**** BEGIN LOGGING AT Sun Nov 30 02:59:57 2008 Nov 30 03:01:10 ok now wired gives me a max of 5up Nov 30 03:01:37 avg 4.5 Nov 30 03:01:44 acceptable Nov 30 03:01:57 its ok but could be better Nov 30 03:02:08 behold the bullsh** of marketing hype! "54G" indeed. Nov 30 03:02:23 should I have a swap on each disk? Nov 30 03:02:44 yeah 54g.. Nov 30 03:02:49 lol Nov 30 03:03:16 that's what wireless G can handle ... not what your routers can actually support Nov 30 03:03:28 and of course, the farther you get from them the less speed you do get Nov 30 03:03:54 I have my swap on an old 256MB flash drive arthur92710 Nov 30 03:04:01 Im not more then 10ft away Nov 30 03:04:14 And put a 11b node in the mix, and it gets even worse Nov 30 03:05:29 down I get avg 7.5 Nov 30 03:05:41 llb? Nov 30 03:06:00 547mb down in 1min 27sec Nov 30 03:06:09 not bad Nov 30 03:07:33 is that like 11 in 802.11b? or something else? Nov 30 03:07:50 1l same Nov 30 03:08:55 like one1 and lama Nov 30 03:09:08 yeah, the presence of an 802.11b node will slow down the 54g nodes Nov 30 03:09:35 I run mixed Nov 30 03:09:41 b/g Nov 30 03:09:45 11 is the sub-part of the 802 standard Nov 30 03:09:54 the damn psp only gets b Nov 30 03:10:15 11g is another part Nov 30 03:10:24 I know Nov 30 03:11:16 Well, if you didn't need the 11b nodes, then you could set the 54g nodes to ignore the 11b traffic, and speed them up a bit. But either way, in the end, wired ethernet still beats wireless hands down. Nov 30 03:12:12 hands down of course Nov 30 03:12:17 or move them to different channels Nov 30 03:15:48 ok Nov 30 03:15:55 I did disconnect Nov 30 03:16:10 Can you put them on different channels? Nov 30 03:16:49 only if they aren't gonna talk to each other Nov 30 03:17:40 why would they? Do you mean as one AP? Nov 30 03:30:24 Sorry I DC'ed Nov 30 03:30:46 I changed it to g only but it does not help much Nov 30 03:31:22 I tried to ftp on to my xbox, and I get the same speed as the slug Nov 30 03:31:34 wireless* Nov 30 03:34:03 thanks alot for helping clearing this up. Nov 30 04:45:34 * joat is away: Away Nov 30 04:47:50 way away or just away away Nov 30 19:16:11 slug's not appearing after upgrade from etch to lenny, any suggestions? Nov 30 19:17:02 eth seems to be coming up, but i dont think it has an address Nov 30 19:26:14 ah, got it working, appears that my old pan to lan bridge stopped working... Nov 30 19:28:08 hmm, by the way, is there a way to mount swap by any form of unique ID? Nov 30 19:28:39 somehow my drive keeps jumping around and I have to manually identify and activate the swap at boot Nov 30 19:31:55 look at the wiki Nov 30 19:32:01 there is a mount by UUID article Nov 30 19:32:06 about mounting by uuid? Nov 30 19:32:11 yup Nov 30 19:32:17 yeah, problem is my swap partition doesnt turn up with a uuid Nov 30 19:32:28 everything else does and works, except the swap Nov 30 19:33:15 hmm, seems like its supposed to have a uuid, what gives? Nov 30 19:33:23 hmm Nov 30 19:33:30 let me reboot mine as it doesnt want to let me connect Nov 30 19:33:33 only ping response Nov 30 19:34:08 check /etc/blkid/blkid.tab (or something like that) to find the UUID for your various partitions. Nov 30 19:35:08 mwester-laptop: nope, no uuid entry for it, only /dev/sda1 Nov 30 19:35:19 which almost certainly changes on every reboot Nov 30 19:38:28 Right, with usb-attached storage there is no guarantee that the /dev/sd is constant. Nov 30 19:38:51 yeah, which makes the whole swap thing irritating Nov 30 19:39:04 since without swap, the slug more or less dies on bootup Nov 30 19:39:23 i have to login as soon as network and sshd comes up and do a swapon Nov 30 19:40:19 Well, you'll need to troubleshoot then -- I can't help you with debian. But swap space DOES have a UUID, you just need to find out how to use that. (it works fine with SlugOS AFAIK, so maybe you can check out how SlugOS does it?) Nov 30 19:41:11 hmm, i'm not sure why though... i see an old post about old swap versions not having a uuid, maybe that's why Nov 30 19:42:47 xieliwei: Can you set up a swap file so that it will at least boot? Nov 30 19:42:55 hmm Nov 30 19:43:00 mine doesnt actually have a uuid for the swap Nov 30 19:43:16 hmm, good idea, swap file Nov 30 19:43:31 Reedy: yeah, weird huh? Nov 30 19:43:42 Can you put a label on a swap partition? Nov 30 19:44:05 yes, but i dont think it appears too Nov 30 19:44:06 http://www.nslu2-linux.org/wiki/HowTo/MountDisksByLabel Nov 30 19:44:16 i remember adding a label when i created the swap Nov 30 19:45:06 interestingly, this link says swap partitions dont have uuids: http://www.unix.com/linux/47919-how-give-uuid-swap-partition.html Nov 30 19:46:06 hmmm.... Nov 30 19:46:19 * mwester-laptop begins to doubt the SlugOS mechanism Nov 30 19:46:33 (appears the upgrade broke pan and openvpn) Nov 30 19:47:07 i do see a few posts with fstab with UUIDs attached to swap Nov 30 19:47:13 could be a new feature Nov 30 19:51:50 okay, seems like i do have an "outdated" swap Nov 30 19:51:53 # mkswap --help Nov 30 19:51:58 Usage: mkswap [-c] [-v0|-v1] [-pPAGESZ] [-L label] [-U UUID] /dev/name [blocks] Nov 30 19:53:49 PAN as in bluetooth? Nov 30 19:56:21 yeah Nov 30 19:58:24 I haven't spent the time to get the latest bluetooth working, but it no longer uses pand -- and it seems that you can't do anything without dbus command (i.e. no config files; a GUI is implied). Nov 30 19:59:49 yeah, it requires a lot of lengthy dbus commands now Nov 30 20:00:15 problem isn't the configuration, it was the bridging of the pan interface with eth0 Nov 30 20:00:30 somehow i'm getting a file exists or something Nov 30 20:00:54 That's the problem I encountered. Nov 30 20:01:08 hmm, you too? Nov 30 20:01:30 latest 3.x bluez package (on SlugOS) Nov 30 20:02:06 You can't start pand until you remove the network.so file in /usr/something-or-other/...bluetooth/.../ Nov 30 20:02:07 i'm not sure what's mine, but i suspect the same version since i just did a dist-upgrade Nov 30 20:04:33 # find -iname bluetooth -print Nov 30 20:04:37 ./lib/bluetooth Nov 30 20:04:43 nothing there Nov 30 20:07:24 mwester: this page http://www.nslu2-linux.org/wiki/DSMG600/SerialConsole mentions temporary serial access but doesn't go into how this would be done and has '(fill-in-url-here)' as if someone forgot to fill in the URL... Nov 30 20:08:28 Yeah, that page was never written, due to general lack of interest in the DSM-G600 Nov 30 20:08:45 * scant has much interest Nov 30 20:08:57 The basics are pretty simple: Aquire a serial cable with proper connector, then Nov 30 20:09:10 mwester-laptop: if you will, i can fill in the details, if you advise me of them Nov 30 20:09:35 cut off the connector you DON'T need on that cable. Nov 30 20:09:57 Strip the outer jacket to expose wires, strip short bit of insulation from each wire, then Nov 30 20:10:15 use a meter to identify which color wire matches with the important pins on the serial connector. Nov 30 20:10:24 (TX, RX, and Ground) Nov 30 20:11:24 Then the part I never go to is to figure out what size wire will fit tightly into the holes for the serial header on the circuit board. Nov 30 20:11:39 k Nov 30 20:11:49 But basically, find solid (not stranded) wire that would fit in those holes in the circuit board. Nov 30 20:12:17 mwester: this page, I think is outdated, the part where it discusses flashing back to dlink firmware: http://www.nslu2-linux.org/wiki/DSMG600/Reflashing Nov 30 20:12:26 Cut three lengths. Connect the ground wire on the cable to one of the ground holes. Nov 30 20:13:01 The do the same with the PC's RX pin (recieve data INTO PC) and the TX hole on the circuit board. Nov 30 20:13:35 Start up a term emulator on the PC, with 115200 baud, and boot the DSM -- you should see the boot data if all is well. Nov 30 20:14:03 this depreciated page discusses in brief extracting the dlink firmware, but the new one doesn't: http://www.nslu2-linux.org/wiki/DSMG600/InstallingOpenSlug Nov 30 20:14:14 sorry to poke in, doesn't it need a level converter? Nov 30 20:14:24 Once you can see the data, hook up the remaining solid wire from the hole to the other pin, so that you can send data - start with a -C and you should be able to interact. Nov 30 20:14:37 xieliwei: Nope, the DSM-G600 is unique in that regard. Nov 30 20:15:01 wow, integrated level shifter, almost sounds like its inviting to be hacked Nov 30 20:15:53 scant: no need to extract the firmware, as it turns out. Nov 30 20:15:54 anyway, it seems like remaking my swap partition gave it a uuid, so i'm not happy Nov 30 20:16:19 xieliwei: Yep, and the serial header matches the pinout for the PC-XT serial adaptors. Nov 30 20:16:45 hmm, maybe to ease factory debug i guess Nov 30 20:16:47 not happy? or "now happy"? Nov 30 20:16:58 whoops, *now Nov 30 20:17:14 xieliwei: Yeah, the Rev A device was superceded rather quickly with a Rev B that cut memory and flash in half, etc. Nov 30 20:17:24 but i still have problems with openvpn and bluetooth Nov 30 20:17:54 mwester-laptop: argh, that sounds so much like linksys's wrt54 series Nov 30 20:18:13 the wiki pages also doesn't discuss the menu that appears when holding down the reset button and powering up, it's mentioned, but not discussed here: http://forum.openwrt.org/viewtopic.php?pid=21798 Nov 30 20:20:33 AFAIK nobody has ever looked at what that menu/mode does. Nov 30 20:22:22 mwester-laptop: my biggest questions revolve around the when things go wrong, it says burn the kernel image to reflash back to dlink firmware, but no procedure is given on how to do that, same thing with reflashing within slugos, it has a 'insert procedure' example Nov 30 20:23:05 and, if SlugOS is up and running from a ramdisk can I reflash back to DLink firmware? Nov 30 20:24:04 yes, but you'll need the reflash utility if its not present on the SlugOS you flash. Nov 30 20:24:39 and if SlugOS won't boot and the person never got serial access working that page suggests to dump the device, but isn't that kind of a quick judgement or something, shouldn't it be, get serial access working, if you can't or won't your device is a flower pot? Nov 30 20:25:59 It's intended to be tongue-in-cheek. Of course someone should attempt to recover the device. The idea is that if you are asking for detailed step-by-step instructions, then perhaps you shouldn't be doing it.;) Nov 30 20:27:57 mwester: i'm back, i may have not received the last things you typed Nov 30 20:32:25 running from ramdisk is: turnup ramdisk, and turnup ramdisk != turnup memstick, so does running from ramdisk involve a memstick at all? Nov 30 20:33:44 No Nov 30 20:34:22 you were saying something about SlugOS version may not have the flash command, does 4.8 beta have it? is it easily accessable when running from ramdisk? Nov 30 20:34:27 turnup ramdisk sets things up so that at next boot, the flash rootfs is copied to a ramdisk, and the system switches to the ramdisk instead. Nov 30 20:35:23 I don't remember if the latest SlugOS has the reflash command or not; you'll have to check. If not, it's in the OE repository in its entirety -- just copy it to the flash rootfs. Nov 30 20:36:48 mwester: and if I copy the reflash command into the flash rootfs and i'm running from ramdisk it will also be copied over upon every boot? Nov 30 20:43:33 D-Link firmware includes these files: ip-ramdisk, rootfs.gz, usr.cramfs, & version.msg, if I want to completely revert back to D-Link firmware and not save any settings, what do I do? Nov 30 20:44:15 oh, and roofs.gz has inside, rootfs Nov 30 20:46:51 the reflash command will probably take the d-link firmware image file as downloaded from d-link. Nov 30 20:47:13 I know I added that functionality at one point. Nov 30 20:48:00 like, the zip file? Nov 30 20:48:20 or the bin inside the zip file? Nov 30 20:48:26 the bin file Nov 30 20:48:38 opps, sorry my mistake,as downloaded, it's a bin Nov 30 20:49:14 check the help message printed by reflash, it should tell you about flashing full images (as opposed to just a kernel or rootfs at a time) Nov 30 20:52:50 mwester: I'm still trying to locate reflash, i don't think it's on my SlugOS 4.8beta, and my cross-dev environment is busted right now... you may be seeing me in here for help sooner than later... eeek! Nov 30 20:53:02 mwester: the ipkg-find page says reflash is in slugos-init Nov 30 20:55:54 It's a script, in plain script form -- you'll find it in packages/something-or-other/files/reflash; you don't even need to build anything, just copy it in. Nov 30 20:57:01 see above... i think it can be retreived with ipkg from the slugos-init package... Nov 30 21:02:42 If its not in your rootfs already, then the newest slugos-init package doesn't actually contain it any longer. Older versions won't have the dsmg600 support. So the one in the OpenEmbedded repo is the one you'll probably need. (an executive decision was made that deprecated that package at the last moment before release). Nov 30 22:21:54 anyone here use rsync a lot? Nov 30 22:22:26 * mwester-laptop loves rsync Nov 30 22:23:25 i would ask in the rsync channel, but i don't want my head to be bitten off, but how do you minimize data corruption being synced from the source to the target drive? Nov 30 22:24:23 i wish there was a rsync option to... not sync a blank source drive, that would solve perhaps 1/2 of the problem Nov 30 22:32:21 There's not much you can do with rsync to address that problem. Nov 30 22:32:42 You need a wrapper with some heuristics that can decide if you should rsync or not. Nov 30 22:33:30 By not deleting anything on the target that no longer exists on the source, you can make rsync a little more resiliant, but only at the risk of filling up the target disk way to quickly (because you'll have to manually delete data). Nov 30 23:01:09 mwester: do you do any of that kind of stuff? Nov 30 23:04:15 No. I gave up. Nov 30 23:05:39 I have a pair of drives that are rsync'd completely automatically, and I manually rsync the important directories to a third backup device (and ultimately to DVD-R) manually. Nov 30 23:05:49 mwester-laptop: hmmm, that's what i was thinking Nov 30 23:06:14 mwester: instead of error-checking, using a 3rd drive and sync that perhaps every month **** ENDING LOGGING AT Mon Dec 01 02:59:58 2008