match |
| 01:54.32 | rwhitby-away | I really don't understand the drive for replacement bootloaders and changing partition layout and so forth ... |
| 01:54.33 | *** join/#nslu2-linux dyoung-web (www-data@netblock-66-159-209-60.dslextreme.com) |
| 01:54.50 | rwhitby-away | what's really wrong with the layout exactly as it comes stock? |
| 01:55.25 | rwhitby-away | seems to me that there is much more avenue for improving the useability of the device in normal changes, rather than changing the basic flash infrastructure on people .... |
| 01:55.55 | rwhitby-away | anyway, back later .... |
| 02:07.00 | danwalkowski | no, I didn't do it manually, but I did foolioshly try tyo install crosstool like a week ago. perhaps then they were not correctly pathed? |
| 02:20.16 | ka6sox | ~seen jacques |
| 02:20.17 | jbot | jacques is currently on #intimate (8h 20m 11s) #nslu2-linux (8h 20m 11s) #oe (8h 20m 11s) #elinux (8h 20m 11s) #handhelds.org (8h 20m 11s) #bzflag (8h 20m 11s). Has said a total of 143 messages. Is idling for 30m 52s |
| 02:21.05 | *** join/#nslu2-linux kergoth (~kergoth@ip71.churchill-park.dfw.ygnition.net) [NETSPLIT VICTIM] |
| 02:27.49 | jacques | ka6sox, you rang? |
| 02:32.32 | ka6sox | hi...a question about crossbuilds. didn't you guys run into something that required a x86 binary to build? |
| 02:34.35 | jacques | yeah appweb |
| 02:34.41 | *** join/#nslu2-linux dyoung-web (www-data@netblock-66-159-209-60.dslextreme.com) |
| 02:34.57 | ka6sox | was it for native or cross? |
| 02:35.09 | jacques | for no good reason IMHO - really lowered my opinion of those devs |
| 02:35.23 | ka6sox | yeah...heh |
| 02:35.35 | jacques | it ships with x86 binaries in the tarball and it tries to use them when it configures itself |
| 02:35.53 | ka6sox | thats stupid |
| 02:36.04 | jacques | very |
| 02:36.41 | [g2] | well maybe they are trying to make it easier on x86 devs :) |
| 02:37.32 | ka6sox | at the expense of everyone else. :P |
| 02:37.46 | jacques | i found a thread on their support list some guy trying to build on a uclibc system - the main dev who answered seemed rather clueless |
| 02:40.40 | dyoung-web | Whats an arm? |
| 02:40.54 | ka6sox | something like a leg |
| 02:45.45 | CIA-3 | 03djf * 10unslung/make/bzflag.mk: added configure patch to allow bzadmin to build |
| 02:47.00 | CIA-3 | 03djf * 10unslung/sources/bzflag/ (bzflag-configure.patch gl.h glu.h): patch to allow bzadmin to build and some header files it requires |
| 02:47.13 | ka6sox | LOL...fun on the Slug! |
| 02:47.45 | jacques | any bzflag players here? |
| 02:48.56 | ka6sox | NWN and WC3 |
| 02:49.12 | *** join/#nslu2-linux Tiersten (tman@ideal.trejan.com) |
| 02:49.48 | jacques | well none of those are going to be running on the slug anytime soon |
| 02:55.42 | dyoung-web | back later |
| 03:45.24 | *** join/#nslu2-linux dyoung-web (www-data@netblock-66-159-209-60.dslextreme.com) |
| 03:46.02 | dyoung-web | Is there anyone here that is actually using appweb? |
| 03:51.57 | rwhitby-away | I tested it, but am not running it now |
| 03:52.14 | dyoung-web | For some reason it never starts for me. |
| 03:52.21 | dyoung-web | The process is running; but it doesnt DO anything. |
| 03:52.28 | dyoung-web | so I was gonna ask to see your error.log file. |
| 03:52.37 | rwhitby-away | haven't tested it lately. did you see the talk about /dev/random? |
| 03:52.46 | dyoung-web | in irclogs? |
| 03:52.48 | dyoung-web | not yet. |
| 03:56.39 | rwhitby-away | There's a slugbug on it too |
| 04:02.28 | dyoung-web | the random bit didnt affect my appweb problem. |
| 04:05.13 | dyoung-web | ooh, I guess I wasnt patient enough; it evertually started after the random change. |
| 04:07.06 | dyoung-web | now I can give a go to cgiirc. |
| 04:07.25 | dyoung-web | Cool. |
| 04:07.32 | kergoth | does it work well? it looks pretty nice |
| 04:07.39 | kergoth | nice and modular |
| 04:08.10 | dyoung-web | Not sure yet, this is the first time it actually did anything for me. I'll let you know later though. I do like that client-side ssl cert functionality. |
| 04:09.34 | kergoth | it supports client side ssl certs? |
| 04:09.38 | kergoth | thats really nice |
| 04:10.53 | rwhitby-away | yep - the other alternative is stunnel |
| 04:10.59 | rwhitby-away | (for client side certs) |
| 04:14.02 | *** join/#nslu2-linux rt (~markv@c-24-7-79-31.client.comcast.net) |
| 04:14.35 | rt | or, more precisely, i can't seem to get a download of unslung from sourceforge. |
| 04:14.44 | *** join/#nslu2-linux tjyang (~Administr@c-67-175-74-12.client.comcast.net) |
| 04:18.29 | tjyang | http://ipkg.nslu2-linux.org/feeds/unslung/cross/crosstool-native_0.28-rc37-3_armeb.ipk is still not ready |
| 04:19.27 | tjyang | I tried to follow Rod's post to try out crosstool-native |
| 04:19.48 | tjyang | I mean here http://groups.yahoo.com/group/nslu2-linux/message/3774 |
| 04:27.30 | *** part/#nslu2-linux tjyang (~Administr@c-67-175-74-12.client.comcast.net) |
| 04:33.47 | dyoung-web | rwhitby, it seems to work; though I cant fully test it from here. with lynx it works until it gets to the frames. yes I know I should be using elinks. |
| 04:34.01 | dyoung-web | I'll have it tested in around 3 hours. |
| 04:35.56 | dyoung-web | back htne. |
| 05:39.17 | CIA-3 | 03rwhitby * 10unslung/Makefile: Promoted bzflag |
| 05:47.23 | CIA-3 | 03rwhitby * 10unslung/Makefile: Demoted bzflag, cause it doesn't build cross. |
| 05:51.41 | *** join/#nslu2-linux ByronT (~byron-por@1Cust1180.an4.tpa9.da.uu.net) |
| 05:53.53 | ByronT | whew... almost done with README and new file... |
| 05:54.29 | rwhitby-away | ByronT: what's the new file? |
| 05:54.31 | ByronT | http://www.toddcomp.com/NSLU2/Unslung-3.1x-beta-README.txt |
| 05:54.54 | rwhitby-away | (I'd rather keep everything as one file in the distro, and put everything else on the wiki) |
| 05:55.07 | ByronT | and http://www.toddcomp.com/NSLU2/Unslung-3.1x-beta-Notes.txt |
| 05:56.09 | ByronT | that README then takes a bunch of stuff out... gone "sigh ya naw ra" |
| 05:57.13 | ByronT | I've gotten down to the Diversion Scripts in the notes.txt file... |
| 05:57.35 | ByronT | I was thinking that the README is install and install only |
| 05:57.46 | rwhitby-away | ByronT: agreed. Can we just put the Notes on the wiki, and point to it in the README ? |
| 05:57.53 | ByronT | and notes is basic stuff to get you running after readme... |
| 05:58.05 | ByronT | without having to have internet access |
| 05:58.49 | rwhitby-away | Hmm - interesting point. How did they get the download if they don't have internet access? |
| 05:59.17 | rwhitby-away | (BTW, excellent work on the content of the two files) |
| 05:59.22 | ByronT | well... for me... I usually burn stuff to a cd or to an external hd at work... and then bring home to sucky dialup |
| 05:59.43 | ByronT | Thank you... |
| 06:00.17 | ByronT | unfortunately... I was particularly proud of the first portion of the Notes.... |
| 06:10.07 | rwhitby-away | ok, let's leave Notes.txt as it is and we will include both files in the distro |
| 06:12.36 | CIA-3 | 03djf * 10unslung/make/ncurses.mk: added ncurses.h link in staging |
| 06:16.58 | CIA-3 | 03rwhitby * 10unslung/make/mtr.mk: Fixed cross-compilation. |
| 06:17.52 | rt | woohoo! |
| 06:18.14 | rwhitby-away | rt: congrats, you're unslung! |
| 06:19.05 | rt | so, running /sbin/unsling does what, precisely? |
| 06:19.44 | rwhitby-away | symlinks /opt and /usr/lib/ipk onto the attached disk |
| 06:20.06 | rt | ah. will the system still boot if the drive isn't attached? |
| 06:20.11 | rwhitby-away | you can read the script to see precisely, it's very straight-forward |
| 06:20.13 | *** join/#nslu2-linux Ximinez (~repvik@ti500710a080-1569.bb.online.no) |
| 06:20.49 | rwhitby-away | you have a choice - you can *not* unsling, and then your packages will be stored in internal flash (very limited space, and if you fill it you will have to reflash) |
| 06:20.59 | rwhitby-away | or you can unsling and store the packages on disk. |
| 06:21.21 | rt | well, i guess I'll unsling it for now. |
| 06:21.23 | rwhitby-away | because the init.d files for the packages are also on that disk, then if it is not attached, then those packages just don't start |
| 06:21.58 | rwhitby-away | as long as you're not silly enough to put /opt/bin/bash as root's shell, or something like that, then it will boot fine without the disk |
| 06:22.39 | rt | is it complicated to get the development environment hosted on this thing? |
| 06:23.04 | rwhitby-away | you on the mailing list? |
| 06:23.40 | rwhitby-away | (I sent an announcement about that very topic last night) |
| 06:25.01 | rwhitby-away | you can ipkg install everything you need |
| 06:25.57 | rt | nope, not on the list. |
| 06:27.22 | rt | the yahoo group? or elsewhere? |
| 06:27.34 | rwhitby-away | Yahoo nslu2-linux, but you can read it at gmane if you despise the yahoo interface |
| 06:27.52 | rwhitby-away | (or your own mail client, of course) |
| 06:31.11 | rt | what was the title of the messag, nothing I see here looks right. |
| 06:31.28 | rwhitby-away | Announce: ... crosstool-native .... |
| 06:31.56 | rwhitby-away | http://groups.yahoo.com/group/nslu2-linux/message/3774 |
| 06:32.28 | rt | ah, thanks. |
| 06:32.41 | rwhitby-away | Hmm - it didn't include the link to http://www.nslu2-linux.org/wiki/HowTo/NativelyCompileUnslungPackages |
| 06:32.53 | rwhitby-away | or http://www.nslu2-linux.org/wiki/Unslung/AddAPackageToUnslung |
| 06:37.00 | rt | hmmm. nameservice/gateway seems not right. |
| 06:37.13 | rwhitby-away | you need to set that in the web interface |
| 06:37.29 | rt | doh, but of course. |
| 06:37.33 | rwhitby-away | Is that not in the README? |
| 06:39.07 | rt | oh, probably. |
| 06:39.27 | rt | Most of the time I'm right. :-) |
| 06:39.46 | rwhitby-away | heh - let's see you change the root password on the NSLU2 then .... ;-) |
| 06:40.06 | rwhitby-away | (and make it stick, without locking yourself out of the device requiring a reflash) |
| 06:40.21 | rt | well, perhaps that can wait till tomorrow. |
| 06:40.57 | rwhitby-away | it's already tomorrow here .... |
| 06:41.06 | rt | out of curiousity, has anyone gotten python to run? |
| 06:41.27 | rwhitby-away | jacques has on OpenSlug, but I don't think anyone has on Unslung |
| 06:41.39 | christopher | I have once, but I didn't test it too much |
| 06:41.43 | rt | hmm. maybe that would be a good thing for me to do then that's useful. |
| 06:41.49 | christopher | I was working on making a package |
| 06:42.00 | christopher | but I've been swamped with school lately |
| 06:42.14 | rwhitby-away | a python package would be cool |
| 06:42.21 | jacques | I also built it on unslung, but it had a lot more test failures than on openslug |
| 06:42.22 | christopher | ok |
| 06:42.46 | rwhitby-away | christopher: perhaps you can check in your .mk file and rt can finish it off :-) |
| 06:42.56 | christopher | jacques: what did you use to test it? |
| 06:43.26 | christopher | rwhitby-away: I'll need to check where I left off, but that's a good idea |
| 06:43.27 | rt | hmm. this perl download seems to be taking its own sweet time. |
| 06:43.29 | jacques | its own testsuite |
| 06:43.35 | christopher | jacques ok |
| 06:46.28 | christopher | i'm running the make now to see what it gives me |
| 06:47.27 | christopher | ok, it makes python and libpython, but then make tries to run the python it just made which it can't do cross-copmiling |
| 06:47.39 | christopher | it's going to need a patch that removes that "feature" |
| 06:48.19 | christopher | or the patch that I have for configure could be modified |
| 06:48.32 | christopher | I'll push it out there, rt it you want to look at it you can |
| 06:48.41 | christopher | if I get some time this next week I'll look at it |
| 06:50.27 | rt | well, it will take some time for me to catch up, i'm sure. |
| 06:50.31 | rt | well, it will take some time for me to catch up, i'm sure. |
| 06:50.40 | rt | doh, sorry. bounced the key. |
| 06:52.12 | christopher | how do I "un-add" a file in CVS? |
| 06:52.25 | rt | so, after I hooked up my harddrive, the default root password no longer works. this is because the backup copy on the hdd supercedes it, right? |
| 06:52.32 | christopher | yes |
| 06:52.36 | rt | cvs remove? |
| 06:52.39 | christopher | tried it |
| 06:52.55 | rt | sorry, my brain confuses p4 and cvs commands now. |
| 06:52.59 | rwhitby-away | christopher: if you haven't committed yet, then cvs remove does the trick |
| 06:53.18 | christopher | ah, it didn't like the wildcard in the filename |
| 06:53.23 | christopher | got it |
| 06:54.07 | CIA-3 | 03edmondsc * 10unslung/ (make/python.mk sources/python/python-cross-compile.patch): python checkin, not complete out there for others to work on |
| 06:54.10 | christopher | there it is |
| 07:01.38 | rt | oh well, time for bed. |
| 07:01.47 | rt | i'll tweak this tomorrow. |
| 07:02.02 | rt | by then, this will all be installed. |
| 07:05.29 | rwhitby-away | anyone know how to find all hard links in a directory tree? |
| 07:05.42 | rwhitby-away | seems that ipkg will not unpack hard links in a tar file. |
| 07:13.12 | rwhitby-away | ok, I did it manually |
| 07:13.26 | rwhitby-away | look for crosstool-native -4 coming in a couple of hours |
| 07:14.08 | *** join/#nslu2-linux jipi (~jipi@cm187.gamma181.maxonline.com.sg) |
| 07:39.41 | *** join/#nslu2-linux kolla_ (kolla@halbrend.uninett.no) |
| 07:57.44 | *** join/#nslu2-linux LarsAC (~chatzilla@134.130.124.227) |
| 08:15.31 | *** join/#nslu2-linux dyoung (~dyoung@cpe-24-94-83-103.hawaii.rr.com) |
| 08:18.34 | CIA-3 | 03rwhitby * 10unslung/ (make/bash.mk sources/bash/control): Added a symlink for /bin/bash |
| 08:20.46 | *** join/#nslu2-linux perlguru-work (www-data@netblock-66-159-209-60.dslextreme.com) |
| 08:21.07 | perlguru-work | morning |
| 08:21.15 | CIA-3 | 03rwhitby * 10unslung/ (8 files in 2 dirs): Replaced the missing g++ hardlink with a symlink |
| 08:23.02 | rwhitby-away | that's a new version of crosstool-native |
| 08:23.12 | rwhitby-away | morning europe! |
| 08:25.58 | perlguru-work | ERROR: Cannot satisfy the following dependencies for crosstool-native: |
| 08:26.04 | *** part/#nslu2-linux dyoung (~dyoung@cpe-24-94-83-103.hawaii.rr.com) |
| 08:26.33 | perlguru-work | but: Upgrading crosstool-native-arch-bin |
| 08:26.38 | perlguru-work | :/ |
| 08:27.46 | rwhitby-away | yeah, I'm in the middle of that too |
| 08:28.06 | rwhitby-away | I dunno how else to specify the dependencies ... |
| 08:28.26 | rwhitby-away | cause it definitely does need the same ipk version, so <= or >= is not good enough |
| 08:28.56 | perlguru-work | i'll wait and see then.... |
| 08:29.34 | perlguru-work | what changed in bash? |
| 08:41.34 | perlguru-work | strange: ls: proc-armo: Input/output error |
| 08:47.26 | *** join/#nslu2-linux dyoung (~dyoung@cpe-24-94-83-103.hawaii.rr.com) |
| 08:53.25 | perlguru-work | welcome, dyoung |
| 08:56.11 | dyoung | hi |
| 08:56.46 | perlguru-work | any ideas about creating/packaging/installing/configuring kernel modules? |
| 08:57.04 | perlguru-work | there should be a howto for the developers |
| 08:57.05 | dyoung | creating is easy. |
| 08:57.12 | perlguru-work | yep |
| 08:57.47 | perlguru-work | take the printer stuff as example. bob_tm and christopher did great work |
| 08:58.06 | perlguru-work | but it is a bit user unfriendly |
| 08:58.20 | dyoung | thats a understatement ;-) |
| 08:58.24 | perlguru-work | you need to be an expert |
| 08:58.29 | dyoung | I gree |
| 08:58.30 | perlguru-work | to install it |
| 08:58.32 | dyoung | agree. |
| 08:58.44 | dyoung | okay the strategy might be this. |
| 08:58.46 | perlguru-work | so, lets think of a procedure.... |
| 08:58.52 | dyoung | (looking for it, just moment) |
| 08:59.05 | perlguru-work | web-irc is a bit slow in response.... |
| 09:00.23 | dyoung | rwhitby suggests we should copy/modify package.bbclass to meet our needs to package modules as ipkgs we can use. |
| 09:00.45 | dyoung | as package_unslung_ipkg.class |
| 09:00.55 | dyoung | ipk thta is |
| 09:00.56 | perlguru-work | is there a way to determine if standard or able is used? |
| 09:01.10 | dyoung | when> |
| 09:02.03 | perlguru-work | ~seen bob_tm |
| 09:02.04 | jbot | bob_tm <~chatzilla@254.80-203-34.nextgentel.com> was last seen on IRC in channel #nslu2-linux, 6d 21h 39m 9s ago, saying: 'Good - thanks.'. |
| 09:03.09 | rwhitby-away | we will have an openembedded feed |
| 09:03.53 | rwhitby-away | oops. perlguru-work: the reason why the dependencies are wrong is cause I typed in the wrong rc (27 instead of 37). |
| 09:03.58 | rwhitby-away | will fix and upload -5 :-) |
| 09:05.11 | rwhitby-away | hmm - actually all I need to upload is a new version of the smallest -4 one, so I will do that. |
| 09:05.49 | perlguru-work | ipkg: /opt/armeb/armv5b-softfloat-linux/include/asm/proc-armo/system.h: Input/output error |
| 09:05.54 | perlguru-work | strange |
| 09:06.01 | rwhitby-away | got enough disk space? |
| 09:06.08 | perlguru-work | it is on data |
| 09:06.24 | perlguru-work | i suspect a disc error... |
| 09:06.53 | rwhitby-away | me too |
| 09:06.58 | perlguru-work | ls: /opt/armeb/armv5b-softfloat-linux/include/asm/proc-armo: Input/output error |
| 09:07.35 | perlguru-work | yesterday it clobbert a directory of a normal user :( |
| 09:09.54 | rwhitby-away | good, it's working. |
| 09:11.22 | rwhitby-away | perlguru-work: try it now |
| 09:13.38 | perlguru-work | Configuring crosstool-native |
| 09:13.57 | perlguru-work | that went fast :) |
| 09:14.26 | rwhitby-away | yep - the small one only has the symlinks in it |
| 09:14.40 | rwhitby-away | (and the all important dependencies) |
| 09:14.47 | dyoung | anyone know offhand what the memory footprint of appweb is? |
| 09:15.06 | rwhitby-away | nope |
| 09:16.03 | dyoung | it goes away on me after "a while" with no errors in the error.log |
| 09:16.14 | dyoung | this is on a flashkey only box. |
| 09:17.02 | perlguru-work | dyoung: appweb goes away, as in, not running anymore? |
| 09:17.21 | perlguru-work | i noticed it too |
| 09:17.31 | dyoung | yeah, as in it dies for no apparent reason |
| 09:17.52 | perlguru-work | it is snowing :) |
| 09:23.03 | CIA-3 | 03rwhitby * 10unslung/sources/crosstool-native/control: Fixed the dependencies. |
| 09:34.13 | perlguru-work | EXT3-fs error (device sd(8,1)): ext3_free_inode: bit already cleared for inode 1687553 |
| 09:35.29 | dyoung | back later. |
| 09:35.39 | *** part/#nslu2-linux dyoung (~dyoung@cpe-24-94-83-103.hawaii.rr.com) |
| 10:05.05 | *** join/#nslu2-linux highbury (~steve@81-1-73-196.homechoice.co.uk) |
| 10:05.47 | Ksilebo|Kweh | Its a toasty -16.8 C out... |
| 10:10.11 | *** join/#nslu2-linux dyoung (~dyoung@cpe-24-94-83-103.hawaii.rr.com) |
| 10:11.50 | rwhitby-away | 37 C here today |
| 10:11.56 | siddy-away | does anybody know, how to mount an jffs2-image in unslung? |
| 10:12.13 | rwhitby-away | image or fs ? |
| 10:12.18 | siddy-away | image |
| 10:12.50 | rwhitby-away | [g2] was saying something about that earlier |
| 10:12.58 | dyoung | -o loop |
| 10:13.01 | dyoung | but you need the mtdram |
| 10:13.21 | siddy-away | ? |
| 10:14.18 | dyoung | the short story is its not quite ready to do it with the tools we have in unslung. |
| 10:14.38 | siddy-away | ah, k |
| 10:14.46 | dyoung | reboot time, back in a few |
| 10:14.50 | *** part/#nslu2-linux dyoung (~dyoung@cpe-24-94-83-103.hawaii.rr.com) |
| 10:21.25 | *** join/#nslu2-linux dyoung (~dyoung@cpe-24-94-83-103.hawaii.rr.com) |
| 10:24.30 | siddy-away | dyoung: is there any way to mount that gddam jffs2-image with an x86-linux? I can't even find the jffs2-fs in my kernel-config... |
| 10:25.00 | dyoung | not as far as I know. |
| 10:25.24 | dyoung | never really tried to though. |
| 10:25.52 | siddy-away | so... how can I have a look at the contents of my sweet openslug-image? no way? |
| 10:26.05 | dyoung | you just bitbaked it? |
| 10:26.12 | siddy-away | y |
| 10:26.21 | dyoung | ls tmp/rootfs ? |
| 10:26.31 | siddy-away | hm... too easy... ;) |
| 10:27.01 | dyoung | if you give me a few minutes I'm sure I can come up with something more annoying. ;-) |
| 10:27.19 | siddy-away | hehe, thx so far |
| 10:52.34 | CIA-3 | 03rwhitby * 10unslung/Makefile: Promoted openssl and openssh to cross and native. |
| 10:54.46 | rwhitby-away | thanks to perl and autoconf ... |
| 10:57.45 | jacques | yey! |
| 10:58.11 | rwhitby-away | looking for other freebies now ... |
| 10:58.19 | jacques | and for the record, both openssl and openssh pass all their tests when built native |
| 10:58.35 | rwhitby-away | how about when built cross? |
| 10:58.57 | jacques | haven't tested that - would have to tar up build dir from pc |
| 10:59.23 | jacques | automake comes from when autoconf is installed |
| 10:59.36 | rwhitby-away | just overwrite exes from cross ipk into native build dir and run the tests? |
| 10:59.37 | jacques | s/from/free/ |
| 11:00.28 | rwhitby-away | hmm - not for me |
| 11:00.35 | jacques | what's the error? |
| 11:00.39 | rwhitby-away | configure: error: The installed version of autoconf does not work. |
| 11:01.08 | jacques | I definitely don't get that :-\ |
| 11:01.15 | rwhitby-away | exec: /opt/bin/autom4te: No such file or directory |
| 11:01.28 | rwhitby-away | looks like autoconf ipk is not complete. |
| 11:02.33 | jacques | ah :-\ oops. it's possible I did a 'make install' in autoconf rather than install the ipk - good thing I have you around :-) |
| 11:02.51 | rwhitby-away | aha - autom4te is a script with #!/usr/bin/perl |
| 11:02.58 | rwhitby-away | looks like we need a symlink in the perl ipk |
| 11:03.43 | jacques | oh yea, I had made that symlink manually in anticipation of that type of issue :-\ |
| 11:04.00 | jacques | before you even packaged perl :-) |
| 11:04.39 | jacques | BTW I recommend a /usr/bin/perl5 link as well - some scripts look for that |
| 11:05.16 | jacques | (at least I've seen configure scripts look for it) |
| 11:05.38 | rwhitby-away | dunno about that one |
| 11:06.22 | rwhitby-away | ... /usr/bin/perl5 does not exist on my build host ... |
| 11:06.57 | jacques | hmm mine either, OK never mind - openssh configure looked for it, but settled for perl anyway |
| 11:07.29 | jacques | I thought I remembered it, but that was probably when perl5 was new |
| 11:08.44 | jacques | soon we'll have perl6 synlinks to worry about :-) |
| 11:09.00 | CIA-3 | 03rwhitby * 10unslung/ (make/perl.mk sources/perl/control): Added /usr/bin/perl symlink |
| 11:09.45 | jacques | do you remember why busybox is cross-only? |
| 11:10.12 | jacques | ah |
| 11:10.20 | jacques | lex.zconf.c: In function `zconf_fopen': |
| 11:10.20 | jacques | lex.zconf.c:3593: error: `PATH_MAX' undeclared (first use in this function) |
| 11:10.46 | jacques | I wish I knew the right way to fix that |
| 11:14.31 | rwhitby-away | bummer - I have to completely rebuild perl to fix that symlink, cause perl's configure puts hard coded paths in the Makefiles to the system include files ... |
| 11:16.47 | jacques | I don't understand |
| 11:20.10 | jacques | lol, dhcp build *fails* because it can't find nroff to make the man page |
| 11:23.47 | rwhitby-away | an xemacs and two compiles, and only 17MB swap used |
| 11:27.39 | jacques | I still say we need fatslugs :-) |
| 11:49.29 | rwhitby-away | is "ed" in any of the packages? |
| 11:51.36 | jacques | no, I built it to try to build appweb before I found out how braindead its configure is (x86 binaries) |
| 11:57.23 | rwhitby-away | CIA is slow tonight .... |
| 11:58.00 | jacques | what did you update? |
| 11:58.09 | CIA-3 | 03rwhitby * 10unslung/ (Makefile make/ed.mk sources/ed/control): Added ed |
| 11:58.17 | jacques | ah heh |
| 11:59.47 | jacques | why did you add it? |
| 11:59.59 | rwhitby-away | why not? |
| 12:00.12 | rwhitby-away | ed.ipk is uploaded |
| 12:00.39 | jacques | i just thought you needed it for something |
| 12:00.45 | perlguru-work | i did a school assignment ones with ed :) |
| 12:00.55 | rwhitby-away | I saw it go past - it may well have been appweb |
| 12:01.42 | rwhitby-away | patch looks for ed too |
| 12:02.03 | rwhitby-away | (not that anyone ever uses that patch option) |
| 12:02.06 | jacques | does you ed also print /lib/locale on stderr for no good reason? |
| 12:02.20 | jacques | your |
| 12:06.39 | perlguru-work | whooo! |
| 12:07.00 | *** join/#nslu2-linux LarsAC_ (~chatzilla@134.130.124.227) |
| 12:07.00 | *** join/#nslu2-linux dyoung (~dyoung@cpe-24-94-83-103.hawaii.rr.com) [NETSPLIT VICTIM] |
| 12:07.00 | *** join/#nslu2-linux highbury (~steve@81-1-73-196.homechoice.co.uk) [NETSPLIT VICTIM] |
| 12:07.00 | *** join/#nslu2-linux Tiersten (tman@ideal.trejan.com) [NETSPLIT VICTIM] |
| 12:07.01 | *** join/#nslu2-linux kergoth (~kergoth@ip71.churchill-park.dfw.ygnition.net) |
| 12:07.01 | *** join/#nslu2-linux garpinc (~garpinc@pool-141-154-232-220.bos.east.verizon.net) [NETSPLIT VICTIM] |
| 12:07.01 | *** join/#nslu2-linux cpage_ (~cpage@ip68-6-39-177.sb.sd.cox.net) [NETSPLIT VICTIM] |
| 12:07.01 | *** join/#nslu2-linux billwernet2 (~NSLU2-IRC@h73.100.140.67.ip.alltel.net) [NETSPLIT VICTIM] |
| 12:07.01 | *** join/#nslu2-linux Ksilebo|Kweh (russ@d47-69-56-108.try.wideopenwest.com) [NETSPLIT VICTIM] |
| 12:07.01 | *** join/#nslu2-linux glc-work (~glc@www.thelabelservice.com) [NETSPLIT VICTIM] |
| 12:07.01 | *** join/#nslu2-linux siddy-away (~sid@2001:5c0:857f:0:0:0:0:1) [NETSPLIT VICTIM] |
| 12:07.01 | *** join/#nslu2-linux rwhitby-away (~rwhitby@CPE-144-136-199-164.sa.bigpond.net.au) [NETSPLIT VICTIM] |
| 12:07.01 | *** join/#nslu2-linux ka6sox-zzzz (~tking@adsl-66-159-198-116.dslextreme.com) [NETSPLIT VICTIM] |
| 12:07.01 | *** join/#nslu2-linux grimreap (~grimreap@c-24-7-207-20.client.comcast.net) [NETSPLIT VICTIM] |
| 12:07.01 | *** join/#nslu2-linux dyoung-logger (~dyoung-lo@cpe-24-94-83-103.hawaii.rr.com) [NETSPLIT VICTIM] |
| 12:07.01 | *** join/#nslu2-linux VoodooZ_log (~VoodooZ@CPE00c0f021712c-CM014110003785.cpe.net.cable.rogers.com) [NETSPLIT VICTIM] |
| 12:07.01 | *** join/#nslu2-linux Exion (~jon@airplane1.suphammer.net) [NETSPLIT VICTIM] |
| 12:07.02 | *** join/#nslu2-linux noud (noud@infernoz.speedxs.nl) [NETSPLIT VICTIM] |
| 12:07.02 | *** join/#nslu2-linux giel (~giel@145.116.4.60) [NETSPLIT VICTIM] |
| 12:07.02 | *** join/#nslu2-linux peteru-home (~peterHome@adsl-53-18.swiftdsl.com.au) [NETSPLIT VICTIM] |
| 12:07.02 | *** join/#nslu2-linux CIA-3 (~CIA@flapjack.navi.cx) [NETSPLIT VICTIM] |
| 12:07.02 | *** join/#nslu2-linux cbrake_away (~cbrake@mail.accelent.com) [NETSPLIT VICTIM] |
| 12:07.02 | jacques | hey, groff builds native :-) |
| 12:07.02 | perlguru-work | welcome back all! |
| 12:07.15 | jacques | :-\ |
| 12:07.30 | jacques | wb |
| 12:07.30 | jacques | rwhitby-away, in case you missed because of split, groff builds native |
| 12:23.04 | CIA-3 | 03rwhitby * 10unslung/make/ed.mk: Fixed the install target |
| 12:31.19 | rwhitby-away | jacques: this "ed" doesn't print out /lib/locale as far as I can see ... can you test it? |
| 12:32.07 | rwhitby-away | my native groff fails: configure: error: a working C++ compiler is required |
| 12:35.45 | jacques | is there a config.log? that's usually a good way to find out what stupid thing the configure script tried to do |
| 12:36.24 | rwhitby-away | Just added libstdc++ to the list of packages that must be installed for native compilation .... |
| 12:38.11 | jacques | oh yeah, not too many c++ things build without that |
| 12:39.18 | perlguru-work | how can I do rsync over openSSH? |
| 12:40.31 | jacques | hmm I've done it before. I think it's an env var like RSYNC_RSH or something |
| 12:40.31 | jacques | but I can't remamber |
| 12:40.44 | perlguru-work | i mean server side |
| 12:40.54 | dyoung | rsync -e ssh blahblah |
| 12:40.55 | rwhitby-away | --rsh=COMMAND |
| 12:41.10 | perlguru-work | that is client side |
| 12:41.13 | rwhitby-away | yeah, -e is the short option |
| 12:41.24 | rwhitby-away | do you mean rsyncd? |
| 12:41.26 | perlguru-work | i mean: how do i set it up on serverside |
| 12:41.51 | perlguru-work | yes, and how to 'link' it to ssh |
| 12:43.20 | perlguru-work | something in sshd_config? |
| 12:43.59 | rwhitby-away | do you have ssh access to the server? |
| 12:44.08 | perlguru-work | Subsystem? |
| 12:44.20 | perlguru-work | the server is the slug ;) |
| 12:44.38 | perlguru-work | i can ssh in |
| 12:44.43 | rwhitby-away | so use rsync -e "ssh" ... |
| 12:44.52 | perlguru-work | from work ;) |
| 12:45.03 | perlguru-work | i will, if the server works |
| 12:45.17 | rwhitby-away | you don't need a server. |
| 12:45.24 | perlguru-work | got rsync installed |
| 12:45.44 | rwhitby-away | rsync uses ssh to log in as the user, then runs rsync on the server machine. No rsyncd need be involved. |
| 12:45.47 | perlguru-work | ssh is on port 443 |
| 12:46.20 | dyoung | "ssh -p 443" then. |
| 12:46.20 | rwhitby-away | have you tried rsync -e "ssh" ... ? |
| 12:46.57 | rwhitby-away | you use the same command you use to ssh in normally |
| 12:47.06 | perlguru-work | i getr a connection time out |
| 12:47.37 | perlguru-work | rsync error: error in rsync protocol data stream (code 12) |
| 12:47.37 | rwhitby-away | paste your command line |
| 12:48.52 | rwhitby-away | and check syslog on the server |
| 12:49.43 | perlguru-work | 13:27:27 rsyncd[4233]: rsyncd version 2.6.3 starting, listening on port 873 |
| 12:49.53 | perlguru-work | this is the last line |
| 12:50.23 | rwhitby-away | don't use a double colon |
| 12:50.49 | rwhitby-away | hostname:path is what you want (single colon) |
| 12:51.02 | rwhitby-away | double colon means run a server, which won't be listening on the right port for you |
| 12:51.21 | rwhitby-away | single colon means transfer everything over the existing socket connection (ssh port 443 for you) |
| 12:51.57 | perlguru-work | single colon does not work either :/ |
| 12:52.26 | rwhitby-away | what's the error, and syslog entry? |
| 12:52.32 | perlguru-work | (just want to get a list) |
| 12:52.52 | perlguru-work | syslog says nothing |
| 12:53.11 | rwhitby-away | ssh ls -l foo :-) |
| 12:53.14 | perlguru-work | error is what i pasted earlier |
| 12:53.33 | rwhitby-away | did you give a pathname ? |
| 12:53.41 | rwhitby-away | hostname:path ... |
| 12:54.29 | rwhitby-away | I didn't know you could just get a list .... |
| 12:54.55 | perlguru-work | it may be a proxy problem. normal ssh does not work either |
| 12:55.10 | rwhitby-away | ah |
| 12:55.27 | rwhitby-away | well, you're one step closer by using a single colon |
| 13:04.10 | rwhitby-away | night all |
| 13:04.37 | jacques | `night rwhitby-away |
| 13:07.12 | *** join/#nslu2-linux perlguru-work (www-data@netblock-66-159-209-60.dslextreme.com) |
| 13:18.53 | CIA-3 | 03djf * 10unslung/make/man.mk: initial checkin - builds and packages native - needs testing |
| 13:19.16 | perlguru-work | now i get: protocol version mismatch - is your shell clean? |
| 13:19.50 | CIA-3 | 03djf * 10unslung/sources/man/ (control man-configure.patch man2html-Makefile.in.patch): initial checkin |
| 13:20.38 | perlguru-work | btw we are using an old ssh |
| 13:20.39 | perlguru-work | <PROTECTED> |
| 13:20.53 | perlguru-work | (from openssh site) |
| 13:21.11 | perlguru-work | from ssh -V: OpenSSH_3.8p1, SSH protocols 1.5/2.0, OpenSSL 0.9.7d 17 Mar 2004 |
| 13:22.48 | jacques | lol |
| 13:22.59 | jacques | what? |
| 13:25.07 | perlguru-work | rsync gives me: protocol version mismatch - is your shell clean? |
| 13:25.08 | perlguru-work | client and host are same version rsync :/ |
| 13:26.32 | perlguru-work | mmm, rsync -e "ssh -p 443" root@nslu-host: |
| 13:26.37 | perlguru-work | works fine |
| 13:30.18 | *** join/#nslu2-linux [1]garpinc (~garpinc@216.118.190.9) |
| 13:31.59 | *** join/#nslu2-linux [2]garpinc (~garpinc@pool-141-154-232-220.bos.east.verizon.net) |
| 13:35.26 | *** join/#nslu2-linux siddy (~sid@slug.6bone.siddy.org) |
| 13:59.56 | CIA-3 | 03djf * 10unslung/make/man-pages.mk: initial checkin - needs testing |
| 14:00.38 | CIA-3 | 03djf * 10unslung/sources/man-pages/control: initial checkin |
| 14:07.39 | *** join/#nslu2-linux [1]garpinc (~garpinc@216.118.190.9) |
| 14:23.20 | *** join/#nslu2-linux [2]garpinc (~garpinc@pool-141-154-232-220.bos.east.verizon.net) |
| 14:25.36 | CIA-3 | 03djf * 10unslung/sources/bzflag/glext.h: initial checkin - forgot this one before |
| 14:25.37 | *** join/#nslu2-linux gradlman (~Miranda@DSL01.212.114.206.8.NEFkom.net) |
| 14:27.51 | *** join/#nslu2-linux [g2] (~g2@rdu57-8-035.nc.rr.com) |
| 14:28.12 | CIA-3 | 03djf * 10unslung/sources/bzflag/control: added deps: ncurses, openssl |
| 14:32.47 | VoodooZ_work | morning |
| 14:32.57 | gradlman | afternoon... |
| 14:33.55 | Tiersten | Hi |
| 14:34.32 | [g2] | morning |
| 14:34.41 | Tiersten | Afternoon |
| 14:34.43 | perlguru-work | hi |
| 14:34.47 | [g2] | ~ugt |
| 14:34.48 | jbot | extra, extra, read all about it, ugt is Universal Greeting Time. It states that it is always morning when person comes in, and it is always late when person leaves. Local time of any member of channel is irrelevant. |
| 14:34.58 | [g2] | heh |
| 14:35.13 | CIA-3 | 03djf * 10unslung/make/bzflag.mk: stages openssl, stages gl headers in build dir, does not install contents of share, strips binaries |
| 14:35.15 | [g2] | hey perlguru-work |
| 14:37.58 | *** join/#nslu2-linux kitno455 (~man@tfhs.net) |
| 14:38.07 | kitno455 | re |
| 14:39.52 | Tiersten | Hey |
| 14:41.26 | Tiersten | Hmm. Need to trim the AllRecentChanges page |
| 14:41.38 | Tiersten | It's massive at the moment |
| 14:47.42 | *** join/#nslu2-linux [1]garpinc (~garpinc@pool-141-154-232-220.bos.east.verizon.net) |
| 15:04.26 | *** join/#nslu2-linux garpinc (~garpinc@pool-141-154-232-220.bos.east.verizon.net) |
| 15:08.14 | VoodooZ_work | Anybody know of a machine vision library (like OpenCV) package in OE? |
| 15:09.17 | Tiersten | no clue |
| 15:09.46 | Tiersten | For your robot, you using the standard kernel or RT? |
| 15:09.54 | VoodooZ_work | Standard kernel. |
| 15:10.00 | VoodooZ_work | It's plenty for now. |
| 15:10.08 | VoodooZ_work | But I'd love to get the RT patches in later. |
| 15:10.33 | Tiersten | Ah okay. Don't want your robot to crash if suddenly it ignores your app for a few ms :) |
| 15:10.47 | VoodooZ_work | hehe |
| 15:11.10 | VoodooZ_work | Yeah, That's why I need to do more testing to see how linux behaves as it's my first time using it. |
| 15:11.48 | Tiersten | Interrupt latency is quite variable if you're not using a real time patch |
| 15:12.01 | VoodooZ_work | my code spawns two threads, one tcp server and a higher priority one that does the AI |
| 15:12.43 | VoodooZ_work | It basically loops at around 25-30Hz. |
| 15:12.53 | VoodooZ_work | fetching updates from the sensors and processing them. |
| 15:13.34 | VoodooZ_work | I'm starting to play with vision and it's real interesting. It's not a priority for my project but I'd definitely like to use it for other projects now that I have a working USB cam |
| 15:14.22 | VoodooZ_work | My old robot did a very primitive blob tracking in an 8K 16MIPS micro so I couldn't do much. |
| 15:14.23 | *** join/#nslu2-linux jacques (~jacques@c-67-171-198-193.client.comcast.net) |
| 15:14.35 | Tiersten | My old university specialised in machine vision |
| 15:15.02 | VoodooZ_work | and with 2K RAM I had to process pixel as they came in as I couldn't store the full 128x128 frame! :( |
| 15:15.36 | VoodooZ_work | which University? |
| 15:15.36 | Tiersten | RHUL in the UK |
| 15:15.36 | *** join/#nslu2-linux ByronT (ByronT@adsl-212-196-43.bhm.bellsouth.net) |
| 15:15.36 | VoodooZ_work | cool. Did you do some? |
| 15:15.49 | Tiersten | Nope :) |
| 15:16.02 | Tiersten | The lecturers were all into AI and machine vision |
| 15:16.04 | VoodooZ_work | I'm learning so I'm asking for ideas at this point. |
| 15:16.14 | kitno455 | ideas on what? |
| 15:16.24 | kitno455 | doing cv on linux? |
| 15:16.27 | VoodooZ_work | how to approach my vision problem. |
| 15:16.32 | Tiersten | Get glasses :) |
| 15:16.32 | VoodooZ_work | yep |
| 15:16.33 | ByronT | can someone who logs tell me what rwhitby said for resolution of the broken README into 2 text files? |
| 15:16.35 | VoodooZ_work | heheh |
| 15:16.50 | VoodooZ_work | Check the logs on the web page. |
| 15:16.58 | ByronT | ah |
| 15:16.59 | kitno455 | VoodooZ_work, what is your task this time, still blowing out candles? |
| 15:20.53 | VoodooZ_work | yep. for now. I'm just building a base which is flexible enough to be reused for other tasks. |
| 15:21.45 | VoodooZ_work | Using an IR bandpass filter helps eliminate the noise and capture just the candle but you still have to segment the blob(s). |
| 15:22.10 | VoodooZ_work | I might start checking out the OpenCV lib to see if It can save me time. |
| 15:22.21 | Tiersten | I take it smoking near your robot wouldn't be a good thing then? |
| 15:22.49 | VoodooZ_work | Nope! The UV sensor would think there's a candle near by and hunt you down! :) |
| 15:23.45 | kitno455 | you have separate IR and vis cameras? |
| 15:24.47 | VoodooZ_work | nope. there's a UV sensor that only indicates the presence of a flame within 45degrees and a camera that filters visible light out. |
| 15:25.45 | kitno455 | so do you start in the room with the candle? |
| 15:25.58 | VoodooZ_work | of course not. |
| 15:26.12 | Tiersten | How you detecting walls? |
| 15:27.26 | VoodooZ_work | the expert level has a random maze |
| 15:27.40 | VoodooZ_work | using sharp IR rangers for now but a laser ranger would have been better. |
| 15:27.50 | kitno455 | how wide are maze halls? |
| 15:27.57 | VoodooZ_work | around 50cm |
| 15:28.04 | VoodooZ_work | in a 3mx3m maze |
| 15:28.07 | kitno455 | any wide opens? |
| 15:28.09 | VoodooZ_work | with a second floor. |
| 15:28.13 | *** join/#nslu2-linux [1]garpinc (~garpinc@pool-141-154-232-220.bos.east.verizon.net) |
| 15:28.15 | kitno455 | two stories? |
| 15:28.16 | VoodooZ_work | wide open? |
| 15:28.29 | kitno455 | big areas, 1m squaRE |
| 15:28.32 | Tiersten | Shame they've got rules to prevent you going over and through walls :) |
| 15:28.35 | VoodooZ_work | well, it's there's a 15 degrees max ramp to the second floor |
| 15:28.56 | VoodooZ_work | the rules define all that but I don't think you can have such a big room. |
| 15:29.03 | kitno455 | you got a means to locate the ramp? |
| 15:29.10 | VoodooZ_work | as there's a minimum of 2-3 rooms. |
| 15:29.37 | VoodooZ_work | I was thinking of using an accelerometer as a tilt sensor but then again the robot doesn't need to know. |
| 15:29.44 | VoodooZ_work | It's just another hall really. |
| 15:29.50 | kitno455 | not exactly |
| 15:30.23 | kitno455 | if you are using a graph to keep track of places you have been, it is good to know that you are on a different floor plan |
| 15:31.33 | Tiersten | detect it by your incline? |
| 15:31.33 | kitno455 | stupid mercury bubble could do that if there are no bumps |
| 15:32.22 | VoodooZ_work | yeah, but there's carpets and all so... |
| 15:32.48 | VoodooZ_work | That's why I would prefer to use a tilt-sensor averaged over a few seconds to raise a flag. |
| 15:33.02 | kitno455 | right |
| 15:33.22 | VoodooZ_work | Plus, the second flood is not directly over the first one so it's really like a long hall to an extension of the base floor. |
| 15:33.33 | kitno455 | oh, thats lazy |
| 15:33.36 | VoodooZ_work | It's quite challenging. |
| 15:33.50 | VoodooZ_work | I know but they have to build-up the challenge slowly every year. |
| 15:33.59 | kitno455 | hope the ramp has a rail :) |
| 15:34.30 | VoodooZ_work | heheh Last year only 3 groups participated and none of them could even navigate the whole maze so... |
| 15:34.46 | kitno455 | what does your logic look like? |
| 15:35.19 | kitno455 | right turns with measured distance, followed by a left? |
| 15:37.10 | VoodooZ_work | well, it's not exactly defined yet but I'm aiming for a behaviour-based approach. With a second layer planning and mapping. |
| 15:37.58 | VoodooZ_work | There are cycles in there so you can't just blindly explore as you could get stuck |
| 15:39.59 | VoodooZ_work | the local robotic club has a nerf-bot competition too so I'd like to eventually use vision for that too. |
| 15:40.08 | VoodooZ_work | I have big plans for that little slug! |
| 15:40.14 | Tiersten | what is a nerf-bot? |
| 15:40.59 | ByronT | off topic warning - any reason to get WRT54GS rather than WRT54G (linksys wireless aps)? |
| 15:41.14 | Tiersten | Faster CPU and more RAM |
| 15:41.27 | Tiersten | Speedbooster can be enabled on later revisions of the WRT54G hardware tho |
| 15:41.29 | [g2] | ByronT, twice the memory and Flash, NO faster CPU |
| 15:41.42 | Tiersten | It's faster than a v1 non GS :) |
| 15:41.42 | VoodooZ_work | a modded nerf-gun on an autonomous robot hunting each other in an arena. |
| 15:41.46 | VoodooZ_work | No remote control! |
| 15:41.59 | ByronT | at pcmall.com only $20 difference |
| 15:42.22 | [g2] | Tiersten, V1's have'n been around for 6-9 months unless you're buying off e-bay :0 |
| 15:42.24 | [g2] | :) |
| 15:42.24 | VoodooZ_work | anyways I'll stop boring you guys with my OT stuff. |
| 15:42.42 | Tiersten | [g2]: You've not seen the ancient crap coming out of Amazon UK then :) |
| 15:42.50 | Tiersten | I bought a G about 3 months ago and it was a v1 unit! |
| 15:43.09 | Tiersten | grr |
| 15:43.12 | [g2] | cool, then you've got the minipci connector |
| 15:43.18 | *** join/#nslu2-linux LarsAC (~chatzilla@pD9501154.dip0.t-ipconnect.de) |
| 15:43.29 | Tiersten | Yeah. Don't know what to do with it however |
| 15:43.52 | ByronT | Tiersten, you're supposed to feel *special* just having it |
| 15:44.02 | [g2] | that's like 4 hw generations ago |
| 15:44.18 | Tiersten | I'm overflowing with Ctbertan based APs at the moment |
| 15:44.35 | Tiersten | Got 2 Buffalo ones and 2 Linksys ones. All v1 |
| 15:44.56 | ByronT | my apple airport extreme has decided to go flaky on me, so it looks like I'll go with the GS |
| 15:45.08 | Tiersten | hmm? whats wrong with it? |
| 15:45.15 | Tiersten | it overheating? |
| 15:45.29 | ByronT | it doesn't want to connect two G clients at same time |
| 15:45.35 | Tiersten | weird |
| 15:45.54 | ByronT | so the iMac G5 and this laptop fight.... |
| 15:48.06 | *** join/#nslu2-linux NSLU2-IRC (~NSLU2-IRC@c-24-10-48-131.client.comcast.net) |
| 16:01.17 | kitno455 | VoodooZ_work, not bored, on a break :) |
| 16:02.25 | VoodooZ_work | hehehe. I'm not. But It's not too busy right now so I'm reading up on OpenCV. |
| 16:02.42 | VoodooZ_work | I wonder how slow it would be on the slug. |
| 16:02.47 | rt | okay, stupid question, but the conf partition on my drive (formatted prior unslung) is too small to hold crosstools-native. Did I skip a step while putting this thing together? |
| 16:16.56 | rt | never mind, found the answer on the wiki. |
| 16:25.36 | VoodooZ_work | darn, a new openslug error on nslu2-linksys-firmware: substr outside of string at /local/openslug/build/tmp/staging/i686-linux/bin/slugimage line 206. |
| 16:26.54 | kitno455 | what kernel is default in openslug? |
| 16:27.36 | VoodooZ_work | 2.6.9 |
| 16:27.52 | VoodooZ_work | I'd love to have 2.6.10 but I can't find the time to do it. |
| 16:36.32 | *** join/#nslu2-linux ByronT-Laptop (ByronT@adsl-220-251-246.bhm.bellsouth.net) |
| 16:41.52 | ByronT | I couldn't check the logs because the conversation I was looking for must be in today's log |
| 16:43.49 | ByronT | anyways, I now have three files for the README... two files are for one set - README and Notes.txt, the other is all in one README.. |
| 16:44.23 | ByronT | I'll let you guys make the choice between the two... |
| 16:44.52 | VoodooZ_work | Sorry, my log doesn't seem to go that far. |
| 16:45.31 | ByronT | the set is http://www.toddcomp.com/NSLU2/Unslung-3.1x-beta-README.txt and http://www.toddcomp.com/NSLU2/Unslung-3.1x-beta-Notes.txt |
| 16:45.49 | ByronT | the single (all in one) file is http://www.toddcomp.com/NSLU2/Unslung-3.1x-README.txt |
| 16:46.24 | ByronT | in the set, the README.txt is complete, but the Diversion Scripts forward of the Notes.txt is being worked on |
| 16:46.53 | VoodooZ_work | sorry, I've never used unslung so I can't really help. |
| 16:47.10 | ByronT | in the all-in-one README, I'm working on the context to make sense and again the Diversion Scripts forward is being worked on |
| 16:47.41 | ByronT | that's ok... I'm just getting that stuff in so that rwhitby or others can review when possible.... |
| 16:48.09 | VoodooZ_work | ic |
| 16:48.20 | ByronT | I'm going to get back and finish the Diversion stuff and the context so that the doco will be *complete* for the time being |
| 17:02.40 | ByronT | heh, I can tell that the original doco was written by a coder... all "Ifs" followed by "thens".... |
| 17:02.59 | ka6sox-zzzz | lol |
| 17:08.26 | Tiersten | Ick. A BASIC programmer! :) |
| 17:09.02 | ka6sox-zzzz | OR vhdl. |
| 17:09.56 | Tiersten | Nah. There would be processes :) |
| 17:10.06 | ka6sox-zzzz | true |
| 17:10.37 | ka6sox-zzzz | if (gck0); |
| 17:15.25 | ByronT | can someone real quick check the enable telnet on boot diversion script at bottom of this page: http://www.nslu2-linux.org/wiki/Unslung/DiversionScripts |
| 17:15.39 | ByronT | if that one is good, I'm putting it into README... |
| 17:17.13 | *** join/#nslu2-linux gradlman1 (~Miranda@DSL01.212.114.206.8.NEFkom.net) |
| 17:21.36 | rt | looking at the output of "ps", it appears to be running /usr/sbin/thttpd twice. Is that right? |
| 17:27.58 | rt | woohoo! |
| 17:31.11 | jacques | rt, it's quite common for webservers to run multiple processes |
| 17:31.23 | Tiersten | Or threads |
| 17:31.53 | jacques | yeah. I dunno which ps he's using and how it displays threads |
| 17:33.09 | rt | I'm virtually certain that thttpd doesn't use threads. |
| 17:33.15 | Tiersten | Depending on the thread library used, you either see them as separate processes or don't see them at all |
| 17:33.34 | Tiersten | Shouldn't be anything to do with ps unless they try to identify what is a thread and what isn't |
| 17:37.40 | rt | hmmm. |
| 17:38.31 | ByronT | the two separated doco files finito.... now just finishing context of the all in one README to make sense... |
| 17:45.28 | ByronT | ok, context finished... working on Suggested Changes wiki page |
| 17:45.47 | rt | heh. it's weird to use a machine this slow after using 3ghz dual xeons. :-) |
| 17:46.33 | Tiersten | Yeah but the fans in that are probably drawing more power than the entire NSLU2 |
| 17:46.40 | rt | heh. |
| 17:46.45 | rt | Indeed. |
| 17:46.53 | rt | that's one of the reasons I bought one of these to play with. |
| 17:47.16 | jacques | lol maybe *one* of the fans even |
| 17:47.43 | jacques | well not counting the attached drive |
| 17:49.35 | rt | is the sed that is part of slingbox especially stripped down in some way? |
| 17:49.44 | jacques | yes |
| 17:49.57 | Tiersten | it's the busybox one isn't it? |
| 17:50.07 | rt | grrr. |
| 18:04.24 | rt | well, that worked out of the box. |
| 18:07.54 | kitno455 | does openslug keep the sysconf partition? |
| 18:09.17 | jacques | it doesn't overwrite it |
| 18:09.41 | jacques | [g2] is working on code for openslug to be able to read it to set the ethernet parameters |
| 18:10.22 | kitno455 | ok, the redboot musings page on the wiki talked about trashing it. |
| 18:10.35 | kitno455 | i was going to ask about where are you gettign the mac from... |
| 18:10.58 | jacques | well, some ppl would like to have a much more flexible bootloader which would allow that |
| 18:11.16 | jacques | I'm one of them |
| 18:11.30 | jacques | that's camp4 |
| 18:11.33 | kitno455 | sure. place that makes the most sense might be in the first block with the bootloader |
| 18:11.36 | *** join/#nslu2-linux jipi (~jipi@cm187.gamma181.maxonline.com.sg) |
| 18:12.11 | jacques | i dunno about that - you would have to erase the whole block to change the settings |
| 18:12.14 | kitno455 | that would be APEX, or mod'd redboot |
| 18:12.29 | jacques | or uboot or blob or... |
| 18:12.34 | jacques | LAB |
| 18:12.53 | kitno455 | i played with uboot a bit a few months ago, not on the slug though |
| 18:13.42 | kitno455 | i myself am a little freaked about redboot running a telnetd, etc. |
| 18:14.04 | kitno455 | i think i might put a slug on a hostile network |
| 18:14.19 | rt | hmm. is there a library with tgetent/tgets/tputs in it? |
| 18:14.35 | jacques | you mean like ncurses ? |
| 18:14.47 | jacques | libtermcap might have all that too |
| 18:14.55 | jacques | and is smaller if that matters |
| 18:16.23 | ByronT | rwhitby-asleep, my work on doco complete - for the time being... http://www.nslu2-linux.org/wiki/Unslung/SuggestedChangesToTheReadme for review |
| 18:16.42 | rt | hmmm. |
| 18:16.46 | ByronT | I'm taking a break... whew and blah.. |
| 18:17.33 | rt | hmmm. links okay, but can't find the shared library when run. |
| 18:17.46 | rt | what's that silly environment variable to add things to the dynamic link path? |
| 18:18.15 | ByronT | My NSLU2 todo list reads: DiversionScript wiki, Changelog (readme and wiki), and Features (wiki) |
| 18:18.31 | jacques | LD_LIBRARY_PATH but we recommend against using it |
| 18:18.53 | rt | well, what's the right way to add /opt/lib ? |
| 18:18.59 | jacques | LDFLAGS=-L/opt/lib -Wl,-rpath,/opt/lib |
| 18:20.39 | rt | hmmm. okay. |
| 18:20.48 | rt | well, I've got tcsh built. :-) |
| 18:20.58 | jacques | heh |
| 18:21.18 | jacques | i used tcsh for over a decade, but finally forced myself to switch to bash |
| 18:21.28 | jacques | got tired of never having it on strange machines |
| 18:22.53 | jacques | so rt when are we going to see tcsh.mk committed to cvs? |
| 18:24.38 | rt | the changes needed are minor, but I haven't ever created a .mk. What does it entail? |
| 18:25.38 | jacques | checkout unslung cvs, copy template.mk to tcsh.mk and make a few changes, get cvs write access, check it in |
| 18:26.22 | jacques | i would paste a wiki page but I dunno exactly where that is |
| 18:26.57 | jacques | unslung cvs works for both cross and native compilation |
| 18:27.29 | jacques | the template is quite nice - you wouldn't have had to worry about "-L/opt/lib -Wl,-rpath,/opt/lib" for example |
| 18:28.04 | jacques | BIAB |
| 18:28.33 | rt | hmmm. seems to work. uses about 500K of memory. Not great, but not terrible. |
| 18:28.58 | rt | and now with the friendly "%" prompt character. |
| 18:29.11 | kitno455 | you csh guys scare me |
| 18:30.04 | rt | what scare me are people who run bash because it is "small", and yet run emacs to edit files. :-) |
| 18:30.09 | kitno455 | hahahah |
| 18:30.15 | kitno455 | i wont touch emacs ever |
| 18:30.19 | Tiersten | bah. who needs a shell? emacs can do it all! :) |
| 18:30.22 | jacques | i run both bash and emacs on my slug |
| 18:30.38 | kitno455 | who needs an os when you have emacs |
| 18:30.47 | jacques | and my bash processes are currently between 1MB and 1.3MB |
| 18:31.13 | rt | so, presumably I could just create a new thttpd.conf file and run another instance of the installed daemon to serve files (on another port) |
| 18:31.28 | kitno455 | i run low 14's in the quarter |
| 18:31.40 | Tiersten | rt: Yeah |
| 18:31.46 | Tiersten | In reference to thttpd |
| 18:32.41 | jacques | rt, I've been running top for weeks |
| 18:32.41 | kitno455 | guys, do any of the bootloaders have any leaning toward supporting the usb natively? |
| 18:32.54 | Tiersten | I think uboot does |
| 18:33.05 | jacques | rt, ipkg install procps |
| 18:33.06 | Tiersten | somebody mentioned ages ago that it supported console on a USB serial dongle |
| 18:33.16 | kitno455 | thats a start |
| 18:33.29 | kitno455 | i am thinking about upgrade and config via flash drive |
| 18:33.51 | Tiersten | LAB would as well if anybody can get it going |
| 18:38.36 | *** join/#nslu2-linux ka6sox (~tking@netblock-66-159-209-52.dslextreme.com) |
| 18:45.58 | rt | heh. http://vandewettering.net is now moved to my slug. :-) |
| 18:47.25 | rt | (probably has some dead links, but the page itself is moved) |
| 18:47.46 | giel | you belgian? |
| 18:48.01 | rt | moi? no, just an ugly american. |
| 18:48.07 | rt | of dutch ancestry. |
| 18:48.17 | giel | dutchmen rock :P |
| 18:48.42 | rt | well, i don't know about the rest of them, but I know I certainly do. |
| 18:48.47 | rt | :-) |
| 18:48.51 | giel | i know i do |
| 18:48.53 | giel | too |
| 18:50.41 | rt | when I install procps, it bitches that kill and uptime are alredy installed by coreutils. |
| 18:50.47 | rt | should I force an overwrite? |
| 18:52.44 | *** join/#nslu2-linux [cc]smart (~smart@gw.ptr-62-65-149-158.customer.ch.netstream.com) |
| 19:05.24 | jacques | rt, up to you. there's no pretty way to handle that issue yet. |
| 19:09.58 | dyoung | BASIC Programmer?! |
| 19:10.53 | jacques | you lost me there dyoung :-) |
| 19:15.00 | ka6sox | Tiersten's reference to if, then statements earlier |
| 19:15.18 | ByronT | lots of them in the original documentation |
| 19:15.55 | dyoung | sigh |
| 19:16.16 | dyoung | Luckily we have a professional looking at it now. |
| 19:16.37 | ByronT | ? |
| 19:17.26 | dyoung | (I wrote the original if-then-else like document) |
| 19:17.29 | ByronT | heh... |
| 19:17.45 | ByronT | sorry.. ;) |
| 19:18.27 | dyoung | np. ;-) |
| 19:18.55 | ByronT | but I'm only professional in the sense of "professional user" |
| 19:22.41 | jacques | aah it all makes sense now |
| 19:23.07 | jacques | dyoung, was (rightfully) offended at being called a BASIC programmer |
| 19:24.41 | *** join/#nslu2-linux rt_ (~markv@omnipotus.pixar.com) |
| 19:24.44 | rt_ | *sigh* |
| 19:25.29 | rt_ | power apparently glitched at home, and now my machine isn't talking to me anymore. |
| 19:25.51 | rt_ | i'm surprised that my irc session hasn't timed out. |
| 19:26.42 | rt | there we go. |
| 19:26.54 | rt | Damnit, now I can't play with my nslu2 box until I go home tonight. |
| 19:27.27 | kitno455 | have not done the always-on hack? |
| 19:27.37 | rt | nope. |
| 19:27.47 | VoodooZ_work | I hate when that happens. Especially after flashing a bad load. arrrghh! |
| 19:27.47 | rt | irritatingly enough, my main webserver also appears down. |
| 19:36.35 | *** join/#nslu2-linux rt (~markv@c-24-7-79-31.client.comcast.net) |
| 19:37.08 | rt | ah, well, my main webserver is back at least. |
| 19:37.28 | rt | but the slug is probably powered off. |
| 19:37.30 | rt | *sigh* |
| 19:40.07 | jacques | i hate it when that happens :-( |
| 19:41.07 | rt | so, this power switch mod seems to be unreliable? |
| 19:44.45 | jacques | I dunno current status on it |
| 19:45.23 | VoodooZ_work | I hope it's not unreliable as I was just about to do it on my slug for my 'bot. |
| 19:45.36 | VoodooZ_work | perhaps I'll find a better way of doing it. |
| 19:46.44 | rt | how are you wiring i/o to the slug from the robot? |
| 19:47.33 | VoodooZ_work | I take out the sda/scl pins from the rtc chip. |
| 19:48.03 | VoodooZ_work | because it's an open collector bus I can talk to 5V i2c devices. as long as the master pulls it up at 3.3V |
| 19:48.16 | rt | ah, cool. |
| 19:50.36 | VoodooZ_work | it's not as good as a hardware i2c module but it will do fine. |
| 19:51.32 | VoodooZ_work | damn right! |
| 20:00.16 | Tiersten | They lose money |
| 20:00.32 | rt | yes, but how much? |
| 20:00.47 | rt | if they really were selling xboxes at a profit, how much would they cost? |
| 20:01.04 | Tiersten | They don't sell them at a profit |
| 20:01.05 | VoodooZ_work | Don't worry, now that they've got everybody hooked (by the balls!) they'll crank up the price of Xbox2 |
| 20:01.16 | VoodooZ_work | They wanted to gain marketshare, and they did. |
| 20:01.38 | Tiersten | The way it works is that you sell it at a huge loss initially but as you refine the design you reduce the costs |
| 20:01.53 | VoodooZ_work | I hate sony as much but I hope they come up with a killer console too. competition is great for us. |
| 20:02.10 | Tiersten | Apparently they lost over $100 on each console. That's what I heard anyway. May have changed now |
| 20:02.18 | rt | mucking with the nslu2 has made me think about other boxes that could be repurposed to more useful ends. |
| 20:02.55 | Tiersten | The power mod is unreliable yes |
| 20:03.18 | Tiersten | My solution (when I get around to it) is to use a PIC or something to do it |
| 20:03.41 | Tiersten | I'll also wire it into the I2C bus so I can get a few extra IOs pins and use it to control the pullups on the USB slave port |
| 20:04.04 | rt | yeah, I was thinking one of those tiny atmel procs I have lying around would work okay. |
| 20:04.21 | VoodooZ_work | yep! attiny or mega8. |
| 20:04.26 | Tiersten | If you just need the auto power on ability then yeah, it'd be easy |
| 20:04.58 | Tiersten | Just make the uC send a pulse on one of it's IO pins and wire that to the flipflop |
| 20:07.07 | rt | maybe I could get that to work with the slug. :-) |
| 20:07.19 | Tiersten | VoodooZ_work has gotten some other USB cam to work |
| 20:09.13 | Tiersten | Assuming the driver works with your variant of the camera and it doesn't have dumb assumptions like everything is a x86 then it should work |
| 20:10.22 | VoodooZ_work | yeah, I got a used Intel Me2Cam for cheap and it was on the ov5xx supported list so I jumped on it! |
| 20:10.33 | *** join/#nslu2-linux NSLU2-IRC (~NSLU2-IRC@c-24-10-48-131.client.comcast.net) |
| 20:10.59 | Tiersten | A PS2 EyeToy apparently is also supported as well if you can find one cheap |
| 20:11.02 | VoodooZ_work | The funny thing is: There's no Win2k/xp drivers for it so it only works in linux! |
| 20:11.03 | Tiersten | It's another OV chip inside |
| 20:11.06 | Tiersten | heh |
| 20:11.11 | VoodooZ_work | yep. same driver. |
| 20:11.15 | rt | that's hilarious. |
| 20:11.17 | VoodooZ_work | built-in 2.6 kernel too. |
| 20:11.22 | Tiersten | Must be ancient then if it doesn't have new drivers for it |
| 20:11.32 | VoodooZ_work | I guess it was very old as it only had win9x drivers. |
| 20:11.36 | Tiersten | 2K/XP changed the USB subsystem |
| 20:11.50 | VoodooZ_work | works fine for me though. 352x288 colour. |
| 20:11.54 | Tiersten | You can try installing the older drivers but chances are it won't work or just crash |
| 20:12.12 | VoodooZ_work | nah. I rip the board out for installing on my robot. |
| 20:12.18 | VoodooZ_work | it works fine. |
| 20:12.31 | Tiersten | Only 352x288? |
| 20:12.34 | VoodooZ_work | I just hope I can disable the auto-exposure or control it |
| 20:12.41 | VoodooZ_work | that's plenty for my app. |
| 20:13.07 | VoodooZ_work | I don't want to kill the slug processing all these pixels. |
| 20:13.11 | rt | so, you just compiled the standard driver and it worked? |
| 20:14.06 | NSLU2-IRC | nick CrazyEddie |
| 20:14.21 | VoodooZ_work | hehehe |
| 20:14.25 | VoodooZ_work | yep. |
| 20:15.06 | VoodooZ_work | with OE it's even simpler, i just changed the defconfig file openslug-kernel and it built it. |
| 20:15.13 | Tiersten | Make a 2 wheel balancing robot |
| 20:15.17 | Tiersten | They look quite neat |
| 20:15.27 | VoodooZ_work | And there's even a nice cam package called camsource in OE. |
| 20:15.42 | VoodooZ_work | yeah, but they don't do anything more. |
| 20:15.56 | Tiersten | Save on wheels? :) |
| 20:16.03 | VoodooZ_work | I'd rather spend time giving them "intelligence". |
| 20:16.21 | VoodooZ_work | yeah, I have two with a dual caster suspension. |
| 20:18.48 | VoodooZ_work | scanline: Are you there? |
| 20:44.10 | *** part/#nslu2-linux gradlman1 (~Miranda@DSL01.212.114.206.8.NEFkom.net) |
| 20:48.38 | ka6sox | Tiersten:SoW? |
| 20:50.09 | Tiersten | ? |
| 20:50.50 | ka6sox | Save on Wheels |
| 20:52.12 | VoodooZ_work | ? |
| 20:55.30 | ka6sox | [12:18] <Tiersten>Save on wheels? :) |
| 20:59.48 | VoodooZ_work | he was refering to the balancing robot thing. |
| 21:01.44 | scanline | VoodooZ_work: hmm? |
| 21:02.10 | VoodooZ_work | hey, |
| 21:02.40 | VoodooZ_work | kergoth told me you were looking into a CV lib package? |
| 21:02.56 | VoodooZ_work | I was wondering if there was one already in OE, like OpenCV for example. |
| 21:03.56 | *** join/#nslu2-linux [1]garpinc (~garpinc@pool-141-154-232-220.bos.east.verizon.net) |
| 21:05.05 | *** join/#nslu2-linux grimreap- (grimreap@c-24-7-207-20.client.comcast.net) |
| 21:07.02 | CIA-3 | 03djf * 10unslung/make/perl.mk: updated Errno_pm.PL sed to match new toolchain path |
| 21:08.14 | VoodooZ_work | scanline: my question above. |
| 21:14.01 | scanline | VoodooZ_work: I've used OpenCV, but not on any embedded hardware |
| 21:18.09 | *** join/#nslu2-linux [1]garpinc (~garpinc@pool-141-154-232-220.bos.east.verizon.net) |
| 21:18.20 | VoodooZ_work | i see. umm. I wonder how possible to cross-compile that would be. ouch. |
| 21:19.02 | VoodooZ_work | I want to do basic CV on my slug-based robot with a usb webcam so It would be nice not to have to reinvent the wheel. |
| 21:26.47 | scanline | yeah. an OpenCV package for OE would be quite nice |
| 21:27.29 | VoodooZ_log | How big is it approx? |
| 21:27.51 | VoodooZ_log | I doubt I have the knowledge or time do convert it to a package though. |
| 21:28.10 | VoodooZ_log | the whole cross-compiling thing makes my head spin. |
| 21:29.14 | VoodooZ_log | unless our good friend kergoth would do it? ;) |
| 21:29.50 | *** join/#nslu2-linux dJOEk (~djoek@kitchen.sync.be) |
| 21:29.53 | dJOEk | hail |
| 21:30.06 | dJOEk | has anyone ever tried to connect to their slug in dos ? |
| 21:32.59 | christopher | not that I've heard of |
| 21:54.33 | kitno455 | dJOEk, you mean via samba? |
| 22:13.43 | *** join/#nslu2-linux [1]garpinc (~garpinc@pool-141-154-232-220.bos.east.verizon.net) |
| 22:33.39 | *** join/#nslu2-linux dyoung (~dyoung@cpe-24-94-83-103.hawaii.rr.com) |
| 22:38.16 | *** join/#nslu2-linux grimreap- (grimreap@c-24-7-207-20.client.comcast.net) |
| 23:34.19 | *** join/#nslu2-linux ByronT-Away (ByronT@adsl-220-251-246.bhm.bellsouth.net) |