**** BEGIN LOGGING AT Wed Aug 28 02:59:58 2013 Aug 28 03:30:56 build #243 of octeon is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/octeon/builds/243 Aug 28 06:02:00 build #300 of au1000 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/au1000/builds/300 Aug 28 07:39:10 nbd r37854 trunk/target/linux/generic/patches-3.10/220-gc_sections.patch * kernel: fix booting issues on imx23 and imx6 targets Aug 28 09:29:26 hi Aug 28 09:30:52 i tried to flash an unfi ap pro with openwrt and tried it with the patched and the unpatches image - described here : http://wiki.openwrt.org/toh/ubiquiti/unifi - under "installing a new firmware image" - but i allways end up with this: Aug 28 09:30:56 http://paste.debian.net/31258/ Aug 28 09:31:20 some months ago i was able with the patch described in the wiki to flash a device but now it doesnt work anymore Aug 28 09:32:03 any ideas how i could flash it? - i'd like to avoid using the serial console as i have to flash some devices Aug 28 09:33:59 <__trine> tripolar, cant you just tftp a new image Aug 28 09:35:31 as far as i remeber. this device doesnt have a tftpd running Aug 28 09:36:03 the wiki also doesnt write something about tftpd Aug 28 09:36:04 <__trine> all I have to go on is that my nanostation does have tftp Aug 28 09:36:22 <__trine> and I would suspect you unifi also would have it Aug 28 09:36:28 <__trine> your* Aug 28 09:36:48 <__trine> to get into tftp mode you start with the power off Aug 28 09:37:24 press the reset button power on again and wait until the light is blinking? Aug 28 09:37:42 <__trine> and then press and keep pressed the reset button for about 15 seconds then you may see the leds flashing a little differently Aug 28 09:37:58 <__trine> this should be tftp mode Aug 28 09:38:12 <__trine> YMMV Aug 28 09:39:17 <__trine> please let me know if it works Aug 28 09:39:26 <__trine> and update the wiki accordingly Aug 28 09:49:31 __trine: doesnt work Aug 28 09:49:46 normally tftp announces itself but here is no announcement Aug 28 09:49:52 tftpd Aug 28 09:50:14 i used wireshark Aug 28 09:54:27 <__trine> hmm its odd that ubnt have tftp in the uboot on the nano but not on your unifi Aug 28 09:56:35 i think they removed it Aug 28 09:56:42 i also read thsi somewhere Aug 28 09:58:36 <__trine> http://wiki.ubnt.com/Firmware_Recovery Aug 28 09:59:24 <__trine> this seems to be only airos Aug 28 09:59:49 i know the problem is now - the device doesnt accept the latest image i created Aug 28 10:00:04 i created an image some time ago and this one was accepted Aug 28 10:02:41 __trine: for ap pro i didnt find a image update by now - but the controller software allows the check a box which says something like "update devices automatically" Aug 28 10:02:51 <__trine> brainslayer says http://www.dd-wrt.com/phpBB2/viewtopic.php?p=530554 Aug 28 10:04:05 __trine: fhlasing with tftp through the bootloader works but i dont want to open all devices to do this Aug 28 10:05:08 <__trine> why would you need to open them for tftp Aug 28 10:05:17 <__trine> I am not suggesting serial Aug 28 10:06:07 <__trine> did you try to send the tftp to 192.168.1.20 Aug 28 10:06:23 <__trine> thats the default IP for ubnt Aug 28 10:08:21 <__trine> this post may mean it does not have TFTP http://www.dd-wrt.com/phpBB2/viewtopic.php?t=45474 Aug 28 10:10:28 <__trine> I have found a further post by brainslayer which now says the unifi does NOT have TFTP Aug 28 10:10:39 <__trine> http://www.dd-wrt.com/phpBB2/viewtopic.php?t=164199 Aug 28 10:10:50 i know i tried everything - so i was right Aug 28 10:11:04 <__trine> but you can flash them Aug 28 10:11:08 <__trine> with scp Aug 28 10:11:19 it doesnt accept the image Aug 28 10:11:35 <__trine> then it may be a bad image Aug 28 10:11:40 __trine: http://paste.debian.net/31258/ Aug 28 10:13:48 <__trine> I think this is worth a moan to ubnt on the forum that they dont have tftp in the bootloader Aug 28 10:14:59 i think they walk the way of tp-link all the devices i know from them lack tftp Aug 28 10:15:21 the only cool think about the ap pro is that it supports poe Aug 28 10:15:45 <__trine> so does the nanostation m2 Aug 28 10:16:45 i also have the nanostation m2/m5 here :) Aug 28 10:19:14 the unifi pro is also dual band Aug 28 10:19:25 i know Aug 28 10:19:32 one of the main reasons to buy it Aug 28 10:20:27 03:15 tripolar: the only cool think about the ap pro is that it supports poe Aug 28 10:20:40 two cool things ;-) Aug 28 10:20:43 <__trine> I have a nano m5 for more than a year and have yet to hear anyone else other than myself on the band Aug 28 10:21:28 walls++ Aug 28 10:22:38 <__trine> I have had it pointing out of my windows in various directions,, result = nothing Aug 28 10:25:47 <__trine> I had a dream lastnight I dreamt I had a nano m2 walking down the street hand in hand with netgear router dont judge the nano by the colour of its case judge by its contented openwrt firmware Aug 28 10:25:54 we use 5ghz for point-to-point links, but only where we have LOS Aug 28 10:27:22 <__trine> imagined speech by Robert J. 'Luther' Pera :0 Aug 28 10:27:50 <__trine> he ruined ubnt IMHO Aug 28 10:29:01 * russell-- had no idea he owned a basketball team Aug 28 10:29:33 <__trine> yes the fool bought it Aug 28 10:35:37 <__trine> I wonder if this could be done for a nano http://wiki.openwrt.org/toh/d-link/dir-600#usb.mod Aug 28 10:36:14 <__trine> usb port Aug 28 10:38:52 <__trine> a usb port on the nano would be excellent the nano uses an Atheros AR7241 CPU Aug 28 10:40:30 __trine: do you know where the header is defined? ar7240 should be ar934x Aug 28 10:41:12 __trine: found also in the makefile Aug 28 10:43:25 i mangged to flash it :D Aug 28 10:43:28 __trine: Aug 28 10:43:34 at least it accepted the image Aug 28 10:43:43 i will add the changes to the wiki Aug 28 10:44:52 <__trine> I have just examined the actual CPU on a nano and there is nothing attached to pins 73 and 74 Aug 28 10:52:46 cyrusff: how can I activate the norelease option for odhcp6c without it sending a release now ? I did kill -9 on the PID of the one running, norelease is already in /etc/config/network, but it restarts automatically without the -k option Aug 28 10:57:21 ah, maybe I'll add a firewall rule to block DHCPv6 traffic, then ifdown wan6 ; ifup wan6; then remove the rule again Aug 28 10:57:32 I really want to keep this lease now :) Aug 28 11:16:20 nbd: could it be possible that https://dev.openwrt.org/changeset/37483 causes issues (latency) with online gaming over a WDS bridge (both client and AP use ath9k, 5GHz link between 2 Mikrotik R52n - ar9285) ? Aug 28 11:19:19 when I moved to my appartment end June, it wasn't impossible for me to get a decent score, in the last few weeks it was so horribly frustrating that I started using an UTP between the 2 floors, which helped a lot Aug 28 11:34:02 stintel: there were some issues in that commit which i fixed in later commits Aug 28 11:34:35 stintel: did you test recent builds to see if the issue is still there? Aug 28 11:34:38 running r37825 now Aug 28 11:35:24 and it has the same issue? Aug 28 11:35:52 I'll give it another try tonight Aug 28 11:35:55 ok Aug 28 11:37:14 it's difficult to say really, sometimes it's just a bad host, but my average KD increased significantly since using UTP Aug 28 11:38:17 KD? Aug 28 11:38:21 kill-death Aug 28 11:38:28 ah Aug 28 11:38:30 ;) Aug 28 11:38:35 gamer slang ;) Aug 28 11:40:05 then there is this other very weird thing, happens for a long time, happens in different call of duty version, maybe even in other games (not sure), when I am the host (so other players in the same match connect all to my xbox), at the end of the first round, many of the connected players their xbox crash Aug 28 11:41:36 this didn't happen over the WDS bridge, started happening again when I switched to UTP Aug 28 11:41:36 that's really a WTF issue Aug 28 11:41:46 ah, I should test if this happens when I put my xbox in the WAN vlan Aug 28 11:42:13 so it bypasses openwrt Aug 28 11:44:33 and one more thing, that WDS link, 5GHz, with just a ceiling between, seems to be stuck on MCS rates <= 7, so only using a single spatial stream. card support 2x2 and both devices have 2 antennas. is ~-70dB signal strength too low? Aug 28 12:06:29 where on the gpl sources could gpio assigments be? I mean, is there a trivial location? ralink by the way.. Aug 28 12:13:02 drivers/char/ralink_gpio.c Aug 28 12:13:07 in the kernel folder Aug 28 12:13:35 yay Aug 28 12:14:49 firing up VM Aug 28 12:27:44 blogic thanks Mr! I'll take a closer look at this, extensive code heheh Aug 28 12:28:52 blogic do you have some pointers about available values for "ralink,portmap" and "ralink,led-polarity"? afaik, the latter accepts 0 or 1, right? Aug 28 12:30:37 stintel: 5ghz doesn't propagate as well through walls and ceilings as 2.4 ghz Aug 28 12:31:13 stintel: if you don't get MCS > 7, it means there's no spatial diversity that it can use Aug 28 12:31:14 ralink,portmap is the lan to wan port mapping Aug 28 12:31:28 stintel: so it can only transmit a single stream on the same frequency Aug 28 12:31:28 ralink,led-polarity if the leds are active high or low Aug 28 12:31:39 ralink,led-polarity can be ignored in most cases Aug 28 12:32:11 stintel: i might be able to tell you more if you show me a dump of the rate control stats Aug 28 12:34:02 nbd: ok. too bad I unplugged the WDS client last night, so can't show those right now Aug 28 12:34:40 k Aug 28 12:34:56 I also tried the WDS link on 2.4GHz but iirc that didn't result in higher MCS rates Aug 28 12:35:16 but there are quite a few 2.4GHz networks around Aug 28 12:35:37 appartments and wifi suck :) Aug 28 12:35:44 ;) Aug 28 12:36:45 blogic: have you received an experia box yet? Aug 28 12:39:16 maybe I'll ask the landlord if there are some free pipes between the floors :) Aug 28 12:41:07 blogic I can't, for the life of me, configure ethernet leds properly, I've changed ralink,portmap from 0x1f to 0x2f and vice-versa, in one case all ports remain lit without cable, except for port 1 which works as expected, on the other case, ports 4 & 2 remain lit without cable while port 3 blinks together with wireless led, and again, port 1 works correctly Aug 28 12:41:31 The_Lizard_: nope Aug 28 12:41:38 friday it should arrive Aug 28 12:41:52 Devastator: rt5350 ? Aug 28 12:41:56 this is driving me nuts, because I don't know where to get proper values and apply the famous "try & error" technique :) Aug 28 12:42:02 blogic yup Aug 28 12:42:15 ok, nice, I still have vacation, so I might be at home so I could give some support ;) Aug 28 12:42:38 got my glass fiber connection today ;) Aug 28 12:42:43 Devastator: known bug, wait till the pinctrl driver is in trunk Aug 28 12:43:06 oh, that's a relief Aug 28 12:45:10 blogic are you the writer of this driver? if so, ping me when you want a tester Aug 28 12:49:24 Devastator: ok Aug 28 12:49:32 which board are you using ß Aug 28 12:49:34 which board are you using ? Aug 28 12:49:46 D-Link DIR-610 A1 Aug 28 12:50:29 let me build an image for oyu in a sec Aug 28 12:50:38 do you have serial access to the unit ? Aug 28 12:50:43 yes sir Aug 28 12:50:51 if so i will build a ramdisk image Aug 28 12:50:59 you can use option 1 in the bootloader to run it Aug 28 12:51:08 no problem :) Aug 28 12:51:15 just one little thing Aug 28 12:51:30 can you add a flashchip patch I forgot to send? Aug 28 12:51:56 ok Aug 28 12:51:58 send it Aug 28 12:53:57 http://codepad.org/mbN0cfcT Aug 28 13:05:00 I still doesn't know how to patch the kernel with git, this patch I've made when I was using svn and quilt, blergh hehe Aug 28 13:06:25 get a git checkout of the kernel Aug 28 13:06:31 Devastator: the kernel uses tabs instead of spaces Aug 28 13:06:52 Devastator: git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git Aug 28 13:07:30 larsc: thanks! noted Aug 28 13:08:13 blogic thanks, will do on the next time, can you use the pastebin one just this time? Aug 28 13:08:38 pretty please :) Aug 28 13:44:56 __trine: do you also know a way to find out which pin is used for rx/tx/vcc? it seems a have a new version its v1.1 - i think before i had v1 and it looked like this: http://wiki.openwrt.org/_detail/toh/ubiquiti/unifi_serial.png?id=toh%3Aubiquiti%3Aunifi Aug 28 13:45:27 now there is no pin description beside the pins anymore Aug 28 13:46:32 nbd: any idea how i can find which pins are used for rx/tx/vcc? Aug 28 13:46:47 nbd: any idea how i can find out which pins are used for rx/tx/vcc? Aug 28 13:47:41 i'd recommend not using vcc Aug 28 13:47:50 if you have a usb<->ttl-serial adapter Aug 28 13:48:14 first make sure you get gnd right Aug 28 13:48:42 then you can figure out your board's tx first by trial and error without damaging your adapter Aug 28 13:48:57 then afterwards do the same with your board's rx Aug 28 13:51:31 okay Aug 28 13:52:20 thanks i will try it Aug 28 13:53:55 nbd: if i put rx and tx together shouldn't i be able to see the stuff i typed into the terminal? Aug 28 13:54:01 <__trine> I just came back to the KB Aug 28 13:54:39 tripolar: If you connect your USB<->TTL serial adapter's TX to its RX: yes. Aug 28 13:54:41 <__trine> nbd is right if you're using a usb type serial connector Aug 28 13:55:19 tripolar: but if you connect your TX to device RX, and device TX to your RX, you'll only see what you typed if the device actively echoes it back Aug 28 13:55:38 <__trine> the way I would do it is to use eyesight and a multimeter to first determine ground Aug 28 13:55:42 okay i think my ttl is dead :( Aug 28 13:55:58 tripolar: check if minicom (or whatever you're using) has flow control enabled Aug 28 13:56:00 __trine: ground is connected to the backplane Aug 28 13:56:01 if so, disable it Aug 28 13:56:07 <__trine> ground by looking usually has the most copper Aug 28 13:56:18 nbd: i use creen /dev/ttyUSB0 115200,cs8,-parenb,-cstopb,-hupcl Aug 28 13:56:24 screen /dev/ttyUSB0 115200,cs8,-parenb,-cstopb,-hupcl Aug 28 14:00:40 nbd: when i flash a device like the ap pro is uboot also inside and factory image or is it never touched? Aug 28 14:01:11 it seems they changed the hw in v1.1 or is ar8327 a ar71xx arch? Aug 28 14:01:39 __trine: do you know if ar8327 is a ar71xx arch? Aug 28 14:07:29 <__trine> tripolar, I dont know Aug 28 14:07:40 <__trine> cant you look it up at the FCC Aug 28 14:07:54 tripolar: ar8327 is a switch chip afaik Aug 28 14:08:35 http://www.qca.qualcomm.com/technology/technology.php?nav1=48&product=102 Aug 28 14:11:12 <__trine> that was the first google hit :) Aug 28 14:11:36 <__trine> https://dev.openwrt.org/ticket/13352 Aug 28 14:14:57 dammit, no space left to checkout kernel.. Aug 28 14:21:52 luka: yea or nay? http://patchwork.openwrt.org/patch/3966/ Aug 28 14:26:32 luka: it's in mainline now: https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/log/drivers/net/usb/hso.c Aug 28 14:32:38 well, I really hope the patch I pastebin is enough for now, I cannot make room to checkout kernel and patch it properly, unfortunately :( Aug 28 14:47:17 __trine thanks Aug 28 14:52:03 <__trine> yw Aug 28 15:15:08 __trine: so they really changed the device - v1 (i call it like this because there is no version printed on the board) has a ar7241 chipset and the v1.1 has a ar9350 chipset, is ubiquiti releasing the changes they did to openwrt somewhere? Aug 28 15:15:44 they also added more antennes in v1.1 Aug 28 15:48:05 build #332 of rb532 is complete: Failure [failed shell_12] Build details are at http://buildbot.openwrt.org:8010/builders/rb532/builds/332 Aug 28 15:54:53 q Aug 28 15:54:56 sorry Aug 28 15:55:04 build #332 of ppc44x is complete: Failure [failed shell_12] Build details are at http://buildbot.openwrt.org:8010/builders/ppc44x/builds/332 Aug 28 16:20:33 florian r37855 trunk/target/linux/mvebu/Makefile * mvebu: remove maintainer ship role Aug 28 16:53:06 build #345 of uml is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/uml/builds/345 Aug 28 16:57:18 build #364 of orion is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/364 Aug 28 17:00:23 build #314 of sibyte is complete: Failure [failed shell_12] Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/314 Aug 28 18:16:00 okay i upgraded the ram and flash of wrt54g v7 Aug 28 18:16:10 now which bootloader to use? Aug 28 18:16:30 its a 4 MiB flash and 16 MiB ram Aug 28 18:16:37 its a 4 MiB flash and 32 MiB ram Aug 28 18:16:39 damn it. Aug 28 18:26:37 what do you mean by which bootloader to use? afaik you can only use CFE Aug 28 18:28:39 CFE for atheros? Aug 28 18:29:45 <[florian]> is not the wrt54g using a bcm47xx SoC? Aug 28 18:32:04 wrt54g v7 is atheros Aug 28 18:36:03 so [florian] how does a broadcom cfe work? Aug 28 18:36:06 humm, then I don't know.. Aug 28 18:36:19 hmmmm Aug 28 18:38:48 <[florian]> quackquack: this obviously won't work Aug 28 18:38:59 <[florian]> there should be some redboot sources for atheros though Aug 28 18:39:33 just confirmed, wrt54g v7.0 is broadcom indeed Aug 28 18:39:54 NO its atheros. Aug 28 18:40:03 you mean 7.2 Aug 28 18:40:08 not v7.0 Aug 28 18:40:19 actually, wrt54gs v7 Aug 28 18:40:23 sorry Aug 28 18:40:25 [florian], compile redboot from scratch? Aug 28 18:41:25 about zero info here: http://infodepot.wikia.com/wiki/Linksys_WRT54G_v7.0 Aug 28 18:41:45 <[florian]> quackquack: no idea if there are some binaries available Aug 28 18:41:52 <[florian]> it just requires a mips-elf toolchain Aug 28 18:42:38 aah Aug 28 18:43:46 http://en.wikipedia.org/wiki/Linksys_WRT54G_series Aug 28 18:43:59 7.0 is the only atheros based device. Aug 28 18:44:03 and i got 2 of them. Aug 28 18:44:14 [florian], thanks i will try it soon. Aug 28 18:44:48 <[florian]> does it come with Linux by default? or is it vxworks? Aug 28 18:45:58 vxworks Aug 28 18:46:12 8 mb ram and 2 mb flash. Aug 28 18:46:18 i upgraded both ram and flash. Aug 28 18:46:27 now i have the buspirate ready. Aug 28 18:48:19 <[florian]> sounds fun Aug 28 18:49:52 i did check with jtag. Aug 28 18:50:09 but i preffer to use buspirate Aug 28 18:50:15 its faster than jtag Aug 28 18:50:21 and its spi Aug 28 18:51:18 <[florian]> does it directly access the SPI flash? Aug 28 18:58:09 yes Aug 28 19:00:59 /usr/local/sbin/flashrom -p buspirate_spi:dev=/dev/ttyUSB0,spispeed=8M -V -c "MX25L3206E" -w /path/to/firmware.rom Aug 28 19:01:10 build #324 of avr32 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/324 Aug 28 19:01:24 /usr/local/sbin/flashrom -p buspirate_spi:dev=/dev/ttyUSB0,spispeed=8M -V -c "chipname" -w /path/to/firmware.rom Aug 28 19:38:46 can I put [RFC][PATCH] when sending a patch to the list or [RFC] is enough? Aug 28 19:42:11 hi Aug 28 19:42:21 which image do i need for the nanostation m2? Aug 28 19:48:07 all tutorial talk about openwrt-atheros-ubnt2-squashfs.bin but this file isn't created when i build all images for ubiquiti devices Aug 28 19:51:18 __trine: do you know which image i have to use? Aug 28 19:51:36 is it the openwrt-ar71xx-generic-ubnt-unifi-outdoor-squashfs-factory.bin image? Aug 28 19:59:54 __trine: soled it its openwrt-ar71xx-generic-ubnt-nano-m-squashfs-factory.bin Aug 28 20:26:36 <__trine> tripolar, I compile my own Aug 28 20:41:20 __trine: me too Aug 28 20:41:33 i solved it its just the nano-m image Aug 28 20:41:49 <__trine> I compile special ones :) Aug 28 20:42:15 mine are also modded Aug 28 20:42:54 <__trine> I think my nano's are the most useful interesting routers I have Aug 28 20:43:15 <__trine> they are great for long distance connections Aug 28 20:43:48 the devices are nice, the bullets are also nice but the newer devices like unifi ap pro suck Aug 28 20:45:00 <__trine> I use mine to connect the the free hotspots Aug 28 20:45:55 <__trine> I have a guest wlan0 AP and a client so I can connect then reshare for my other devices Aug 28 20:46:32 nice Aug 28 20:46:37 <__trine> I also feed the rj45 from the nano into a wrt160nl so I can connect other things to the wrt160nl Aug 28 20:46:49 <__trine> all on free wifi Aug 28 20:46:53 the routers i use are used for free wifi hotspots :) Aug 28 20:47:08 <__trine> where are you Aug 28 20:47:13 i mean they are deployed to be used as free wifi hotspots ;) Aug 28 20:47:28 austria Aug 28 20:47:36 <__trine> UK here Aug 28 20:48:17 <__trine> I want to share wifi but I am not sure of the legal implications if someone did something illegal Aug 28 20:48:50 yeah you can run into legal problems when someone downloads stuff with your ip Aug 28 20:49:14 <__trine> and the cops here hate me Aug 28 20:51:22 __trine: not the cops are the problem but the lawyers from copy right holders that do nothing meaningfull but sue people and squeeze money out of them Aug 28 20:51:48 because someone copied stuff no one would buy for money :D Aug 28 20:53:19 __trine: stories like this: http://www.dw.de/lawyers-make-big-money-from-file-sharers/a-16202429 Aug 28 20:53:42 <__trine> in my personal case its the hefty but worthwhile price I pay for being an activist,, I just treat as the rent I pay to be alive Aug 28 20:54:36 <__trine> if it was a football match it would be about 12-0 to me :) Aug 28 20:54:42 <__trine> heh Aug 28 20:55:02 activist? what do you do? Aug 28 20:55:17 <__trine> oh thats too long a story for here Aug 28 20:56:09 what a pity Aug 28 20:58:26 <__trine> I may sometime in the future write a book about it titled 'OUT OF their CONTROL' Aug 28 21:01:42 __trine: is today much is out of control the book will be everything else than a short story :( Aug 28 21:11:42 Anyone look at Julian's patch 3828? I quite like it. I'm surprised it hasn't yet been accepted. Aug 28 22:27:59 philipp64: I took a look at it but its not clear what parts of it are still applicable after the host-record patch in r36943 Aug 28 22:28:24 philipp64: the domain / hostname / ptr record stuff gets quite convoluted and as I see there are three or so different ways to do the same stuff now Aug 28 22:28:45 so at some point I have to take a step back and look at whats still needed and what can be thrown out Aug 28 22:28:54 also the ui needs changes for this etc. Aug 29 01:13:48 jow_laptop: well, I liked the fact that his patch obviated the need for a separate "config domain" section... if we have a "config host" that should imply a matching "config domain" section, IMO. Aug 29 01:14:17 but yeah, that would change the UI a bit I guess... Aug 29 01:14:56 did you get a chance to look at my "needs-broadcast" fix for broken devices that can't handle unicast replies to DISCOVER packets? **** ENDING LOGGING AT Thu Aug 29 02:59:58 2013