irclog2html for #htc-blueangel on 20051203

00:23.35cp15Good night
09:28.30*** join/#htc-blueangel cp15 (i=kljbol@p549A377A.dip0.t-ipconnect.de)
09:28.30*** mode/#htc-blueangel [+o cp15] by ChanServ
09:28.42cp15Good morning
10:39.20oho_good morning
10:40.59cp15Hi oho_, at least one who is here... Feeling a bit lonely in this channel
10:56.46oho_agree would be nice to have more people participating
10:57.07oho_I am just finishing some math off and then I will work on the gomunicator
13:46.49*** join/#htc-blueangel iclem_away (n=iclem@iclem.net)
14:06.33*** join/#htc-blueangel iclem (n=clement@iclem.net)
15:06.15*** join/#htc-blueangel oho_ (n=oho@dhcp2934.ukc.ac.uk)
17:11.23ohocp15: could you please point me towards a library which has a fifo queue for strings. Or do you have to implement one my self ?
17:30.31cp15Hmm... I will check
17:30.47oho_thanks
17:31.44cp15http://www.gtk.org/api/2.6/glib/glib-Doubly-Linked-Lists.html
17:31.58cp15You can use the glib doubly-linked lists
17:32.25bhimagood morning.
17:32.55oho_thanks - my C is quite rusty, but it is getting better
17:32.56cp15Just insert the command in the list via g_list_prepend and get the oldest command in the list with g_list_last and remove it with g_list_remove
17:32.59oho_hi bhima
17:33.06cp15Good morning bhima
17:33.23oho_cp15: thanks
17:33.37oho_I hope I will get most of it done today
18:12.38*** join/#htc-blueangel rob_w (n=bob@p85.212.147.80.tisdip.tiscali.de)
18:12.38*** mode/#htc-blueangel [+o rob_w] by ChanServ
18:54.52*** join/#htc-blueangel JuNuKN (n=JuNuKN@Ka4b9.k.pppool.de)
19:03.05oho_I am off meeting a few friends - see you all
19:39.24JuNuKNHello together, - hello icelem
19:40.04JuNuKNsome progresses, that I missed for Himalaya and BA by the lack of the current IRC-logs ?
19:45.11iclemHi JuNuKN
19:45.36iclemno great progress
19:46.19JuNuKN... so there was no real progress on SD access via Asic3 ? sad
19:46.51iclemAFAIK llange has extended Haret to watch ASIC3 GPIOs
19:47.24iclembut I don't know much
19:48.45iclemhe may have found something, but i've not talked to him since 1 week
19:49.02JuNuKNYes I saw it in the wiki ... hope it will help! CR2 always made some updates in the wiki , - also for Himalaya, - do you know him or his ident in irc ? I never saw him here! It seems to me, that he has nearly all of the devices, because he updates data for all of them!
19:53.33iclemI've seen his post in the forum, but i don't know if he's interested by Himalaya
19:55.21JuNuKNIt seems, that he is very familar with the Haret tool. Probably he is able to find out more Details on the SD access next weeks ... and  with the differences to BA. Probably we have to ask him in the forum.
19:57.24JuNuKN@cp15, I read in the IRC logs, that you where able to reduce the powerconsumption in standby nearly to the value of CE! great! So how long does the battery stay up now?
19:59.41cp15Don't know exactly... The values are from development, and I haven't installed it yet for everyday use
19:59.47cp15But I assume about 100 hours
20:02.00cp15JuNuKN: cr2 is never in irc, he said it makes him addicted
20:04.40JuNuKN100 hours ! Great! That should be enought for most days! Gratulation!
20:05.17JuNuKNAny othe progresses, that I missed in the logs? ... They are very small meantime!
20:05.33JuNuKN<PROTECTED>
20:06.14cp15Yes, I am also a bit disappointed. There are really few people which actually want to help the project
20:06.50cp15But llange does a good job, it is just not that easy to get everything working
20:08.41cp15Oh wow, the logs form yesterday are containing only me and oho saying good morning and a unknown person saying yes
20:10.22JuNuKNYes I know! Its clear to me, and I admire all of you!
20:10.34bhimaGreat job.
20:11.17cp15Ah, bhima, as a usability specialist, maybe you can give me some help
20:11.37cp15I have speeded up the suspend and resume of the device, but now something unwanted happens:
20:11.53bhimaOk. First, look at WinCE. Make a careful list of what they do. Now, for each item, figure out the best approximation to the opposite of it. Do that. :)
20:12.25cp15If the device is suspended, and you press the power button, it will wake up, get an event that the power button was pressed and, because of that, will go to sleep again
20:12.43bhimaAhh, I see. Amusing problem.
20:13.11cp15Well, I could just ignore the power button press after wakeup, but maybe the power button could have a different function whether pressed long or short
20:13.31cp15Do you think we need such a differentiation?
20:13.46bhimaI think it would be useful to do that, yes.
20:14.07bhimaI think having a flag for "device was asleep" might be one hack - not sure how the key events are handled.
20:14.27bhimaIf you can easily add that flag to the event, then you wouldn't really need that much special handling.
20:14.30rob_wsorry to "throw in " but  i believe you should "discharge" that event
20:14.52bhimaDo you know why you woke up or not...?
20:15.00cp15bhima: Yes, I know
20:15.19cp15There is a register which shows an active bit for the wakeup source
20:15.38cp15But I think I know a solution which is pretty and easy
20:15.56cp15I just send a different key if the power button was used to wake up the device
20:16.00bhimaOk, good. So, it depends on the architecture and where you want to make this sort of decision. I would tend to think that a sleep flag would let your code work well on other devices.
20:17.05bhimaHaving it send a different button based on it being the power switch means that code knows which switch is power.
20:17.14bhimaSo it depends on where this code is.
20:18.04cp15It is actually pretty easy. There is a programm called keylaunch, which as a config file where you can define keysyms and actions which get executed when they were pressed long or short
20:18.22cp15For the power button, it is just apm --suspend
20:19.00cp15Ah, there is even a appropriate keysym defined: XF86XK_WakeUp
20:19.05rob_wi think you should discharge the event in a blueangel specific driver , in order to dont mess with to high level gpe apps
20:19.55cp15Hmm... When keylaunch grabs the key, other applications never see it
20:20.50rob_whmm , hard to think that through for me ?!?
20:21.15bhimaSo, right, the real question is: where is it appropriate to put BA-specific stuff in?
20:21.39cp15And maybe the differentiation between long and short press is really useful. Example: Short Press to power on the device, Long press to change between silent and loud alarm
20:22.30bhimaOr maybe I want it to do something when I turn it on with the power button vs. other buttons?
20:23.38rob_whehe , i just have a similar issue on my tablet pc  . Coming back from S3 ( which works great ) then issues a shutdown ,
20:23.41cp15Yeah, that is another think... So maybe the key which woke up the device doesn't get another meaning, but something like a shift state
20:25.28bhimaWell, conceptually it really _is_ a modifier.
20:25.47bhimaThat, IMHO, is an accurate description of it.
21:03.56rob_wcp15, what app did you saw triggers the battery all the time ?
21:04.56cp15keylaunchd, but it does only occur with a newer kernel, not yet in cvs
21:05.41rob_wi meant the bug that the battery module cracks the touchscreen
21:06.26cp15Ah, ok... Just a moment
21:06.50cp15ipaq-sleep
21:08.28rob_whmm , true
21:09.23cp15BTW: I don't think there is a really exact way to get the battery charge
21:09.32rob_w? why
21:09.41rob_wmy battery says 2275
21:09.47cp15I am assuming WM does really count the charge and discharge time
21:10.24cp15Because last time I booted WM with a battery fully charged (from linux) and it said it is 0% full
21:10.46rob_wwell , that would be strange
21:10.57rob_wnormally battery`s at least say their percentage
21:11.17rob_wbut i even saw some reporting actual rates
21:11.50rob_wnow the touch is broken again , although i didnt had ipaq-sleep running and it did work for some minutes
21:12.22cp15Maybe there is another process which looks into /proc/apm
21:12.31cp15The gpe battery icon?
21:13.06rob_wit really feels like we do overtress something ,, as it kinda works again and then after some stressfull touching it stops again
21:13.15rob_wno icon running !
21:15.57cp15Hmm... Then I don't know
21:17.19rob_wi saw you fixed a "usb bug "
21:17.50cp15Not yet fixed... But I know the cause
21:20.10rob_wllange send me patches which included something from you
21:20.49cp15Ok, this is a himalaya specific thing... But it might also occur on BA
21:21.07cp15But I don't know if the fix is really right this way
21:23.24rob_whmm feels like usb is not working for me anymore  * before these patches even !
21:24.25cp15Was usb connected during boot?
21:25.27rob_wno
21:25.41cp15Then it doesn't work
21:26.30cp15If your touchscreen works, do a rmmod g_ether ; rmmod pxa2xx_udc ; modprobe g_ether
21:26.49rob_wi just did only a modprobe g_ether
21:31.38rob_wnice , it sleeps apart if i connect it on the docking
22:00.10rob_wone last question ,, how is your writing to the doc going , cp15  `?
22:00.21cp15Not at all
22:01.06cp15I hoped I can use the doc drivers in the kernel, but they are only for doc m+. The BA has a doc g3, which is a successor of doc m+ and uses a different protocol
22:01.26rob_wdo we have specs ?
22:02.09cp15Even worse, the error correction algorithm has been improved for g3, this will be the hardest part
22:02.47cp15No, not really useful ones. There is a datasheet on the m systems web page, but it is much too high-level
22:02.49rob_wso we have specs but they have probiatory parts ?!?
22:03.51cp15Yes, and the propritary parts make about 90%
22:04.17rob_wnice
22:05.11cp15But the driver is contained in the boot loader, in a windows dll and in a linux kernel module, so 3 possibilities to reverse engineer
22:05.40rob_whehe , so they have a precompiled module
22:06.11cp15Actually cr2 found one in the motorola firmware. But it is for kernel 2.4 only
22:07.22cp15Maybe we can get around with the older error correction, making it incompatible to windows mobile
22:08.21rob_woh oh i hear my kids
22:10.05rob_wwhy does "Enter pin " never dissappear
22:10.41cp15Is it a bug or a feature? Ok, it is a bug
22:11.00rob_wnot wanna blame here !
22:11.15rob_wi am the one who is to blame first
22:11.29cp15It should be replaced by "Registering on Network..." and when registering is complete, by the operator name
22:12.32cp15The problem is that the registering usually takes longer than the command timeout and so displaying the operator name doesn't work
22:12.57cp15If you quit gomunicator and restart it, the operator will be displayed
22:13.24rob_wwow
22:13.44rob_wfirst time i see that !
22:19.11rob_wcp15, why dont you wanna check in some new code , that could help me get motivated again
22:21.18cp15Why doesn't somebody else check in some new code, that could help me get motivated again
22:21.52cp15But ok, I was planning to commit my kernel changes today
22:22.19cp15Some cleanups, left leds working, deep suspend partially working
22:22.26rob_wyeah that is the proplem of beeing "the best " at one point !
22:22.54cp15That has nothing to do with being the best
22:23.07rob_wi never thought i would actually use the phone stuff before march 06 ,, and look at it now !!!
22:23.41cp15Immolo disappeared over a week ago, he promised to make the gpe image... Well, there was a first cut, but it has to be improved
22:24.22cp15bwegh was also not here for now over a week. He wanted to make an installation system
22:25.33cp15Of course you could say: It takes more time. But if the BA linux is useable only in 2 years, it is worthless since the BA will be an absolutely obsolete device by then
22:25.34rob_wyeah, well what we need more drivers
22:26.12rob_wwell look at it from the other side ,
22:26.45rob_wso it was for all the time with gpe ... only recently they where able to pickup in the time frame
22:27.09cp15Yeah, and compared to opie gpe is still a piece of crap
22:27.12rob_wand now that nokia almost used gpe on a hot product it gets better
22:28.06rob_wunfortunately opie is using qt  ..
22:28.48cp15More unfortunate I find its strange storage concept with xml files
22:29.00cp15But I have read recently that you can also choose sqlite now
22:29.40rob_wit defenitly uses "nicer" grafics
22:29.53cp15And the applications are much more complete
22:30.53cp15And much faster also
22:31.28rob_wstill i like gpe and his x-server setup
22:31.46cp15Yeah, the x-server might be an advantage
22:32.21rob_walso the gpe people always mention the "Trolltech " problem
22:32.25rob_wgpe is really GPL
22:32.54cp15And dbus is probably much better than qcop
22:33.39rob_whehe yet all the little linux mobile stuff is nice  *geek*
22:35.05rob_whehe , you probably dont know that , but when you restore and delete and restore and delte wm2003 , your mail folders go crazy ... that is really odd new for ms
22:37.13rob_wwe should be able to "force " the source for the doc g3 from motorola , dont you think ??
22:37.43cp15No, they are within GPL
22:38.06rob_wso then we should have it ?!
22:38.44cp15No, Linus Torvalds allowed binary only modules to be loaded into the kernel
22:39.34rob_whmm is that the same way to load firmwares like the ipw2200 or our wifi ?
22:40.29cp15No, firmware are a different thing. Even GPLed drivers have binary firmware
22:41.13rob_wah right
22:43.45rob_wok i am tired ,, cp15 dont be frustrated !  u are the biggest mover here !   and if the BA is outdated then we met again at another device !
22:44.12cp15Which will take just as long
22:44.49rob_wmaybe , maybe not , it feels for me that linux gets better each time
22:45.13cp15Maybe... But the devices are also getting more complex each time
22:45.33cp15The BA actually is more complex than an old PC
22:45.45rob_wyeah
22:47.46rob_wi also felt like linux is too "slow " for such device life time `s ... and there is alot ppc crap flying around already so you could serve yourself with that ,...still the linux idea attracted me again and startet the chat and so on here !
22:49.14cp15Well, lets see... Maybe I will meet some developers on the CCC
22:49.22rob_wyeah !
22:49.42rob_wget some funding !!!  i gladly spent my carrier on these things !
22:50.27cp15Funding isn't that much of a problem. BA development costs me already some 1000 Euro each month
22:51.00rob_w?? really
22:51.07rob_win your time ?
22:51.23cp15Yeah, every time I work for the BA I can't work for a customer
22:51.35rob_whehe that is different !
22:51.46cp15No, it isn't
22:52.08rob_wwell , would you pay me to work on it the whole day ?
22:52.20rob_w*knowing that i not as good as you !
22:52.31cp15How much?
22:52.34cp15How good?
22:53.23rob_whmm beeing able to concentrate on it < 6 hours a day would be great and healthy for my coding !
22:54.16cp15And how much money do you want?
22:56.23rob_wi am a whore !!
23:20.52*** join/#htc-blueangel cp15__ (i=yvigud@p549A377A.dip0.t-ipconnect.de)
23:21.03*** part/#htc-blueangel cp15__ (i=yvigud@p549A377A.dip0.t-ipconnect.de)
23:29.07*** join/#htc-blueangel cp15_ (n=549a377a@www.it-schaller.de)
23:29.39*** part/#htc-blueangel cp15_ (n=549a377a@www.it-schaller.de)
23:36.05cp15Good night

Generated by irclog2html.pl by Jeff Waugh - find it at freshmeat.net! Modified by Tim Riker to work with blootbot logs, split per channel, etc.