**** BEGIN LOGGING AT Sun Aug 30 02:59:57 2009 Aug 30 03:21:00 hi Aug 30 03:21:57 im trying to install shr on a gta-01, where must I extract the module in order to avoid the "unable to mount root fs on unknown-block(0,0)" error ? Aug 30 04:13:24 hi need help with my freerunner Aug 30 04:14:33 * ndnihil just started opkg upgrade Aug 30 04:14:40 wonder how this will|wont work Aug 30 04:14:58 hi Aug 30 04:15:02 howdy Aug 30 04:15:03 hows ur phone doing Aug 30 04:15:14 its good for now Aug 30 04:15:21 but I could be breaking it Aug 30 04:15:30 if so I'll just flash it again Aug 30 04:15:32 why? Aug 30 04:15:57 just doing an opkg upgrade Aug 30 04:16:00 when u first got it could you even place a call? or you still have to upgrade it something else? Aug 30 04:16:16 it worked just fine out of the box for me Aug 30 04:16:27 still waiting for my phone to be shipped parisoma only has 2 phones left Aug 30 04:16:40 ah Aug 30 04:16:46 it'll work Aug 30 04:16:51 what distro are you planning to use? Aug 30 04:16:52 but you'll probably want to upgrade Aug 30 04:17:00 I've been using shr-unstable Aug 30 04:17:27 hmmm sounds new to me. . . any plans in porting debian on it? Aug 30 04:21:05 debian is already available for it Aug 30 04:21:21 I'll probably go that route eventually Aug 30 04:21:28 I've got debian on everything else Aug 30 04:22:11 cool... Aug 30 04:28:05 you have no problems in using your wifi? Aug 30 04:28:15 quite the contrary Aug 30 04:28:19 many problems with wifi Aug 30 04:28:22 well Aug 30 04:28:28 it works fine for open APs Aug 30 04:28:29 once Aug 30 04:28:58 but trying it twice will often lock the device up Aug 30 04:29:15 and connecting to protected APs just wont happen Aug 30 04:29:21 not for me anyway Aug 30 04:29:27 it's a hit or miss thing Aug 30 04:29:36 some people get it to work, some dont Aug 30 04:29:53 I think the wifi kernel module is buggy Aug 30 04:38:28 hmmm sounds strange Aug 30 04:38:40 i thought it was all worked out... Aug 30 04:38:53 hmm maybe upgrading it would fixed it? Aug 30 04:39:13 dunno, that's what I'm trying now Aug 30 04:39:13 or would it fix if you'll use a usb wifi card? Aug 30 04:39:35 have you checked out www.ipwn.com Aug 30 04:39:38 usb dongle would work, but be very inconvenient to carry around, and drain the battery Aug 30 04:40:10 hmm Aug 30 04:40:18 for wardrvng hee hee Aug 30 04:40:42 kismet+gpsdrive is fine for wardriving Aug 30 04:40:54 we better get those nokia batteries on thnksgvng Aug 30 04:41:07 have u tried it on ur phone? Aug 30 04:41:29 nope Aug 30 04:41:47 if but for no other reason, the antenna on the cards in the laptop is far better Aug 30 04:42:30 what about pdf? Aug 30 04:42:40 ? Aug 30 04:42:52 can u read pdf clearly? Aug 30 04:42:59 oh, haven't tried Aug 30 04:43:10 on the phone damn cant wait to get my hands drty Aug 30 04:43:32 if you want to play videos, get the mplayer with glamo patch Aug 30 04:43:37 ur familiar with backtrack? Aug 30 04:43:47 nope Aug 30 04:43:54 oh wait Aug 30 04:44:05 that pentest dist/bootable cd? Aug 30 04:46:24 yeah Aug 30 04:46:51 tried it out once Aug 30 04:46:53 seemed ok Aug 30 04:47:04 i think someone ported most of the appz on the fone Aug 30 04:47:15 selling his dstro for 99$ Aug 30 04:47:23 but I dont see any reason to use that rather than just installing those tools on your normal dist Aug 30 04:47:47 perhaps some issues with kernel Aug 30 04:48:02 not really Aug 30 04:49:06 dunno why Aug 30 05:38:28 ndnihil: have you tried the latest kernel with the patch applied? Aug 30 05:39:47 tag8tag: looks like some other problem, probably with your u-boot env. Aug 30 05:40:44 PaulFertser: which patch? Aug 30 05:41:29 DocScrutinizer: Tarkovsky's "Stalker" is really different, the other mood/ideas completely. Aug 30 05:41:55 Undrwater: the one on the kernel ML, removes one "unnecessary" lin. Aug 30 05:42:30 oh? :) Aug 30 05:42:35 Undrwater: i'm pushing it today since Werner's complaints are "resolved" and nobody gives any feedback. Aug 30 05:42:49 Screw that, ftw people complain so much about the wifi and nobody's tried it already? Aug 30 05:42:49 i haven't been following Aug 30 05:43:37 i'll try it...which is the kernel ML? Aug 30 05:44:11 i didn't see it on the community ML, which seems to have more activity Aug 30 05:44:36 Undrwater: http://lists.openmoko.org/pipermail/openmoko-kernel/2009-August/010494.html Aug 30 05:44:45 Undrwater: i mentioned it on community ML Aug 30 05:47:41 ah...is there a kernel already built with that patch? Aug 30 05:48:31 Undrwater: hm Aug 30 05:49:09 I mean that's a stupid question, of course i have that (since i tested it) but my upload speed is very slow and so uploading modules will take ages. Aug 30 05:49:28 :P Aug 30 05:49:51 should be able to be hosted somewhere whre the kernels are typically found. Aug 30 05:50:07 But upload speed is the limiting factor Aug 30 05:50:39 It'd be easier to build it on shr buildhost. I just can't understand wtf there's so many users complaining and nobody's tried the patch already. Aug 30 05:51:48 Building the kernel is trivial. Considering one's done that already numerous times (and iirc back in the days, just 10 years ago, even when Linux was out of infacy already one still built the kernel himself) Aug 30 05:52:46 i do that here...but i can't build a kernel for arm on my machine without cross compile or other setup, correct? Aug 30 05:53:21 PaulFertser: is that patch now an alternative to budfive's original wlan patch? Aug 30 05:53:44 lindi-: no, it's to solve others problem Aug 30 05:54:16 lindi-: budfive's patch is trying to bring some consistency to the driver. But it's not the way as i see it. Aug 30 05:54:49 lindi-: rather ar6000_drv.c should be rewritten, i can't see how it can be fixable, the code is too messy :( Aug 30 05:55:17 lindi-: my one-liner is just a bandaid to allow driver to properly function after unbind/bind (and i guess also suspend/resume). Aug 30 05:55:48 Undrwater: emdebian installs arm cross-compiler in a matter of minutes without any complex steps involved. Aug 30 05:56:22 PaulFertser: is that liveCD or something installed? Aug 30 05:56:34 nevermind...i'll look, thanks for the pointer Aug 30 05:57:01 Undrwater: no, that's just a convenient way to crosscompile on Debian/ubuntu. Aug 30 05:57:17 i'm on gentoo Aug 30 05:58:09 Undrwater: there it's easy too, with crossdev Aug 30 05:58:29 ahh...OK Aug 30 05:58:44 Undrwater: for compiling the kernel even armv5 toolchain is ok, but you won't be able to compile properly Qi or other programs with it. Aug 30 05:59:18 Undrwater: unfortunately, i haven't found a way to build armv4-compatible toolchain with crossdev. Gentoo folks employed different hacks but no proper solution iirc. Aug 30 05:59:21 max_posedon: ping Aug 30 06:01:17 lindi-: i have a suspicion that with budfive's patch "ifconfig eth0 down" won't disable the device and hence you'll lose power. Aug 30 06:02:35 PaulFertser: is there any easy way to test that? Aug 30 06:04:05 openmoko: 03werner * r5578 10/trunk/eda/fped/ (TODO fped.c kicad.c postscript.c): (log message trimmed) Aug 30 06:04:05 openmoko: - postscript.c: added some code for frame display, but I'm not happy with it Aug 30 06:04:05 openmoko: - postscript.c: moved tedious "... maxfont maxfont ..." into new function Aug 30 06:04:05 openmoko: "boxfont" Aug 30 06:04:05 openmoko: - postscript.c: "hcenter" was way off for diagonal text Aug 30 06:04:08 openmoko: - postscript.c: "realsize" didn't take into account the effect of rotation and Aug 30 06:04:10 openmoko: returned only the x component in the original coordinate system Aug 30 06:04:37 lindi-: probably adding a printk to HTCStop to see if it's called or not can show that. OTOH if my guess that this function is responsible for turning off the device is incorrect, it doesn't help much. But i think the concept of changing wireless properties on a disabled interface is wrong from the beginning. Aug 30 06:04:58 openmoko: 03werner * r5579 10/trunk/gta02-core/ (Makefile modules/mkloe): Aug 30 06:04:58 openmoko: - Makefile: use "make loe" instead of running mkloe directly, so that we can Aug 30 06:04:58 openmoko: add local fancy stuff Aug 30 06:04:58 openmoko: - modules/Makefile: shortcut for mkloe, Poscript handling, and cleanup Aug 30 06:04:58 openmoko: - modules/mkloe: added -ps option to convert all modules to Postscript Aug 30 06:05:12 lindi-: are you sure that's needed? On many cards i saw until you do "ifconfig blablabla up" you can't even initiate the scan. Aug 30 06:06:38 PaulFertser: i only know that budfive's patch "works for me" very well :) Aug 30 06:07:21 lindi-: ok, do i understand it right that your particular problem is that the driver crashes if you try to change any wireless property on an inactive interface? Aug 30 06:07:46 PaulFertser: no, the problem is that if I "ifconfig eth0 down" then I can not use wifi at all anymore Aug 30 06:08:12 lindi-: that's not what your latest message on the ticket says Aug 30 06:08:54 PaulFertser: it says 'I can ifconfig up/down as many times as I want' Aug 30 06:09:41 lindi-: i meant the latest message describing the problem. With steps to reproduce, the last step is sudo iwconfig eth0 power off Aug 30 06:11:08 PaulFertser: well i was just looking for ways to crash it since crashes are much easier to analyze than "wifi does not work after foo" Aug 30 06:12:13 lindi-: ok, so you want me to try to "ifconfig down/up" several times and then to try to use wifi... Aug 30 06:12:24 PaulFertser: for example yes Aug 30 06:15:12 lindi-: why do you consider this usecase important? Aug 30 06:15:42 PaulFertser: I stop wifi by doing 'ifdown eth0' Aug 30 06:15:49 PaulFertser: that does ifconfig eth0 down internally Aug 30 06:15:57 lindi-: sure Aug 30 06:16:13 lindi-: i propose to unbind the driver whenever it's to put down. Aug 30 06:16:17 also i've heard wicd does 'ifconfig eth0 down' Aug 30 06:18:08 lindi-: proper power-saving is guaranteed only if you rmmod the module or at least unbind it. So whenever wifi is not used the user is supposed to do that, not ifconfig eth0 down. I agree it'd be nice to have it working that way too but is it really worth the trouble when you can simply unbinding/bind it? Aug 30 06:18:39 lindi-: and sure, i can reproduce that it doesn't work at all after ifconfig down. Aug 30 06:18:56 PaulFertser: well werner's comment is "unbind/bind did not solve the problem (!)" Aug 30 06:19:32 PaulFertser: having to patch all these programs to do unbind for the specific wifi driver is going to be tricky too Aug 30 06:19:37 lindi-: without that one-liner it didn't, yes Aug 30 06:19:51 upstream surely won't take such fixes Aug 30 06:20:04 lindi-: why patching? I propose to simply wrap wpa_supplicant invocation. Aug 30 06:20:08 lindi-: which upstream? Aug 30 06:20:13 PaulFertser: of wicd for example Aug 30 06:20:34 PaulFertser: or dhclient? any program that brings the interface up or down Aug 30 06:20:52 i'm not using wpa_supplicant in unencrypted networks Aug 30 06:21:03 lindi-: are you sure dhclient will bring the interface down if it's already up? Aug 30 06:21:22 lindi-: (not using wpa_supplicant for open nets) why? Aug 30 06:22:59 PaulFertser: one less process :) Aug 30 06:23:39 # IP address changed. Bringing down the interface will delete all routes, Aug 30 06:23:39 # and clear the ARP cache. Aug 30 06:23:39 ifconfig $interface inet 0 down Aug 30 06:23:43 from ifconfig sources Aug 30 06:23:50 from dhclient sources Aug 30 06:24:29 why would one use wpa_supplicant for an open network Aug 30 06:25:16 Ke: for consistency and auto-configuration. Aug 30 06:27:54 lindi-: i've just tried, here dhclient didn't put the interface down after the first invocation, wifi still works... Aug 30 06:30:13 lindi-: what about another proposal: revert 9c4451ff31b937a478f3d3eabef30b71cbe12b12 and declare we don't care about rfkill? Aug 30 06:35:49 lindi-: or what about using the driver of pre-rfkill times, 4836494d233cd34fe194ff2040d62e4109067d79 (also d5119bd8f7a606c9e05b022a703b9df7dcc14730 and some other minor fixes might be needed). Aug 30 06:37:09 lindi-: i'm all in favor of fixing the driver properly but since nobody wants to spend enough time no it we have to compromise... Aug 30 06:44:06 PaulFertser: dhclient puts the interface down when the ip address changes Aug 30 06:44:11 PaulFertser: what's your lease time ? Aug 30 06:46:36 PaulFertser: i really don't know the internals Aug 30 06:46:54 PaulFertser: i just know that either using b8b36e5ec3db71d5 from january or using current andy-tracking with budfive's patch works for me (TM) :P Aug 30 06:47:00 lindi-: i agree that if the lease time is short the user will get in trouble. Aug 30 06:48:51 lindi-: if you view "git log drivers/ar6000" you'll see how much of untested/wip code went there... Aug 30 06:51:24 lindi-: so what do you say about using the driver with 9c4451ff31b937a478f3d3eabef30b71cbe12b12 reverted? Probably it's the cleanest solution... Too bad Michael will be upset but hell, we're not guilty he's using non-standard stuff. Aug 30 06:52:08 PaulFertser: well if it works then why not. again, i really don't understand the driver Aug 30 06:52:39 lindi-: do you think it's "ok" if we declare all rfkill-related races as WONTFIX then? Aug 30 06:52:50 PaulFertser: no idea, i have never used rfkill Aug 30 06:52:57 at least knowingly Aug 30 06:53:17 lindi-: all i know is that Michael always reports those because he's android dev. Aug 30 06:53:19 Does anyone else get: "unable to enumerate USB device on port *" Aug 30 06:53:25 ? Aug 30 06:53:44 AlexC: when? Aug 30 06:54:00 when you connect your neo to a usb port Aug 30 06:54:03 in dmesg Aug 30 06:54:17 i can't connect to my neo because of it Aug 30 06:54:52 AlexC: are you doing something fancy? If not, check your cable/usb hub etc. Aug 30 06:55:10 nothing fancy Aug 30 06:55:16 internal hub Aug 30 06:55:24 standard cable from openmoko Aug 30 06:55:32 brand new neo Aug 30 06:55:40 AlexC: is the system fully booted? Aug 30 06:56:08 yes Aug 30 06:56:16 AlexC: try another cable please Aug 30 06:56:22 i'm rebooting with factory settings now Aug 30 06:56:27 i don't have another Aug 30 06:56:30 What's factory settings? Aug 30 06:56:42 system defaults Aug 30 06:56:51 AlexC: on FR? Are there any? Aug 30 06:57:01 for the os Aug 30 06:57:07 qtopia Aug 30 06:57:29 it came pre-installed Aug 30 06:57:47 AlexC: does dfu-util sees your device when you boot to NOR u-boot? Aug 30 06:58:34 what is dfu-util? Aug 30 06:58:52 and "NOR u-boot"? Aug 30 07:00:04 AlexC: read the wiki please Aug 30 07:00:32 which section? Aug 30 07:00:39 AlexC: booting freerunner Aug 30 07:00:45 AlexC: flashing freerunner Aug 30 07:04:52 Okay i'm on nand boot menu Aug 30 07:05:09 which option do i select? Aug 30 07:06:16 AlexC: no option Aug 30 07:06:18 set console to usb Aug 30 07:06:22 oh Aug 30 07:06:24 AlexC: just connect to your host Aug 30 07:06:26 too late :/ Aug 30 07:06:30 AlexC: and see if it enumerates Aug 30 07:06:45 AlexC: the one you selected doesn't matter, connect it to host Aug 30 07:08:52 okay i connected, thankyou Aug 30 07:09:16 AlexC: so, since you device is brand-new i suggest you do the following: Aug 30 07:09:36 AlexC: upgrade your GSM firmware (GSM/Firmware wikipage). Aug 30 07:09:55 AlexC: then upgrade your NAND bootloader (to the latest u-boot or Qi, see the relevant wiki pages). Aug 30 07:10:21 AlexC: then choose some decent software and use it either by flashing to NAND or by loading from uSD. Aug 30 07:10:33 okay Aug 30 07:10:37 thanks for your help Aug 30 07:10:39 :) Aug 30 07:13:53 AlexC: (decent software) afaik you have a choice between an OE-based SHR unstable and Debian. Others are more exotic, you'll find out about them yourself. Aug 30 07:19:39 debian is a bit unfinished even the installer breaks Aug 30 07:23:57 Ke: who cares? Installation by hand is simple enough. Aug 30 07:24:29 simple yes, painful beyond description, yes Aug 30 07:25:26 Ke: why painful? I did it even before the installer supported Qi and it was the first system i installed on FR and the first debootstrap i did in my life. Aug 30 07:25:44 my debian install worked fine too Aug 30 07:25:56 no idea when that was Aug 30 07:26:16 the debootstrap part is trivial, getting something to install with is not Aug 30 07:26:34 you'll still need some other distro than debian to do this Aug 30 07:26:45 and all the other infrastructure Aug 30 07:27:18 Ke: i used OM2007 that came pre-installed on my device, no changes made. Aug 30 07:27:35 Ke: or i could have used my laptop's cardreader with the same result. Aug 30 07:27:35 gfy Aug 30 07:27:48 with qemu perhaps Aug 30 07:27:54 I have x86 laptop Aug 30 07:27:56 Ke: why qemu? Aug 30 07:28:02 =oP Aug 30 07:28:16 Ke: why do you think it's necessary to run "native" to do debootstrap? Aug 30 07:28:35 because you are? Aug 30 07:28:41 there is foreign Aug 30 07:28:53 but you'll still need to do the second stage Aug 30 07:29:26 Ke: can you already boot something after the first stage? Aug 30 07:29:32 PaulFertser: not really Aug 30 07:30:13 definitely not in the future when there will be initramfs Aug 30 07:30:24 (that won't get built until you do second stage) Aug 30 07:30:25 Ke: well, anyway even OM2007 was enough to install Debian without much trouble. Aug 30 07:30:27 PaulFertser: anyways I probably made some silly choices, but all of those choises were quite the norm on other platforms Aug 30 07:31:10 like using wlan instead of usb Aug 30 07:31:27 which had to be fixed Aug 30 07:33:47 om2009 autosuspends => had to be fixed Aug 30 07:34:12 om2007 has no proper terminal/keyboard trivially available Aug 30 07:35:10 I guess debian is trivial, once you are well off with anything else Aug 30 07:36:07 Ke: i used ssh over usb, connected to om2007 without issues etc. Aug 30 07:36:32 still wlan should have worked Aug 30 07:37:46 PaulFertser: so you tried installing debian just as you got the device with no prior experience on openmoko? Aug 30 07:38:09 Ke: basically yes Aug 30 07:38:13 ok Aug 30 07:38:17 Ke: other than reading the community ml for some time Aug 30 07:38:24 =oP Aug 30 07:39:02 I tried the install expecting things to work as documented (or assumed from properly working platforms) Aug 30 07:39:48 Ke: i knew that install doesn't support Qi so i didn't have any false hopes. Aug 30 08:54:54 PaulFertser: (stalker vs roadside-picnic) hmm, I don't think the diffs are so much more significant than with any other movie based on a book Aug 30 08:55:58 DocScrutinizer: sucky movies are closely following books but it doesn't make sense usually. But have you read the book to compare with Stalker movie? Aug 30 08:56:33 loong long time ago, yes Aug 30 08:56:48 It's just really-really different Aug 30 08:57:53 I should get me the book to re-read Aug 30 08:58:12 for sure always worth it Aug 30 09:36:09 motd Aug 30 10:43:19 what's again the name of this tiny X which SHR uses? Aug 30 10:43:30 kdrive? Aug 30 10:45:07 Ke: yeah, thanks :) Aug 30 10:45:13 naeg: yes, kdrive but it's deprecated, everybody's adviced to use x.org now Aug 30 10:45:47 we use x.org anyway Aug 30 10:46:01 but i thought we could provide both, but if you say it's not a good idea Aug 30 10:48:32 naeg: btw, i'm still puzzled why my right-click hack didn't work for you Aug 30 10:50:00 naeg: kdrive is reallyreally dead by now. Probably some barely supported device like iPAQ hx4700 still needs it but not gta01 or gta02. Some kdrive features are "ported" to X.org already and upstream deprecated kdrive several years ago. Aug 30 10:50:32 PaulFertser: i will open a ticket at our site, may the others like it the way you are doing it and then i will come back to you ;) Aug 30 10:51:07 naeg: but have you already tried it yourself? It definetely worked here. And i can try on my host it too. There's just no reason for it to not work ;) Aug 30 10:51:44 naeg: i know there was some nice panel applet in GPE to do something like that but it's unsupported by now. Aug 30 10:52:28 naeg: i guess either something like that should be employed or for those who prefer "long" click to simulate the right button a tslib modification can be done. Aug 30 10:53:30 PaulFertser: i don't think that tslib modification is the right thing for us, since we support more than one device and may someone else takes any other driver, we don't force anything, you know ;) Aug 30 10:53:51 so a kernel/x.org thing would be better Aug 30 10:53:59 but i have to talk to the others anyway Aug 30 10:54:45 naeg: simulating right click inside the kernel? Hm, not sure. Aug 30 10:55:06 naeg: same with X server, probably even more weird idea. Aug 30 10:56:08 PaulFertser: it would be also much effort to find it our for every driver Aug 30 10:57:03 naeg: tslib solution is generic enough Aug 30 10:57:27 naeg: my hack or a good application like GPE's would be too Aug 30 10:58:01 PaulFertser: but as i said, maybe not all are using tslib Aug 30 10:58:12 then we need to find it out for all drivers Aug 30 10:58:32 naeg: all drivers of what? Touchscreens? Why can a touchscreen driver not use tslib? Aug 30 11:00:21 naeg: also what is "other drivers" anyway? Touchscreen is exposed by kernel in a generic standard way. Like an input device. With EV_SYN EV_KEY EV_ABS capabilities. Aug 30 11:00:55 PaulFertser: i can't name one now, but there are others, i am quite sure Aug 30 11:01:19 naeg: others drivers for touchscreen? Aug 30 11:02:23 PaulFertser: evtouch Aug 30 11:02:24 naeg: kernel-level or X-level? Or what? Or is that some non-X FB-only stuff that uses kernel interface but not tslib that you want to support? Aug 30 11:03:21 we want to support everything with an arm processor ;) Aug 30 11:03:46 but non-X things like qtmoko can also use tslib Aug 30 11:03:59 naeg: (everything with arm) even arm7? :-O Aug 30 11:04:15 PaulFertser: we even plan to make repos for every armv* Aug 30 11:04:34 but at first there must be people who need it ;) Aug 30 11:05:02 the openmoko neofreerunner is our main device, but we support atm beagleboard and nokia n8*0 as well Aug 30 11:05:13 naeg: so what about arm7? Aug 30 11:05:38 well, armv4 optimized works on there too, doesn't it? Aug 30 11:05:48 naeg: those chips lack MMU Aug 30 11:06:31 well, then they aren't supported yet Aug 30 11:06:58 but would this be solved by compiling everthing again but optimized for armv7? Aug 30 11:07:41 "porting" from armvX to armvY shouldn't be as hard as x86 to arm Aug 30 11:09:01 naeg: you clearly have no idea about noMMU :( Aug 30 11:10:54 i don't get why the devs should say "armv7 need no MMU" Aug 30 11:11:10 needs* Aug 30 11:11:23 mmu adds cost and complexity to the chip Aug 30 11:11:32 Many apps don't need it Aug 30 11:12:00 but why only armv7? does armv9 lack mmu also? Aug 30 11:13:21 naeg: i'm not talking about "armv7", i'm talking about arm7 Aug 30 11:14:44 ahhh, this changes the situation ;) Aug 30 11:15:21 we just talked about supporting armv4-armv9, since these are the only ones we had to work with yet Aug 30 11:15:43 but not that much people want to use it yet, since we don't have that much to offer yet Aug 30 11:19:00 hi there :) Aug 30 11:21:10 arm7 is in the nintendo gameboys, isnt it? Aug 30 11:24:33 Using mokonnect on my newly flashed phone just hung the whole phone. Aug 30 11:24:58 Inserting the usb-kable made it go into kernel panic. Aug 30 11:25:40 Is that a feature? ;) Aug 30 11:28:23 openmoko: 03alvieboy * r5580 10/trunk/gta02-core/ (components/wm3236aq.lib wlan.sch): WLAN fixes (refs and chip symbol) Aug 30 11:28:41 Daramarak, you should use more recent distros Aug 30 11:29:09 max_posedon: Like the SHR-unstable from 16 aug? Aug 30 11:29:12 this bug was fixed some time ago (kernel panic when connecting usb to windows) Aug 30 11:29:37 hm... shr-unstable from 16 aug should work. Aug 30 11:30:07 max_posedon: Yes thought so. But this was probably connected to the mokonnect hanging. Aug 30 11:31:19 max_posedon: Otherwise I can connect and disconnect the usb as much as I like. Aug 30 11:38:27 openmoko: 03alvieboy * r5581 10/trunk/gta02-core/ (AUTHORS components/STATUS components/mm8430-2610.lib): New component CON8001 (mm8430-2610) Aug 30 11:38:38 openmoko: 03alvieboy * r5582 10/trunk/gta02-core/components/mkloe: New component CON8001 (mm8430-2610) Aug 30 11:42:37 PaulFertser: did you try this xmodmap thing under e17? Aug 30 11:43:01 naeg: no, only with matchbox-wm. Aug 30 11:43:19 i tried it under e17...i will give it a try with fluxbox Aug 30 11:43:20 naeg: can you tell me if "xmodmap -e 'pointer = 2 1'" works for you at all? Aug 30 11:43:32 naeg: there shouldn't be any wm-related differences. Aug 30 11:43:47 but i am not sure if where i can rightclick under illlume Aug 30 11:43:49 -l Aug 30 11:43:58 -if Aug 30 11:44:01 o.O Aug 30 11:44:15 naeg: start xterm, there you'll see if it's right click or left click. Aug 30 11:44:53 i tried to rightclick in xterm, afair Aug 30 11:47:23 PaulFertser: is matchbox fingerfriendly, or do you use something to touch? Aug 30 11:50:50 naeg: i don't remember, i'm using illume currently. Aug 30 11:51:29 naeg: iirc matchbox was finger-nail-almost-acceptable Aug 30 11:51:49 PaulFertser: do you rightclick there? seems like i deleted your scripts :( or at least i can't find them, sry. can you tell me it once more? Aug 30 11:53:06 no ok, its in the logs Aug 30 11:57:10 PaulFertser: seems not to work in fluxbox too Aug 30 11:58:12 I seem to be able to reproduce the mokonnect hanging too. Aug 30 11:58:35 Is it just me or has the Wifi become more and more unstable lately? Aug 30 12:02:21 xmodmap -e 'pointer = 2 1' works Aug 30 12:04:38 i know why now, waitclick doesn't work, it changes the buttons and before i do anything it changes them again Aug 30 12:05:04 naeg: are you using waitclick exactly the way i did? ;) Aug 30 12:05:41 PaulFertser: yes, but i think i don't have input-event Aug 30 12:06:01 naeg: aha, that makes difference indeed. I got it with input-utils from debian. Aug 30 12:07:17 well we don't have that packaged yet, nor xev Aug 30 12:07:46 any other way to do it? or do i have topackage input-utils Aug 30 12:12:53 PaulFertser: other dev showed intertest in your rightclick script, but he want to use it when e.g. pressing AUX Aug 30 12:16:10 naeg: xev and input-utils are very handy to have Aug 30 12:16:55 naeg: no problem in binding it to AUX... Aug 30 12:17:51 PaulFertser: i will at first package patched tslib with rightclick emulation, then i will package xev and try to get the script working, then binding keys...power button isn't binded also Aug 30 13:06:28 naeg: looks like evtouch has ONEANDAHALFTAP which simulates right click. Aug 30 13:08:16 * SpeedEvil wonders about a ring containing an accellerometer for morse input. Aug 30 13:11:52 PaulFertser: ONEANDHALFTAP? Aug 30 13:12:25 * SpeedEvil ponders a keyboard using one and 34/105 tap for 'e'. Aug 30 13:13:10 naeg: http://www.conan.de/touchscreen/libtouch.html Aug 30 13:19:56 openmoko: 03alvieboy * r5583 10/trunk/gta02-core/ (AUTHORS components/STATUS components/wte06016.lib): Add ANT8001 (WTE06016) Aug 30 13:20:14 openmoko: 03alvieboy * r5584 10/trunk/gta02-core/wlan.sch: Add antenna and small fixes Aug 30 13:24:28 PaulFertser: thx, will packge evtouch too to test it Aug 30 13:27:44 naeg: hm, it's what you give me as an example of TS driver that doesn't use tslib. Aug 30 13:27:48 naeg: ;) Aug 30 13:28:47 yeah, but i just googled "linux touchscreen driver" to find any other tsdriver ;) Aug 30 13:29:04 we are a good team ;) Aug 30 13:36:04 I've a problem with my FR running SHR unstable after an update. In settings I'm missing some options because "fso or ophonekitd" isn't running. I would prefer if this could be fixed without flashing the device again. Aug 30 13:37:47 When I try to start ophonekitd manually, it looks like the service-script just stops, since I get no status (ok/fail) after the command. Aug 30 13:42:30 hi Aug 30 13:42:45 im trying to install shr on a gta-01, where must I extract the modules in order to avoid the "unable to mount root fs on unknown-block(0,0)" error ? Aug 30 14:12:25 Hi all Aug 30 14:12:27 :) Aug 30 14:28:35 hi Aug 30 14:35:43 Does anyone know why after an opkg upgrade some of the packages are still listed under opkg list_upgradable? Aug 30 14:55:31 Have anyone started on a "screensaver" that displays phonestatus as a marqee or something like that? Aug 30 14:55:56 Daramarak: having the display on uses lots of power Aug 30 14:56:55 SpeedEvil: Yes but when I am at work or am home the phone is "docked". Then it would be nice if it just displayed information for me. Aug 30 15:01:30 k Aug 30 16:33:53 hi :) Aug 30 16:34:15 I'm new to OpenMoko, and the enlightenment WM is giving me a hard time Aug 30 16:34:54 every now and then I have to power off the phone by removing the battery because of unresponsiveness. Aug 30 16:35:31 as a result, some settings in enlightenment seem to get corrupted and afterwards I am unable to start it again Aug 30 16:36:10 (I keep getting the "Enlightenment error - This is very bad, Enlightenment SEGV'd") Aug 30 16:36:31 lanjoe9: remove your e directory Aug 30 16:36:40 mm Aug 30 16:36:43 where is it? Aug 30 16:36:53 ah I see Aug 30 16:36:54 .e ? Aug 30 16:36:59 yes Aug 30 16:37:04 ok!! Aug 30 16:37:25 rm -rf ~/.e/e/config maybe is enough Aug 30 16:37:26 don't know Aug 30 16:38:16 haha Aug 30 16:38:24