**** BEGIN LOGGING AT Fri Apr 10 02:59:57 2009 Apr 10 03:05:38 http://www.hardwaresecrets.com/article/370 Apr 10 03:20:58 http://www.youtube.com/watch?v=00apfR97g40# <--- YOU GO TO HELL AND YOU DIE! Apr 10 07:52:00 Sleep_Walker: hello Apr 10 07:52:38 hey emfox Apr 10 07:52:56 tmzt: hi Apr 10 07:53:07 anything new on fso Apr 10 07:54:27 tmzt: I've ported fso-ogsmd to treo, yet not finish and the code is ugly, but it's really work Apr 10 07:54:54 what distro? Apr 10 07:54:57 debian? Apr 10 07:55:25 yes Apr 10 07:55:31 and is that the python version or the vala version Apr 10 07:55:44 but it's not distro-specific Apr 10 07:55:57 cool, I have a debian sd card I use on motoq and raph500 Apr 10 07:56:29 the libgsmmux is vala and ogsmd is python Apr 10 07:56:58 python2.6? Apr 10 07:57:04 gsmd of vala version is not usable yet Apr 10 07:58:06 I do not know and not care, since after 2.3 the basic api has been so stable Apr 10 07:58:19 right Apr 10 07:58:37 but it makes a diference for implementing modems Apr 10 07:59:03 although for you if mux is working I guess the basic multiline should work Apr 10 07:59:05 could you be detail? Apr 10 07:59:42 just that the language choice affects anyone implementing extensions to the code Apr 10 08:00:05 assuming by api you meant the dbus interface Apr 10 08:00:41 hmmm, I'm not sure, I use python 2.5 in sid Apr 10 08:01:30 oh, I guess I have experimental for pkg-fso Apr 10 08:02:14 armel right? Apr 10 08:02:22 where? in your own repos? Apr 10 08:02:29 yes, armel Apr 10 08:02:57 no, I use pkg-fso debian repo from fso project Apr 10 08:03:09 but it has outdated packages Apr 10 08:03:17 me too Apr 10 08:03:45 I seem to remember it requiring experimental Apr 10 08:03:46 right, so yesterday I'm asking pkg-fso people to updated it Apr 10 08:04:00 hi emfox Apr 10 08:04:22 Sleep_Walker: :) Apr 10 08:04:56 tmzt: but nobody response even if I would to help Apr 10 08:05:10 emfox: your still interested in a way to power on/off modem in linux? Apr 10 08:05:30 emfox: I still can't make working GSM when not powered in PalmOS, but when you power it up in PalmOS, you can do everything you want Apr 10 08:06:13 I separated that into 3 GPIOs (power, reset, wake) and I've got one another now to have a look Apr 10 08:06:34 Sleep_Walker: well, that's enough for me, since we haven't got a real bootloader today Apr 10 08:06:55 emfox: well, yes, for now it's ok Apr 10 08:07:26 tzmt: yes, of course. but I'm not kernel developer so I just wait for Sleep_Walker :) Apr 10 08:07:38 the freezing issue does not occur if powered by palmos Apr 10 08:07:47 ? Apr 10 08:08:18 tmzt: right Apr 10 08:08:36 tmzt: then gpio57 can act as wake GPIO Apr 10 08:09:03 otherwise it's in cyclic sleep Apr 10 08:10:03 tmzt: do u know any pkg-fso peoples? I'd like to join.at least I could upload new pkgs since I'm a DD Apr 10 08:10:05 but why does that not happen if powered in palmos Apr 10 08:10:25 emfox: ask lindi- Apr 10 08:10:58 it's an alioth project but I've never seen any discussion Apr 10 08:11:41 a little strange... Apr 10 08:11:41 tmzt: well, I found another GPIO yesterday, but I had to take care of my wife Apr 10 08:11:58 tmzt: I think that this one can do something Apr 10 08:12:10 gpio10, out Apr 10 08:12:41 emfox: so you know how to make qemubuilder work? I have a fast laptop so I could build packages but only for myself Apr 10 08:12:54 I'm not a debian developer Apr 10 08:13:35 tmzt: sorry, I no little about qemu Apr 10 08:13:51 usually I use chroot Apr 10 08:13:56 it plugs in to dbuild or whatever that's called Apr 10 08:14:03 with qemu-arm? Apr 10 08:14:20 or how do you build armel packages Apr 10 08:15:13 tzmt: I just build it on my little treo :) as my DD work I use chroot and real debian sid Apr 10 08:15:31 oh Apr 10 08:15:36 that's slow Apr 10 08:15:52 and debian has many developer's machine in different arch I could use Apr 10 08:16:32 not so slow, for python, building is nearly copying Apr 10 08:17:44 yes, but there's a lot of c and vala code in later versions Apr 10 08:17:49 as well as shr Apr 10 08:18:13 a serious debian port would require those I think Apr 10 08:19:19 well, I will use developers' machines Apr 10 08:19:28 ok Apr 10 08:21:01 what's your concern about packaging? will you participate in it too? Apr 10 08:21:50 I'm not sure how I would, I know about as much as dpkg-buildpackage -b Apr 10 08:22:32 that'll be enough :) Apr 10 08:22:34 my concern is I would greatly prefer to use debian to oe or gentoo Apr 10 08:23:05 we should look at the effort required to package ms5.5 Apr 10 08:23:45 and my hope is that as the various pieces come together in the different phone related projects Apr 10 08:23:53 components are ready Apr 10 08:24:19 I know treos/centro are touchscreen but are they all 320x320? Apr 10 08:24:46 treos are Apr 10 08:25:09 not centro? Apr 10 08:25:16 I don't know centro Apr 10 08:25:26 ok Apr 10 08:27:58 there's a few other issues with debian Apr 10 08:28:13 one is that the enlightenment packages are old Apr 10 08:28:58 another is the state of X.org for arm devices, particularly xserver-xfbdev Apr 10 08:29:06 and other kdrive servers Apr 10 08:29:12 the most significant problem is many components are written for a specific phone Apr 10 08:29:25 odeviced is Apr 10 08:29:29 and ophoned Apr 10 08:30:39 for the e17, I have asked if I could package them one or two years ago, raster said it's not mature enough... Apr 10 08:31:05 I do not know what the status now Apr 10 08:31:19 emfox: you need to be careful packaging Apr 10 08:31:20 are you in #openmoko-cdevel? Apr 10 08:31:26 as thing will break/change Apr 10 08:31:45 it could stay with pkg-fso for now? Apr 10 08:31:58 so as to not break desktop use Apr 10 08:32:02 raster:ahha Apr 10 08:32:44 I think there's a pkg-e or such alioth project Apr 10 08:33:15 another one I didn't see much discussion as to direction/status Apr 10 08:35:08 tzmt: it's sad Apr 10 08:35:47 for debian's culture we could not even do help if they don't response Apr 10 08:36:18 the real problem with debian is that debian pretty much dont talk to us (uptsream) Apr 10 08:36:24 they dont pass patches etc. Apr 10 08:36:38 and in the end all the support ends up on our doorstep Apr 10 08:36:42 but it should be possible to put everything needed in pkg-fso? Apr 10 08:36:51 for often packages that are months older than even our own released snapshots Apr 10 08:39:05 you know packaging is tiring and become annoying day by day Apr 10 08:42:26 tmzt, WyrM: do you have idea, how to enable or disable keeping GSM powered during LPM? Apr 10 08:42:50 LPM? Apr 10 08:42:57 Low Power Mode Apr 10 08:43:59 Sleep_Walker: I think you can't do this. Apr 10 08:44:23 but im sure there is a way to awake it when it is sleeping :) Apr 10 08:44:45 still awake? Apr 10 08:45:04 and also, very likely it will signal the AP when it goes to sleep . Apr 10 08:45:08 emfox: yeah. packaging does. thats why we do a fair bit for them Apr 10 08:45:13 we have debian/ dirs in svn Apr 10 08:45:17 and .spec files for rpm Apr 10 08:45:23 *I* update the .bb files for oe Apr 10 08:45:30 also efl is highly consistent packaging-wise Apr 10 08:45:37 what about dependencies, raster? Apr 10 08:45:38 its very easy to just upgrade packages Apr 10 08:45:45 tmzt: well, I meant, when GSM is on, I'd like to keep in on during suspend (MFP_LPM_DRIVE_HIGH) Apr 10 08:45:56 openembedded does it al lthe time - i regularly bump the svnrev to use and simply say "rebuild!" with zero other changes Apr 10 08:46:03 and when it is off, I'd like to have it also off during suspend (MFP_LPM_DRIVE_LOW) Apr 10 08:46:05 dependencies rarely change Apr 10 08:46:20 oe picks them up automaticlly with ldd Apr 10 08:46:32 and runtime depsn u cant check are explicitly in the package recipies Apr 10 08:46:34 its easy Apr 10 08:46:38 (once you do it right) Apr 10 08:46:40 debian Apr 10 08:46:52 of course debain "policy" makes packaging thing slike e and efl that move quickyl REALLY HARD Apr 10 08:46:53 the debian dirs are current with dependencies? Apr 10 08:47:13 their policies reauire releases, uploads to ftp maintainers, approvals, reviews and so on Apr 10 08:47:15 Sleep_Walker: I don't understand what you want to do. Apr 10 08:47:24 they really stifle the pipelien and MAKE packaging a pain in the arse Apr 10 08:47:37 debian dirs should be current Apr 10 08:47:42 but i dont fully track them Apr 10 08:47:52 and if you have say libefl, dpkg-build it with the svn debain/ that is valid for enlightenment's debian/ dir Apr 10 08:48:56 i havent tried as i dont actualluy touch the .deb pakcages Apr 10 08:48:58 you can try Apr 10 08:49:02 but most work should be done Apr 10 08:49:37 raster: well, if it's easy and no need to figure out deps or api changes many hours, I'd like to sponsor e17 for debian Apr 10 08:50:04 lutin is already doing stuff there Apr 10 08:50:05 I just probably need 'pxa2xx_mfp_config' to have in code and reconfigure pin setting again Apr 10 08:50:08 as are others i think Apr 10 08:50:24 sorry Apr 10 08:50:25 of course I must contact those pkg-e people first Apr 10 08:50:33 kaethorn and quaker Apr 10 08:51:11 the problem reall yis just gettign the pieplien working Apr 10 08:51:16 ie build packgews direct from svn Apr 10 08:51:19 pick a svnrev Apr 10 08:51:26 then get it working Apr 10 08:51:29 once u do it right Apr 10 08:51:37 updating to a new rev is a single number chnage Apr 10 08:51:44 then "run the scripts" Apr 10 08:52:15 so you haven't got such a pipeline right? Apr 10 08:52:41 does ms5.5 use a particular svn revision of efl? Apr 10 08:57:28 Sleep_Walker: btw, I searched and found just a "function block diagram" for bcm2133. it's useless, right? Apr 10 08:59:57 emfox: I don't know Apr 10 09:00:05 emfox: I'd rather collect all DOC Apr 10 09:00:09 (since Marvell) Apr 10 09:00:17 can you point me to? Apr 10 09:01:09 emfox: currently no such pipe Apr 10 09:01:14 as debian has a big stall in it Apr 10 09:01:40 don't remember where it is, I just googled it. I can mail it to u Apr 10 09:02:04 raster: what the stall is? Apr 10 09:05:38 Sleep_Walker: something fan Apr 10 09:05:48 Sleep_Walker: just search pxa27x Apr 10 09:06:18 tmzt: ? Apr 10 09:06:43 I thought you were looking for pxa docs Apr 10 09:07:00 since you mentioned Marvell Apr 10 09:07:23 I already have all stored Apr 10 09:07:34 ok Apr 10 09:07:48 function block won't tell you much Apr 10 09:08:04 if the pm is internal to 2133 might be nice to know Apr 10 09:08:13 but it's proably a pmic Apr 10 09:25:50 emf upload to ftpmasters Apr 10 09:25:52 get approvals Apr 10 09:25:56 reviews Apr 10 09:26:07 go into experimental Apr 10 09:26:11 then unstable Apr 10 09:26:14 then testing... Apr 10 12:48:28 hi Apr 10 12:48:45 anyone use mcafee site advisor Apr 10 12:49:28 ??? Apr 10 12:53:36 wtf? Apr 10 12:54:14 commercials over IRC? Apr 10 12:55:19 increasing efficiency on IT people? : Apr 10 12:55:20 :b Apr 10 12:56:51 no I would like to know what you think I would like to sue mcafee Apr 10 12:57:43 I think you can't read topic or at least you're confused. Apr 10 12:58:05 oh sorry Apr 10 13:32:15 wow! Apr 10 13:32:24 GPIO10 caused awake Apr 10 13:37:11 ? Apr 10 13:38:16 I had disabled GSM in PalmOS and I enabled it in Linux Apr 10 13:38:40 ah nice Apr 10 14:22:51 emfox: hi Apr 10 14:23:01 emfox: wait a moment, I've got new kernel for you Apr 10 14:24:08 ok Apr 10 14:24:12 emfox: btw. it seems that GPIO10 can help with powering on in Linux Apr 10 14:24:22 great Apr 10 14:25:07 have you created keymap for xmodmap? Apr 10 14:25:28 yes, just ten minutes ago Apr 10 14:25:54 but it's a little difficult to press some of the symbols Apr 10 14:26:22 you should hold both option and shift, the press another key Apr 10 14:27:02 something is better than nothing Apr 10 14:27:23 emfox: btw. what about '%' on brightness control? Apr 10 14:27:47 I haven't figured out how to use alt+key mapping, maybe there's no such mapping Apr 10 14:28:08 emfox: probably because of alt as navigation key Apr 10 14:28:23 emfox: I can make it some other modifier if you tell me which Apr 10 14:29:26 I'll send you it later Apr 10 14:29:44 % for brightness is ok Apr 10 14:30:21 I think better than just number levels Apr 10 14:30:57 ? Apr 10 14:31:11 heh Apr 10 14:31:15 misunderstanding Apr 10 14:31:28 I'm still talking about keymap Apr 10 14:31:37 there is option+p unused Apr 10 14:31:48 (brightness is not accessible this way) Apr 10 14:32:26 if you want percent for brightness, it can be easy fix, but more values means more positions of brightness control... Apr 10 14:32:35 I understand Apr 10 14:32:50 but will we use that key later? Apr 10 14:33:26 if we could map it to a brightness-changing app Apr 10 14:33:37 at least in X Apr 10 14:35:13 emfox: will you use it? Apr 10 14:35:17 I don't care Apr 10 14:35:24 I just missed '%' Apr 10 14:35:31 and this wasn't used Apr 10 14:36:20 I think 'shift+.' is free Apr 10 14:36:52 and 'option+.' and 'alt+.' Apr 10 14:37:01 all free :) Apr 10 14:37:55 If you ack, I'll send the new version with the x keymap to you tomorrow Apr 10 14:39:00 fine with me Apr 10 14:39:16 I'll export GPIO10 through /sys too Apr 10 14:39:44 and could you tell me what symbols are missed from current mapping except %. I will add them together Apr 10 14:41:12 no problem with that, but it'd be better use a understandable name Apr 10 14:43:22 '%', Esc Apr 10 14:44:19 roger Apr 10 14:44:24 thanks :) Apr 10 14:44:48 kernel is ready Apr 10 14:45:30 the origin place? Apr 10 14:45:59 s/origin/same/ Apr 10 14:49:53 http://sleepwalker.hackndev.com/release/T680/linux-2.6-arm/partition/3/ Apr 10 14:49:55 yes Apr 10 14:50:45 so I go download it. I hate single-threading of palmos ... Apr 10 14:51:07 bye Apr 10 15:19:15 mekius: sorry to say that, but it seems that my kernel (link above) still works well on Centro I have.. Apr 10 15:20:21 I'll try bluetooth, IrDA and GSM now Apr 10 15:23:25 keyboard map has to be altered (home and power button are switched) Apr 10 15:25:21 bluetooth is not working (I'll try to find right GPIO) Apr 10 15:25:31 Sleep_Walker: I will try your kernel later tonight and see how it goes Apr 10 15:25:39 Sleep_Walker: I can use 1944 id in cocoboot even? Apr 10 15:25:46 sure Apr 10 15:26:25 there are 2 machine definition in treo680.c - one for Centro, one for Treo680 Apr 10 15:26:51 I will add differences easily with machine_is_centro() Apr 10 15:29:16 Sleep_Walker: so, you found the ap->bp wakeup pin, great! :) Apr 10 15:29:49 WyrM: yes, but there are two :b Apr 10 15:30:03 two pins that can wakeup bp? Apr 10 15:30:04 WyrM: I managed to wake it up with weird behaviour Apr 10 15:30:14 or 2 that you need to toggle at the same time? Apr 10 15:30:32 while true; do gpio 10 0 ; sleep 0.2 ; AT; gpio 10 1; done Apr 10 15:30:41 I think so Apr 10 15:30:48 after a while it started to respond Apr 10 15:31:23 heh, you really need to setup an interrupt for the other pins, so you can understand what is happening... Apr 10 15:31:24 WyrM: http://sleepwalker.hackndev.com/GPIO/GPIO-tel-on-good2.html if you have a look on GPIOs 10 and 14 Apr 10 15:31:32 generally, wake up is.. Apr 10 15:32:06 1. bp->ap wake. Apr 10 15:32:07 mekius: IrDA doesn't work either - also have to find right GPIOs Apr 10 15:32:17 err, 1. ap->bp wake Apr 10 15:32:23 2. bp->ap rdy Apr 10 15:32:31 3. data.... Apr 10 15:33:11 WyrM: yes, I can imagine that Apr 10 15:33:30 WyrM: but that evil BCM chip doesn't speak with me so easily Apr 10 15:33:41 I think it does.. Apr 10 15:33:48 but you are not seeing it ;) Apr 10 15:33:51 mekius: be careful when suspending Apr 10 15:34:12 WyrM: it's a LOT of GPIOs I don't set up yet... Apr 10 15:35:50 Sleep_Walker: if it was me, I would setup BE interrupts on all unknown pins that are set to IN on palmos. :) Apr 10 15:35:59 with just a printk. Apr 10 15:36:40 and see what happens when I mess with the OUT interrupts.. Apr 10 15:36:59 Sleep_Walker: Well right now all I'm worried about is the screen blanking not screwing up :D Apr 10 15:37:03 s/interrupts/gpios/ Apr 10 15:37:27 Sleep_Walker: But I think someone said that issue was fixed with sw3 or something, I will have to test Apr 10 15:37:41 mekius: :b Apr 10 15:37:59 mekius: I probably touch some GPIO which is different on both machine Apr 10 15:38:29 WyrM: it's about 50 GPIOs, but yes, you're right Apr 10 15:39:30 wow, 50 pins set to IN that you don't know what they are for? Apr 10 15:39:36 this is a lot! :) Apr 10 15:39:42 I was expecting something like 10. Apr 10 15:40:11 WyrM: I have to check Apr 10 15:40:23 it's maybe just 40 :b Apr 10 15:40:46 heh, still a lot ;) Apr 10 15:40:50 that's the reason I'm not happy with this approach :b Apr 10 15:48:25 :) Apr 10 15:51:07 mekius: btw. try to boot with GSM powered on and with GSM powered off Apr 10 15:51:16 I have CDMA ;) Apr 10 15:51:41 I wish it was GSM :/ Apr 10 15:52:18 mekius: can you call with that? Apr 10 15:57:52 WyrM: ok, so it's not that much Apr 10 15:57:54 14 Apr 10 15:58:01 better :) Apr 10 15:58:05 yes Apr 10 15:58:09 Sleep_Walker: Well yeah, it does calls while in PalmOS, but doubt I could just use the GSM methods with it in Linux :) Apr 10 15:58:15 Unless the modems have the same interface Apr 10 15:58:23 mekius: I think you can :) Apr 10 15:58:33 with some "minor" adjustments. :) Apr 10 15:58:46 "minor" :D Apr 10 15:58:56 I just haven't messed with it yet Apr 10 16:00:18 mekius: just please make sure that your CDMA/GSM/whatever is off when starting Apr 10 16:01:21 Sleep_Walker: won't work otherwise? Apr 10 16:01:40 would be awesome if I could send/rcv calls while in Linux :D Apr 10 16:01:52 I'm guessing the rcv is quite a bit harder then the send heh Apr 10 16:02:06 mekius: when you'll be able to boot, we can move further :b Apr 10 16:05:17 WyrM: btw. I don't know many low GPIOs... that ones which can wake ap Apr 10 16:48:49 Sleep_Walker: Yeah, I guess for now I'll have to keep using raster's illume rootfs that I was using before, I can't get oe to cooperate :P Apr 10 16:49:14 mekius: congratulations! you're cured! Apr 10 16:56:08 :) Apr 10 17:16:20 hey fokls Apr 10 17:16:38 will the resume also work for TX miska? Apr 10 17:21:55 msd2: sure Apr 10 17:22:35 ok, I added some buggy code to init all unknown IN iterrupts :b Apr 10 17:24:26 * mekius predicts failure Apr 10 17:27:58 it boots ;b Apr 10 17:28:03 and works Apr 10 17:28:46 msd: hi again Apr 10 17:29:02 msd: I'm sure that |miska| wants to have his PDA suspending Apr 10 17:30:02 changin 79 OUT changes 90 IN Apr 10 17:31:22 changin 40 OUT also changes 90 IN sometimes Apr 10 17:33:11 Sleep_Walker: after the IN changes, do you note any difference on the behaviour of the AT interface? Apr 10 17:35:25 not sure yet Apr 10 17:35:31 I have to go shopping now Apr 10 17:35:38 bb in 1-3 hours Apr 10 17:37:33 hmm, also changing GPIO10 (the new one) generates changes of 90 IN Apr 10 17:37:52 I made new more verbose variant, rebooting and leavign Apr 10 17:39:53 good! :) Apr 10 17:40:03 see you later then. :) Apr 10 17:40:08 good shopping Apr 10 18:07:40 * flyback bbl Apr 10 18:15:31 I have a TX with full flash memory Apr 10 18:15:45 I might send it to miska if hes interested for testing Apr 10 20:33:49 * flyback heads out for dinner Apr 10 20:34:16 WyrM: very good shopping, travelling across city with 30kg of food :b Apr 10 20:42:40 (and notebook and plenty of PDAs and smartphones :b ) Apr 10 20:46:45 mekius: could you send me your GPIO registr dumps? Apr 10 20:46:54 mekius: I think you already did some for Marex... Apr 10 20:47:43 I'll try to create some CDMA friendly version of kernel Apr 10 20:49:50 Sleep_Walker: Will have to be later Apr 10 20:50:27 mekius: your Centro, your time, your problem :) Apr 10 20:50:47 mekius: please, send it to my mail Apr 10 20:50:58 (so I can find it in future) Apr 10 20:53:34 not sure I remember you email :) Apr 10 20:55:19 my_nick hackndev. com Apr 10 20:59:51 k Apr 10 21:00:04 i'll grab them again and let you know Apr 10 21:01:59 thank you Apr 10 22:46:08 * Sleep_Walker just found some weird behaviour about red led Apr 10 22:46:15 it looks like 2 state Apr 10 22:46:33 low light, high light Apr 10 23:26:01 WyrM: weird, I can't wake again Apr 10 23:26:19 WyrM: after reset, I just wait Apr 10 23:26:33 gpio14 in is 0 Apr 10 23:26:48 after first AT it becomes high Apr 10 23:27:03 and I can't take it to low Apr 10 23:30:26 hm, I tried to reset it again, now gpio14 in is 0 and command doesn't change anything Apr 11 01:35:20 * flyback thinks 3000mg of tyneol has already managed to ko his liver, ugh Apr 11 01:49:29 fuck Apr 11 01:50:16 that's what they all say Apr 11 01:58:07 sigh...taxes done Apr 11 01:58:09 there is uncertain group of people who should be ass-kicked for providing such Internet connection Apr 11 01:58:33 heh Apr 11 01:58:55 my early morning curry is almost ready :) Apr 11 02:03:48 * flyback is sure he's going to die now, oh well Apr 11 02:05:20 flyback: I think I'll see you tomorrow here :b **** ENDING LOGGING AT Sat Apr 11 02:59:57 2009