IRC log for #htc-linux on 20080411

00:06.57psionfenixhave you played with it on your kaiser yet?
00:08.23mercmessing with it now, someone just released a new img that has keys mapped so you can navigate menus
00:11.20psionfenixnice
00:12.25Kevin2~pong BabelO
00:12.26aptBabelO: PONG!!!!!!!!!!!!!!!!!!
00:12.33psionfenixi should have bought a damn kaiser.
00:12.55psionfenixi dont get reva where i'm at anyway
00:13.12Kevin2(RTT 4 hours 9 minutes 27 seconds)
00:15.02Kevin2psionfenix: What phone?
00:15.26psionfenixtitan
00:15.35Kevin2Do you know where titan has a cpld located?
00:16.25psionfenixerm...the memory map shows it at  0x90000000
00:18.50Kevin2Okay - one typically does an mmutrace on the range the cpld is thought to be at.
00:19.08Kevin2Once you understand the pattern of accesses, one can add the specific registers to GPIOS and IRQS.
00:19.44psionfenixmmutrace as opposed to dump mmu?
00:20.46Kevin2Yes.
00:22.33psionfenixok
00:22.51psionfenixso command is   mmutrace 0x90000000
00:23.32psionfenixer, unknown command
00:24.35Kevin2No, you can see some examples at: http://www.handhelds.org/moin/moin.cgi/HaRET_20Documentation
00:24.54Kevin2It would be "addlist mmutrace 0x90000000 0x40"  and then "wirq 1".
00:26.24psionfenixhmm
00:26.30psionfenixreturned a whole bunch of the same
00:27.08mercquestion, how to I commit a diff file using git?
00:28.09Kevin2merc: git commit
00:28.55mercwow so ez
00:31.44Kevin2cr2: haret boot bundles work for me.
00:31.53Kevin2make linload KERNEL=../../kernel/herm-ltg/arch/arm/boot/zImage
00:36.43psionfenixscratches his head.
03:15.39*** join/#htc-linux dzo (n=dzo@121.98.128.127)
03:15.50psionfenixi just got hit with a ton of errors trying to rebuild the kernel :/
03:19.35dzoHi all
03:19.45psionfenixhey dzo
03:20.04psionfenixim too determined for my own good, just so we know.
03:20.23dzohow are you getting on with titan kernel.
03:20.34psionfenixsame old.
03:20.39psionfenixA N D R O I D _
03:20.41psionfenixthat's it
03:21.36dzoReally the only option is to get the filesystem into flash and run from there I think.
03:21.54dzoThe msm_nand flash driver works I think.
03:22.05psionfenixhow would i go about that? (new to porting, still)
03:23.46dzoif you could get a contiguous section of flash memory allocated with fs image and give the kernel the offset, then it may work. Not so easy though.
03:24.12psionfenixyeah. i'd probably brick it somehow
03:24.17psionfenixwith my luck recently
03:25.33dzoI had a look at the sd hardware but a driver should go into android kernel some time so best wait for that.
03:26.15psionfenixyeah
03:26.16psionfenixmake ARCH="arm" CROSS_COMPILE="/bin/arm-none-linux-gnueabi-" menuconfig
03:26.22psionfenixwhy is that destroying me...lol
03:26.54dzois the toolchain installed OK?
03:27.07psionfenixyeah
03:27.16psionfenixshould be, i guess i could try it again
03:27.32dzowhat error does it give?
03:29.36psionfenixa ton
03:31.05dzook, keep trying, I'm off, work to do...
03:31.05psionfenixtake it easy
03:34.34psionfenixmake[1]: *** [scripts/kconfig/lxdialog/checklist.o] Error 1
03:34.34psionfenixmake: *** [menuconfig] Error 2
03:34.37psionfenixthat's the last two...
03:34.40psionfenixoh...
03:56.23psionfenixwell i give up.
03:56.40psionfenixi'll just wait and buy a new phone that can run linux.
04:38.18Marexapt-get install ncurses-dev ?
04:40.19psionfenixtried it
04:40.23psionfenixthanks though :)
04:46.45*** join/#htc-linux goxboxlive (n=goxboxli@195.159.97.196)
04:52.29Marexpastebin whole errorlog
04:57.24psionfenixk
05:00.00psionfenixhttp://pastebin.com/d7329f5d8
05:03.59Marexscripts/kconfig/lxdialog/dialog.h:32:20: error: curses.h: No such file or directory
05:04.04Marexisnt it selfexplaining ?
05:04.27Marexeither use make config or install ncurses headers (and libs of course)
05:06.35psionfenixyeah, i went with the make config
05:06.51psionfenixsorry, i'm just not seeing things as i should anymore. been screwing with this thing for way too long
05:15.18Marex7am here ;-)
05:15.34Marexjust have some sleep and start again tomorow ;)
05:21.47psionfenixbut i want to give up. lol
05:22.06Marexhm ok ...
05:25.21psionfenixbut now that it's going i can't.
05:51.18*** join/#htc-linux kiozen (n=kiozen@p54929B1F.dip0.t-ipconnect.de)
06:21.01dcordeshi
06:21.09dcordesand bye ^^
06:46.28*** join/#htc-linux Marex (n=marex@vasut.kolej.mff.cuni.cz)
06:58.51*** join/#htc-linux rmoravcik (n=rmoravci@gtsgw.ttc.cz)
07:26.54*** join/#htc-linux marbalon (n=marbalon@212.122.223.98)
07:34.49psionfenixmarbalon.
07:34.50psionfenixi love you.
07:39.21marbalonpsionfenix: why ?
07:39.31psionfenixyour initrd gave me something seemingly workable
07:39.52psionfenixwell, partly. gives me a sh error, but then drops me into bash (it looks like)
07:42.11marbalonyes  there are some errors but no critical
07:42.51psionfenixwell, i dont get a gui or anything
07:42.52marbaloni didn't fix it becouse i dont need linux working
07:43.02psionfenixand my letters type gibberish.
07:43.09psionfenixbut i'm going to work on that (hopefully)
07:43.12marbaloni need only printk from kernel
07:44.04marbalon(sorry for my english)
07:46.27psionfenixno problem man, i can understand you
07:47.07marbaloni hopeso
07:47.11marbalon;)
07:47.59marbalonthere is a problem with keyboard matrix
07:48.16psionfenixa problem in general? or just in your initrd?
07:48.33marbaloni thought when i prex one button 2 gpios should change
07:49.29marbalonthe problem is in kernel gpio module
07:51.00marbalonthe gpio_input driver shows only one gpio change
07:53.54marbalonmaybe "PXA27x matrix keyboard" in htc universal will be helpfull
07:54.55psionfenixyeah
07:58.06marbalonhave you no problem with start initrd-android ?
07:58.40marbalonit start few times on my device but now hangs every time :(
07:59.34marbalonproblem is a huge initrd file but why sometimes its starts sometimes not ???
07:59.58*** join/#htc-linux marbalon (n=marbalon@212.122.223.98)
08:00.56psionfenixi was never able to get the initrd-android to work
08:01.00psionfenixi'm using a titan
08:01.07psionfenixnot enough ram. hah
08:09.05marbalonone moment i need restart app
08:14.05*** join/#htc-linux marbalon (n=marbalon@212.122.223.98)
08:15.50*** join/#htc-linux marbalon (n=marbalon@212.122.223.98)
08:45.58*** join/#htc-linux marbalon (n=marbalon@212.122.223.98)
08:47.58*** join/#htc-linux marbalon (n=marbalon@212.122.223.98)
08:48.13psionfenixis it normal to have semi different numbers when you push buttons during the watch gpios process? (even if same letter)
08:48.34*** join/#htc-linux marbalon (n=marbalon@212.122.223.98)
08:52.59*** join/#htc-linux marbalon (n=marbalon@212.122.223.98)
09:06.46*** join/#htc-linux raph_ael (i=raphael@kikoolol.orbus.fr)
09:07.56raph_aelhi
09:49.36*** join/#htc-linux skodde (n=skodde@unaffiliated/skodde)
09:56.50*** join/#htc-linux BabelO (n=Fabrice@unaffiliated/babelo)
10:02.16*** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) [NETSPLIT VICTIM]
10:02.16*** join/#htc-linux raph_ael (i=raphael@kikoolol.orbus.fr) [NETSPLIT VICTIM]
10:02.16*** join/#htc-linux marbalon (n=marbalon@212.122.223.98) [NETSPLIT VICTIM]
10:02.16*** join/#htc-linux cr2 (n=konversa@crpl6.physik.uni-wuppertal.de) [NETSPLIT VICTIM]
10:02.17*** join/#htc-linux univac (n=univac@xmms2/user/univac) [NETSPLIT VICTIM]
10:02.17*** join/#htc-linux Dindon (i=Euroserv@193.27.79.0) [NETSPLIT VICTIM]
10:02.17*** join/#htc-linux psionfenix (n=fenix@host-69-146-196-11.chy-wy.client.bresnan.net) [NETSPLIT VICTIM]
10:02.17*** join/#htc-linux JEEB (n=kanakana@a88-112-94-153.elisa-laajakaista.fi) [NETSPLIT VICTIM]
10:02.17*** join/#htc-linux _who (i=who@194.145.250.184) [NETSPLIT VICTIM]
10:02.17*** join/#htc-linux swetland (n=swetland@sunnyvale.frotz.net) [NETSPLIT VICTIM]
10:02.17*** join/#htc-linux dase (i=dase@p1atin.de) [NETSPLIT VICTIM]
10:02.17*** join/#htc-linux tg (i=tg@digit.drk.hu) [NETSPLIT VICTIM]
10:02.17*** join/#htc-linux PCrazee (i=pcrazee@static.88-198-43-214.clients.your-server.de) [NETSPLIT VICTIM]
10:02.17*** join/#htc-linux toi (n=pleemans@d5152D3B4.access.telenet.be)
10:02.17*** join/#htc-linux Raven_coda (n=wtg@64.50.35.215)
10:12.40*** join/#htc-linux rob_w|laptop (n=rob@M9553.m.pppool.de)
10:20.43*** join/#htc-linux marbalon (n=marbalon@212.122.223.98)
10:37.14*** join/#htc-linux dcordes_ (n=snoopdog@f054217184.adsl.alicedsl.de)
10:45.41*** join/#htc-linux Nickbee (n=Dheeraj@122.163.43.35)
10:50.00*** join/#htc-linux patoche (n=patoche@77.200.40.21)
10:51.27BabelOhi
11:07.42BabelOgoxboxlive: around ? ;)
11:41.58*** join/#htc-linux marbalon (n=marbalon@212.122.223.98)
11:50.58*** join/#htc-linux marbalon (n=marbalon@212.122.223.98)
12:02.28*** join/#htc-linux Marex (n=marex@vasut.kolej.mff.cuni.cz)
12:31.43dcordes_marbalon: do you have matrix kbd working on htckaiser?
12:32.24goxboxliveBabelO: yes, for few sec. Im about leaving work
12:32.54dcordes_what is a softkey speaking of htc device buttons?
12:33.17BabelOgoxboxlive: no, i found, it is ok now, i just upload a clear live image :)
12:33.35goxboxliveok
12:36.03marbalonsome of softkeys use the same gpio like kaybord, so i think it's include in keyboard matrix
12:36.33marbalonbut seme buttons have own gpio
12:37.16dcordes_did you check Kaiser_CPLD on xda wiki?
12:38.53marbalonyes
12:40.38dcordes_good
12:41.01*** join/#htc-linux marbalon (n=marbalon@212.122.223.98)
12:41.54marbalondcordes_: but i didn't find anything keyboard related
12:43.02*** join/#htc-linux marbalon (n=marbalon@212.122.223.98)
12:43.22dcordes_marbalon: http://wiki.xda-developers.com/index.php?pagename=Kaiser_IRQ there I sorted out all irqs I could find
12:43.59*** join/#htc-linux marbalon (n=marbalon@212.122.223.98)
12:45.39marbalondcordes_:this is more interesting if we use kbd irq combined with gpio we have keyboard working ;)
12:45.59*** join/#htc-linux marbalon (n=marbalon@212.122.223.98)
12:46.14marbalondcordes_:this should be very similar to HTC Univerrsal
12:46.59*** join/#htc-linux marbalon (n=marbalon@212.122.223.98)
12:50.41dcordes_marbalon: did you checkout universal keyboard source?
12:50.48marbalonyes
12:51.19dcordes_hillarious: the probably most important thing in android works: popping out the wallpaper menu ;)
12:52.45dcordes_marbalon: did you try accesing the nand?
12:54.09marbalondcordes_:no, I started enjoy with linux+kaiser yesterday evening ;)
12:54.53marbalondcordes_:why do you need access to nand ?
12:55.06dcordes_I'm 'enjoying' it since november 2007 that's when I got it and started investigating its guts with the help of all the brilliant people in this channel
12:55.32dcordes_marbalon: to boot a filesystem from it?
12:55.59raph_aelthere is still no installable images of android ?
12:56.34marbalondcordes_:there is a kernel driver for wince filesystem ?
12:57.15dcordes_we could probably boot a loop filesystem with this?
13:00.11marbalonsorry, i don't understand, I thought we use nand instead mmc (no driver till now)
13:00.21*** join/#htc-linux CVirus (n=GoD@82.201.178.99)
13:00.44dcordes_marbalon: is that what dzo hosted image does?
13:00.50dcordes_sorry
13:02.03*** join/#htc-linux Sliss_ (n=chatzill@212.71.15.163)
13:03.25Sliss_friday afternoon... lets try to boot my htc back into linux ... is openmoko working?
13:04.28Sliss_htc blueangel btw
13:04.59dcordesSliss_: qtopia is cureently best working 'available' image for BA
13:05.12Sliss_ok
13:05.34dcordesSliss_: I'm working on an openmoko image but there are GUI problems I can't fix
13:05.56Sliss_really bad ones of doable?
13:06.15dcordesSliss_: nothing you want to try available, sorry
13:06.25Sliss_i see ;-)
13:06.36dcordesyou should checkout the qtopia image though
13:07.01dcordeswhich is really working great
13:08.09dcordesSliss_: http://linuxbox2.free.fr/blueangel/qtopia/
13:10.49Sliss_downloading...
13:12.47Sliss_dcordes: i do not need the 2 folders right?
13:12.53dcordesright
13:13.54dcordesyou need to have a fat partition as first partition ony your SD card where you put zImage and startup and a second partition in ext2 where you extract the filesystem (Angstrom-Qt...)
13:14.45BabelOSliss_: just copy both Qtopia* file on top of your SDcard and verify you ve at least 200Mb free http://linuxtogo.org/~htcpxa/htcblueangel/Qtopia/
13:14.54BabelOno partition needed
13:14.58dcordesoh :)
13:15.56dcordesBabelO: did you change anything since last image?
13:15.57BabelOi m currently testing my messenger applciation :)
13:16.18Sliss_aah... so is need to remove the partitions then?
13:16.50BabelOdcordes: no, maybe soon this afternoon you get a new one
13:17.15BabelOSliss_: with the .exe you only need a Fat partition
13:17.55dcordesSliss_: wouldn't matter how many partitions of any kind you have afterwards though ;)
13:18.29Sliss_nice speed ;-)]
13:19.35dcordesSliss_: are you trying to insult BabelO's webspace?!
13:20.17dcordesjust kidding ;p
13:20.28Sliss_insult? I am happy about 900k/s
13:20.30Sliss_;-)
13:21.11marbalondoes anyone have a msm7200 datasheet?
13:23.28dcordesmarbalon: no
13:23.55dcordesnot unless you sign NDA and get one from qualcomm
13:24.10dcordesmarbalon: did you look at usb yet?
13:24.41marbalondcordes: no, but will be usefull for rndis
13:25.22Sliss_so both files are on the sd card. just boot?
13:26.31dcordesmarbalon: that's why I'm asking
13:26.44dcordeswifi would also be but that's iirc SD which we don't have
13:26.49BabelOSliss_: yes start th exe file and wait
13:27.00dcordess/SD/SDIO
13:27.07Sliss_my windhoos is first installing :/
13:27.35dcordesgood old wm3 configure-on-boot :D
13:27.42Sliss_s/SD/SDIO/  ;-)
13:28.02dcordesSliss_: nevermind it's about the statement not the bot fun
13:28.07dcordes~lart Sliss_
13:28.07aptgives Sliss_ a good seeing to
13:28.16dcordes~lart apt
13:28.16aptresizes dcordes's terminal to 40x24
13:29.34Sliss_great.. the 200mb root file is in creation
13:30.40dcordesmarbalon: dzo told me usb would work for him
13:30.44dcordes(on vogue)
13:31.36BabelOSliss_: i notice some bugs in this image already
13:31.43marbalonwow maybe it's similar to Kaiser
13:32.57dcordesmarbalon: I connected usb to kaiser but no reaction from host visible
13:33.37Sliss_bugs are normal... but how long does this creation take?
13:36.19marbalonyes i know that, but USB working is in currently posted path ?
13:37.04dcordesmarbalon: can you point me to the forum thread you refer to?
13:37.56marbalonhttp://forum.xda-developers.com/showthread.php?t=383695&page=10
13:38.29dcordeslinux/arch/arm/mach-msm/board-htcvogue.c <- there are definetly usb related modifications to this file from current patch ( i fetched patch 5 mins ago)
13:38.38Sliss_BabelO>i have the feeling the root creation hangs...
13:39.14marbalonits sounds good, see you in the evening i'm @ work, now -> home
13:39.28dcordesmarbalon: alright see you
13:39.51BabelOSliss_: no it is very long
13:40.07Sliss_BabelO: ah ok... like an hour?
13:40.42Sliss_what is it doing? compiling gentoo style?
13:40.44BabelOSliss_: no, like 10 mn ;)
13:41.09BabelOSliss_: no, it  just unpack tar.bz2 file into 200Mb loopback file
13:41.25BabelOSliss_: and i think i va cpufreq activated ;)
13:42.08dcordesBabelO: you know what FIQ is?
13:42.46BabelOdcordes: no, where do you see FIQ ?
13:43.04dcordesBabelO: configuration of kernel for kaiser
13:43.05dcordesEnable installation of an FIQ handler. (MSM_FIQ_SUPPORT) [N/y] (NEW)
13:43.38Sliss_btw I need the 200 mb free on the phone or the card?
13:45.16BabelOSliss_: on the card, and you need also /Storage Card not a translation
13:45.24dcordesit's really nice seeing tux on kaiser after all the hassle :D
13:46.09dcordesBabelO: why is /Storage Card hard coded?
13:47.14Sliss_i guess my phone os broken now... after a softboot windows installation isnt starting up anymore
13:47.55Sliss_the screen light is comming up the first time
13:48.20Sliss_and I head the pling-pling but then nothing
13:48.25BabelOdcordes: because it is in haret, and we need to found a wince command or environmùent variable that get it
13:48.35BabelOmaybe we need to read simply the registry ;)
13:50.25Sliss_someone did get this problem before?
13:51.28BabelOSliss_: yes screen light come down sometimes in wince, long press on power button, or use a light to look at lcd
13:52.56Sliss_no .. there is no wince yet... it want to install so I have to tab the corners copy paste etc... but the screen stays white
13:53.05Sliss_bright white
13:53.31dcordesBabelO: Sliss_ after the calibration?
13:53.47dcordesSliss_: i mean
13:53.51Sliss_it doesnt even get that far
13:55.10dcordesdid you power it off for a short time? My is often better after I remove the battery and ac power for a while
13:55.19dcordeslet it lay like that few minutes and retry
13:56.39Sliss_i hope its not broken now...
13:57.18Sliss_I still have an other one with a broken screen but it is not easy to openup the device
13:59.55Sliss_shoot.. its not good.  after a few minutes I started it again i get a white screen and after a few sec a little darker one (at the time the calibration normally starts)
14:01.18*** join/#htc-linux SmallR2002 (n=SmallR20@79-77-91-161.dynamic.dsl.as9105.com)
14:02.20dcordesSliss_: you must have done something seriously wrong
14:02.56Sliss_I guess.. but i do not believe I did
14:04.42Sliss_the BabelO's exe file was extracting and hang at one monent. After 10 min I did put the device in its dongly for ac power... and linux comfirmed that I did... thats all
14:07.19dcordesno idea I have a BA but I didn't break it yet, sorry
14:07.30dcordesonly thing I can say is the battery thing
14:07.43dcordesBabelO: http://pastebin.ca/981016 <- can you look
14:08.31dcordesI get undefined reference to `msm_proc_comm' en lines 85 92 98
14:11.01Sliss_ok i am going now... a little sad without phone ;-) so have a good weekend
14:11.31Sliss_I will leav the phone without battery for a little and try again tonoght
14:11.36Sliss_later...
14:12.59dcordesthat's really frustrating. hope this guy will get it working again.
14:13.33dcordescr2: ping
14:47.47BabelOdcordes: it is just a reboot needed, on mine with an old wm5 image, i often lost the backlight, each time i ve to look at it better
14:49.51dcordesBabelO: I leave just the old wm3 rom on my BA
14:51.51BabelOdcordes: mine was due to backlight settings too low on wm5
14:52.23dcordesmaybe that is sliss' problem
14:57.24*** join/#htc-linux TimRiker (n=timr@68.27.25.129)
15:05.04*** join/#htc-linux rob_w|laptop (n=rob@M9553.m.pppool.de)
15:07.40dcordesBabelO: could you look at the code I pasted?
15:09.08BabelOdcordes: and ?
15:10.11BabelOdcordes: look in same dir at proc_comm.h
15:12.00*** join/#htc-linux patoche (n=root@lns-bzn-44-82-249-207-137.adsl.proxad.net)
15:12.43dcordesBabelO: ok. what's wrong in proc_comm.h?
15:13.24BabelOfunction msm_proc_comm'  is maybe in
15:13.39BabelOlook at #ifdef around
15:24.01dcordesBabelO: msm_proc_comm is in? what you mean?
15:25.07BabelOdcordes: do you see it in the .h file ?
15:43.00*** join/#htc-linux patoche_ (n=root@lns-bzn-44-82-249-207-137.adsl.proxad.net)
15:47.45dcordesBabelO: yes there is int msm_proc_comm(unsigned cmd, unsigned *data1, unsigned *data2);
15:47.53dcordesand 2 commented lines:
15:48.06dcordes#ifndef _ARCH_ARM_MACH_MSM_MSM_PROC_COMM_H_
15:48.07dcordes#define _ARCH_ARM_MACH_MSM_MSM_PROC_COMM_H_
15:48.16BabelOdcordes: that is ok
15:49.13dcordesBabelO: why do I get the error then?
15:50.53BabelOdcordes: you miss an option maybe ?
15:51.18BabelOdcordes: can you pastebin the h file ?
15:51.34*** join/#htc-linux merc (n=xxmercur@24-181-29-60.dhcp.smyr.ga.charter.com)
15:54.11dcordeshttp://pastebin.ca/981151
15:54.34dcordesBabelO: this is msm git tree w/ dzo's patches for htcvogue btw
15:56.56BabelOdcordes: sorry no idea, i t look like it canot find the function
15:57.14dcordesok thank you
15:57.24dcordesI sent an email to dzo
16:46.59*** join/#htc-linux goxboxlive (n=goxboxli@208.84-48-176.nextgentel.com)
16:47.28*** join/#htc-linux marmotta (n=skodde@unaffiliated/skodde)
16:54.44*** join/#htc-linux CVirus (n=GoD@82.201.178.99)
17:14.18BabelOcr2: linload, liveramdisk works fine
17:16.12*** join/#htc-linux CVirus (n=GoD@82.201.178.99)
17:22.13*** join/#htc-linux diogene31_ (n=rj@mur31-2-82-243-122-54.fbx.proxad.net)
17:26.54*** join/#htc-linux tsdogs (n=tsdogs@net-84-18-144-39.c.bkom.it)
17:27.56*** join/#htc-linux JohnnyK (n=johnnyk@nfx-nat-229.pilsfree.net)
18:01.25*** join/#htc-linux kiozen (n=kiozen@rgnb-4db1df3a.pool.einsundeins.de)
18:07.34BabelOhi kiozen
18:07.46kiozenhi :)
18:23.43cr2kiozen: i've read your GT docs
18:24.21cr2BabelO: strange. does not work for me on n560. something must be wrong.
18:25.17BabelOcr2: did you set correct startup.txt ? with ramboot ;)
18:26.41kiozenhi cr2
18:27.08BabelOcr2: you want jjust ramboot not the liveramdisk ?
18:36.01BabelOcr2: http://git.linuxtogo.org/ i use haret from here
18:48.07cr2BabelO: i use the same setup i've always used before.
18:49.25cr2kiozen: i don't understand several things there.
18:49.51kiozensome others don't either :)
18:49.55cr2kiozen: where does the +towgs84= for ED50 come from ?
18:50.53cr2also, it's a wishful thinking to have a GK grid on a map with numbers and other data.
18:50.56kiozencr2: "pointed me to a document that " document is a link
18:51.17kiozenhm, but a broken one :/
18:51.22cr2afaik it's only done by the military in most countries in the world .
18:52.35cr2kiozen: the +towgs84= business is the most confusing story in the georeferencing.
18:53.20kiozenhttp://maps.la.fh-weihenstephan.de:8080/habitalp/documents/HABITALP-WP9-Technical_Report-CRS_and_Transformation.pdf
18:53.34kiozencr2: yes but it's necessary
18:53.54cr2there is this list http://earth-info.nga.mil/GandG/publications/tr8350.2/wgs84fin.pdf which is referenced by many people
18:54.20cr2then i've saved the similar list from ozi
18:55.25cr2and they differ
18:56.00cr2then there are some +towgs84= values in /share/gdal/gcs.csv
18:56.25cr2but they were autogenerated from the EPSG db, and have bugs too.
18:57.02cr2finally, there is a very interesting list here
18:57.05cr2ogr/ogrsf_frmts/mitab/mitab_spatialref.cpp
18:57.22cr2i think it's the most complete one
18:57.37cr2which makes most sense to me.
18:57.40cr2but
18:58.19cr2the ozi "datum list" includes the datum numbers from it as a first column
18:58.43cr2but the +towgs84= params are not the same as in ogr/ogrsf_frmts/mitab/mitab_spatialref.cpp
18:59.03cr2so in the end we have an unbelieavble mess ;)
18:59.54cr2to import the ozi .map data one probably needs to use the ozi datum_list, but i'm not quite sure.
18:59.56kiozenso..... what do you suggest to make it easy for a noob :) I would say: use google, someone will have ask for the parameters already
19:00.20kiozens/ask/asked/
19:00.35cr2i suggest to stick to the ogr/ogrsf_frmts/mitab/mitab_spatialref.cpp table
19:00.41cr2where possible
19:01.05cr2it's created by clever RE of mapinfo
19:01.15cr2and seems the most accurate to me.
19:01.50cr2at least for the nonwgs84 data i know.
19:04.05cr2and we need to do something about the 2 types of Lambert/Albers conic projections.
19:04.28cr2they are (were) very often used for "lowzoom" maps.
19:06.00cr2kiozen: i think we'd be as simple as ozi, but use the accurate data with clear references.
19:06.23kiozencoeffs from mitab_spatialref.cpp looks good.
19:06.30cr2{28, "European_Datum_1950",         4, -87,  -98,  -121,0, 0, 0, 0, 0},
19:07.01cr2i'd certainly use these values.
19:07.25cr2an 4 is
19:07.27cr2{ 4,"International 1924",                       6378388.0,      297.0}
19:08.55kiozenI wonder if I can access these table in a sane way via gdal api
19:10.35kiozenor I just convert them into a table on the webpage
19:12.11cr2it's a global table
19:12.14cr2MapInfoDatumInfo asDatumInfoList[] =
19:13.02cr2nm /usr/local/lib/libgdal.so | grep asDatum
19:13.02cr200000000006ca8e0 D asDatumInfoList
19:13.32cr2but it's certainly a hack to access it directly
19:13.42cr2but there is no better away.
19:14.05cr2only extracting the table and hardcoding it in the program
19:14.32kiozenthe second problem I see is that a combobox with all these shifts is as scaring to a noob as qgis
19:14.39cr2there is also a MapInfoSpheroidInfo asSpheroidInfoList[] =
19:15.10cr2ozi (de-facto) uses a combobox on the second column
19:15.14cr2of this table
19:15.28cr2with a bit different names
19:16.03cr2since ozi also supports some mapinfo compatibity features, i guess it was the origin anyway
19:16.42*** join/#htc-linux SmallR2002 (n=SmallR20@79-77-5-121.dynamic.dsl.as9105.com)
19:17.56kiozenlet's see, I try to do this list as a combo box...
19:18.59cr2good idea.
19:20.44kiozencurrently cleaning up M, trying to get a status quo between ts and keys
19:21.43cr2yes, it'd be possible to switch between the "modes" with arrow keys, not only with the ts.
19:23.52kiozenwhat modes
19:24.07cr2after you press the "mode" button
19:24.39kiozenyou mean menu?
19:24.41cr2ok, i'll check the latest version
19:24.42cr2yes.
19:25.06kiozenthat would work if the kernel/qtopia/keymaping would be ok
19:25.27cr2ok.
19:32.54kiozencr2: how is the resume working?
19:32.54cr2looking through the heaps of .asm ;)
19:33.17kiozenpiling up  instructions :)
19:33.49cr2my only weekend business :)
19:34.28cr2now i need to find out what's wring with linload.exe too
19:34.44cr2hmm. i've not yet tested it with +2MB zimage...
19:35.01cr2this is standard now.
19:35.21BabelOcr2: did you try the one from git ?
19:36.16cr2BabelO: are there any diffs to CVS ?
19:36.41cr2i see that git is 3 weeks old too
19:37.07BabelOcr2: no more cvs here
19:37.10cr2wow. i use 5 systems now.
19:37.24cr2cvs,svn,monotone,hg and git
19:37.49BabelOcr2: you miss dark :)
19:37.56*** join/#htc-linux k3asd (n=k3asd@host-62-10-242-164.cust-adsl.tiscali.it)
19:38.06cr2and arch
19:38.20kiozenand copy-n-paste :P
19:38.26BabelOlol
19:38.52kiozenour consultant for iso9001 suggested to do it that way
19:39.25kiozennew binary for M, nothing really new, just a clean up
19:40.14cr2i've already booted.
19:40.52cr2BabelO: btw, how we can solve the \Storage Card problem ?
19:42.09BabelOcr2: i m not sure from where it is comming
19:42.54cr2there is some c++ code to read its name from registry.
19:43.10cr2and there is some .inf method for creating .cab
19:43.23cr2but i don't know if the last works with free software.
19:43.40cr2i think we'd document and check both.
19:47.59*** join/#htc-linux kiozen_ (n=kiozen@rgnb-4db1dd9f.pool.einsundeins.de)
19:55.11*** join/#htc-linux SmallR2002 (n=SmallR20@79-77-109-196.dynamic.dsl.as9105.com)
19:57.57cr2kiozen: the icons are a bit too small for n560
19:58.07cr2i guess they are ok on qvga
19:58.47cr2the battery meter shows 2 lines. how does it work ?
19:59.04BabelOdoes not work i think ;)
19:59.42cr2ok. this is also something where we need tome unified /sys approach
19:59.52cr2s/tome/some/
20:01.05cr2the keyboard buttons are not used in the menu too
20:03.04cr2kiozen: if i've selected the map (highlight), how why is the 'map' display not centered on it ?
20:11.05*** join/#htc-linux lpotter (n=ljp@CPE-121-208-61-29.qld.bigpond.net.au)
20:14.44*** join/#htc-linux SmallR2003 (n=SmallR20@79-77-76-79.dynamic.dsl.as9105.com)
20:16.56cr2~ping Kevin2
20:16.57aptpong Kevin2
20:18.16*** join/#htc-linux patoche (n=root@lns-bzn-44-82-249-207-137.adsl.proxad.net)
20:18.43BabelOcr2: you click woth ok button to selectthe maop ?
20:19.01cr2no
20:19.27cr2where is the ok button ?
20:19.49BabelOcr2: center of joypad ?
20:19.58cr2does not help too
20:21.04*** join/#htc-linux wireddd (n=wired@unaffiliated/wireddd)
20:24.03cr2BabelO: how do you compile linload ?
20:24.09cr2make linload KERNEL=~/linux/kernel26/arch/arm/boot/zImage INITRD=/dev/null SCRIPT=n560_2linload.txt
20:24.12*** join/#htc-linux pH5 (n=ph5@e178216153.adsl.alicedsl.de)
20:24.45BabelOmake linloag KERNEL=./zImage SCRIPT=./startup.txt
20:24.50BabelOhi pH5
20:25.02pH5good evening
20:25.48BabelOpH5: if you want to try something else than gpe or openmoko, just take a look at Qtopia for BA ;) i just add a live image
20:26.23BabelOcr2: before i was thinking it does not works, because i ve used boot2 and not ramboot
20:27.19cr2BabelO: the script
20:27.22cr2set RAMADDR 0xa0200000
20:27.22cr2set RAMSIZE 62*1024*1024
20:27.22cr2set CMDLINE "root=/dev/mmcblk0p2 console=tty0 rootdelay=5 psplash=1"
20:27.22cr2ramboot
20:28.01BabelOcr2: mine ->
20:28.03BabelOset CMDLINE "installsize=200 installprefix=QtopiaPhone-image mem=128M rootdelay=2 psplash=1"
20:28.04BabelOramboot
20:28.04cr2oh. it works :)
20:28.21cr2but with this complex +2MB setup.
20:28.49BabelOcr2: resume works too now ?
20:29.17cr2the kernel offset may be a generic pxa option.
20:29.24cr2i'll check
20:30.00cr2<PROTECTED>
20:30.31cr2no. i get the splash and resume into wince.
20:30.51cr2something seems to be broken.
20:31.07cr2the code looks horrible, see the CVS ;)
20:34.31cr2somehow wince sees that the code at +2MB is not a wince kernel :) i must fix it.
20:37.50cr2BabelO: http://handhelds.org/cgi-bin/cvsweb.cgi/~checkout~/linux/kernel26/arch/arm/mach-pxa/looxc550/looxn560_pm.c?rev=1.2&content-type=text/plain
20:39.05BabelOcr2: lol, but it isthe wince code ;)
20:39.38BabelOcr2: do you know what will be the next linux kernel generation ? 2.8 or 3.0 ?
20:40.25cr2i'd be happy about 2.6.25 first
20:40.57cr2although it'd be easy for n560 and athena.
20:41.04cr2they don't use asic3.
20:41.10BabelOcr2: yes me too
20:41.40cr2but ati sd is very similar to asic3_mmc
20:49.51cr2what a pita.
20:50.24cr2it seems if i'll do it, we can flash the kernel into mdoc ;)
20:51.38*** join/#htc-linux lpotter (n=ljp@CPE-121-208-61-29.qld.bigpond.net.au)
20:59.48cr2pH5: are you still here ?
21:02.58*** join/#htc-linux zyxell (n=John@chello089078117046.chello.pl)
21:03.09zyxellhi all
21:06.22zyxelldoes someone with starting linux in Titan on this channel ?
21:10.38pH5cr2: kind of
21:10.44pH5BabelO: http://linuxtogo.org/~htcpxa/htcblueangel/Qtopia/Qtopia-Linux-htcblueangel.exe <-- this one?
21:11.12cr2pH5: i'm completely confused about the resume on n560
21:11.31BabelOpH5: yes, with the bz2 :)
21:11.51cr2it checks if the 0x bit in RCSR is set
21:11.55cr20x4
21:12.26cr2then it loads the SPL to 0xa0100000 from mdoc and jumps there.
21:13.05cr2on universal and other similar devices if the 0x4 bit in RCSR is set, then IPL jumps to "wince"
21:13.08cr2wtf ?
21:13.57pH5interesting
21:14.11pH5but what does the SPL do then?
21:14.30cr2it's a very complex thing.
21:15.46cr2i'll check it again.
21:16.22pH5I guess it checks for RCSR & SMR, too, somewhere and then jumps to wince somehow.
21:17.51cr2hehe. at first it resets the gpio & such one again.
21:18.21pH5during the resume path?
21:18.23pH5ugh :)
21:19.27cr2yes, but it's a usual business in the IPL too
21:19.36cr2omg.
21:19.57cr2then it enables mmu :(
21:20.36*** join/#htc-linux Foxdie (n=foxdiefu@87-194-16-145.bethere.co.uk)
21:20.54cr2it seems than without reflashing it's not going to resume...
21:22.18cr2i'll check the ipl again, but it's unlikely that i can override the SPL loading into ram.
21:25.57pH5cr2: but what is the difference between SMR set and not set in RCSR in the IPL code? in both cases it loads the SPL to RAM and jumps there?
21:28.26cr2it check 0x1 first, then 0x8, and only then 0x4
21:28.59cr2if 0x1 is not set goto splload
21:29.08cr2<PROTECTED>
21:29.25cr2if 0x4 is  set goto splload
21:29.34pH51 is HWR (hardware/power-on reset)
21:29.42cr2then there are some other checks.
21:29.43pH54 is SMR (sleep-exit reset)
21:29.56pH58 is GPR (gpio reset)
21:31.24cr2so if !pwr_reset, !gpio_reset, sleep goto spl_load
21:32.10cr2then it checks if PSSR bit 0x1 is set
21:32.10pH5ok, strange.
21:32.17cr2sleep by PWRMODE
21:32.28cr2if yes, goto spl_load
21:32.50cr2then there are some other RAM checks
21:33.03cr2and after that there is a jump to "wince"
21:33.15cr2i really don't understand it.
21:36.24cr2EnterOS() cold reset
21:36.40cr2EnterOS() warm reset
21:37.41k3asdnight
21:38.04cr2lol
21:38.40cr2verbatim :) "We cannot found wince signal!!\n"
21:39.32cr2fsck.
21:39.50cr2it looks to 0x43454345 at +0x40
21:39.56cr2ECEC ?
21:40.10pH5yes
21:40.44cr2ok, if found it loads 0xa0200000
21:40.51cr2and jumps there.
21:40.54cr2looks good.
21:41.14cr2but i'm not sure what will be broken
21:41.19cr2and the mmu is on.
21:41.23cr2with the wince layout.
21:41.40pH5so it jumps to a0200000 virt?
21:42.02pH5is it a 1:1 mapping?
21:42.05cr2looks strange...
21:42.24cr2hm. after the jump it does mov pc, r0
21:42.29cr2that's very good
21:42.46cr2i need to find the code path which leads to this function.
21:43.09cr2it's really very complex compared to any other device in hh.org CVS
21:43.44cr2ah, it checks the second parameter
21:43.55cr2and disables the mmu if necessary
21:44.06cr2looks like ramboot to me.
21:46.00cr2i'll try to write the ECEC at +0x40 on suspend too.
21:46.54cr2yes, there is an alternative path with 0xac200040
21:47.07cr2so this code  will run with mmu on and off too.
21:47.10cr2bizarre.
21:47.50cr220525555
21:47.56cr220 is space ?
21:48.04cr2looks like ascii to me
21:48.30pH5'UUR '
21:48.38cr2sramupdatemode, and oemupdate mode.
21:48.59cr252444c55
21:49.10pH5sounds like a twisted maze :)
21:49.53cr2yeah
21:50.19cr2it checks many ram locations at +0x0
21:50.33cr2with the SPL loaded at +0x1MB
21:50.38cr2and mmu turned on.
21:50.43cr2really bizarre
21:50.55cr2htc is much more linux friendly :)
21:51.43cr2lol, they are even checking the battery cover gpio at boot.
21:51.50cr2high security mode ? :)
21:53.24cr2maybe i'll convince the SPL that linux is an oem update.
21:55.11cr2looks like a real console app.
21:59.26cr2RDLU
21:59.47cr2ok, it's "ULDR"
21:59.50pH5no, little-endian
21:59.52pH5ah :)
22:00.51cr2update loader
22:01.03pH5magician has the battery cover hardwired to the reset gpio
22:01.33cr2ok.
22:02.54cr2and what is "UUR " ?
22:03.48pH5no idea
22:04.30cr2then "launch minikernel"
22:04.44cr2hmm. looks like the mdoc update stuff.
22:05.26cr2calculate ULDR checksum.
22:07.37cr2ah, the 25304 is a warm reset flag.
22:08.02cr2hm. but i'm preserving it.
22:08.17cr2and the <2MB is not touched now anyway.
22:09.14cr20xb0000004 - 0x07100000
22:10.40cr2a8f00004
22:10.54cr2looks virtual
22:13.04cr28c+4
22:13.07cr2ac+4
22:13.10cr2b0
22:13.29cr2hm. it's at the end of ram.
22:13.32cr2weird.
22:14.32cr2it should be !=2
22:15.53cr2hm. found a bug in my code.
22:16.51pH5can't follow anymore and heads off to dream
22:25.39cr2hm. still resumes to wince
22:37.43*** join/#htc-linux Zappy (n=zappy@f6.nl)
22:49.05*** join/#htc-linux Marex (n=marex@85-132-216-250-eth3-gwfm10-user.802.cz)
22:58.52*** part/#htc-linux Zappy (n=zappy@f6.nl)
23:06.12Raven_codadoes KERN_ALERT always print to console?
23:13.12cr2Raven_coda: depends on your syslog config
23:18.46Kevin2~pong cr2
23:18.47aptcr2: PONG!!!!!!!!!!!!!!!!!!
23:19.46Kevin2(RTT 3 hours 1 minute 51 seconds)
23:38.52*** join/#htc-linux LunohoD_ (n=alex@e180069083.adsl.alicedsl.de)
23:48.00*** join/#htc-linux nickbee (n=nickbee@117.96.69.43)
23:50.25*** part/#htc-linux nickbee (n=nickbee@117.96.69.43)

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.