00:00.13 | tmzt | it's there |
00:00.29 | tmzt | #define EV_VERSION 0x010000 |
00:00.50 | dcordes | ok |
00:02.03 | tmzt | well, theres other reasons check_fd can return -1 |
00:02.13 | tmzt | including if EVIOCGRAB is not supported |
00:02.36 | tmzt | can you put grab_events=0 in /etc/ts.conf module_raw input |
00:03.11 | tmzt | module_raw input grab_events=0 |
00:04.03 | dcordes | set |
00:04.20 | tmzt | try TSLIB_TSDEVICE=/dev/input/event0 ts_print_raw |
00:05.01 | dcordes | is not a ts I understand |
00:08.05 | tmzt | that message is from the if in input-raw.c |
00:08.14 | tmzt | so it couldn't be grab_events anyway |
00:08.26 | tmzt | bad flags in the kernel driver |
00:08.46 | tmzt | does evtest report the driver as supporting EV_ABS? |
00:09.14 | dcordes | supported events? |
00:09.22 | tmzt | I think so. |
00:09.54 | tmzt | it looks like it does. |
00:10.24 | dcordes | Event type 0 (Sync)\ Event type 1 (Key)\ _Event code 330 (Touch)\ Event type 3 (Absolute)\_ Event code 0 (X) |
00:10.46 | dcordes | ... |
00:10.48 | tmzt | yeah, but in the stuff reported before the events |
00:10.51 | tmzt | if you don't touch it |
00:11.13 | dcordes | that is the "Supported events:" section |
00:11.19 | tmzt | okay. |
00:11.23 | dcordes | I didn't touch anything. it's printed after evtest is ran |
00:11.43 | dcordes | I can output a bit, write to file, paste? |
00:11.55 | tmzt | no |
00:12.25 | tmzt | we could try disabling the check in tslib, or at least figuring out which check fails |
00:12.37 | tmzt | it's one big boolean expression |
00:12.38 | dcordes | event type 3 has also event code 1 (y) and 33 (?) |
00:12.58 | dcordes | and below the several event codes more info is printed |
00:13.02 | tmzt | pressure? |
00:13.28 | dcordes | I can see no pressure |
00:16.00 | tmzt | I don't see 33 at all in linux/input.h |
00:16.58 | dcordes | Value Min MAx of 33 are 0 |
00:17.41 | tmzt | pressure is 0x18, it won't work without it |
00:18.06 | tmzt | but it seems to be set in the driver |
00:28.32 | Marex | dang :S mainline asic3 seems kinda broken |
00:32.03 | *** join/#htc-linux gerard (n=gerard@s55932332.adsl.wanadoo.nl) |
00:40.01 | marajin | meep |
00:40.50 | dcordes | merhaba |
00:42.49 | ginge | why cant I boot from the fcking loopback??? |
00:43.33 | ginge | have you guys had any dma errors on sd? |
00:44.10 | marajin | nope.. |
00:44.13 | marajin | not that I saw |
00:44.26 | ginge | youd know about it |
00:44.42 | marajin | then no |
00:49.21 | marajin | do you have some stuff I could test that is likely to cause it? |
00:57.15 | marajin | ginge: ? |
00:57.31 | ginge | marajin: sorry head in code |
00:57.41 | marajin | haha :) it's ok |
00:57.47 | marajin | that's the way I like you best ;) |
00:58.12 | ginge | you can download dzos android image and put it on partition 1 of a fat sd and boot it |
00:58.35 | marajin | are we worried about the fact that this'll be an SDHC card? :p |
00:58.39 | ginge | you need his initrd in the haret dir and the img files on the root of sd |
00:58.47 | ginge | marajin: no its fubar on both |
00:58.54 | marajin | ok |
00:59.05 | marajin | er, can ya link me to dzo's page I forgot to bookmark |
00:59.44 | ginge | firefox 3 man... just start typing random words you remember. http://it029000.massey.ac.nz/vogue/ |
00:59.58 | marajin | I refuse to use it |
01:00.06 | dcordes | I'm also a fan of that function |
01:00.07 | ginge | get off my lawn! |
01:00.28 | marajin | I was annoyed with FF2 at release, FF3 I'm not even going to try for a while |
01:00.31 | ginge | do you brosw the web only with lynx? |
01:00.40 | marajin | Nah I use FF2 |
01:00.49 | marajin | I distrust FF new releases since 2 |
01:01.06 | marajin | they're valuing features over quality IME |
01:01.06 | ginge | I normally have so many tabs open at once that ff2 vomits |
01:01.11 | marajin | heh |
01:01.19 | marajin | My GF is like that, drives me nuts |
01:01.21 | marajin | 100s of tabs |
01:01.31 | marajin | specially cause FF2 leaks memory |
01:01.36 | ginge | I have 173 tabs split over 4 windows :/ |
01:02.20 | marajin | damn tabusers! |
01:02.51 | marajin | I'm not sure I've ever opened more than 30 tabs at once and it's been for periods < 15 minutes |
01:03.20 | marajin | I guess it's much liket he reason I'm a process nazi on windows |
01:03.33 | marajin | I used to have XP down to ~ 17 processes |
01:03.39 | ginge | I have it sorted though. I have a tab plugin that automatically groups tabs together by task. As long as the tab's url is tagged in google bookmarks tabs I get it grouped with my tasks. |
01:04.10 | marajin | she just claims if she leaves the tab open she won't forget about something |
01:04.20 | marajin | which I think's rubbish with a coupla hundrd tabs :P |
01:04.43 | marajin | 's why FF has bookmark folders and "open folder in tabs" options :P |
01:04.54 | ginge | well I generally have a tab for each of the kaiser functions gpio memmap etc. then some for some patches and some interesting links |
01:05.07 | ginge | I can get it all at hand |
01:05.16 | marajin | yeah I guess you vaguely justify it :P |
01:05.35 | ginge | not in the least, but these things kind of keep my tab problems at a minimum |
01:05.39 | marajin | she just spams all kindsa crap that people have linked her to and whatnot and then I get it in the neck when her laptop runs at snailpace |
01:05.49 | ginge | i.e I found a tab open with slashdot from about 1 month ago |
01:05.54 | marajin | hah :P |
01:06.04 | Marex | cr2: ping |
01:06.06 | marajin | I have 3 tabs open atm |
01:08.43 | marajin | anyway |
01:08.47 | marajin | downloading dzo's stuff |
01:09.20 | marajin | ginge: Using your latest zImage I presume? |
01:09.39 | ginge | using the one you are using at the mo please |
01:09.51 | ginge | you only need that one zip file from him |
01:11.37 | dcordes | btw: does kaiser support usb host? |
01:11.53 | *** join/#htc-linux NiZoX_ (n=none@2a01:e35:8a13:a2b0:21c:c0ff:fe25:ff68) |
01:12.11 | marajin | ginge: Hmm ok, last one I have is your sdhctest ? |
01:12.18 | dcordes | ehci_hcd kind of host |
01:12.20 | ginge | ok |
01:12.36 | ginge | dcordes: it does usb OTG |
01:12.38 | marajin | Jollymuffins.. time to fiddle with it then |
01:14.09 | marajin | damnit, must stop putting down the cards, I can never find the bastards again |
01:17.30 | marajin | annnd copying to the kaiser which might take a while |
01:17.55 | ginge | you only need to copy the initrd.gz |
01:18.02 | ginge | leave the img on the sd card |
01:18.20 | marajin | they're going to the SD card but my system won't read SDHC remember? |
01:18.28 | marajin | so I have to use the kaiser to write the card anyway |
01:18.34 | ginge | that sucks |
01:18.51 | marajin | it's kinda annoying that not much WILL read SDHC directly |
01:19.09 | marajin | and yet plenty of phones and embedded devices read SDHC fine |
01:20.03 | dcordes | that reminds me of the need of hc compatible reader |
01:20.09 | marajin | heh :P |
01:20.39 | marajin | The only thing I know that'll read SDHC is my dad's laptop's inbuilt reader |
01:20.47 | marajin | and I doubt he wants waking at 2:30 to steal his laptop |
01:21.32 | marajin | 'bout 60% done... |
01:21.41 | marajin | the kaiser's wifi is kinda sad |
01:23.04 | dcordes | but first I need an sdhc card. 8gb 10Mbit/9Mbit r/w 40⬠|
01:23.17 | marajin | Not too bad |
01:23.19 | tmzt | I had a cheap usb reader, and somehow sdhc worked |
01:23.25 | marajin | tmzt: Lucky bastard! |
01:23.27 | dcordes | I wonder how much speed the kaiser controller does |
01:23.39 | tmzt | it seems they just used the right chip |
01:23.50 | dcordes | somebody in this channel told me it's merely a software problem with the controllers |
01:23.57 | marajin | I get th feeling that SDHC is mostly software |
01:24.09 | tmzt | I never did find out if my microsd usb reader would though. |
01:24.10 | marajin | I remember they made the gp2x read SDHC with a software update |
01:24.37 | tmzt | Yeah, but the usb readers can only translate mass storage (scsi) to sd cmds |
01:24.37 | marajin | i might have a USB reader somewhere.. it may read SDHC I guess but it's a bit pointless now |
01:24.46 | marajin | yeah |
01:25.27 | tmzt | any sdhci type should be able to with software |
01:26.54 | marajin | I'm on top form tonight |
01:26.57 | marajin | lost my bloody stylus |
01:27.41 | tmzt | it has hsusb, but only internal? |
01:27.48 | tmzt | kaiser |
01:32.32 | marajin | ginge: I think it crashed after haret.. will soft reset and retry |
01:34.13 | marajin | aaand it crashed in haret. |
01:34.17 | marajin | one more time.. |
01:34.43 | *** part/#htc-linux exco (n=exco@e181095143.adsl.alicedsl.de) |
01:35.14 | marajin | freezes at the same point |
01:36.48 | marajin | ginge: Seems pretty fucked |
01:37.03 | ginge | really? doesnt crash for me |
01:37.26 | marajin | ginge: freeze with part blanked screen followed by two freezes at the end of the haret load bar.. |
01:37.44 | ginge | with dzos initrd.gz? |
01:37.45 | ginge | wow |
01:37.54 | marajin | make that 3 |
01:37.58 | marajin | it won't even boot |
01:39.34 | ginge | can you try a known working kernel with sd support |
01:39.35 | Marex | marajin: you have it broken :p |
01:39.51 | marajin | ginge: your sdhc kernel IS working with SD support. |
01:40.14 | marajin | ginge: I can boot angstrom/gpe-image/openmoko-image with it (admittedly from partition) |
01:40.30 | ginge | hrm |
01:40.42 | marajin | and it's the only sdhc enabled kernel I have |
01:41.02 | ginge | well... its not your default.txt is it? |
01:41.16 | marajin | maybe but I don't think so |
01:43.28 | marajin | ... |
01:43.34 | marajin | so I maybe it was |
01:43.44 | marajin | it's trying to init modem.. |
01:44.11 | marajin | wtf was that |
01:44.14 | marajin | something just traced |
01:44.22 | marajin | dma yep |
01:44.25 | ginge | fuck |
01:44.41 | ginge | why? |
01:44.44 | marajin | NFC |
01:45.02 | marajin | maybe something based on file size? |
01:45.18 | ginge | hmmmmmm |
01:45.53 | ginge | it dos it varying sectors |
01:45.59 | ginge | so I guess not |
01:47.02 | marajin | mrm.. |
01:47.25 | marajin | Marex: I'm good at breaking things though! |
01:49.09 | ginge | okay, it doesn't seem speed related |
01:56.46 | dcordes | can somebody think of a key combination to swith vts in console image? |
01:56.56 | dcordes | or is that even possible to configure? |
01:57.25 | marajin | presumeably we could |
01:59.12 | dcordes | ginge: do you have an idea why in evtest pressure is not shown (in the zImage with the android optimized ts) |
01:59.30 | dcordes | ? still using the same image |
01:59.45 | ginge | dcordes: maybe is because the pressure is changed for android |
02:00.09 | ginge | if the ts does level detection, it might be below it |
02:01.47 | dcordes | then I should "try module_raw input grab_events=0" in /etc/tslib.conf again with an image that does not change the pressure thing |
02:02.21 | tmzt | camera button? |
02:02.40 | dcordes | what about it? |
02:02.55 | tmzt | are those zImages from htc-msm? |
02:03.12 | tmzt | for switching vts? |
02:03.20 | dcordes | ginge: the "zImage" I'm using is mostly based on the git code, isn't it? |
02:03.20 | tmzt | if its not used for osk |
02:03.26 | ginge | dcordes: entirely |
02:04.59 | dcordes | ginge: do you still have one image with the raw code around? |
02:05.12 | tmzt | have you tried commenting out the check in tslib and seeing if it works? |
02:05.25 | tmzt | plugins/input-raw.c |
02:05.49 | ginge | dcordes: those two links I posted a whle ago |
02:06.06 | tmzt | tscraw and y-fixed? |
02:06.27 | ginge | yeah |
02:06.40 | dcordes | which of them shell I try? |
02:06.41 | dcordes | 00:55 ginge http://www.headfuzz.co.uk/files/android/zImage-tsc2003raw with the 0-320/240 correction |
02:06.44 | dcordes | 00:55 ginge http://www.headfuzz.co.uk/files/android/zImage-tsc2003-corr-raw with just the y correction |
02:06.51 | ginge | right |
02:07.07 | tmzt | do you have a raw ABS_X ABS_Y ABS_PRESSURE? |
02:07.24 | dcordes | tmzt: not in the zImage I have now. not sure which to pick |
02:07.36 | ginge | dcordes: 2003raw |
02:07.38 | dcordes | only have the 1:1 git code now |
02:08.20 | tmzt | from my understanding of reading tslib code, the only thing preventing it from working at all is the check of the driver flags |
02:08.21 | ginge | night all |
02:08.34 | marajin | night ginge |
02:08.40 | dcordes | tmzt: in the raw image I had on kaiser yesterday I'm almost sure I got absolute values for all three x,y,pressure |
02:08.47 | dcordes | night ginge |
02:09.01 | tmzt | but we thought there were too many of them didn't we? |
02:09.20 | dcordes | not too sure |
02:09.24 | tmzt | you didn't get the "I don't understand this ts"? |
02:09.45 | tmzt | but you didn't have ts_print ts_print_raw bins |
02:09.55 | dcordes | oh yesterday I did not try ts_print (didn't havew it yet) |
02:09.58 | dcordes | right |
02:09.59 | tmzt | do you have that zImage? |
02:10.20 | dcordes | overwrote :/ |
02:10.25 | tmzt | thought so |
02:10.51 | tmzt | well, see if you can comment out the if return 0 in input-raw.c |
02:11.19 | dcordes | ok |
02:11.19 | tmzt | well see if these values are ok |
02:11.58 | tmzt | if sane_fd == -1 { return 0; } |
02:12.02 | tmzt | two lines |
02:12.41 | tmzt | in ts_input_read |
02:14.32 | tmzt | well, it looks like EV_SYN tells it to process the event |
02:14.54 | tmzt | do you still get EV_SYN in evtest? |
02:15.04 | tmzt | (did you) |
02:15.52 | dcordes | in the android linear optimized image that prevents pressure from being recognized? |
02:16.32 | tmzt | that must be for drag events |
02:16.38 | tmzt | EV_SYN |
02:16.51 | tmzt | EV_TOUCH should work also, I think |
02:16.59 | tmzt | BTN_TOUCH |
02:17.19 | dcordes | ok I have the bare configured/patched code now |
02:17.28 | tmzt | ok |
02:17.51 | tmzt | will bitbake let you do that or do you need to write a patch? |
02:18.22 | dcordes | you can tell bitbake with the -c switch to which operation it shall work |
02:18.36 | tmzt | so just make then |
02:18.39 | tmzt | ? |
02:18.47 | dcordes | so if you do -c patch it will exit before the do_compile operation starts |
02:18.59 | dcordes | you can change the code |
02:19.05 | dcordes | then do -c compile |
02:19.33 | tmzt | so just skip the patch then? |
02:19.56 | dcordes | not sure if that's possible because it runs the stuff in order |
02:20.09 | dcordes | but I can easily remove certain patches from the meta file |
02:20.25 | tmzt | yeah, that'll work |
02:20.33 | dcordes | which patches do we want to get rid off? |
02:21.06 | tmzt | no, I just wondered if you edited the file would it overwrite it |
02:21.16 | dcordes | nope |
02:21.27 | tmzt | good |
02:21.56 | tmzt | I tried patching the ubuntu kernel and couldn't figure out how to get it to compile the patched modules |
02:22.10 | tmzt | kept putting them back to orig |
02:22.31 | tmzt | mac80211 stuff |
02:22.38 | tmzt | ubuntu has two copies! |
02:23.36 | dcordes | two copies of the kernel image? |
02:23.53 | tmzt | no two different versions of mac80211, one for intel only |
02:24.01 | tmzt | iwl |
02:24.26 | dcordes | I see no "if return 0" in the input-raw.c |
02:24.47 | dcordes | aah stupid me. I bet I sent you the workdir w/o patches applied |
02:25.16 | tmzt | if (i->sane_fd == -1) |
02:25.30 | tmzt | [tab] return 0; |
02:25.40 | dcordes | I see |
02:26.10 | dcordes | so what to change about it? |
02:26.20 | dcordes | remove the two lines? |
02:26.29 | tmzt | comment it out /* */ |
02:27.18 | dcordes | ok anything else to do? |
02:27.25 | tmzt | probably not |
02:28.05 | dcordes | ok which kernel will be best to try it with? 2003raw one? |
02:28.23 | tmzt | yeah, no y-fix stuff |
02:28.38 | tmzt | not the y-fix |
02:29.01 | dcordes | ok I'm downloading that one to kaiser now www.headfuzz.co.uk/files/android/zImage-tsc2003raw (just to make sure) |
02:29.19 | tmzt | ok |
02:41.10 | dcordes | ok I hope I installed it correctly now |
02:47.02 | dcordes | tmzt: what is the file that has the changes from input-raw.c? |
02:47.32 | tmzt | all of tslib, ts_print, ts_print_raw, ts_calibrate, Xfbdev |
02:47.44 | tmzt | /usr/lib/ts/input-raw.so ?? |
02:48.19 | tmzt | not sure of the name |
02:48.41 | dcordes | well OE outputs 3 packages that contain libts so I installed all 3 of them |
02:49.15 | tmzt | libts-bin? |
02:50.16 | dcordes | lol now I don't have the actual lib |
02:50.36 | tmzt | /usr/lib/libts* |
02:51.05 | tmzt | what works on kaiser now? |
02:51.57 | dcordes | nothing just shutdown to install libts |
02:52.18 | tmzt | no, what hw works in linux now |
02:53.15 | dcordes | hm we lack bluetooth wifi cameras |
02:53.23 | marajin | pm |
02:53.24 | tmzt | wifi is sdio? |
02:53.28 | dcordes | yes |
02:53.44 | tmzt | you closer with that now that sd works? |
02:53.58 | tmzt | acx? |
02:55.42 | dcordes | the sdio wifi should be acx supported |
02:55.51 | tmzt | you have a partition on the sd with ext2 right, you don't have to build a new image? |
02:56.02 | dcordes | yes it's on the partition |
03:01.53 | dcordes | tmzt: ok now I'm quite sure I got the lib with the two commented out lines |
03:03.12 | tmzt | alright |
03:03.46 | tmzt | TSLIB_TSDEVICE=/dev/input/event0 ts_print_raw |
03:05.00 | dcordes | selected device is not a touchscreen I understand but it's giving output now. |
03:05.23 | tmzt | right, we didn't remove the check, only the exit early part |
03:05.32 | tmzt | ts_calibrate |
03:05.47 | tmzt | well, do the Abs look right? |
03:05.47 | dcordes | when I give it some pressure it says unkown event type 0 |
03:06.04 | tmzt | I think thats EV_SYNC |
03:07.08 | tmzt | that's not in tslib that I can find |
03:07.16 | dcordes | evtest info on Absolute Event type: 0(x), 1(y), 33(?) 0/0/0 |
03:07.34 | dcordes | running ts_calibrate now |
03:08.58 | dcordes | looks like the old no-pendown situation as before |
03:09.11 | dcordes | it won't quit, n'either change the crosshair |
03:09.12 | tmzt | so it's stuck? |
03:09.33 | tmzt | did you get BTN_TOUCH on evtest? |
03:09.39 | dcordes | in the background it printed Unkown event type 0 |
03:09.48 | dcordes | I guess everytime I taped/held down |
03:09.57 | dcordes | like in print_raw |
03:10.12 | tmzt | over the framebuffer? |
03:10.36 | dcordes | it printed it to the console (I ran ts_calibrate with a timed kill) |
03:10.50 | tmzt | good, didn't think of that |
03:11.12 | tmzt | can you see if that is in dmesg |
03:11.22 | tmzt | do you have tail? |
03:11.27 | dcordes | the unknown event messages? |
03:11.30 | tmzt | yes |
03:11.42 | tmzt | I grepped tslib and didn't find it |
03:12.00 | dcordes | not present there, only tsts_thread: address 0x48 show up 4 times |
03:12.10 | dcordes | I think that's everytime I run one of the programs that use tslib |
03:12.34 | tmzt | yeah, on opening the device in the kernel |
03:12.41 | dcordes | "tslib: Unknown event type 0" |
03:12.43 | tmzt | it starts the thread |
03:12.51 | dcordes | ..that's the exact message |
03:12.58 | tmzt | capital Unknown |
03:14.23 | tmzt | see if EV_SYN is set on evtest |
03:16.22 | tmzt | whats grab_events set to in /etc/ts.conf |
03:17.58 | dcordes | grab_events=0 |
03:18.06 | tmzt | you don't get Unable to grab select input device from ts_print_raw |
03:18.12 | dcordes | module_raw input grab_events=0 |
03:18.21 | dcordes | let me double check |
03:19.15 | dcordes | if I give it no ts data, it only shows tsts_thread: address 0x48\ selected device is not a touchscreen I understand |
03:19.44 | tmzt | yeah, that doesn't seem to be it |
03:20.36 | dcordes | the pressure values are infamous "tslib: Unknown event type 0" message and the other two absolute values get printed correctly |
03:20.45 | dcordes | like in the raw image from yesterday with the old tslib |
03:21.03 | tmzt | you got Unknown event type 0 yesterday? |
03:21.06 | dcordes | the values are around 2000 4000 |
03:21.09 | dcordes | nono |
03:21.17 | dcordes | I was only referring to the x/y, sorry. |
03:21.34 | dcordes | the Unknown event is something new with the commented out thing |
03:21.43 | tmzt | ok |
03:23.45 | tmzt | it appears that Unknown event type can only happen if i->using_syn is false |
03:23.59 | tmzt | 0 is EV_SYN |
03:24.37 | tmzt | we need to know if EV_SYN flag is set in the driver |
03:25.25 | tmzt | you didn't comment out the call to check_fd completely |
03:25.33 | tmzt | just the two lines? |
03:26.11 | tmzt | i've got it |
03:26.25 | tmzt | check_fd still returns -1, and doesn't do the second check |
03:26.37 | tmzt | that sets i->using_syn |
03:26.48 | tmzt | we commented out the wrong part |
03:26.54 | dcordes | oops |
03:27.17 | tmzt | the first if must be false, so the driver flags are still wrong |
03:27.56 | tmzt | okay, we can comment out the first "return -1" in check_fd() |
03:28.33 | tmzt | after selected device s not a touch screen i understand |
03:28.38 | tmzt | the next line |
03:29.47 | tmzt | I guess the syn tells it that the ABS_X ABS_Y and ABS_PRESSURE events are related |
03:30.24 | dcordes | no clue, not quite catching up |
03:31.06 | tmzt | after the line that says "Selected device is not a touchscreen I understand" |
03:31.18 | tmzt | comment out the line below that text |
03:31.34 | dcordes | ok |
03:31.43 | tmzt | in plugins/input-raw.c |
03:36.41 | tmzt | it would be good to add some printf's for the flags, but I'll have to work on that |
03:39.11 | dcordes | tmzt: can I add something funny in the code to make sure I actually use the modified library? |
03:39.49 | tmzt | lets just add fprintf(stderr, "bit: %0x\n", bit); |
03:40.05 | tmzt | and fprintf(stderr, "bit: %0x\n", absbit); |
03:40.17 | dcordes | where? |
03:40.21 | tmzt | you can change the strings if you want |
03:40.35 | tmzt | where return -1 was would be fine |
03:41.06 | dcordes | <PROTECTED> |
03:41.45 | tmzt | yeah, but make those %6x, not %0x |
03:42.03 | tmzt | no, sorry, in the check_fd function |
03:42.12 | tmzt | were we just made the last change |
03:42.39 | tmzt | otherwise, they will print every event I think |
03:42.52 | dcordes | ok |
03:43.17 | dcordes | directly at the beginning of the function, or inside of that if thing? |
03:43.38 | tmzt | at the end of the if before the }; |
03:43.54 | tmzt | under the I don't understand line |
03:44.49 | dcordes | (absbit & (1 << ABS_Y)) && (absbit & (1 << ABS_PRESSURE)))) { |
03:44.49 | dcordes | fprintf(stderr, "selected device is not a touchscreen I understand\n"); |
03:44.49 | dcordes | <PROTECTED> |
03:44.49 | dcordes | <PROTECTED> |
03:44.49 | dcordes | //return -1; |
03:44.59 | dcordes | } |
03:45.02 | tmzt | yes |
03:45.30 | tmzt | the second one should be absbit: |
03:46.55 | dcordes | don't get it, what needs changing? |
03:47.27 | tmzt | the second fprintf shoud have fprintf(stderr, "absbit: %6x\n", absbit); |
03:47.46 | tmzt | otherwise it will print two bit: lines and confuse us |
03:47.54 | dcordes | ok |
03:49.15 | dcordes | tmzt: so anything else we can do? |
03:49.28 | tmzt | don't think so |
03:49.54 | tmzt | we can figure out the flags from the hex, to find out what each zImage wants |
03:50.00 | dcordes | oh wait |
03:50.13 | dcordes | the previously two uncommented lines, do they stay uncommented? |
03:50.24 | tmzt | I don't think it should matter |
03:50.31 | dcordes | s/uncommented lines/commented lines/ |
03:50.47 | dcordes | haah |
03:51.00 | dcordes | ok so I will leave them _commented_ |
03:51.15 | tmzt | if we leave them commented, it will get further if something breaks, and we'll know what broke |
03:51.39 | tmzt | yes |
04:03.24 | dcordes | ok booting |
04:03.33 | dcordes | host needs a reboot too |
04:03.38 | dcordes | brb |
04:10.25 | *** join/#htc-linux dcordes (n=dcordes@unaffiliated/dcordes) |
04:10.38 | tmzt | you're back |
04:10.44 | dcordes | tmzt: ok |
04:10.59 | tmzt | it booted? |
04:11.16 | dcordes | ts_print only shows the fprintfs now |
04:11.27 | dcordes | plus the don't understand msg |
04:11.29 | dcordes | on every event |
04:12.00 | tmzt | what do you have for bit: |
04:12.29 | dcordes | bit: b absbit: 3 |
04:12.31 | dcordes | all the time |
04:13.45 | dcordes | ts_calibrate still doesn't like the pressure |
04:14.21 | tmzt | I missed an obvious fprintf |
04:14.39 | tmzt | i->using_syn |
04:15.04 | dcordes | what does the present one tell us? |
04:16.12 | tmzt | 1011 |
04:16.47 | tmzt | ABS, KEY, SYN |
04:18.04 | tmzt | no ABS_PRESSURE |
04:19.51 | tmzt | using_syn has to be false to get the unknown event message, I think |
04:20.47 | tmzt | ABS_PRESSURE is set in the driver |
04:21.57 | tmzt | theres no input_set_abs_params in the driver source |
04:22.04 | tmzt | for ABS_PRESSURE |
04:22.12 | tmzt | but there is in tsc2200 |
04:23.17 | dcordes | is it similar to 2003.c? |
04:24.18 | tmzt | input_set_abs_params(idev, ABS_PRESSURE, 0, 1, 0, 0); |
04:24.34 | tmzt | I don't know |
04:26.58 | tmzt | I wonder if that sets the flags now in input subsystem |
04:27.05 | tmzt | not the BIT macros |
04:27.17 | dcordes | evtest? |
04:27.35 | tmzt | yeah, do you get ABS_PRESSURE |
04:27.45 | tmzt | at the top |
04:28.18 | dcordes | no the same x y 33(?) |
04:31.26 | tmzt | Input device ID: bus ?? |
04:31.51 | dcordes | 0x0 vendor 0x0 product 0x0 version 0x0 |
04:32.06 | tmzt | Input device name: |
04:32.16 | dcordes | kaiser-ts |
04:32.34 | tmzt | Event type |
04:34.21 | tmzt | Event code |
04:34.42 | tmzt | (working through evtest.c, trying to figure out what it's printing) |
04:35.32 | dcordes | darn I wrote it to a file and ejected card but if tailed to write |
04:35.40 | tmzt | sync |
04:35.48 | dcordes | hm? |
04:36.00 | tmzt | sync command should be in busybox |
04:36.13 | dcordes | I see |
04:36.13 | tmzt | tells the kernel to write data out |
04:36.14 | dcordes | will do next time |
04:36.14 | tmzt | yeah |
04:36.27 | tmzt | let's add the using_syn then |
04:36.44 | tmzt | and some ands for the EV types |
04:37.09 | tmzt | first, can you paste your input-raw.c |
04:37.14 | dcordes | yes |
04:38.39 | dcordes | http://rafb.net/p/YVNVEo69.html |
04:39.55 | tmzt | allright, |
04:40.08 | tmzt | above if (i->using_syn) { |
04:41.05 | tmzt | fprintf(stderr, "i->using_syn: %d\n", i->using_syn); |
04:41.42 | dcordes | line 101? |
04:41.46 | dcordes | 100 |
04:41.57 | tmzt | yeah |
04:42.15 | dcordes | ok done |
04:42.59 | tmzt | 71, after the } |
04:43.14 | tmzt | not inside it |
04:44.43 | tmzt | fprintf(stderr, "abs EV_ABS: %d\m", (abs & (1 << EV_ABS))); |
04:44.50 | tmzt | \n not \m |
04:45.42 | tmzt | fprintf(stderr, "abs ABS_X: %d\m", (abs & (1 << ABS_X))); |
04:45.54 | tmzt | fprintf(stderr, "abs ABS_Y: %d\m", (abs & (1 << ABS_Y))); |
04:46.04 | tmzt | fprintf(stderr, "abs ABS_PRESSURE: %d\m", (abs & (1 << ABS_PRESSURE))); |
04:46.15 | tmzt | all \n not \m |
04:48.04 | dcordes | done? |
04:48.12 | tmzt | yeah |
04:48.29 | tmzt | well, EV_SYN |
04:48.31 | tmzt | hold on |
04:48.37 | dcordes | http://rafb.net/p/pnnTkl79.html |
04:48.57 | tmzt | I messed all those up |
04:49.03 | dcordes | does that fix the 33 (?) problem or only add debug info for us? |
04:49.17 | tmzt | debug, but we should be able to figure out the 33 |
04:49.19 | tmzt | maybe |
04:49.26 | tmzt | its just reporting it |
04:49.36 | dcordes | ok. what's messed up now? |
04:49.46 | tmzt | the first line should have (bit & ... |
04:49.57 | tmzt | the rest should have (absbit & ... |
04:50.26 | tmzt | I should have edited the file, but I don't have cross compiler set up |
04:51.31 | tmzt | fprintf(stderr, "abs EV_SYN: %d\n", (bit & (1 << EV_SYN))); |
04:51.37 | tmzt | that should be the last one |
04:51.51 | tmzt | can you paste that |
04:52.49 | dcordes | now I'm confused. I thought the first should be fprintf(stderr, "abs EV_ABS: %d\n", (bit & (1 << EV_ABS))); and the rest like fprintf(stderr, "abs ABS_X: %d\n", (absbit & (1 << ABS_X))); |
04:53.14 | tmzt | well, we can make the last one come after the first |
04:53.16 | dcordes | oh right the syn is additional line? |
04:53.46 | tmzt | the ones that say EV_ should have (bit & |
04:53.56 | dcordes | ok because it is not absolute data? |
04:54.04 | tmzt | the others (ABS_) should say (absbit & |
04:54.06 | tmzt | yes |
04:54.20 | tmzt | this is not the data though, its the driver flags |
04:54.24 | dcordes | http://rafb.net/p/RRhw8S99.html |
04:54.27 | tmzt | the use the same bitmasks |
04:54.27 | dcordes | oh ok |
04:54.34 | dcordes | no clue what a driver flag is though |
04:54.54 | tmzt | something set in the driver that tslib tests |
04:55.10 | tmzt | to see what events are exposed, is it a mouse, keyboard, joystick, or whatever |
04:55.26 | tmzt | they're all done through input subsystem now |
04:56.08 | tmzt | looks good |
04:56.39 | tmzt | at 114, we can add a fprintf for ev.type |
04:57.22 | tmzt | fprintf(stderr, "ev.type: %d\n", ev.type); |
04:57.59 | tmzt | at 117 and 146 we can add: |
04:58.16 | tmzt | between 145 and 146 |
04:58.32 | tmzt | fprintf(stderr, "ev.code: %d\n", ev.code); |
04:58.38 | tmzt | that will print on every event |
04:58.54 | tmzt | so it might not be good to leave in for Xfbdev |
04:59.31 | tmzt | lets make 114 "key ev.code" |
04:59.39 | dcordes | we can take it out again when we have the events in the test apps? |
04:59.40 | tmzt | and 146 "abs ev.code" |
04:59.43 | tmzt | yes |
04:59.54 | dcordes | 146 is fprintf(stderr, "ev.code: %d\n", ev.code); now |
05:00.11 | tmzt | okay stderr, "abs ev.code ... |
05:00.22 | tmzt | don't change the last part |
05:00.46 | dcordes | add that abs into 146? |
05:00.53 | tmzt | yes |
05:01.15 | dcordes | <PROTECTED> |
05:01.20 | tmzt | yes |
05:01.29 | dcordes | ok what about like 114? |
05:01.32 | dcordes | line |
05:01.36 | tmzt | "key ev.code: %d\n" |
05:01.43 | tmzt | for 114 |
05:02.41 | dcordes | fprintf(stderr, "key ev.code: %d\n", ev.type); |
05:02.59 | tmzt | ev.code |
05:03.14 | tmzt | fprintf(stderr, "key ev.code: %d\n", ev.code); |
05:04.19 | dcordes | good anything else? |
05:04.29 | tmzt | no |
05:04.40 | dcordes | can you look it over again? |
05:04.44 | tmzt | yes |
05:05.20 | dcordes | http://rafb.net/p/Ty8P6M15.html |
05:07.13 | tmzt | 114 should be fprintf(stderr, "ev.type: %d\n", ev.type); |
05:07.36 | tmzt | 117 should be fprintf(stderr, "key ev.code: %d\n", ev.code); |
05:07.58 | tmzt | between 116 and 117 |
05:08.34 | tmzt | 146 is right |
05:09.30 | dcordes | ok got it |
05:09.50 | dcordes | http://rafb.net/p/x9NJOa50.html |
05:10.43 | tmzt | looks good now |
05:10.58 | dcordes | ok building. |
05:24.22 | tmzt | in input.h, input_set_abs_params |= the bit |
05:24.54 | tmzt | I think we need to add that to tsc2003.c, for the zImage |
05:25.01 | tmzt | when somebody can build it |
05:25.18 | dcordes | unfortunately I do not have the source of the zImage I'm running (with the rawts, uncorrected y) |
05:25.25 | dcordes | only got git |
05:25.26 | tmzt | yeah |
05:25.36 | dcordes | ginge has it. I'm in the ts debug image now |
05:25.37 | tmzt | still building? |
05:25.42 | dcordes | with new lib |
05:25.42 | tmzt | ok |
05:25.51 | dcordes | and in print raw |
05:26.15 | tmzt | is using_syn set? |
05:26.44 | dcordes | where |
05:26.49 | dcordes | in evtest? |
05:27.00 | dcordes | oh right |
05:27.02 | tmzt | whats the output from ts_print_raw when you don't touch anything |
05:27.07 | dcordes | i->using_syn:1 |
05:27.11 | tmzt | ok |
05:27.16 | tmzt | you get events ev.code |
05:27.24 | tmzt | ev.type 0 ?? |
05:28.12 | dcordes | yes I see that line |
05:28.19 | tmzt | may times? |
05:28.21 | tmzt | many |
05:28.38 | tmzt | do you see key ev.code |
05:28.44 | dcordes | no, only once |
05:29.09 | dcordes | wait I will give you the output |
05:29.53 | tmzt | 330 is touch! |
05:30.04 | tmzt | that might have been it |
05:31.41 | tmzt | wait, you don't get Unknown anymore? |
05:32.07 | dcordes | it still understands what it must ^^ |
05:32.20 | tmzt | huh? |
05:32.36 | dcordes | oh you mean in evtest? |
05:32.46 | tmzt | no, in ts_print_raw |
05:33.16 | dcordes | I only see the not a touchscreen I understand stuff |
05:33.37 | tmzt | so we did something, using_syn must not have been set |
05:33.42 | tmzt | have you shut it down yet? |
05:34.18 | dcordes | the program? yes |
05:34.25 | dcordes | not kaiser |
05:34.26 | tmzt | the kaiser |
05:34.27 | tmzt | good |
05:34.37 | tmzt | try ts_calibrate with your kill script |
05:35.13 | dcordes | no reaction |
05:35.20 | dcordes | same as before |
05:35.30 | dcordes | what about evtest? |
05:35.46 | tmzt | you only got on ev.code 0 from ts_print_raw when you pressed the screen? |
05:36.03 | tmzt | we did not change it, it uses the kernel api directly |
05:36.11 | dcordes | ok |
05:36.18 | tmzt | you can look for EV_SYNC though |
05:37.56 | dcordes | when I touch the screen there is from ev.code 0 and 1 |
05:38.25 | tmzt | that should be ev.type |
05:38.35 | tmzt | ev.code 0 and 1 should be X and Y |
05:38.48 | dcordes | 3 3 1 0 |
05:38.57 | tmzt | do you get ev.type: 0 |
05:39.32 | dcordes | can I output the whole to a file? |
05:39.42 | tmzt | yeah |
05:39.45 | tmzt | > /sd/file |
05:39.50 | dcordes | doesn't work |
05:40.00 | tmzt | what does mount say |
05:40.20 | dcordes | I can create files in the workdir. |
05:40.38 | tmzt | you're root is on sd, right |
05:40.40 | tmzt | forgot |
05:40.51 | dcordes | it touches the file, but there's nothing in it. |
05:40.58 | dcordes | and the output is also drawn on the screen |
05:41.02 | tmzt | 2> file |
05:41.52 | tmzt | do you get ev.type: 0 |
05:42.35 | dcordes | yes |
05:42.50 | tmzt | so abs ev.code: 0 |
05:42.51 | tmzt | so abs ev.code: 1 |
05:42.59 | tmzt | so ev.type: 0 |
05:43.05 | tmzt | but no pressure? |
05:43.22 | dcordes | abs ABS_PRESSURE: 0 |
05:43.41 | tmzt | so the kernel doesn't set it |
05:44.13 | dcordes | evtest sees pressure stuff I think |
05:44.23 | tmzt | but no ab ev.code: 24 |
05:44.30 | tmzt | but no abs ev.code: 24 |
05:45.04 | tmzt | what abs ev.codes do you get |
05:45.14 | tmzt | 0 1 24, others? |
05:45.58 | dcordes | <PROTECTED> |
05:46.06 | tmzt | just 0 and 1 |
05:46.07 | dcordes | wait |
05:46.28 | tmzt | sync |
05:46.30 | tmzt | first |
05:47.07 | tmzt | you can do init 0 or init 6 first, that's best |
05:47.17 | tmzt | it probably won't shutdown or reboot though |
05:47.18 | dcordes | looks like only o and 1 |
05:47.24 | dcordes | for abs ev.code |
05:47.30 | tmzt | X and Y |
05:47.38 | dcordes | probably there are others but they can hide too well |
05:47.47 | tmzt | that must confuse ts_cal |
05:47.50 | dcordes | x:1 y:2 |
05:47.55 | tmzt | but it needs the zImage |
05:48.15 | tmzt | yeah, those are the bit values, but the defines are the place |
05:48.44 | tmzt | like bit 0 |
05:48.46 | tmzt | bit 1 |
05:49.01 | tmzt | from the right |
05:49.49 | dcordes | huh |
05:50.29 | tmzt | in base ten you have the ones place, the tens place the hundreds place |
05:51.37 | dcordes | sorry can't follow |
05:51.44 | dcordes | what base ten? |
05:51.56 | tmzt | decimal system |
05:52.01 | tmzt | its not important |
05:52.10 | dcordes | oh maths |
05:52.15 | tmzt | yeah |
05:52.17 | dcordes | :> |
05:52.18 | tmzt | binary maths |
05:52.20 | tmzt | fun stuff |
05:52.42 | dcordes | so do you have all the info needed from ts_print_raw? |
05:52.48 | tmzt | yeah |
05:53.07 | tmzt | it really looks like this is going to need the zImage fixed |
05:53.12 | tmzt | but I can't be sure |
05:55.23 | dcordes | darn I was about task ginge for the source of the raw ts before he left |
05:55.35 | dcordes | now I need some rest. will be back later |
05:55.52 | tmzt | i was getting there too, see you later |
05:55.54 | dcordes | thanks very much for the help |
05:56.02 | tmzt | not one problem |
05:56.05 | tmzt | no problem |
05:56.13 | dcordes | bye |
06:32.06 | *** join/#htc-linux Zoolooc (n=fredsiba@nrbg-4dbf916a.pool.einsundeins.de) |
08:08.51 | *** join/#htc-linux axe- (n=axe@217.116.62.236) |
08:18.51 | *** join/#htc-linux lpotter (n=ljp@CPE-124-191-144-181.vic.bigpond.net.au) |
08:50.39 | *** join/#htc-linux axe- (n=axe@217.116.62.236) |
09:12.35 | *** join/#htc-linux meredydd (n=meredydd@cpc4-cmbg2-0-0-cust471.cmbg.cable.ntl.com) |
09:12.51 | meredydd | Heyoo - anyone here working with Kaiser? |
09:13.04 | meredydd | Was wondering whether you had screen rotation working yet |
09:13.52 | meredydd | (also what the ETA was for SMS, but that's a lower priority) |
09:23.15 | *** join/#htc-linux goxboxlive (n=goxboxli@31.84-48-212.nextgentel.com) |
09:23.23 | *** join/#htc-linux pH5 (n=ph5@p5485FC41.dip.t-dialin.net) |
09:28.00 | meredydd | And a more prosaic question: What's the stability like, generally? |
09:28.08 | meredydd | The following chart is somewhat scary: http://spreadsheets.google.com/pub?key=pWRxYmfKRYUKqujEu6fizNg |
09:52.09 | *** join/#htc-linux diogene31 (n=rj@mur31-2-82-243-122-54.fbx.proxad.net) |
09:59.41 | *** join/#htc-linux okias (i=ejabberd@poseidon.kraja.cz) |
10:01.28 | *** join/#htc-linux kiozen (n=oeichler@Mb1e3.m.pppool.de) |
10:03.28 | *** join/#htc-linux exco (n=exco@e181077068.adsl.alicedsl.de) |
10:04.22 | *** join/#htc-linux gerard (n=gerard@s55932332.adsl.wanadoo.nl) |
10:26.09 | *** join/#htc-linux ellisway (n=ellis@80-46-67-47.static.dsl.as9105.com) |
10:30.13 | *** join/#htc-linux BabelO (n=Fabrice@lun34-2-82-238-28-28.fbx.proxad.net) |
10:37.17 | *** join/#htc-linux exco (n=exco@e181077068.adsl.alicedsl.de) |
10:41.56 | *** join/#htc-linux exco (n=exco@e181077068.adsl.alicedsl.de) |
10:43.46 | *** part/#htc-linux okias (i=ejabberd@poseidon.kraja.cz) |
10:45.25 | *** join/#htc-linux okias (i=ejabberd@poseidon.kraja.cz) |
10:49.03 | ginge | morning |
10:50.43 | *** part/#htc-linux okias (i=ejabberd@poseidon.kraja.cz) |
10:51.08 | ginge | meredydd: you still about? |
10:56.41 | *** join/#htc-linux axe- (n=axe@217.116.62.236) |
11:01.08 | *** join/#htc-linux gerard (n=gerard@s55932332.adsl.wanadoo.nl) |
11:33.16 | meredydd | ginge: Yep |
11:33.20 | meredydd | ginge: but only briefly |
11:33.35 | ginge | I just wanted to address your question about status |
11:33.42 | ginge | in short for android it is pretty grim |
11:33.43 | meredydd | sure - what is it, then? |
11:34.34 | meredydd | hmm |
11:34.35 | ginge | there are people here that are workign on an openmoko image that will run stably off sd, but until google release a more stable image, and we complete the kernel support for the hardware, its not workable |
11:34.54 | meredydd | How badly non-workable? |
11:35.01 | meredydd | As in, "userspace is shot"? |
11:35.07 | ginge | eg we can only talk to people to the speakerphone. |
11:35.07 | meredydd | or "kernel is shot"? |
11:35.14 | ginge | in fact, both |
11:35.23 | meredydd | That's...not so good. |
11:35.35 | ginge | the kernel needs a lot more device driver support. We are working to get this stuff going but it takes time. |
11:35.50 | meredydd | I'm mostly interested in stability. |
11:35.59 | meredydd | I don't actually need a working phone - I have one of those |
11:36.07 | ginge | stability will hopefully come with decent userspace |
11:36.15 | meredydd | I want a device to road-test Android apps on |
11:36.31 | ginge | Android is pretty stable if you dont use data |
11:36.45 | meredydd | Oh, dear. That's a pretty bad caveat :P |
11:36.49 | ginge | really |
11:36.51 | meredydd | Any idea where the crashes there are coming from? |
11:37.22 | meredydd | (kernel, Android infrastructure, or apps like the browser?) |
11:37.25 | ginge | we are not too sure yet. Personally I am more focussed on writing drivers for the kaiser. What we need is an android guru to get this shit sorted |
11:37.59 | ginge | all data crashes android on the kaiser, regardless of the app. Data on other X based linux distros are fine |
11:38.23 | meredydd | Oh, joy. So it's definitely userspace. |
11:38.45 | ginge | it would seem so, but our image works stably for other devices |
11:39.01 | meredydd | Ah...so Vogue/Touch is okay? |
11:39.28 | ginge | vogue is mostly stable, dunno about touch |
11:39.36 | meredydd | Aren't they the same thing? |
11:39.51 | ginge | shrugs |
11:40.42 | meredydd | I thought Vogue::Touch = Trion::Hermes or similar |
11:40.57 | meredydd | (uh... Hermes::Trion. What I mean is that it's the marketing name) |
11:40.58 | ginge | I'm not so up on the different flavours of devices and the names |
11:41.01 | meredydd | k |
11:42.15 | ginge | I have no idea what the deal is with the data crash, but I want to look into it today. I am just building an image up that will dump raw logs out over the data connection. Hopefully we can see whats going on. |
11:42.51 | meredydd | okay then. |
11:43.00 | meredydd | Hmm...if you're looking at it today, I'll hold on. |
11:43.36 | ginge | heh. It doesn't mean it will be fixed anytime soon ;) |
11:44.20 | meredydd | Indeed. But equally, I've seen in the forum logs how fast you work. |
11:44.43 | meredydd | I'm planning to buy a device in the next few days, so I'll hang around and see if you pull something out of your hat. |
11:44.54 | meredydd | If not, I'll just hang out for a Vogue. |
11:45.06 | meredydd | (unless someone has Android working on something older and cheaper :P) |
11:45.07 | ginge | I think the vogue is the wiser choice at this stage. |
11:45.12 | meredydd | k |
11:45.25 | ginge | the kaiser is just too different from what google puts out that it causes nonstop headaches |
11:46.03 | ginge | cross you fingers, android data logs coming in now |
11:46.19 | meredydd | crosses all available extremities |
11:46.48 | ginge | hmm. well... data connection is still up, I can talk to the device but android is frozen solid |
11:46.52 | ginge | that tells me nothing |
11:46.57 | ginge | damn |
11:48.50 | ginge | its something android-y going wrong |
11:48.56 | ginge | give us the source! |
11:49.13 | ginge | oh well. I will keep at it. If I make any progress I will make a nte in the forum. |
13:50.16 | *** join/#htc-linux JEEB (n=kanakana@a88-114-192-27.elisa-laajakaista.fi) |
14:36.17 | *** join/#htc-linux exco (n=exco@e181077068.adsl.alicedsl.de) |
15:00.02 | *** join/#htc-linux okias (i=ejabberd@poseidon.kraja.cz) |
15:01.05 | okias | Hi, maybe this acx100 driver can be compiled with lastest kernel: http://gitorious.org/projects/acx-mac80211/repos/mainline |
15:01.56 | okias | last changes are from today |
15:09.51 | pH5 | nope, doesn't compile against 2.6.26 either |
15:10.16 | okias | :-( new changes, but not working with lastest kernel.. this sux :-( |
15:11.23 | pH5 | maybe we should just drop david planella a mail and ask about it. |
15:13.58 | okias | pH5: good idea, but my english is really bad, then it's job for you :-D |
15:15.21 | okias | pH5: i noticed, Universal have acx100 driver, but he using old kernel... |
15:15.59 | okias | pH5: maybe this is reason why Universal using old kernel :-D |
15:16.45 | pH5 | yes, I think the old non-mac80211 acx driver works on hx4700, universal, h4000, blueangel and maybe a few others. no WPA support there, and the hh.org kernel is at 2.6.21... |
15:17.19 | dcordes | pH5: mac80211 acx supports wpa? |
15:20.07 | pH5 | I'll ask them about it. We should pull their changes into the linuxtogo repository once it works. |
15:20.26 | pH5 | dcordes: as far as I know, no |
15:20.51 | okias | pH5: but probably is easier implement WPA support or something like that |
15:21.02 | pH5 | I hope the mac80211 port will get that support one day, the non-mac80211 certainly won't. |
15:21.14 | okias | pH5: or just mainline using mac80211, then must be great :-D |
15:22.38 | pH5 | http://handhelds.org/moin/moin.cgi/HandheldHardwareXref has quite a few acx100 users listed, but I think this list is not complete? |
15:23.45 | dcordes | http://acx100.sourceforge.net/wiki/WPA "...it is possible use ndiswrapper in order to have WPA support." |
15:23.53 | dcordes | that's something |
15:29.23 | pH5 | nah |
15:29.34 | pH5 | I sure hope it's impossible to use ndiswrapper on ARM. |
15:30.10 | *** join/#htc-linux Marex (n=vasum7am@auryn.karlin.mff.cuni.cz) |
15:30.39 | okias | it's sad, but ndiswrapper for arm really not exist :-( |
15:32.58 | cr2 | okias: thanks God it does not exist. |
15:38.15 | okias | cr2: why? |
15:38.30 | okias | cr2: i don't like ndiswrapper, but better than nothing... |
15:39.54 | dcordes | okias: ACK better than using windows to get on your friend's wifi with wpa. |
15:43.05 | cr2 | it completely defeats the purpose of having linux |
15:43.50 | cr2 | not only you have some insecure binary blob in the kernel, but you also need to implement 50% of wince api in the kernel. |
15:44.02 | okias | cr2: but if U temp. don't have driver then what? I used 1 year ndiswrapper on server for rtl8180l and now in 2.6.25 exist lin driver, then i just removed ndiswrapper |
15:45.18 | cr2 | on the server you can just use better hardware. on the phone it's not possible, but after all we have to deal with only 1 acx* chipset. |
15:46.19 | cr2 | the acx100 driver was rewritten 3 times already, but the code quality is still questionable. that#s the real problem. |
15:46.51 | okias | cr2: :-( |
15:52.22 | okias | cr2: we just need someone who write good basic code and try get into mainline kernel... :-D |
16:00.32 | okias | i looked on gitorious and xazz.no-ip.info and looks like they both can just working on one tree... |
16:00.47 | dcordes | ~lart acx100 devs for writing bad quality code and not supporting WPA |
16:00.47 | apt | rm -rf's acx100 devs for writing bad quality code and not supporting WPA |
16:01.30 | okias | lol |
16:14.48 | *** join/#htc-linux kiozen (n=oeichler@rgnb-5d86575f.pool.einsundeins.de) |
16:19.20 | BabelO | hi |
16:19.30 | BabelO | kiozen: M freeze on artemis now :( |
16:20.05 | kiozen | BabelO: that is bad, any idea why? |
16:20.47 | BabelO | kiozen: it freeze when i open menu int gps settings, can't change anything once i m on it |
16:21.13 | kiozen | via key or via ts? |
16:21.24 | BabelO | kiozen: ts |
16:21.56 | kiozen | does it show the sattelite bar graph? |
16:22.22 | BabelO | kiozen: it display the settings, but can't click to change speed baud |
16:22.39 | BabelO | i m looking more, right now |
16:23.01 | kiozen | hm, sound like the event loop is blocked |
16:23.47 | kiozen | i'll be back later, son is bugging me for stuff ... |
16:24.02 | BabelO | ok |
16:25.26 | ginge | dcordes: marajin: just booted android off the sd finally. its really slow :O |
16:27.45 | okias | ginge: hi, what's looking with ts calibration? |
16:28.21 | ginge | okias: its almost ok in android, not perfect though. |
16:29.24 | okias | ginge: :-) great |
16:32.08 | ginge | enough to use the buttons on screen |
16:33.27 | okias | ginge: now i waiting on wifi support and after i see first wifi connection on kaiser i leaving WM... :-) |
16:34.04 | ginge | okias: we need a working acx driver I think |
16:34.39 | okias | ginge: hehe :-D |
16:35.22 | ginge | hey I think I fixed sms in android too |
16:41.30 | marajin | morning peeps |
16:41.38 | marajin | ginge: What was the problem? |
16:42.11 | ginge | the delay in dma was too high for our device, and the irq handling doesn;t need a delay at all |
16:42.24 | ginge | now I am confused as to why it takes 22 minutes to boot android off sd |
16:45.09 | *** join/#htc-linux rob_w (n=bob@M9bd5.m.pppool.de) |
16:45.42 | ginge | I think the signal strength is working now too. I need to wander around a bit |
16:46.36 | marajin | hmm |
16:46.38 | marajin | nifty |
16:47.00 | marajin | ginge: If it works my house is ideal for testing signal strength. I can get from 0 to 100% signal by moving a few feet |
16:48.46 | AstainHellbring | damn marajin got some lead in the walls? |
16:49.03 | marajin | AstainHellbring: I dunno @.@ it just has always been like that |
16:49.20 | marajin | AstainHellbring: TBH some phones are fucked in my place, they just won't get a signal. |
16:49.51 | marajin | AstainHellbring: I was kinda impressed with the kaiser and hermes for how much more reliably they DID get signals here actually. |
16:49.59 | AstainHellbring | nice |
16:50.03 | ginge | yeah I lived in a place like that. According to the landlord they used metal mesh to reinforce the walls so it must act as a faraday cage |
16:51.09 | marajin | ginge: Very possible. That's exactly what they tend to do at places like Supermarkets. Hence why it's hard to get a signal there for most. |
16:51.22 | marajin | tho' my hermes/kaiser are not defeated by the local ASDA! |
16:51.30 | marajin | my old Nokia was |
16:51.40 | BabelO | kiozen: ok was just a strange wince issue, gps was down in wince too |
16:52.13 | marajin | reckon the wall ties or something have a scattering effect ? |
17:00.25 | marajin | ginge: I assume you're still booting from loopfiles? |
17:00.46 | ginge | no I have 2 ext3 partitions |
17:01.09 | ginge | loopfile should work now |
17:01.41 | marajin | got a zImage? I'll give it a shot |
17:01.49 | ginge | just building it now |
17:01.59 | marajin | goodo |
17:02.34 | ginge | http://www.headfuzz.co.uk/files/android/zImage-sd3 |
17:05.35 | marajin | ginge: Are you sitting down? ;) |
17:05.43 | ginge | mmm |
17:05.50 | marajin | dma! |
17:05.59 | marajin | dead as a motherfucking doornail |
17:06.51 | ginge | damn |
17:07.30 | marajin | pretty much after the modem init attempt it traces and dma timeout abort |
17:12.15 | cr2 | marajin: doornail does not start with 'a' |
17:12.51 | marajin | cr2: er? |
17:13.20 | BabelO | apt: ping kiozen |
17:13.20 | apt | pong kiozen |
17:13.26 | cr2 | kiozen: we set all the cpld gpios to zero at boot now. you may set the battery related pins to there usual wince values. |
17:13.51 | cr2 | marajin: your explanation what is 'dma' |
17:14.22 | marajin | cr2: Riiiight, that's a pretty bad joke man :P |
17:14.29 | cr2 | :D |
17:39.36 | *** join/#htc-linux dwaradzyn_ (n=gudz666@damian8155.net.autocom.pl) |
17:40.09 | *** join/#htc-linux tmzt (n=tmzt@adsl-70-228-92-50.dsl.akrnoh.ameritech.net) |
17:40.10 | *** join/#htc-linux LunohoD (n=alex@e180075182.adsl.alicedsl.de) |
17:50.36 | dcordes | lol |
17:51.22 | dcordes | ginge: did you see the result of this "morning" tslib hackery? |
17:51.37 | ginge | there was a result? |
17:51.47 | ginge | I didnt read it, there was quite a lot |
17:52.40 | tmzt | it seems there is a line missing from tsc2003.c, but whether that is the problem we don't know |
17:52.48 | dcordes | sort of. tmzt said we need kenel changing. I think in the raw kaiser ts driver, pressure is not ok |
17:53.07 | ginge | what do you need? |
17:53.15 | ginge | and did you decide raw is best? |
17:53.22 | dcordes | so we need the source. can you paste the raw code w/o y corr? |
17:53.24 | tmzt | input_set_abs_param( |
17:53.37 | tmzt | for ABS_PRESSURE, 0 to 1 |
17:53.44 | ginge | its not really much change. Just pass the result of the reads to report event |
17:54.17 | ginge | <PROTECTED> |
17:54.19 | kiozen | cr2: ok, is there a kernel update to load? |
17:54.29 | ginge | pressure is reported |
17:54.34 | kiozen | BabelO: ~ping |
17:54.48 | tmzt | yes, but the ABS_PRESSURE flag is not set |
17:54.56 | ginge | oh. any clues on how? |
17:55.05 | tmzt | and I'm not sure what the input layer is doing to the flags |
17:55.12 | tmzt | without the set_abs_param |
17:56.03 | ginge | dcordes: to make that reprt the raw events change xc to c yc to y and p to p1 remove corr |
17:56.15 | ginge | s/xc to c/xc to x/ |
17:57.31 | tmzt | 256, 257 |
17:57.44 | BabelO | kiozen: track hang, editing gps settings hang :( |
17:58.02 | tmzt | sorry, 266, 267 |
17:58.10 | kiozen | BabelO: using latest version? |
17:58.13 | tmzt | 265, 266 |
17:58.27 | BabelO | kiozen: latest update one hour ago |
17:58.40 | kiozen | BabelO: ok, what it the cpu usage? |
17:59.16 | BabelO | kiozen: 0% |
17:59.27 | BabelO | kiozen: i check again |
17:59.31 | tmzt | 351 input_set_abs_params(data->idev, ABS_PRESSURE, 0, 1, 4, 2); |
18:01.27 | kiozen | BabelO: do you get position info from GPS? |
18:01.47 | BabelO | kiozen: yes, and it is ok on my new map :) |
18:02.13 | kiozen | BabelO: thus once you switch to setup dialog it hangs? |
18:02.45 | BabelO | kiozen: only if i select something in th setup dialog, yes |
18:03.04 | kiozen | BabelO: that is strange... |
18:03.19 | BabelO | kiozen: yes very strange |
18:03.38 | kiozen | BabelO: can you use it without qpe |
18:03.56 | BabelO | yes i can try |
18:05.59 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
18:06.56 | tmzt | should 346, 347, 348 have |= in tsc2003.c ? |
18:07.50 | ginge | I dont know . I guess noti if it is regiisterig the device. Makes some sense to start clean |
18:08.20 | tmzt | ABS_X is bit 1, and ABS_Y is bit 2, so it looks like it might be clearing it |
18:08.30 | BabelO | kiozen: you have a loop to set date ? |
18:08.40 | tmzt | dcordes: do you have that paste from ts_print_raw? |
18:08.51 | dcordes | tmzt: are you in the git? |
18:08.51 | kiozen | BabelO: yes, now, it uses the gps date |
18:08.58 | tmzt | yes, htc-msm |
18:09.06 | kiozen | BabelO: but only without qpe |
18:09.32 | BabelO | kiozen: ok but is it ok to do it so often ? |
18:10.29 | kiozen | BabelO: it's only done if the clock does not match the gps clock |
18:10.44 | tmzt | dcordes: do you have the git source? |
18:11.07 | dcordes | tmzt: that should be the optimized code in git |
18:11.22 | tmzt | can you build it though |
18:11.34 | dcordes | tmzt: yrp Im doing a fresh checkout this second |
18:25.09 | dcordes | tmzt: ok git it. |
18:32.31 | dcordes | tmzt: I don't really understand what we need changed |
18:33.20 | kiozen | BabelO: does it work with -qws? |
18:34.25 | tmzt | can you paste the tsc2003.c |
18:35.34 | dcordes | ok |
18:36.24 | dcordes | tmzt: I don't know how to copy paste from console. could you look it up in the git? |
18:36.41 | tmzt | yeah |
18:36.52 | dcordes | I could only upload the file to ltg but it's already there |
18:38.13 | kiozen | BabelO: think I can reproduce it with the emulator |
18:38.21 | tmzt | it won't save in links2, going to try firefox |
18:38.45 | dcordes | oh right that is broken. |
18:38.58 | dcordes | you will get a blank file |
18:39.26 | tmzt | email? |
18:40.55 | BabelO | kiozen: i ve a small problem to boot without qpe :) |
18:41.24 | kiozen | BabelO: no worries, I can reproduce, that is half the solution |
18:41.50 | BabelO | kiozen: ahhh :) good, so track should be the same problem |
18:42.14 | kiozen | something seems to block the event q |
18:42.52 | BabelO | kiozen: was working before i think |
18:47.33 | kiozen | BabelO: think I fixed it, svn up |
18:48.27 | BabelO | kiozen: compiling :) |
18:53.05 | dcordes | tmzt: linuxtogo.org/~lgorris/tsc2003.c |
18:54.20 | BabelO | kiozen: perfect ! |
18:54.54 | BabelO | kiozen: so i can test track now :) |
18:55.10 | kiozen | try it , what map are you using? |
18:55.32 | BabelO | kiozen: a new one :) |
18:55.51 | kiozen | BabelO: ok, you have to kill me after you told me? |
18:56.21 | BabelO | kiozen: :) |
18:56.31 | BabelO | the track edit, is not good for my screen |
18:57.15 | kiozen | BabelO: yes, think I have to do different resolutions at compile time |
18:57.43 | BabelO | kiozen: wait i send you my patch |
18:57.48 | dcordes | bbiab |
18:57.57 | BabelO | but don't understand why icon are big again :( |
18:58.13 | BabelO | kiozen: you change the zoom icon ? |
18:58.43 | kiozen | BabelO: yes I increased the size of all icons and buttons |
18:58.58 | kiozen | BabelO: they where too small for 480x640 |
18:59.15 | BabelO | :( |
18:59.50 | kiozen | BabelO: think we have to add Artemis to the common file and make the icons screen resolution dependent |
19:00.58 | tmzt | dcordes: can you email just the tsc2003.c? |
19:01.43 | BabelO | kiozen: email sent |
19:02.06 | BabelO | kiozen: it is not clean, because i change CInfo without if |
19:02.26 | kiozen | uhhh :( |
19:10.32 | kiozen | BabelO: no CKeysSmallButton |
19:11.10 | BabelO | :( poor have to do it with the png too |
19:12.07 | kiozen | BabelO: can you set the screen width in common as define and do a clean patch? |
19:13.01 | *** join/#htc-linux tmzt (n=tmzt@adsl-70-228-92-50.dsl.akrnoh.ameritech.net) [NETSPLIT VICTIM] |
19:13.01 | *** join/#htc-linux Hymie (i=hymie@l8r.net) [NETSPLIT VICTIM] |
19:13.01 | *** join/#htc-linux par (i=par@dipole.idlepattern.com) [NETSPLIT VICTIM] |
19:13.01 | *** join/#htc-linux kaboofa (n=tom@rain.angrycoder.org) [NETSPLIT VICTIM] |
19:14.37 | BabelO | kiozen: i can try, but i want to get the CDialog height corerct and my 3D/2D icon |
19:15.38 | kiozen | BabelO: ok, but anyway a _clean_ patch would be highly appreciated ;) |
19:15.52 | BabelO | kiozen: ok |
19:17.00 | kiozen | BabelO: I use the icons from the nuvola icon set |
19:17.32 | kiozen | BabelO: think we have to switch to the 16x16 icons for smaller devices |
19:18.20 | BabelO | yes |
19:21.47 | kiozen | BabelO: let me rename the icons to make place for the 16x16 versions |
19:22.19 | BabelO | kiozen: ok |
19:26.39 | tmzt | http://blog.iphone-dev.org/post/42858313/thanks-for-waiting |
19:35.40 | dcordes | tmzt: linuxtogo.org/~lgorris/tsc2003.c but I can also email it |
19:36.05 | tmzt | ok, I'm getting the htc-msm git anyway |
19:36.11 | tmzt | I found my usb cord |
19:43.51 | kiozen | BabelO: what size do you need the track recording icons? |
19:44.28 | BabelO | kiozen: euh... |
19:44.41 | kiozen | now they are 48x48 |
19:44.42 | BabelO | kiozen: track icon on map ? |
19:44.49 | BabelO | don't know i did not move enough :) |
19:44.50 | kiozen | BabelO: no dialog |
19:45.07 | BabelO | kiozen: 32 should be enough |
19:45.14 | kiozen | ok |
19:46.30 | dcordes | tmzt: alright. which usb cord? |
19:46.47 | tmzt | no, for my usb hdd |
19:49.01 | tmzt | http://rafb.net/p/l11Ztf85.txt |
19:49.07 | tmzt | regular diff |
19:49.08 | tmzt | not git |
19:53.31 | kiozen | BabelO: svn up |
19:53.51 | kiozen | BabelO: Try to match the naming scheme with new icons |
19:54.14 | BabelO | kiozen: ok |
19:54.15 | kiozen | BabelO: no idea how to switch iconsize |
19:54.22 | kiozen | BabelO: other than ifdefs |
19:54.56 | dcordes | tmzt: Hunk #1 FAILED at 265. Hunk #3 FAILED at 352. |
19:55.18 | BabelO | kiozen: there is no scale for icon ? i think i use that somewhere |
19:55.18 | tmzt | not sure how to get a unified diff |
19:55.30 | ginge | -u |
19:55.55 | tmzt | and the filename at the top? |
19:56.03 | dcordes | if you are in the git tree you can do git-diff origin/htc-msm tsc2003.c > tsc2003.diff |
19:56.18 | tmzt | let me see if it has downloaded yet |
19:56.41 | kiozen | BabelO: IMHO it would be best to link the icons depending on the device selected |
19:57.00 | tmzt | no |
19:57.01 | kiozen | BabelO: but don't know if qtopiamake is able to do so |
19:57.07 | tmzt | i'll paste the file for now |
19:57.40 | BabelO | kiozen: else we can name the image in qrc file under a constant directory |
19:57.51 | BabelO | so the whole can be generic |
19:58.34 | dcordes | ok |
19:59.00 | tmzt | http://rafb.net/p/pBrYGW98.txt |
19:59.10 | kiozen | BabelO: you mean doing the string like: ICONDIR "apply.png" |
19:59.29 | kiozen | BabelO: with ICONDIR defined to ":/icons/32x32/" |
19:59.56 | BabelO | kiozen: yes something like that, so only in the structure detecting the size |
20:00.39 | kiozen | BabelO: the only problem is the qrc file. It has to keep all icons in mem, but IMHO this is not too bad |
20:01.06 | BabelO | kiozen: ok, maybe lpotter have another idea ? |
20:01.13 | BabelO | apt: ping lpotter |
20:01.13 | apt | pong lpotter |
20:01.45 | kiozen | BabelO: sure it will be: store the icons in the qpe file hirachy |
20:02.05 | kiozen | BabelO: but I prefere monolitic apps as long as possible :) |
20:02.13 | BabelO | kiozen: yes, same than you |
20:02.17 | BabelO | or at compile time |
20:02.22 | kiozen | BabelO: easier to maintain :p |
20:03.13 | kiozen | BabelO: recorded a 2hour track today |
20:03.46 | BabelO | kiozen: good test :) |
20:04.29 | tmzt | it was p1 for the pressure, right ginge? |
20:04.30 | kiozen | BabelO: just need a better gps mouse that static navigation make me nuts |
20:04.48 | ginge | tmzt: yes |
20:05.22 | BabelO | kiozen: that is strange because here, my house point move all over |
20:05.53 | kiozen | BabelO: because you do not have static navigation |
20:06.04 | BabelO | kiozen: but i have it :) |
20:06.18 | kiozen | BabelO: than you have a bad reception |
20:06.47 | BabelO | kiozen: 9 satellite. or maybe because it is not locked |
20:07.10 | *** join/#htc-linux dwaradzyn_ (n=gudz666@damian8155.net.autocom.pl) |
20:08.28 | dwaradzyn_ | which key constant in kernel/include/linux/input.c represents the tilde key? |
20:09.05 | dcordes | dwaradzyn_: thanks for testing the polaris kernel |
20:09.14 | BabelO | dwaradzyn_: not here, depend of the keyboard language setting |
20:09.45 | dwaradzyn_ | dcordes: no problem |
20:09.50 | tmzt | KEY_GRAVE, shifted? |
20:10.06 | ginge | ~ |
20:10.14 | ginge | depending on keymap |
20:10.17 | dcordes | tmzt: arch/arm/mach-msm/tsc2003.c:268: error: âABX_Yâ undeclared (first use in this function) |
20:10.21 | dwaradzyn_ | tmzt: i'll try, thanks |
20:10.24 | ginge | ABS_Y |
20:10.25 | tmzt | ABS_Y |
20:10.30 | dcordes | oh of course |
20:11.25 | tmzt | are you using loadkeys? |
20:15.10 | dcordes | ok built let's see.. |
20:26.57 | *** join/#htc-linux rmoravcik (n=rmoravci@ip-89-102-255-171.karneval.cz) |
20:31.46 | *** join/#htc-linux pronto (n=pronto@pool-71-179-232-167.bltmmd.fios.verizon.net) |
20:32.05 | dcordes | tmzt: alright I'm in the rootf |
20:32.14 | dcordes | with patched kernel |
20:32.17 | tmzt | ts_print_raw |
20:32.31 | tmzt | you have the tslib from yesterday right? |
20:32.39 | dcordes | yep nothing changed |
20:33.00 | dcordes | ok what ts_print_raw output are we interested in? |
20:33.03 | tmzt | see if you get ABS_X ABS_Y and ABS_PRESSURE in the flags now |
20:33.11 | tmzt | don't press anything |
20:33.18 | dcordes | pressure show's a value |
20:33.28 | dcordes | 16777216 |
20:33.41 | tmzt | what is absbit: |
20:33.59 | dcordes | tap screen? |
20:34.02 | tmzt | no |
20:34.07 | tmzt | just start ts_print_raw |
20:34.19 | dcordes | did it. no absbit: |
20:34.25 | tmzt | tap it once |
20:35.38 | *** part/#htc-linux okias (i=ejabberd@poseidon.kraja.cz) |
20:35.49 | dcordes | hm nothing |
20:36.26 | dcordes | the ABS_PRESSURE: stayed 16777216 after I taped (was printed 3 times) |
20:36.33 | dcordes | is that ok? |
20:37.39 | BabelO | kiozen: ok, doing to do a track now:) |
20:37.58 | kiozen | BabelO: don't get lost :) |
20:38.00 | BabelO | s/doing/going/ |
20:39.49 | BabelO | kiozen: i send you my actual map :) |
20:40.02 | kiozen | ok |
20:40.43 | dcordes | tmzt: can't see no "absbit: " |
20:46.45 | *** join/#htc-linux tmzt (n=tmzt@adsl-70-228-92-50.dsl.akrnoh.ameritech.net) |
20:47.13 | tmzt | can you try evtest, just the top |
20:49.10 | dcordes | tmzt: ok there are 4 Abs type event codes now |
20:49.24 | tmzt | I guess those ranges are long, not float, I just though 0 to 1 was floating point |
20:49.32 | dcordes | 0 (X), 1 (Y), 24 (Pressure), 33 (?) |
20:49.36 | tmzt | I guess we will have to figure out the range |
20:49.57 | dcordes | btw ts_calibrate exits now again on 5 taps |
20:50.11 | dcordes | but that was also in the bare git code |
20:51.04 | tmzt | can you get some values from evtest |
20:51.22 | dcordes | which do you need? |
20:51.42 | tmzt | Abs X, Abs Y, Pressure, BTN_TOUCH |
20:51.46 | dcordes | it's shitty to read with 240x320 instead of rotated |
20:51.48 | tmzt | top left, bottom right |
20:52.24 | tmzt | cat 3 > /sys/class/graphics/fbcon/rotate ? |
20:52.49 | dcordes | oh cool I thougt that needed rebooting |
20:52.59 | tmzt | it worked |
20:53.01 | tmzt | ? |
20:53.42 | tmzt | 1 should make it portrait again |
20:53.53 | tmzt | or 0, maybe |
20:54.13 | dcordes | yep works. no I also used 3 before to get the right rotation |
20:57.00 | dcordes | ok now to the pressing top left output. Abs: code 0 (X) 671, code 1 (Y) 2096, code 24 (Pressure) 864 Key: code 330 (Touch), value 1 |
20:58.05 | tmzt | ts_calibrate exits after you tap 5 times, but the targets don't move? |
20:58.33 | dcordes | Abs: code 0 (X) 3711, code 1 (Y) 3728, code 24 (Pressure) 1584 Key: code 330 (Touch), value 1 |
20:58.39 | dcordes | that is bottom right. |
20:58.45 | dcordes | tmzt: yes. |
20:59.13 | dcordes | erm of course, in both corners, after release: Key: code 330 (Touch), value 0 |
20:59.14 | tmzt | can you move /etc/pointercal and try again |
20:59.24 | dcordes | move? |
20:59.29 | tmzt | or delete |
20:59.37 | dcordes | k |
20:59.52 | dcordes | then run calib? |
21:00.07 | tmzt | yes |
21:00.14 | dcordes | ok. same behaviour |
21:00.26 | tmzt | what is pointercal now |
21:00.27 | dcordes | is the output not interesting? |
21:00.28 | tmzt | cat |
21:00.48 | dcordes | depending on where I press I think, the output after ending (pressing 5 times) ts_calibrate changes. |
21:01.23 | tmzt | did you tap in five different places? |
21:01.34 | tmzt | last one is center |
21:01.51 | dcordes | 5518 -1388 2515776 -9255 18985 -29450112 65536 |
21:02.02 | dcordes | that are values with random 5 taps |
21:02.11 | dcordes | I can tap where the crosshair _should_ move |
21:02.15 | tmzt | yes |
21:02.17 | dcordes | as I know it from gpe calibration gui |
21:02.31 | tmzt | we might try xtscal after this |
21:02.56 | tmzt | It looks like it gets its events through an x ext from tslib |
21:03.31 | dcordes | ok pointercal content now after pseudo correct calibration? |
21:03.36 | tmzt | yes |
21:04.05 | dcordes | -223 325 7394112 -1050 14262 -29735776 |
21:04.08 | tmzt | did you get any debug output from ts_print_raw? |
21:05.03 | dcordes | what's the debug messages again? |
21:05.13 | tmzt | is EV_SYN set? |
21:05.18 | tmzt | at the top |
21:05.20 | dcordes | yes |
21:05.23 | dcordes | 1 |
21:05.35 | tmzt | do you get ev.type 0 messages |
21:05.41 | BabelO | kiozen: done :) |
21:06.00 | BabelO | kiozen: perfect on the road :) |
21:06.01 | dcordes | yes and 1 and 3 |
21:06.06 | kiozen | BabelO: btw is that map scanned? |
21:06.26 | BabelO | kiozen: noo , chutt |
21:09.37 | tmzt | do you get a timestamp: X Y pendown? from ts_print_raw? |
21:10.32 | dcordes | it shows no pendown but abs ABS_PRESSURE: 16777216 |
21:10.35 | BabelO | kiozen: rebooting in wince, i ve a small but with usb ethernet :) |
21:11.34 | tmzt | ts_cal output, do you get xres=320 yres=240 ? |
21:11.36 | tmzt | or reverse? |
21:12.16 | tmzt | then, top left: X = ?? Y = ?? |
21:13.23 | tmzt | not the debugging stuff, the line from tslib |
21:13.25 | BabelO | kiozen: sent |
21:13.51 | tmzt | there should be lines that say <timestamp>: X Y [0 or 1] |
21:13.57 | tmzt | at least for my hpc |
21:14.08 | dcordes | tmzt: where now? |
21:14.09 | tmzt | in ts_print_raw |
21:14.15 | tmzt | when you tap |
21:14.42 | dcordes | in ts_cal output, I see no xres= yres= because the screen is flooded by the fprintfs on exit |
21:14.54 | tmzt | don't exit |
21:15.01 | tmzt | oh |
21:15.06 | dcordes | how'd I see anything w/o exiting |
21:15.06 | tmzt | 2 > tmp |
21:15.11 | dcordes | won't work |
21:15.12 | tmzt | 2> tmp |
21:15.17 | dcordes | and then? |
21:15.21 | tmzt | 2>tmp |
21:15.23 | tmzt | less tmp |
21:15.34 | BabelO | kiozen: why two file ? |
21:16.03 | dcordes | tmzt: I don't understand what that does, but the resulting file is empty |
21:17.13 | tmzt | (TSLIB_TSDEVICE=/dev/input/event0 ts_calibrate >tmp1 2>tmp2) && (sleep 30; killall ts_calibrate) |
21:17.53 | tmzt | 2 is the stderr, which is where the messages are, I think |
21:18.08 | kiozen | BabelO: that is just on the device |
21:18.31 | BabelO | kiozen: ok |
21:18.58 | BabelO | kiozen: i have some issue with GT if i play too much with zoom +/- |
21:20.03 | dcordes | ok trying that |
21:20.09 | BabelO | kiozen: track once downloaded look very good |
21:20.31 | kiozen | BabelO: did you download it to GT? |
21:20.51 | kiozen | BabelO: can you send a gpx? |
21:20.52 | BabelO | kiozen: i send you the file but i download it too :) |
21:21.04 | BabelO | kiozen: ah yes, i can |
21:21.15 | kiozen | BabelO: yes the two files ar M internal only |
21:21.42 | kiozen | you cant load them directly into GT |
21:21.54 | BabelO | kiozen: ah ok |
21:22.19 | BabelO | kiozen: where is the file on my disk ? |
21:22.35 | kiozen | BabelO: you have to save it |
21:22.44 | kiozen | right now it's just in mem |
21:23.58 | BabelO | shit :( |
21:24.19 | kiozen | you missed to save it? |
21:25.04 | BabelO | kiozen: yes because i see the download all menu |
21:25.10 | BabelO | i was thinking it save it for me |
21:25.23 | dcordes | tmzt: and now? |
21:25.26 | BabelO | saving it right now :) |
21:25.43 | dcordes | ok got the output in tmp1 |
21:25.56 | dcordes | x240 y320 |
21:27.59 | BabelO | kiozen: save menu ? |
21:28.02 | kiozen | yes |
21:28.09 | BabelO | where ? |
21:28.18 | kiozen | file -> save geo data |
21:29.10 | BabelO | kiozen: sent |
21:30.43 | kiozen | BabelO: I have a small n-s offset, you too? |
21:30.50 | BabelO | kiozen: ok looks good, but too fast |
21:31.18 | BabelO | kiozen: no at zoom x1 it looks ok |
21:31.54 | kiozen | BabelO: does it fit the roundabout perfectly? |
21:32.29 | BabelO | kiozen: ah roundabout is a "ronf point" in frecnh ? i was looking for the word :) |
21:33.05 | BabelO | kiozen: sorry "rond point" |
21:33.12 | kiozen | the one at N43° 42.234 E004° 07.337 |
21:35.22 | kiozen | BabelO: does it fit, or does it have an offset? |
21:36.06 | BabelO | kiozen: this one is ok |
21:36.13 | BabelO | kiozen: i m on the red line |
21:36.36 | kiozen | BabelO: can you send a screenshot? |
21:36.43 | BabelO | kiozen: yes |
21:36.49 | kiozen | BabelO: think my GDAL is buggy |
21:37.18 | ewon | any windows programmers in the house? |
21:37.34 | ewon | wondering if there is a callback function that tells you if a given machine is connected / disconnected to the internet |
21:37.37 | marajin | ewon: We sedatd them. |
21:37.45 | ewon | heh :) |
21:38.13 | marajin | They kept screaming "developer! developers! developers! developers!" |
21:38.16 | marajin | It had to be done |
21:38.54 | BabelO | kiozen: sent |
21:39.45 | kiozen | BabelO: ok, seems to be my GDAL version |
21:39.58 | dcordes | lol that old cokaine addict |
21:40.12 | BabelO | kiozen: ok, because i check the whole track and it looks ok |
21:40.13 | kiozen | BabelO: which one do you use (about->copyright) |
21:40.16 | dcordes | makes me wanna see http://youtube.com/watch?v=KMU0tzLwhbE |
21:40.33 | marajin | the song remix was fun |
21:40.38 | BabelO | kiozen: i do this maps with gdal version under windows and fwtools, and read under linux :) |
21:41.06 | BabelO | kiozen: gdal 1.4.2.0 from 27/06/2007 |
21:41.20 | BabelO | kiozen: proj4 is 460 |
21:41.32 | kiozen | BabelO: Intersting, I have to check tomorrow |
21:41.44 | BabelO | kiozen: ok :) |
21:42.00 | kiozen | ok time to sleep, 6:00 is quite close, bbl |
21:42.04 | BabelO | kiozen: so now we have a nice tool |
21:42.09 | BabelO | kiozen: good night |
21:51.34 | tmzt | what did you get from top left, bottom left, center? in tmp1 |
21:53.43 | dcordes | tmzt: what do you mean, tap those locations before? |
21:55.07 | tmzt | from ts_calibrate, in tmp1, you should get values for the five points, at least mine does |
21:55.26 | dcordes | oh got you |
21:56.13 | dcordes | what is n in "Took n samples..." ? |
21:57.26 | tmzt | it always takes more than one because you can't hold the stylus there steady enough |
21:57.31 | tmzt | how many samples? |
21:58.09 | dcordes | 3 2 2 2 2 |
21:58.14 | tmzt | let's try Xtscal and then I need to go for a few hours |
21:58.18 | dcordes | I taped tl tr bl br m |
21:58.29 | dcordes | want the values? |
21:58.35 | tmzt | yeah |
21:58.54 | dcordes | x = 1295 y = 2400 tl |
21:59.44 | dcordes | x = 3007 y = 3384 |
22:00.53 | tmzt | when you get back on bitbake, can you save the input-raw.so we made and the revert the changes to input-raw? |
22:01.22 | tmzt | save the changes in a backup .c file |
22:01.29 | tmzt | input-raw-debug.c |
22:02.53 | tmzt | we will need to change that zImage tsc2003.c to the right range for ABS_PRESSURE too |
22:03.10 | tmzt | do you have the five values? |
22:03.59 | dcordes | tmzt: you mean revert OE stuff back to normal? |
22:04.03 | dcordes | bitbake |
22:04.44 | tmzt | yes, save the changes to input-raw-debug.c |
22:04.53 | tmzt | and the input-raw.so file as input-raw-debug.so |
22:05.35 | tmzt | if you can, add it to the image as /usr/lib/ts/input-raw-debug.so |
22:09.43 | tmzt | can you do (TSLIB_TSDEVICE=/dev/input/event0 Xfbdev :0) & DISPLAY=:0 xtscal & (sleep 30; chvt 1) & |
22:10.40 | *** join/#htc-linux meredydd (n=meredydd@cpc4-cmbg2-0-0-cust471.cmbg.cable.ntl.com) |
22:11.27 | dcordes | tmzt: in the debug image? |
22:11.33 | dcordes | didn revert libts back yet |
22:11.44 | tmzt | yes, that is for later |
22:12.49 | meredydd | Argh. |
22:13.05 | dcordes | & DISPLAY=:0 xtscal or ; DISPLAY...? |
22:13.34 | dcordes | meredydd: arrrr |
22:14.24 | tmzt | (TSLIB_TSDEVICE=/dev/input/event0 Xfbdev :0) & (DISPLAY=:0 xtscal) & (sleep 30; chvt 1) & |
22:14.38 | meredydd | So, fool that I am, it takes me a while to figure out that the Vogue/Touch, while best-supported for Android, is a CDMA device |
22:15.29 | dcordes | don't worry |
22:15.44 | meredydd | So, given that I need GSM/UMTS...could anyone recommend the next-best one for Android support? |
22:15.57 | dcordes | hmmm kaiser? |
22:16.02 | meredydd | I see Kaiser and Polaris have the most noise being made about them |
22:16.12 | dcordes | in xda forum probably, yes |
22:16.26 | dcordes | but that's first of all no userland question, but a kernel one |
22:16.31 | meredydd | Well, I was here earlier and ginge was saying that Android userspace crashes on data send, for reasons he can't quite figure out |
22:16.36 | tmzt | any 7200a yet? |
22:16.39 | dcordes | tmzt: oops messed up the command didn't quit. |
22:16.48 | meredydd | dcordes: Indeed. |
22:16.54 | tmzt | quit? |
22:17.01 | tmzt | are yoou in xtscal? |
22:17.11 | meredydd | dcordes: ...except that it would appear that for some reason the Kaiser stuff plays especially poorly with the Android userspace. |
22:17.13 | dcordes | um I see a white underscode top left on black bg |
22:17.27 | meredydd | Anyway, was wondering how stable/featureful Polaris has become |
22:17.29 | dcordes | tjat |
22:17.36 | dcordes | *_* |
22:17.47 | tmzt | crosshairs? |
22:18.01 | dcordes | nah |
22:18.22 | dcordes | resetting the beast |
22:18.24 | tmzt | is this an openmoko image |
22:18.34 | dcordes | yes the one I'm on all the time |
22:18.39 | tmzt | see if you can make those changes, I will be back |
22:18.41 | tmzt | later |
22:18.43 | dcordes | and xtscal is present I checked before |
22:18.54 | dcordes | tmzt: which changes? |
22:19.05 | tmzt | is there a gpe image you can try? |
22:19.12 | dcordes | no but I sure can bake one |
22:19.23 | dcordes | you mean so I get the calib on bootup? |
22:19.38 | tmzt | just copy that plugins/input-raw.so to plugins/input-raw-debug.so |
22:19.49 | dcordes | I fear I have doubt it will start up anything but a blank xfbdev, just like openmoko one |
22:20.01 | tmzt | copy plugins/input-raw.c to plugins/input-raw-debug.c |
22:20.14 | dcordes | and then build me a gpe image? |
22:20.17 | tmzt | and restore plugins/input-raw.c to the oe version |
22:20.40 | tmzt | am wondering about the gtk version with xtscal |
22:20.55 | dcordes | I really doubt it will start |
22:21.05 | tmzt | yeah, probably |
22:21.13 | dcordes | in the moko image, when I run xserver-nodm which should normally start the whole gui, only blank xfbdev starts.. |
22:21.19 | dcordes | but will try anyway, |
22:21.26 | tmzt | let's just try reverting that, backup the tslib-1.0.0 or whatever and just restore from oe |
22:21.51 | tmzt | the whole directory like you sent me |
22:21.59 | dcordes | oks no problem |
22:22.02 | tmzt | but we want the modified module in the final image |
22:22.11 | tmzt | so we can change between them. |
22:22.17 | dcordes | ok will do |
22:22.30 | tmzt | I'll right, see you later than (a few hours) |
22:23.00 | dcordes | ok cya. not sure if I'm on then |
22:23.10 | dcordes | thanks again for the help |
22:55.04 | *** join/#htc-linux goxboxlive_ (n=goxboxli@31.84-48-212.nextgentel.com) |
23:00.56 | *** join/#htc-linux amiral_ (n=root@gob75-2-81-56-65-10.fbx.proxad.net) |
23:21.09 | *** join/#htc-linux marmotta (n=skodde@unaffiliated/skodde) |
23:24.32 | *** join/#htc-linux exco (n=exco@e181077068.adsl.alicedsl.de) |
23:25.04 | exco | hi |
23:26.34 | dcordes | moin |
23:29.13 | exco | slaps himself for upgrading to Intrepid Alpha |
23:29.34 | dcordes | on what system? |
23:29.53 | exco | Ubuntu 8.04 -> 8.10 |
23:30.14 | dcordes | I'm installing a debian on my future htpc |
23:30.19 | exco | my daily use notebook |
23:30.25 | dcordes | new ubuntu brings to much hassle to my desktop use |
23:31.08 | exco | It really looks ... 'interesting' |
23:31.26 | exco | and I hadn't planned on doing that today |
23:32.03 | dcordes | I dist upgrad'D my ubuntu once and never did again |
23:32.13 | exco | so bad? |
23:32.20 | dcordes | it destroyed the whole system |
23:32.27 | dcordes | rendering it unusable |
23:33.07 | exco | well... took me the last couple of hours to get it sort of working |
23:35.26 | exco | I just wanted to upgrade network-manager because it kept pissing me off loosing the passwords.... next thing you know ... 30 packages dl by hand to satisfy the dependencies ... messed up half my system and thought "what the heck" |
23:37.35 | *** join/#htc-linux dcordes_ (n=dcordes@unaffiliated/dcordes) |
23:52.01 | dcordes_ | exco: oh right the dist-upgrade also broke my apt |
23:53.25 | BabelO | good night |
23:53.38 | dcordes_ | night BabelO |
23:54.16 | *** join/#htc-linux pigeon (n=pigeon@60-241-137-179.static.tpgi.com.au) |
23:55.34 | *** join/#htc-linux exco1 (n=exco@e181077068.adsl.alicedsl.de) |
23:56.15 | dcordes_ | that debian net install is not too fast. looks like de2 was not the best mirror to pick |
23:56.32 | tcccp | dcordes_: esslingen is quite fast |
23:56.54 | tcccp | dcordes_: I tend to use it - especially when we have broadband (100MBit/s or more) |
23:57.01 | dcordes_ | thx |
23:57.04 | tcccp | np |
23:57.09 | dcordes_ | will keep that in mind |
23:57.38 | exco1 | I only get 800k upgrading... and I'm too lazy too call alice for my full 16mbit instead of only 8 |
23:58.52 | dcordes_ | exco: will they upgrade free? |
23:59.38 | exco1 | we'll I'm already paying for 16 |