**** BEGIN LOGGING AT Mon Feb 07 23:59:56 2005 Feb 08 00:11:32 goodmorning! Feb 08 00:11:51 hey\ Feb 08 00:12:14 it's bad when you have to tell the technical support person to stop asking silly questions Feb 08 00:13:14 Linksys EZXS16W... the thing is dead as a door nail.. and Linksys support :Katrina Baldo: Which port did you connect the pc? ids the led lit? Feb 08 00:14:04 the thing won't power up at all... and she asks me to plug something into it... Feb 08 00:15:16 follow up question: Katrina Baldo: Can you still network pc using the switch Feb 08 00:15:22 hah! Feb 08 00:15:45 "just give me fcking RMA, you st00pid " Feb 08 00:16:12 this Katrina Baldo isn't very good at english, is she now? ;) Feb 08 00:18:32 yes, she's a moron and now she's blaming me for confusing her Feb 08 00:20:03 she says that maybe the LEDs are defective... but switch doesn't work... Feb 08 00:29:32 well, even if it's just a defective led, then you should get an RMA anyway Feb 08 00:29:51 well, good luck with it, got to go to a meeting now **** ENDING LOGGING AT Tue Feb 08 00:30:09 2005 **** BEGIN LOGGING AT Tue Feb 08 00:33:04 2005 Feb 08 00:35:29 this is ridiculous... they're refusing to process a rma without receipt **** ENDING LOGGING AT Tue Feb 08 00:36:12 2005 **** BEGIN LOGGING AT Tue Feb 08 00:39:05 2005 **** ENDING LOGGING AT Tue Feb 08 00:39:55 2005 **** BEGIN LOGGING AT Tue Feb 08 00:46:05 2005 Feb 08 01:30:49 g'morning Feb 08 01:32:03 hy Feb 08 01:32:39 got a quick question about the device Feb 08 01:32:59 I got the redboot prompt up, and then power cycled the device Feb 08 01:33:08 now it Feb 08 01:33:24 it's just showing up with an amber status led Feb 08 01:34:20 er... okay, never mind, it's blinking away cheerfully now Feb 08 01:35:00 ~nslu2-general Feb 08 01:35:01 #nslu2-general is the IRC channel specifically set up for end-user questions about the NSLU2 device. Questions that are not related to the nslu2-linux development projects should not be asked in #nslu2-linux, but should be asked in #nslu2-general instead. Feb 08 01:35:08 :) Feb 08 01:35:16 danke. :) Feb 08 01:35:23 but it is good to hear that you're ok now Feb 08 01:46:19 Bye bye 300MB of old test Unslung and pre-unslung firmware images. Feb 08 01:46:29 You served me well! Feb 08 01:46:46 dyoung, that your local repo? Feb 08 01:47:24 No those were hacked, split, built, binary sed-ed, and various other firmware images dating back to version pre-1.0 . Feb 08 01:47:35 before the concept of "local repo" Feb 08 01:47:36 heh Feb 08 01:47:55 nvm... it's almost 4AM and my brain is working on 20 some odd hours of "awake" Feb 08 01:48:33 thats was the beginnings of the project. Feb 08 01:48:59 pre firmware/Makefile even ... Feb 08 01:49:04 Rwhitby and I hacked the thing before he even had the device in hand. Feb 08 01:49:28 dyoung was my test dummy :-) Feb 08 01:49:30 you burned it to a cd for history's sake Feb 08 01:49:38 uh... ? Feb 08 01:49:41 NO I just put in in my rubbish can. Feb 08 01:50:12 I can easily reproduce all of it if someone needs a history lesson. But now its just taking up valuable space. Feb 08 01:50:38 but cd's are cheap. amybe I should do that. Feb 08 01:50:40 okay done. Feb 08 01:51:45 more important would be the irc logs .... Feb 08 01:51:54 Yeah. Feb 08 01:52:23 Those are in duplicate; and I was eventually gonna sync it up to nslu2-log. But I didnt get that far yet. Feb 08 02:14:32 jacques: did you get the new libjpeg to build shared cross? Feb 08 02:16:19 ah - hang on, it's just the dependency which is wrong. if only everyone typed "make" just one more time when everything is built to make sure that things do not rebuild every time you type make ... Feb 08 02:18:52 03rwhitby * 10unslung/make/libjpeg.mk: Fixed the .buit dependencies. Feb 08 02:26:32 Hmm - I wonder if installing automake1.9 on builds (to get pkgconfig to cross-compile) will break anything else .... Feb 08 02:26:40 * rwhitby-away starts a make distclean all on builds .... Feb 08 02:40:08 rwhitby-away, yes it built cross and native for me Feb 08 02:43:37 yeah, but did you type make again, and did it rebuild :-) Feb 08 02:44:14 (everyone forgets that test for bad Makefile dependencies) Feb 08 02:47:17 hi all, I have an external USB2 250GB maxtor HDD, and I want to put it on a NSLU2, but I want multiple shares that are restricted to different users, will the default f/w do this or do I need to delve into changing the f/w please? Feb 08 02:47:35 ~nslu2-general Feb 08 02:47:36 #nslu2-general is the IRC channel specifically set up for end-user questions about the NSLU2 device. Questions that are not related to the nslu2-linux development projects should not be asked in #nslu2-linux, but should be asked in #nslu2-general instead. Feb 08 02:47:48 :) Feb 08 02:47:54 ty Feb 08 02:50:01 ~botsnack Feb 08 02:50:01 thanks, dyoung Feb 08 02:50:35 DCC CHAT chat 1145815817 36691 Feb 08 03:19:56 good night... er morning.. DARN after 5AM and I'm not home... Feb 08 03:20:01 later all Feb 08 03:20:08 nite Feb 08 03:20:37 night Feb 08 05:29:13 hallo in die runde, bin neu hab seit gestern ein nslu2 Feb 08 05:29:21 sure Feb 08 05:31:45 kennt sich von euch jemand mit webcam und nslu aus? Feb 08 05:31:52 heu Feb 08 05:32:04 can u try in english please? Feb 08 05:32:12 shure! Feb 08 05:32:41 oh Feb 08 05:32:49 i've already unsluged my nslu. just have a problem with my old philips vesta webcam.. Feb 08 05:32:50 u want to plug a webcam on the nslu2 Feb 08 05:33:15 yes, seems to be compatible but does not get recognised - strange... Feb 08 05:33:52 maybe to need some kernel modules no? Feb 08 05:34:13 i've them both installed... Feb 08 05:34:43 hum Feb 08 05:34:44 weird Feb 08 05:35:35 have u try to ask on #nslu2-general ? Feb 08 05:35:40 i don't have a harddisk, i use an usb-stick on port1 Feb 08 05:36:06 um, no, this ist my fist visit in one of the nslu-chats Feb 08 05:36:14 ok Feb 08 05:38:20 it is a Vesta Pro (PCVC680K) with SAA8115HL and seems to be supported. is there a difference which port i use...? Feb 08 06:02:31 I am using xchat-text. Feb 08 06:03:55 using xchat command line to just get on irc channel behind firewall Feb 08 06:05:01 execuse me for doing the test here. Feb 08 06:11:32 xchat has a text client? cool I didn't know that Feb 08 06:12:34 thanks, jacques for your reponse. now the test is complete. I can see the dialog. Feb 08 06:13:10 :-) Feb 08 06:14:24 I I have to telnet back to my home from office and run xchat-text. I am happy now. I can on irc from home and office(via home) ;) Feb 08 06:16:06 i haven't been able to contribute codewise looks like I will contribute moneywise. Feb 08 06:16:45 I plan to donate 3 nslu2 for native compilation farm that rod posted recently. Feb 08 06:19:16 I was able to use my two nslu2 with distcc ipkg to have them work together to compile c codes. but the distcc packages need to be fixed and distccmon-text doesn't work. Feb 08 08:08:36 jacques: Do you know if XIP (eXcute-In-Place) would be viable on the slug? kernel 2.6.10 apparently support it now. Feb 08 09:18:06 bye Feb 08 09:22:33 msg nickserv register blafasel Feb 08 12:01:36 rwhitby-away: What about man-pages for dropbear and gawk? Can I add them? In general: Should we go for uncompressed man pages or gzip'ed ones? Feb 08 12:15:29 bob_tm: just out of curiosity: I thought the aim of nslu2-linux is to provide an OS for an embedded device, i.e. for a device, which is _very_ low on resources. so... why does suddenly everybody start to add full-blown man-pages to every package? Feb 08 12:16:11 Siddy-away: Which scarce resources do these man-pages drain? Feb 08 12:16:24 my sweet little 8MB flash-drive for instance. Feb 08 12:16:52 Ah, make a cron-job that wipes /opt/man every 8 hours. Feb 08 12:17:20 lol, why not just split the packages into xzy and xyz-doc ? Feb 08 12:17:33 It is an idea. Feb 08 12:18:22 siddy-away: good point Feb 08 12:18:41 I don't now if it is worth it in the sense that it will cause more work than it saves. Feb 08 12:20:39 agreed, it would be quite a lot of work... but adding the man-pages to existing packages would be equally laborious Feb 08 12:23:14 Yeah - but it is less work than man page fans each downloading the source and extracting the man pages themselves. Feb 08 12:23:32 We could have a ipkg setting for this? Feb 08 12:24:21 Like a "man-pages yes" or "man-pages no" in /etc/ipkg.conf Feb 08 12:24:47 and let ipkg filter them out for those who run on mem sticks or just find man pages useless. Feb 08 12:26:01 hmz... before loading ohci_hcd, do i need to load ixp400 for it to work? Feb 08 12:26:43 bob_tm: hm... i don't think ipkg can do this. Feb 08 12:27:10 It can't at present. Feb 08 12:27:53 You think it would be hard to add? Feb 08 12:29:56 I don't know, never looked at the ipkg-source. But.. how about a slight modification of the build-system, like letting "make xyz-ipk" create a package xyz and xyz-doc (from some named man-pages) Feb 08 12:30:08 ? Feb 08 12:38:59 Yes - I'm up for any simple solution that makes most people happy. Feb 08 12:50:08 we don't make mods to ipkg unless they are made upstream Feb 08 12:53:07 03jp30 * 10unslung/make/gtk+.mk: correct gtk+ package description, dependencies Feb 08 12:53:11 siddy-away: I would recommend adding a 256MB USB flash key giving you a sweet package installation area Feb 08 12:55:52 rwhitby-away: I know, flash is cheap, but that's not the point. I'd like to fill the flash with more sensible things than man-pages... I see no need for them on the slug. Feb 08 12:56:59 siddy-away: others do. It's more work to take them out, or split them into a separate ipk, than it is to just leave them in. Feb 08 12:57:06 perhaps you can symlink /opt/man to /dev/null ? Feb 08 12:57:19 03jp30 * 10unslung/make/ (ice.mk sm.mk xauth.mk xmu.mk xt.mk): initial checkin of xauth, xmu, xt, sm, and ice Feb 08 12:57:31 03jp30 * 10unslung/sources/xauth/autofoo.patch: initial checkin of xauth, xmu, xt, sm, and ice Feb 08 12:57:34 CompBrain: ping Feb 08 13:08:47 rwhitby: before loading ohci_hcd, do i need to load ixp400 for it to work? Feb 08 13:09:13 No. Feb 08 13:09:23 hmz weird Feb 08 13:09:43 rwhitby: must have been my mistake that I thought the NSLU2 was a low-end embedded device. Looking at the checkins of the last days, the nslu2-project seems to be turning that little piece of hardware into a full-featured desktop-machine, with XServer and -- of course -- man-pages... I just have to find out, where to connect my flatscreen to this thing... ;) Feb 08 13:10:05 nslu2, the minimac killer Feb 08 13:10:10 lol Feb 08 13:10:44 siddy-away: the packages go where the contributors want to take them ... Feb 08 13:11:56 if you wanted lean and mean, then you'd be using OpenSlug instead of Unslung anyway (so that the Samba NTFS password shared library wasn't taking 1MB of your sweet 8MB of internal flash. Feb 08 13:12:47 rwhitby: your advice is to late, I'm the glorious #13 in the openslug-list! ;-p Feb 08 13:13:09 and you may be able to use http://justiceforall.free.fr/pvrusb2.html to connect the flatscreen .... Feb 08 13:13:57 (although I'm not sure if it's record-only or not) Feb 08 13:14:41 someone should really try using a minimal read only squashfs flash for openslug and union mount the hard disk on top of it, or something Feb 08 13:14:54 * kergoth somehow expects this someone will end up being him Feb 08 13:15:35 kergoth: is the unionfs OE package still stable as a rock? Feb 08 13:15:45 seems to be working great Feb 08 13:15:50 cool Feb 08 13:15:57 the scripts for a union root need a bit of tweaking, but they do get the job done Feb 08 13:16:00 I dunno, I blew away my manpages. Feb 08 13:16:32 for the "no moving parts" production slug. Feb 08 13:16:39 the annoying part of a union / is the odd mtab and /proc/mounts Feb 08 13:16:42 apps can get confused Feb 08 13:16:58 Maybe that'll be Unslung 5.x .. get unsling to do a union mount instead of symlinking .... Feb 08 13:17:10 yeah, thatd be smooth Feb 08 13:19:11 siddy-away: anyone is free to split any package into bits, and -doc would be the convention to use. I'm not going to do the work, and I'm not going to force package contributors to do the work (although I'd be happy to accept a patch to template.mk which shows contributors how to do it easily). Feb 08 13:19:41 our efforts would be better focused on getting OE packages to work on Unslung .... Feb 08 13:20:06 (cause OE has a much finer-grained system of ipks) Feb 08 13:20:53 rwhitby: what's the current problem with oe-packages on unslung? the non-opt-paths? Feb 08 13:21:47 oe lets you override paths, but not all .bb's obey it. fixing that wont take long Feb 08 13:21:47 http://www.nslu2-linux.org/wiki/Unslung/OEPackagesForUnslung Feb 08 13:23:03 lo Feb 08 13:23:18 Theres the whole stuff with ldconfig too. Feb 08 13:23:36 Oh thats #4. Feb 08 13:26:02 rwhitby: you could build the ldconfig binary static. its static by default.. its our patch that enables a dynamic one. thatd ensure that it functions Feb 08 13:26:33 i can write you a bbclass to manage path manipulation on a per .bb basis Feb 08 13:26:36 so thats minor Feb 08 13:26:54 i'd say the big ones are startup script handling, and making sure all the packages obey the prefix/etc vars correctly Feb 08 13:27:16 bad afternoon all! Feb 08 13:27:45 ~today Feb 08 13:27:46 I don't know what today is, but it probably sucks. Feb 08 13:28:28 my feelings exactly! Feb 08 13:28:33 ~botsnack Feb 08 13:28:34 ByronT: :) Feb 08 13:35:38 03ptweety * 10unslung/ (Makefile make/lynx.mk make/postfix.mk sources/lynx/control): Lynx ready for testing. Feb 08 13:38:40 03ptweety * 10unslung/sources/postfix/ (postinst rc.postfix): Postfix postinstall and start script added. Feb 08 13:41:48 rwhitby: Does the wiki contain the rational for changing 4.x unslug to load root fs externally and the pros and cons of each posiible way. Seems it as changed from 2->3 and it will be changed back in 4 so I was trying to get some understanding. Feb 08 13:41:56 blegh Feb 08 13:42:34 i'm trying to mount usbcore, {o,e}hci_hcd, usb_storage, scsi_mod and sd_mod from within linuxrc, but only usbcore and the two scsi modules get loaded Feb 08 13:43:08 "Rule: Temporary files created by applications should in general be placed in /opt/tmp. If files are very small, /tmp (which is a ramdisk) can be used." Feb 08 13:43:17 so /opt/tmp is more like /var/tmp than /tmp, correct? Feb 08 13:43:56 Lof folksI've built myself an unslung image by following the HOWTOs on the nslu2-linux site, is there anyway I can test this without actually flashing the slug? Feb 08 13:44:26 bleh, mind my typing. Feb 08 13:46:31 Am I right in thinking the image also contains redboot? Is it possible (and safer) to flash just the kernel/jffs2 parts of the firmware so I don't need to worry about getting out the soldering iron to fix any badness? Feb 08 13:46:50 Thanks in advance... Feb 08 13:46:55 It's part of the image but it's not flashed Feb 08 13:47:05 Ah cool Feb 08 13:47:34 Well upslug worked no problem, so I guess I have nothing to worry about then :) Feb 08 13:47:46 Also - you can test it via TFTP via redboot Feb 08 13:47:58 (I believe) Feb 08 13:48:31 Right, I saw the HOWTO about testing a kernel/ramdisk, I wasn't sure how that related to unslung3s jffs2 Feb 08 13:48:49 I'll go look into doing that. Thanks! Feb 08 13:48:53 that doesn't apply to jffs2 Feb 08 13:49:12 ah ok Feb 08 13:49:23 Can't you test the kernel? Feb 08 13:49:26 You can only test the kernel and ramdisk in RAM Feb 08 13:50:41 For testing Unslung3, is the best we can do flash, then upslug if it fails? Feb 08 13:51:37 Pretty much Feb 08 13:51:55 ok, I can live with that. Thanks! Feb 08 13:55:52 garpinc: 4.x is a superset of 2.x and 3.x functionality Feb 08 13:57:11 hi all. is there a way (easy?) to throttle bandwidth usage of nslu2? Feb 08 13:57:40 kergoth: what's your recommendation regarding /tmp, /var/tmp (which is still a ramdisk on Unslung at the moment) and /opt/tmp ? Feb 08 13:58:28 i'd recommend, in general, following the FHS regarding path _behavior_, and just manipulating structure inside opt for whats what Feb 08 13:58:50 i.e. /tmp contents not guaranteed to survive reboot, /var/tmp contents expected to survive reboot, etc Feb 08 13:59:02 right now /var/tmp is the major violation of the FHS by oe built distributions Feb 08 14:01:37 I think we'll end up (in Unslung 4.x) symlinking /var over to the disk instead of having it a ramdisk Feb 08 14:01:50 we need to rethink how to handle /var, in general Feb 08 14:01:55 some areas in there could be a ramdisk Feb 08 14:01:57 others not Feb 08 14:02:31 agreed. safest for Unslung is just to have all of /var be real disk (so people don't fill the memory by writing to /var) Feb 08 14:02:56 the reason i recommended following fhs guidelines is, those guidelines are generally done based on application expectations Feb 08 14:03:02 and -that- is what we need to be concerned with Feb 08 14:03:16 nod Feb 08 14:03:31 anyway, hi ho, hi ho, ..... Feb 08 14:03:47 have fun Feb 08 14:04:14 almost there... Feb 08 14:04:15 (as dilbert advises): I will, at the expense of my co-workers .... Feb 08 14:06:48 yes! i have hdd access from the linuxrc in switchbox Feb 08 14:06:58 now i can use the hdd as root filesystem Feb 08 14:45:58 I really really wish I could get bk working on Windows XP Feb 08 15:29:22 tjyang: what's this I read about donating three slugs? :-) Feb 08 15:29:31 Slugfarm! Feb 08 15:30:02 (or as we say out here...ranching) Feb 08 15:30:32 can you guys talk in the other channel for a sec ? Feb 08 15:31:56 you mean the nslu2-general? Feb 08 15:31:59 other channel? Feb 08 15:32:07 ~nslu2-general Feb 08 15:32:08 #nslu2-general is the IRC channel specifically set up for end-user questions about the NSLU2 device. Questions that are not related to the nslu2-linux development projects should not be asked in #nslu2-linux, but should be asked in #nslu2-general instead. Feb 08 15:40:47 Hello all, I am loading the developer packages to the slug and the coreutils won't install because of busybox. I tried to remove busybox with ipkg, but it keeps telling me "No packages removed." what is the best way to remove busybox from my slug? Feb 08 15:41:05 ~nslu2-general Feb 08 15:41:06 #nslu2-general is the IRC channel specifically set up for end-user questions about the NSLU2 device. Questions that are not related to the nslu2-linux development projects should not be asked in #nslu2-linux, but should be asked in #nslu2-general instead. Feb 08 15:42:03 oh, sorry about that Feb 08 15:42:11 np Feb 08 15:43:23 he's having a tough time making it into channel over there Feb 08 15:44:12 seems like it Feb 08 15:44:20 next revision README... point to #nslu2-general Feb 08 15:44:42 actually that'll fit into NOTES Feb 08 15:44:49 good. Feb 08 15:45:21 this week sometime.. faster if I could get bk to work with windows... Feb 08 15:47:08 I can pull and clone in windows, but I can't get push to work Feb 08 15:47:29 same here... it has to do with authentication according to bitmover Feb 08 15:47:42 i c Feb 08 15:47:44 and they really don't want to do much - I was told that they are in between releases Feb 08 16:00:26 tjyang: just wanted to clarify - the donations for the single nslu2 build machine is not for public access, it's for the single official package build machine which uploads to the feed. Feb 08 16:01:22 so we don't really need more than one (assuming that native packages don't come in faster than we can compile them) - especially since we're only compiling native-only packages on it in the future (which are few, cause most packages will cross-compile) Feb 08 16:02:41 now, if someone else wanted to donate a slug for a more open developer access, then we could look into that ... Feb 08 16:15:24 ka6sox: what is the vision for the slugfarm.org domain? it's not for the official build hosts, right (as they need to be tightly controlled) - they will stay in the nslu2-linux..org domain as a restricted resource. Feb 08 16:17:34 that idea is a distributed CC environment to harvest the build abilities of distributed compilations. Feb 08 16:19:26 there are a lot of slugs that are idling most of the time. Feb 08 16:19:57 and also to look at a model for distCC compilations for embedded systems. Feb 08 16:20:01 (native) Feb 08 16:20:22 my slugs are idling right now Feb 08 16:20:42 distCC isn't good for WAN model... Feb 08 16:20:54 something else would have to be used... Feb 08 16:22:12 distCC sends the files over the "lan" that are to be compiled Feb 08 16:22:37 http://boinc.berkeley.edu/ Feb 08 16:22:38 the slugs should go get the source themselves under the direction of the "master" slug. Feb 08 16:22:46 bwahh ha ha haaa!!! Feb 08 16:23:28 My slugs are idling now too. Feb 08 16:23:41 ones a doorstop. Heh. Feb 08 16:25:49 jeez, someone should teach that kid the rules of logos Feb 08 16:25:55 http://boinc.berkeley.edu/hatfield.png Feb 08 16:26:15 I like that logo. Its... Different in a busy sort of way. Feb 08 16:26:30 yes. Feb 08 16:27:13 the problem with low powered and embedded systems is that native builds are s l o w Feb 08 16:27:27 ka6sox: so slugfarm.org is made up of individuals donating some of their own slug's time Feb 08 16:27:41 the slugs remain distributed around the world Feb 08 16:27:46 correct.. Feb 08 16:27:50 Actually now that I think about it a bit, its kind of a nice logo. It embodies all of the projects currently using "Boinc Technology". Feb 08 16:28:11 there might be some benefit to exploring distCC in a "local" environment too. Feb 08 16:28:26 so 3 or more slugs connect via NFSRoot. Feb 08 16:28:31 locally. Feb 08 16:28:32 did i just say that out loud? oops. Feb 08 16:28:41 you did Feb 08 16:29:24 I think I discovered my problem with cvs and sf... Feb 08 16:31:09 baby, wanna come back to my place and see my BOINC implementation? Feb 08 16:31:35 "If you application replies on 100% correctness, you shouldnt use Boinc" Feb 08 16:31:56 ugh - I missed that :-( Feb 08 16:32:04 well, that's because ppl can game it Feb 08 16:32:08 s/replies/relies/ Feb 08 16:32:12 rwhitby-web: there is a benefit of using distCC for native builds for the project. so if additional slugs were made available they could be part of the official www.nslu2-linux.org package build "cluster" Feb 08 16:32:36 "Although redundant computing can be used to minimize the problem...." Feb 08 16:33:48 right, the problem (I think) is incorrect results returned either due to problems with the individual nodes or ppl intentionally sending wrong results Feb 08 16:35:05 yeah Feb 08 16:39:02 ka6sox: I would rather spend the donations on a faster cross-compilation machine (now that we have bought the native compilation slug) rather than multiple slugs. We still intend for cross-compilation to be the default mode of compiling packages for the feeds, and only resort to native compilation when cross cannot be made to work in a practical timeframe. Feb 08 16:39:37 I agree that we need a faster Cross compilation machine. Feb 08 16:40:48 this cobbled together machine is too slow. Feb 08 17:02:25 cobbled together from misc bits. Feb 08 17:02:34 "what did you expect for free" Feb 08 17:03:37 I expect the best software in the world for free Feb 08 17:03:57 :) Feb 08 17:04:09 oh yeah Feb 08 17:04:19 "what did you expect for minimal effort?" Feb 08 17:04:35 (I'm assuming the effort was minimal at least) Feb 08 17:06:28 it took a couple of hours. Feb 08 17:06:45 heh - faster to build the machine than to build the packages :-) Feb 08 17:06:51 "Hardware should be FREE!" - Steve Ballmer. Feb 08 17:07:11 at least I think it was him. Feb 08 17:07:13 Well, we could synthesize a ppc in a V4.... Feb 08 17:07:22 slap some ram on it. Feb 08 17:07:40 it would be "free" in a non-monetary sense. Feb 08 17:08:38 You don't synthesize a PPC in a V4 Feb 08 17:08:48 It's a hardcore and actually part of it anyway Feb 08 17:08:54 true. Feb 08 17:19:14 didnt I read something about someone having a ppc405 or something in a v4? Feb 08 17:19:26 oh, its REALLY part of it? Feb 08 17:19:39 weird. Feb 08 17:19:45 Yeah Feb 08 17:19:57 The larger parts contain multiple PPC cores Feb 08 17:20:20 neat. Feb 08 17:28:46 rwhitby-web is it possible to debug a core file with cross compile tool chain. Feb 08 17:29:00 Is there a gdb that can read the core for nslu2 Feb 08 17:30:55 Does anyone know? Feb 08 17:31:52 yes Feb 08 17:32:40 So yes I can debug with cross compile toolchain? Feb 08 17:33:19 Well I don't know about the cross compile toolchain... Feb 08 17:33:45 but gdb seems to compile cleanly with the native compiler with a few minor Feb 08 17:33:47 changes Feb 08 17:34:50 Do you need other stuff or just the executable? Feb 08 17:35:41 I haven't tried much with it yet. I only got it to compile today. Feb 08 17:36:17 Can you send it to me? Feb 08 17:37:50 Well with the disclaimer that it might not work properly or at all... Feb 08 17:38:58 Sure... Feb 08 17:39:18 nslu2 at leclub.ws Feb 08 17:41:19 The compressed binary is about 1Mb is that OK? Feb 08 17:41:30 I need to head home.. a few knowledgable souls in #nslu2-general for newbie questions would help that channel start to take off Feb 08 17:42:13 Yes... Feb 08 17:44:55 Binary just sent Feb 08 17:49:39 shenson. Got it... So far I see 2 problems: error while loading shared libraries: libncurses.so.5 Feb 08 17:49:49 had to set LD_LIBRARY_PATH Feb 08 17:50:07 And second: dlopen failed on 'libthread_db.so.1' Feb 08 17:50:29 Yes I get the second one. Don't know how to fix it. It didn't matter with the small core Feb 08 17:50:50 dump I checked with it though. Feb 08 17:50:52 does anyone know if there is a ipkg install for tcsh shell? Feb 08 17:53:38 ask rt Feb 08 17:58:50 BTW if anyone can grant me CVS write access I'll have a go at getting Feb 08 17:59:40 gdb into an ipkg. Source forge id is shenson and I can be mailed steve at openssl dot org Feb 08 18:05:33 * rt blinks awake. Feb 08 18:06:00 * rt got tcsh to compile, but never debugged his problem with sourceforge, so alas, it has not benefitted the world at large. Feb 08 18:06:44 it wasn't hard, once you have the non-busybox sed compiled. Feb 08 18:15:47 rt: what was the sf.net problem? Feb 08 18:33:27 rwhitby-web is there something special I have to do with toolchain to retain debug info for a shared library? Feb 08 18:38:23 Read the man pages for GCC Feb 08 18:38:54 Well I know about -g option Feb 08 18:39:00 isn't that it? Feb 08 18:46:53 I tried both -g and -ggdb and the output complains about missing files rather than showing line numbers. Feb 08 18:48:10 gdb that is.. Feb 08 18:51:02 Do we have nslu2 pkg that includes 'libthread_db.so.1' Feb 08 19:16:52 heh, it cant show line numbers unless the source is available, in the location thats hardcoded in the binary Feb 08 19:16:59 if it is, then it should work Feb 08 19:19:10 So if it says: Feb 08 19:19:11 (gdb) l 290 Feb 08 19:19:11 285 in as_search_man.c Feb 08 19:19:33 then I have to copy as_search_man.c to my slug in the directory gdb is run? Feb 08 19:53:05 back late Feb 08 19:53:07 r Feb 08 19:53:15 happy lunar new year everyone. Feb 08 19:54:29 you lunatic you Feb 08 19:54:48 tjyang: ping Feb 08 19:54:55 I'm a lunatic alright. ;-) Feb 08 19:54:58 rod, I am back. Feb 08 19:55:01 Lunartick. Feb 08 19:55:47 looks like there is discussion about distcc and slugs ;) Feb 08 19:57:56 hi all, is there a utility to allow me to hook up an ext3 slug drive to a windows xp pc, but be able to wrute with cirrect permissions? if not, can I use one of the normal utilities that writes as root, then somehow change the permissions once back on the slug? Feb 08 19:58:14 FYI: it looks like gdb, shenson compiled works famously. I was able to debug a core file on the slug... Feb 08 20:00:50 my words are out and logged, I plan to donate 3 nslu2 to the community to make me feel better that I am giving too. Feb 08 20:05:31 rod, are you here ? Feb 08 20:07:10 tjyang, he's at work now I think Feb 08 20:07:24 probably busy :-\ Feb 08 20:07:47 jac, that is ok. let me when I should pay ;) Feb 08 20:08:14 let me know when I should transfer the money. Feb 08 20:08:44 tjyang, you made a very generous offer :-) I'll let rwhitby-web give you the details (he's in charge of the donations) Feb 08 20:10:42 jacques, sure. I will try to catch him too. Feb 08 20:11:35 tjyang, do you have PayPal? Feb 08 20:11:39 nslu2 open my eye to embedded system programming. Feb 08 20:12:13 tjyang, yes it's an excellent learning platform Feb 08 20:12:17 yes, I do have paypal and I will wire the money via paypal. Feb 08 20:12:54 this message explains how to do it. Feb 08 20:12:55 http://groups.yahoo.com/group/nslu2-linux/message/4154 Feb 08 20:13:15 I just wish we can have nslu2 farm to utilize distcc. Feb 08 20:13:40 that is a wish not a requirement. Feb 08 20:13:47 for general use or for core use? Feb 08 20:14:08 yeah once I get access to more than one slug running same dist I wanna play with distcc Feb 08 20:14:25 I don't care. Feb 08 20:14:33 ka6, I don't care. Feb 08 20:14:50 thank you for your generous offer. Feb 08 20:15:02 I have two nslu2 now and I can see two of them compile code together with distcc ;) Feb 08 20:15:40 it might be a good way to do development of embedded systems natively. Feb 08 20:16:26 it is wierd to me that nslu2 developer community prefer cross-compilation environment over native one. Feb 08 20:16:52 why? crosscompilation is faster and works as well. saves developer time and energy Feb 08 20:16:52 tjyang: are you planning to donate the cash equivalent of three nslu2 to the current donation drive (which is intended to buy a faster cross-compilation host, not another NSLU2), or are you making a proposal for a different type of donation of hardware to slugfarm.org for more general developer use? Feb 08 20:17:06 at work, cross-complation is only to generate the native gcc. Feb 08 20:17:30 at my work, everything is crosscompiled. same for all my embedded linux open source projects Feb 08 20:18:07 ok, let's not go down the cross-vs-native argument again. we need both for nslu2-linux Feb 08 20:18:19 hehe Feb 08 20:18:20 rod,you tell me what you prefer. Feb 08 20:18:40 no argument from here. Feb 08 20:18:43 * ka6sox looks at his cobbled together cross build machine. Feb 08 20:19:03 i am a sysadmin, not qualify for programming debate ;) Feb 08 20:19:20 tjyang: wait for a week or so, and see how we are going with respect to getting the cash for an updated cross-host. if we are rolling in cash at that time, then we might look at a separate developer-access slugfarm ... Feb 08 20:20:03 or if you trust us to make the right judgement, then just donate the cash via paypal now, and we'll sort out how to spend it :-) Feb 08 20:20:41 I need new hiking boots. Feb 08 20:20:42 rod, just let me know, just count me in for donation of 3 nslu2 worth of money. Feb 08 20:20:43 :-D Feb 08 20:21:59 but the priorites are (1) a native build host (which has now been bought from donations), (2) an updated cross-compilation host (specs dependent upon how much we achieve in the donation drive), (3) ensuring that domain and hosting costs are covered for at least two years into the future, (4) slugfarm for package developer use, (5) rod's new Treo 650 fund ... Feb 08 20:22:25 5 should be my hiking boots :-( Feb 08 20:22:25 lol! Feb 08 20:22:37 and a new mini Feb 08 20:22:43