| 00:01.20 | cr2 | tsdogs: we need to fix the "places" support. |
| 00:01.40 | cr2 | it's the only thing that i'm painfully missing in roadmap. |
| 00:01.44 | tsdogs | cr2: yes, what should we do there? |
| 00:02.11 | cr2 | it's done as an addon now. |
| 00:02.23 | cr2 | with some strange ascii input format |
| 00:02.46 | cr2 | i'd like to use .dbf as input |
| 00:02.53 | tsdogs | well, paul talked about taking utf8 from editor branch |
| 00:03.03 | cr2 | then the shapelib dbf support can be reused. |
| 00:03.16 | tsdogs | btw, does the editor support places? |
| 00:03.23 | cr2 | no. |
| 00:03.41 | cr2 | editor supports driving directions and routing. |
| 00:03.42 | tsdogs | cr2: ok but generating a dbf from some other format ... |
| 00:03.56 | cr2 | dbf can be generated with bash |
| 00:04.03 | tsdogs | ok |
| 00:04.09 | cr2 | using the tools included with shapelib |
| 00:04.15 | tsdogs | never looked really at it's internals :) |
| 00:04.19 | cr2 | if you have shapelib you have all these tools. |
| 00:04.48 | tsdogs | ok |
| 00:04.50 | cr2 | you can even add points/lines/polylines from the command line with these tools. |
| 00:05.28 | cr2 | i think paul's problem is the non-utf8 encoding of the data. |
| 00:05.48 | tsdogs | yes |
| 00:05.53 | cr2 | and the indexer does not support utf8 because of some bizarre pointer arithmetic. |
| 00:06.10 | cr2 | qt4roadmap supports utf8 natively. |
| 00:06.18 | cr2 | we did everything right. |
| 00:06.37 | tsdogs | cr2: so we should add some sort of new layer on top of the map to draw pois from dbf (something like the editor is doing with routing?) |
| 00:06.40 | cr2 | since the internal format is ucs-2 anyway |
| 00:07.00 | cr2 | no, the "places" will be added to .rdm |
| 00:07.11 | cr2 | it's routing that can be an addon |
| 00:07.23 | cr2 | that requires only minimal core support |
| 00:07.37 | tsdogs | cr2: with places u mean the data that I render now with roadmap_pois ? |
| 00:08.07 | cr2 | the routing plugin uses its own database, and asks to draw a polyline from point A to point B |
| 00:08.21 | cr2 | yes, but it's limited to the polygons now. |
| 00:08.40 | cr2 | we need some extra db. |
| 00:08.54 | cr2 | the "points" are already there. |
| 00:09.09 | tsdogs | ok so we must fight against one point labeling |
| 00:09.22 | cr2 | so we only need an additional dictionary with "place" names, and a link to the point db |
| 00:10.05 | cr2 | and it's necessary to fight against the label overlap |
| 00:10.26 | tsdogs | yes, that too. |
| 00:10.27 | cr2 | and polygon label rendering on all zoom levels. |
| 00:11.08 | tsdogs | cr2: u tryed latest code? I think I fixed some overlaping issue, though it's not a solution... |
| 00:11.53 | cr2 | i can easily adjust the zoom levels to use the osm/google zoom schema, but the real problem is to switch from cos(lat) scaling to mercator. |
| 00:12.00 | cr2 | i think yes. |
| 00:12.06 | tsdogs | ok |
| 00:12.18 | cr2 | at least the polygon colors and buffering work as expected. |
| 00:12.31 | cr2 | s/colors/label colors/ |
| 00:13.18 | cr2 | i think you should not draw them if the string bbox exceeds the polygon bbox |
| 00:13.30 | cr2 | independently of the label overlap. |
| 00:13.58 | tsdogs | cr2: it's already checked that one, there is a gap (don't remember how much in a define) |
| 00:14.22 | cr2 | i'll cvs up |
| 00:14.49 | tsdogs | ho, I made it even better :) it has a configuration option |
| 00:14.58 | tsdogs | detaults to 10 pixels |
| 00:15.33 | cr2 | yes, i'm missing some changes. |
| 00:16.09 | cr2 | you can also fix the files needed for the full install. |
| 00:16.09 | tsdogs | cr2: ok, I though that also if the label does not fit, we could draw a spot or something :) |
| 00:16.27 | *** join/#htc-linux CVirus (n=GoD@62.135.96.245) |
| 00:16.33 | tsdogs | u mean make install? |
| 00:16.40 | cr2 | then we'll bundle it with qlandkarteM by kiozen |
| 00:16.43 | cr2 | yes. |
| 00:17.31 | tsdogs | cr2: I have made a .pro file so we can build it with qmake and qtopiamake, though it builds only qtroadmap and qtroadgps |
| 00:17.36 | cr2 | if i'll fix the utf8 index, and the place will be supported it will be a very useful program already. |
| 00:18.04 | cr2 | ok, the installed files are a complete mess. even i run without toolbar icons ;) |
| 00:18.23 | cr2 | because i'l lost the track which files i should edit, and where is should put them |
| 00:18.47 | cr2 | even not talking about the broken right button / long click menues. |
| 00:18.49 | tsdogs | cr2: what I did not understand was: will the dbf be used for generating the rdm, or been kept as a separate file? |
| 00:19.13 | tsdogs | cr2: I think that having everything in .roadmap whould be nice :) |
| 00:19.19 | cr2 | i'm generating the map out of 3 shapefiles now. |
| 00:19.45 | cr2 | yes, .roadmap should be able to override everything. like with X11 |
| 00:20.16 | tsdogs | ok so paths and install are a task. |
| 00:20.36 | tsdogs | now what about my previous question on dbf |
| 00:20.36 | cr2 | 3 shapefiles are segments (=routable roads), lines (railroads, rivers, power lines et al.) and polygons(lakes, seas, forests, etc.) |
| 00:21.10 | cr2 | a shapefile has the geometry (points and point arrays = lines and borders) |
| 00:21.17 | cr2 | and a dbf |
| 00:21.24 | tsdogs | ok |
| 00:21.26 | cr2 | where you keep all the "attributes" |
| 00:21.37 | cr2 | like layers and names |
| 00:21.37 | tsdogs | I saw it when converting from mp |
| 00:21.55 | cr2 | so dbf db is supported natively |
| 00:21.57 | cr2 | already |
| 00:22.01 | dcordes | tsdogs, I tried the same inittab and statup.txt as they worked in the console image now in x11-image. But now it says can't open initial console after mounting the rootfs and I end up in the system messages w/o terminal. is it possible that now the terminal is reserved for x? |
| 00:22.26 | dcordes | all modules load fine though |
| 00:22.30 | tsdogs | dcordes: no idea |
| 00:23.01 | dcordes | tsdogs, I will just try a different terminal then. Do I do this by changing in inittab last line or in the startup? |
| 00:23.03 | cr2 | tsdogs: so i may need an extra .dbf/.shp for poi |
| 00:23.30 | tsdogs | cr2: ok but then I read data from rdm with a new structure for pois, right? |
| 00:23.59 | tsdogs | dcordes: yes |
| 00:24.17 | dcordes | I mean which of the both? |
| 00:24.26 | tsdogs | inittab |
| 00:24.51 | cr2 | tsdogs: not really. i think you'll only get a 'places' table |
| 00:24.56 | cr2 | for pois |
| 00:25.39 | tsdogs | cr2: ok so we need to render this new table |
| 00:25.39 | cr2 | where you can pick the names from some dictionary, and the point coordinates from the usual 'point' table, where all road/line/polygon points are stored |
| 00:26.00 | tsdogs | ok like I'm doing with polygons ... |
| 00:26.01 | cr2 | yes, like you do it with the 'landmarks' table |
| 00:26.16 | cr2 | yes, landmarks, not the polygons themself. |
| 00:26.37 | cr2 | you render the landmarks after all. |
| 00:26.39 | tsdogs | ok, still confusing me about rdm landmarks and gui ones :) |
| 00:26.49 | cr2 | because they are points and keep the names. |
| 00:27.08 | cr2 | gui landmarks are 'personal landmarks' |
| 00:27.19 | tsdogs | yes ;) |
| 00:27.24 | cr2 | they are stored as .gpx waypoints in your ~/.roadmap |
| 00:27.37 | cr2 | and these are 'polygon landmarks' |
| 00:27.49 | tsdogs | ok. |
| 00:27.52 | cr2 | there is one 'polygon landmark' for each polygon |
| 00:27.57 | cr2 | always |
| 00:28.11 | cr2 | it's the data model. |
| 00:28.20 | cr2 | very similar to the model used by grass |
| 00:28.53 | tsdogs | cr2: I think lines have landmarks too (street names) or are them polygons too after all? |
| 00:29.23 | tsdogs | cr2: ok I need to study rdm model :) |
| 00:29.27 | cr2 | hehe. i need to fix the declutter levels some time. |
| 00:29.34 | tsdogs | lol |
| 00:29.51 | cr2 | tsdogs: no, the streets don't have landmarks. |
| 00:30.19 | cr2 | they have their own labeling schema |
| 00:30.26 | tsdogs | hmm, ok |
| 00:30.36 | cr2 | only the points don't have labels, and we need to fix that. |
| 00:30.59 | tsdogs | cr2: btw, what do u think about gui moving around the map change I made? |
| 00:32.10 | tsdogs | actually it was already there, but used only by AGG version ... |
| 00:32.35 | cr2 | it's nice, but has some funny side effects |
| 00:33.11 | tsdogs | yeah I think it needs some more work :) |
| 00:33.53 | cr2 | if the polygon area is cut into pieces, you see each subdivision borders when moving the map. |
| 00:33.57 | tsdogs | maybe a config option whould be good. so we can enable it on the PDA |
| 00:34.18 | cr2 | it's nice for the "external" borders, but not for the internal |
| 00:34.36 | cr2 | but i think it's too difficult to make a distinction between them. |
| 00:34.43 | tsdogs | cr2: I was thinking on merging latest pateches, they should also fix this for what I understood |
| 00:35.00 | tsdogs | from mainline |
| 00:36.00 | cr2 | ok. |
| 00:36.27 | cr2 | btw, is it difficult to do the patterned brushes for the polygons ? |
| 00:36.42 | *** join/#htc-linux goxboxlive (n=goxboxli@208.84-48-176.nextgentel.com) |
| 00:37.05 | tsdogs | cr2: no, it's only a matter of responsiveness, but haven't really tested it on an heavy map |
| 00:37.06 | cr2 | it'll make everything even slower, but much nice at close zooms. |
| 00:37.30 | tsdogs | cr2: I think patterns are already there :) |
| 00:37.49 | cr2 | i'm also thinking about switching to the real osm zoom level model. |
| 00:38.04 | tsdogs | BrushStyle |
| 00:38.15 | cr2 | need to locate all the /1000000 places and replace them to LLA2RDM() |
| 00:38.38 | cr2 | but you obviously need some transparent .png for brushes ? |
| 00:38.50 | cr2 | i think GMT has a brush collection |
| 00:39.07 | cr2 | and there may be some user-defined collections. |
| 00:39.14 | tsdogs | hmm there is a Texture option, but I think I need to implement the loading of the texture |
| 00:39.20 | cr2 | like a 16x16*N .pngs |
| 00:39.58 | cr2 | so you load only 1 .png file |
| 00:40.22 | cr2 | and just pick the brush by its index. |
| 00:40.56 | cr2 | so load all brushes at once, it's not that much space |
| 00:40.59 | tsdogs | well brushstyle comes from QT4 so it can take anything from Solid to RadialGradient, though Gradient and Textures are not handled right now |
| 00:41.09 | goxboxlive | tsdogs: i give up. |
| 00:41.31 | goxboxlive | it complaines about missing ipkg-cl for creating a initramfs-image |
| 00:41.34 | cr2 | but we need to take a lot of care of zoom levels, to not overload the renderer. |
| 00:41.42 | tsdogs | goxboxlive: lol, I'm at glibc-intermidiate |
| 00:41.54 | goxboxlive | ohh that takes time :-) |
| 00:42.21 | cr2 | tsdogs: i have swamps as white holes now :) |
| 00:42.24 | tsdogs | Running task 208 of 2090 :) |
| 00:42.37 | tsdogs | cr2: lol |
| 00:42.59 | cr2 | on the light-rosa standard background |
| 00:45.50 | tsdogs | cr2: yes brush is loaded only once for every element to be rendered, then they are used to draw them |
| 00:46.40 | cr2 | it seems we have a layering bug. i got a lake "under" the forest |
| 00:46.50 | *** join/#htc-linux CVirus (n=GoD@62.135.96.245) |
| 00:47.18 | cr2 | it's bad to draw a forest in this way, but it should be easy to draw lakes after the forests. |
| 00:47.52 | tsdogs | :) unless they are islands |
| 00:49.12 | cr2 | strange. |
| 00:49.21 | cr2 | it's a problem only for one lake. |
| 00:50.16 | cr2 | yes, the polygon labels are ok now. |
| 00:50.19 | cr2 | :D |
| 00:50.26 | tsdogs | cr2: cool |
| 00:50.55 | tsdogs | cr2: do u still have the issue of strange chars rendering? |
| 00:52.47 | tsdogs | cr2: qtopia is building, oe is building (thoug I'm almost sure it'll fail) ... I'd say good night |
| 00:52.49 | cr2 | yes, but it's a font problem i gess. |
| 00:53.02 | cr2 | good night :) |
| 00:53.05 | tsdogs | cr2: I had it too (hmm) |
| 00:55.11 | tsdogs | maybe it's a buffer problem, I'll do some tests if i'll still get it. good night |
| 01:10.44 | *** join/#htc-linux dcordes_ (n=snoopdog@unaffiliated/dcordes) |
| 01:29.33 | *** join/#htc-linux Marex (n=marex@gwfm10-3-250.802.cz) |
| 01:33.58 | *** join/#htc-linux the_sys0p (i=the_sys0@gateway/tor/x-03693244e50f1bc2) |
| 02:42.22 | *** join/#htc-linux joschkopf79 (n=joschi@92.227.64.26) |
| 02:45.16 | *** join/#htc-linux TimRiker (n=timr@rikers.org) |
| 02:45.53 | *** join/#htc-linux TimRiker (n=timr@rikers.org) |
| 02:56.13 | *** join/#htc-linux noelcragg (n=noel@adsl-69-232-223-120.dsl.pltn13.pacbell.net) |
| 02:58.29 | noelcragg | cr2: htc sable kernel doesn't suspend correctly on my unit; screen backlight goes off, but some power seems to still be going to lcd |
| 02:58.38 | noelcragg | cr2: its as if half the scan lines have been turned off |
| 03:53.23 | *** join/#htc-linux ImCoKeMaN (n=imcokema@pool-72-70-180-20.hrbgpa.fios.verizon.net) |
| 03:56.05 | *** join/#htc-linux noelcragg (n=noel@adsl-69-232-223-120.dsl.pltn13.pacbell.net) |
| 04:30.26 | dereks | window stick on |
| 05:14.18 | *** join/#htc-linux ndavey (n=neil@60-240-227-199-vic-pppoe.tpgi.com.au) |
| 07:29.23 | *** join/#htc-linux SmallR2002 (n=SmallR20@79-77-75-194.dynamic.dsl.as9105.com) |
| 07:35.21 | *** join/#htc-linux rmoravcik (n=rmoravci@gtsgw.ttc.cz) |
| 09:00.00 | *** join/#htc-linux goxboxlive (n=goxboxli@208.84-48-176.nextgentel.com) |
| 09:18.51 | *** join/#htc-linux Zoolooc (n=fredsiba@nrbg-4dbfbc1f.pool.einsundeins.de) |
| 09:32.07 | *** join/#htc-linux kiozen (n=kiozen@rgnb-4db1dc7b.pool.einsundeins.de) |
| 09:40.12 | *** join/#htc-linux patoche (n=pat@77.197.52.155) |
| 09:56.00 | *** join/#htc-linux hlbot680 (n=adm@iclem.net) |
| 10:09.56 | *** join/#htc-linux BabelO (n=Fabrice@lun34-2-82-238-28-28.fbx.proxad.net) |
| 10:26.56 | *** join/#htc-linux pH5 (n=ph5@p5485E39E.dip.t-dialin.net) |
| 11:22.34 | *** join/#htc-linux rob_w (n=bob@M8b4e.m.pppool.de) |
| 11:35.15 | kiozen | cr2: ? |
| 11:39.00 | *** join/#htc-linux Marex (n=marex@gwfm10-3-250.802.cz) |
| 11:40.27 | *** join/#htc-linux patoche (n=pat@77.197.52.155) |
| 11:47.00 | *** join/#htc-linux tsdogs (n=tsdogs@net-84-18-144-39.c.bkom.it) |
| 12:02.33 | *** join/#htc-linux tsdogs (n=tsdogs@net-84-18-144-39.c.bkom.it) |
| 12:16.04 | *** join/#htc-linux kiozen (n=kiozen@rgnb-4db1dc7b.pool.einsundeins.de) |
| 12:23.37 | *** join/#htc-linux LunohoD (n=alex@e180077072.adsl.alicedsl.de) |
| 12:33.11 | *** join/#htc-linux SmallR2002 (n=SmallR20@79-77-75-194.dynamic.dsl.as9105.com) |
| 12:38.04 | *** join/#htc-linux CVirus (n=GoD@62.135.96.245) |
| 12:39.47 | *** join/#htc-linux patoche (n=patoche@77.197.52.155) |
| 12:55.59 | cr2 | hi kiozen |
| 12:57.10 | BabelO | hi |
| 12:57.20 | cr2 | hi BabelO |
| 12:59.49 | Marex | hi cr2, BabelO |
| 13:00.17 | pH5 | hi BabelO, cr2, Marex |
| 13:00.41 | Marex | hi pH5 |
| 13:00.48 | BabelO | pH5: i m not able to get SPP services on maemo :( |
| 13:04.54 | pH5 | BabelO: SPP is bluetooth serial profile as in DUN? or something different? |
| 13:06.05 | BabelO | pH5: it is different, just a serial profile |
| 13:06.25 | BabelO | but my artemis does not see any profile available on n800 |
| 13:25.12 | *** join/#htc-linux patoche (n=patoche@77.197.52.155) |
| 13:34.49 | *** join/#htc-linux CVirus (n=GoD@62.135.96.245) |
| 13:38.03 | *** join/#htc-linux SmallR2002 (n=SmallR20@79-77-103-47.dynamic.dsl.as9105.com) |
| 13:45.50 | BabelO | goxboxlive: how do you bypass the kernel build in oe ? |
| 14:03.52 | *** join/#htc-linux dcordes (n=snoopdog@unaffiliated/dcordes) |
| 14:10.10 | pH5 | BabelO: I think you'd have to remove the virtual/kernel and kernel-module-* dependencies from all tasks. |
| 14:11.07 | BabelO | ok |
| 14:11.24 | pH5 | or maybe using your own custom kernel recipe "nokernel.bb" http://en.pastebin.ca/951551 could work. |
| 14:12.20 | dcordes | hi all |
| 14:12.32 | pH5 | hi dcordes |
| 14:12.37 | dcordes | are you trying to build OE image w/o kernel? |
| 14:14.20 | dcordes | I came across same problem yesterday because hh won't build for me |
| 14:15.12 | BabelO | pH5: ok, it seem build continue like this, thx |
| 14:15.21 | BabelO | hi dcordes |
| 14:17.31 | BabelO | hm pH5 no, error come again, i think i stop to try OE, i lose to much time on that :( |
| 14:17.54 | dcordes | can't we try fixing hh in OE instead? |
| 14:19.20 | dcordes | BabelO, what machine are you trying to build OE? |
| 14:19.48 | BabelO | dcordes: blueangel |
| 14:20.50 | dcordes | BabelO, me too. Strange is: when I build DISTRO = "angstrom-2008.1" it doesn't work. with DISTRO = "openmoko" it worked |
| 14:22.47 | goxboxlive | pH5: why all this changes in OE. Wassent it good enough how it was? |
| 14:23.20 | pH5 | dcordes: I'm past the the openmoko qvga themes now, no problem reported. |
| 14:24.03 | dcordes | pH5, both icon-theme and theme? |
| 14:24.12 | pH5 | goxboxlive: why all those changes in the kernel ;) |
| 14:24.41 | pH5 | I like the idea of both sysroot and packaged staging a lot, because now the build system has a much more similar structure to a native build |
| 14:25.17 | pH5 | and finally the staging area is organized instead of just a place where every recipe built throws its stuff somehow. |
| 14:25.21 | pH5 | dcordes: both of them |
| 14:26.09 | dcordes | pH5, I entirely rebuild my |
| 14:26.31 | dcordes | whole OE folder from scratch two days ago and got errors |
| 14:27.17 | pH5 | strange. same here, only that I started three days ago (and still not finished) |
| 14:29.39 | dcordes | pH5, sed: can't read {32x32,48x48}/Makefile.am: No such file or directory <- same error as before. btw thos from o-hand included 22x22 icons to the non-qvga icon-theme |
| 14:31.49 | pH5 | dcordes: so try how installing the non-qvga theme works out |
| 14:32.11 | pH5 | you should clean out the svn checkout in the sources dir just to be sure |
| 14:32.48 | dcordes | I will remove both theme's workdir and stamps and try and rebuild them |
| 14:34.20 | pH5 | dcordes: "you should clean out the svn checkout in the sources dir just to be sure" |
| 14:34.31 | pH5 | for me it's in ~/src/oe/sources/svn/openmoko-icon-theme-standard |
| 14:36.23 | dcordes | pH5, is that your "download" dir? |
| 14:38.21 | pH5 | yes, DL_DIR = "${HOME}/src/oe/sources" |
| 14:39.58 | dcordes | pH5, in my OE/download/svn/ there's only svn.openmoko.org |
| 14:40.24 | dcordes | with trunk/src/target |
| 14:42.03 | BabelO | i ve started openmoko build, and i m already at task 1700, look good |
| 14:42.38 | BabelO | shit !!!!!!!! |
| 14:42.41 | BabelO | failed ! |
| 14:43.41 | pH5 | dcordes: hm, then maybe that's leftovers. in this case, it should be under trunk/src/taget/OM-2007.2/artwork/icons |
| 14:43.53 | pH5 | cleans out his DL_DIR |
| 14:44.31 | dcordes | pH5, ok I removed artwork/icons |
| 14:45.41 | dcordes | now rebuild icon theme? |
| 14:45.56 | pH5 | yes, give it a shot |
| 14:47.26 | dcordes | bitbake -c rebuild openmoko-icon-theme-standard2-qvga ERROR: function do_configure failed sed: can't read {32x32,48x48}/Makefile.am: No such file or directory |
| 14:47.59 | pH5 | ?! |
| 14:48.18 | dcordes | rm -rf src/target/OM-2007.2/artwork/icons/ <- before |
| 14:49.02 | pH5 | dcordes: I suggest bitbake -c devshell openmoko-icon-theme-standard2-qvga and have a look inside the work dir to see what's different to the contents of src/target/OM-2007.2/artwork/icons/openmoko-standard |
| 14:49.35 | tsdogs | pH5: why do I get this: NOTE: preferred version 2.6.1 of glibc not available? package was built/installed fine... |
| 14:50.14 | pH5 | tsdogs: ask RP about that :) |
| 14:50.22 | tsdogs | :) |
| 14:50.36 | pH5 | it is an age-old bitbake bug somehow related to the corresponding virtual package |
| 14:50.49 | pH5 | it doesn't seem to do any harm, but I think nobody figured out where this comes from yet |
| 14:51.14 | tsdogs | pH5: isn't it related to kernel build fails which is looking for uclibc? |
| 14:51.17 | pH5 | I certainly didn't understand how this happens when I looked at it. |
| 14:51.23 | *** join/#htc-linux TimRiker (n=timr@70.1.135.7) |
| 14:51.36 | pH5 | tsdogs: no. are you building in ANGSTROM_MODE="uclibc"? |
| 14:51.46 | dcordes | pH5, this is taking quite a while. |
| 14:51.49 | tsdogs | hmm, not in my local conf |
| 14:53.35 | tsdogs | lol: KERNEL_INITRAMFS_DIR = "${TMPDIR}/deploy/uclibc/images/${MACHINE}/" in angstrom-2008.1 |
| 14:57.03 | pH5 | tsdogs: do you read the mailing lists? |
| 14:57.18 | tsdogs | OE? nope |
| 14:57.33 | pH5 | :) might help avoid some surprises |
| 14:58.00 | cr2 | tsdogs: hehe. does not look like a very smart variable |
| 14:58.14 | tsdogs | haven't used it for some time now, but subscribing would be of help yes :) |
| 14:58.47 | tsdogs | cr2: :( nope |
| 14:59.07 | pH5 | the glibc kernel build needs an uclibc initramfs to be included, the whole process now is completely unstraightforward |
| 14:59.22 | pH5 | ANGSTROM_MODE="uclibc" bitbake initramfs-bootmenu-image |
| 14:59.25 | pH5 | bitbake virtual/kernel |
| 14:59.25 | cr2 | pH5: what is 'glibc kernel' ? |
| 14:59.43 | tsdogs | pH5: ok, that sounds good |
| 15:00.42 | tsdogs | is there a way to skip initramfs, I don't use it anyway ... |
| 15:01.04 | pH5 | cr2: the normal kernel, built in glibc mode. bitbake just doesn't support the concept of building against different c libraries and as the libc is a bit interwoven with the toolchain, the kernel is built inside the glibc or uclibc directories. |
| 15:03.04 | dcordes | pH5, NOTE: package openmoko-icon-theme-standard2-qvga-0.1.0+svnr4232-r2: task do_devshell: started <- doesn't seem to do anything |
| 15:03.19 | cr2 | pH5: ok, so it's about the build tools. |
| 15:05.04 | pH5 | cr2: yes, as gcc-cross is built against the glibc/uclibc that is also used for the rest in OE. |
| 15:05.32 | *** join/#htc-linux bernt (n=bernt@dslb-084-061-230-046.pools.arcor-ip.net) |
| 15:05.35 | goxboxlive | pH5: http://pastebin.no/3520 |
| 15:05.36 | pH5 | dcordes: I'd have expected that you get a terminal in the workdir. you can also look at the workdir manually. |
| 15:05.43 | tsdogs | pH5: -devel and -announce and -users, or devel is enough? |
| 15:06.04 | goxboxlive | I have to create a initramfs image with uclibc to have linux-handleds build with glibc. But initramfs image failes, do you know why? |
| 15:06.23 | pH5 | tsdogs: I only subscribed to -devel (angstrom and oe), no idea what vital information I'm missing from -users and -announce :) |
| 15:06.34 | tsdogs | lol |
| 15:07.58 | pH5 | goxboxlive: looks like you're a victim of the ipkg -> opkg transition somehow |
| 15:08.25 | goxboxlive | pH5: yes, but i started this out yesterday with a new tmp directory |
| 15:08.31 | cr2 | pH5: it seems i miss a lot :) what is opkg ? |
| 15:09.12 | pH5 | goxboxlive: try s/ipkg/opkg/ in initramfs-{bootmenu-,}image.bb, somebody just didn't update those yet |
| 15:09.42 | pH5 | cr2: fork of ipkg, they say it's maintained now and includes the patches we applied in oe anyway |
| 15:09.52 | pH5 | http://wiki.openmoko.org/wiki/Opkg |
| 15:10.02 | *** join/#htc-linux swc|666 (n=no@24.143.67.18) |
| 15:10.04 | dcordes | pH5, I just noticed it is rather useless to care about the qvga packaga since normal icons have the qvga now. It would be more efficient to have this defined somewhere so they get installed |
| 15:10.05 | pH5 | partial reason for the rename is probably this strange handhelds.org trademark issue |
| 15:10.33 | pH5 | dcordes: just change the THEMES_htcblueangel in task-openmoko-ui.bb to use the non-qvga icon theme. |
| 15:10.52 | dcordes | pH5, will then the small icons be picked from the non-qvga automagically? |
| 15:11.12 | pH5 | dcordes: no idea. but I'd be happy to hear about it :) |
| 15:11.28 | pH5 | should be used if the qvga gtk theme works correctly.. |
| 15:11.33 | swc|666 | hey guys.. i just accidentally ran make modules_install for a cross compiled kernel without an INSTALL_MOD_PATH on my development machine, now I have a 2.6.21-hh20 folder in /lib/modules.. is this going to screw up my machine? |
| 15:11.41 | goxboxlive | pH5: thanks, that was the trick. |
| 15:12.08 | dcordes | pH5, it is not working at all http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=1283 |
| 15:12.19 | pH5 | goxboxlive: ok, I'll push that change then |
| 15:12.35 | goxboxlive | great |
| 15:12.53 | pH5 | haha, nice |
| 15:12.58 | dcordes | pH5, have a look at the screenshots |
| 15:13.23 | dcordes | somebody gave me an image who managed to build the qvga icon theme so I tried and the result was not so good |
| 15:13.50 | dcordes | I doubt the qvga-theme will include the new icons correctly. first of all this mess has to be fixed I guess |
| 15:19.20 | *** join/#htc-linux swc666 (n=no@24.143.67.18) |
| 15:27.33 | kiozen | cr2: do you know a working tool to edit geotiff tags? |
| 15:28.38 | noelcragg | cr2: sable suspend status (from latest cvs): correctly powers off leds and display backlight, but seems to leave the lcd itself in a strange state |
| 15:29.23 | noelcragg | cr2: seems like every other scan line is still being drawn |
| 15:29.25 | cr2 | noelcragg: yes, i have a similar problem with n560 too... |
| 15:30.03 | *** part/#htc-linux swc666 (n=no@24.143.67.18) |
| 15:30.03 | noelcragg | cr2: the good news is that the power button at least is doing something: the pattern on the display changes slightly after the button is pressed |
| 15:30.04 | cr2 | kiozen: there is geotiffcp in the geotiff lib source |
| 15:30.27 | kiozen | cr2: I tried - not working |
| 15:30.47 | kiozen | no error output either :/ |
| 15:31.08 | cr2 | strange. but i've never used it myself. |
| 15:31.16 | kiozen | target file is just 8 byte |
| 15:31.21 | cr2 | kiozen: i'm writing the GCP tags myself now. |
| 15:31.39 | cr2 | manually. without libgeotiff. |
| 15:31.48 | cr2 | only plain tiff. |
| 15:32.02 | kiozen | I need to move the topleft ref point |
| 15:32.20 | noelcragg | What are people using for gui testing? I've been trying to use the htcblueangel daily builds from angstrom, but the TinyX server therein doesn't support a 240x240 framebuffer. Suggestions? |
| 15:33.06 | dcordes | pH5, did you look at the bug? |
| 15:33.29 | dcordes | BabelO, do sms work for you on BA? |
| 15:33.32 | cr2 | noelcragg: xfb should just read the data from fb itself. |
| 15:33.33 | pH5 | yes |
| 15:33.42 | pH5 | further action postponed until I can reproduce |
| 15:34.17 | cr2 | hmm. qt4 is available under gpl3 too, but qtopia gpl2. strange. |
| 15:34.40 | cr2 | recompiling 4.3.4 |
| 15:34.43 | pH5 | dcordes: I'm going to finish the magician image and test it while building the rest of the BA image, whatever the current differences are. |
| 15:34.43 | noelcragg | cr2: I thought so too, but seems not to work. Trying to get strace compiled to figure out where it's dying. |
| 15:34.47 | tsdogs | cr2: no qtopia is gpl3 from 4.3.x |
| 15:34.54 | tsdogs | and gpl2 also |
| 15:35.14 | cr2 | with all tiff and kernel hacking i've lost the track. |
| 15:35.17 | dcordes | pH5, ok. |
| 15:35.19 | cr2 | recompiling gdal too. |
| 15:35.27 | cr2 | tsdogs: ok. |
| 15:36.14 | cr2 | is gcc-4.3 available for arm ? |
| 15:36.56 | cr2 | some changes in the c++ code may be necessary |
| 15:36.58 | cr2 | http://gcc.gnu.org/gcc-4.3/porting_to.html |
| 15:37.35 | tsdogs | cr2: gcc-cross_4.3.0.bb is in OE |
| 15:38.16 | cr2 | ok. |
| 15:38.22 | cr2 | kiozen: i use TIFFSetField(out, TIFFTAG_GEOTIEPOINTS, ... |
| 15:39.18 | cr2 | kiozen: but you need to activate the private tags first. i don't know if you can overwrite existing tags. |
| 15:39.50 | cr2 | kiozen: it's all done with a normal tiff library. i can send you the program. |
| 15:41.28 | BabelO | dcordes: yes |
| 15:51.14 | *** join/#htc-linux pH5 (n=ph5@p5485E39E.dip.t-dialin.net) |
| 15:55.39 | *** join/#htc-linux the_sys0p (i=the_sys0@gateway/tor/x-0a89b9875e9c18d7) |
| 16:06.51 | BabelO | pH5: is it better to use kdrive or kdrive-w100 on ba ?, my openmoko compile just finish |
| 16:07.31 | dcordes | openmoko uses xserver-kdrive |
| 16:07.43 | dcordes | afaik also on ba |
| 16:09.32 | pH5 | BabelO: I think kdrive-imageon is state of the art, but if that doesn't work you should fall back to kdrive-w100 |
| 16:09.54 | pH5 | kdrive-fbdev doesn't have any acceleration |
| 16:10.22 | dcordes | pH5, did you yet check magician? |
| 16:10.38 | BabelO | pH5: ok good, i try like this first |
| 16:11.03 | BabelO | is looking for generated image, too much time without oe use |
| 16:11.42 | pH5 | dcordes: stop waiting for me, impatient man :) |
| 16:11.47 | *** join/#htc-linux rmoravcik (n=rmoravci@ip-89-102-255-171.karneval.cz) |
| 16:11.55 | pH5 | task 5240 of 5928, I'll get there today :P |
| 16:13.18 | cr2 | lol |
| 16:14.33 | *** join/#htc-linux kiozen (n=kiozen@rgnb-4db1dc7b.pool.einsundeins.de) |
| 16:16.21 | kiozen | cr2: GeoTIFFExaminer rules :P |
| 16:16.44 | cr2 | kiozen: what's that ? |
| 16:17.07 | kiozen | dark software from the other side |
| 16:17.51 | cr2 | ok. |
| 16:19.05 | BabelO | cross fingers |
| 16:20.42 | BabelO | lol, nice bootup screen |
| 16:20.54 | BabelO | but does not fit into |
| 16:25.55 | *** join/#htc-linux Raven-coda (n=wtg@64.50.35.215) |
| 16:28.39 | dcordes | BabelO, psplash ist nof qvga |
| 16:39.15 | cr2 | lol. now i'm missing the gcc-arm |
| 16:39.29 | cr2 | after removing 25GB of oe trash ;) |
| 16:41.22 | *** join/#htc-linux patoche (n=patoche@lns-bzn-56-82-255-254-232.adsl.proxad.net) |
| 16:52.02 | tsdogs | cr2: u can get the fic neo toolchain from qtopia.net |
| 16:52.39 | BabelO | pH5: i link Xw100 -> X then start it but it fail, missing xterm, is it normal ? |
| 16:53.00 | cr2 | tsdogs: good idea. where is it ? |
| 16:53.32 | tsdogs | cr2: http://www.qtopia.net/modules/mydownloads/visit.php?lid=38 |
| 16:53.34 | cr2 | tsdogs: i got a lot of warnings compiling roadmap with qt4-3-4 |
| 16:53.38 | cr2 | thanks. |
| 16:53.49 | tsdogs | cr2: hmmmm |
| 16:53.50 | cr2 | tsdogs: mostly const char * warning. |
| 16:55.05 | BabelO | pH5: hm Xw100 was not a good idea :) |
| 16:55.28 | pH5 | BabelO: no idea, what scripts are you using to start the server? the openmoko-image uses xserver-kdrive-common's Xserver script, so no renaming needed. |
| 16:55.34 | cr2 | buildplace_main.c has the most nasty warning, but this file need a major rewrite anyway. |
| 16:56.18 | BabelO | pH5: yes i ve upadated the script to run Xw100 when it is here |
| 16:56.39 | BabelO | pH5: but ;) Xw100 give me a bad screen with diagonal lines |
| 16:56.42 | pH5 | BabelO: no, it does that already |
| 16:56.46 | pH5 | no change needed |
| 16:57.10 | pH5 | whoops, my bad |
| 16:57.14 | BabelO | pH5: i did not see that in Xserver, no call to Xw100 |
| 16:57.17 | BabelO | ok :) |
| 16:57.19 | pH5 | you have to set IMAGEON="w100" for blueangel |
| 16:57.29 | BabelO | where ? |
| 16:57.44 | pH5 | in the "case `module_id` ..." block |
| 16:58.05 | pH5 | after that you have XSERVER=Xw100 or XSERVER=Ximageon in an if block |
| 16:59.05 | dcordes | BabelO, I did not change anything and x startet fine in openmoko htcblueangel |
| 16:59.33 | pH5 | dcordes: 5852 of 5928, just fyi :) |
| 17:00.13 | BabelO | pH5: hum, no still bad picture with w100 |
| 17:01.01 | dcordes | pH5, hehe easy I'm patient |
| 17:02.47 | cr2 | tsdogs: things like qt_main.cc:284: Warnung: deprecated conversion from string constant to »char*« |
| 17:04.32 | BabelO | pH5: fyi output of Xw100 http://pastebin.ca/951750 |
| 17:05.16 | dcordes | BabelO, is w100 accerlated BA driver? |
| 17:05.31 | BabelO | dcordes: yes |
| 17:05.40 | cr2 | BabelO: Initialized DMA ? |
| 17:06.20 | dcordes | BabelO, ah now I understand why alge image is so fast :D |
| 17:06.28 | BabelO | cr2: yes don't now what is that, screen is bad anyway, not usable, it is scrambled |
| 17:06.59 | BabelO | dcordes: don't think your or alge image use w100 |
| 17:07.04 | BabelO | w100 driver |
| 17:07.17 | dcordes | BabelO, how can I find out? |
| 17:07.17 | cr2 | Xw100 driver |
| 17:07.25 | cr2 | the kernel uses w100fb |
| 17:07.37 | cr2 | which is accelerated |
| 17:07.48 | pH5 | BabelO: no idea, I think Xw100 works more on hx4700, Ximageon does not. |
| 17:07.49 | cr2 | to some extent |
| 17:07.59 | tsdogs | cr2: ok but that's more a g++ thing :) |
| 17:08.05 | BabelO | pH5: ok |
| 17:08.42 | cr2 | tsdogs: this one is bad, but will not be a problem soon. |
| 17:08.44 | cr2 | buildplace_main.c: In function 'buildplace_dsg_add': |
| 17:08.44 | cr2 | buildplace_main.c:181: warning: cast to pointer from integer of different size |
| 17:09.25 | BabelO | cr2: what are correct screen dpi for ba ? |
| 17:09.38 | tsdogs | cr2: ok that's more like a x86_64 thing |
| 17:09.48 | cr2 | tsdogs: can use make compile in the shapelib subdir too ? |
| 17:10.02 | cr2 | BabelO: 3.5" lcd ? |
| 17:10.10 | tsdogs | cr2: no idea, u created that one |
| 17:10.21 | tsdogs | :) |
| 17:10.30 | cr2 | tsdogs: make lib |
| 17:10.35 | BabelO | cr2: yes |
| 17:11.03 | cr2 | tsdogs: and 'make all' for the utils |
| 17:11.19 | cr2 | like dbfadd, dbfdump |
| 17:11.25 | cr2 | dbfcreate |
| 17:11.48 | tsdogs | cr2: yes, but there are already some targets in makefile afair... |
| 17:11.52 | cr2 | shpadd, shpcreate, shpdump and shprewind |
| 17:12.10 | cr2 | are they called from main makefile ? |
| 17:13.13 | tsdogs | yes, runtime, build, but I dunno if they really work as espected |
| 17:13.59 | tsdogs | also install depends on build tools ... |
| 17:17.44 | *** join/#htc-linux kiozen (n=kiozen@rgnb-4db1dc7b.pool.einsundeins.de) |
| 17:30.22 | cr2 | kiozen: -co zlevel=9 is slow, but the file is better compressed. |
| 17:31.04 | dcordes | BabelO, any success with your X? |
| 17:38.59 | BabelO | dcordes: no, i start with standard fbdev |
| 17:44.46 | *** join/#htc-linux LunohoD_ (n=alex@e180064130.adsl.alicedsl.de) |
| 17:46.44 | kiozen | cr2: how much will be the speed penalty when reading the file? |
| 17:47.00 | kiozen | btw GL is slower on Win32 |
| 17:47.14 | cr2 | don't know. we need some objective measurement. |
| 17:47.19 | cr2 | -pg maybe. |
| 17:47.25 | cr2 | GT ? |
| 17:47.51 | kiozen | yes GT :) |
| 17:48.24 | cr2 | btw, even if i add the gcps in a latlong/weird datum as tags, gdalwarp does the right thing while warping. don't know why it does not work for you. |
| 17:48.39 | cr2 | 1.5.0 |
| 17:48.44 | BabelO | pH5: i m close to get the correct setup, but status bar usually on top ( the one with network / bt /battery ) override calendar button / dial and so.. and there is a black line on bottom |
| 17:48.47 | cr2 | i'll compile 1.5.1 now. |
| 17:49.17 | cr2 | kiozen: sorry, i've installed 1.5.1 already :) but it was the same in 1.5.0 |
| 17:50.06 | kiozen | cr2: my problem was the needed towgs84 param. it's not coded into the resulting tif |
| 17:50.36 | kiozen | but if you supply sane gcps, you don't need it :) |
| 17:52.09 | cr2 | kiozen: ok. |
| 17:52.43 | kiozen | removed GeoGridViewer this afternoon, GT is better now |
| 17:52.55 | cr2 | kiozen: somehow some datums already "know" their towgs84, but others not. |
| 17:53.05 | cr2 | lol |
| 17:53.38 | cr2 | it's the program delivered with TOPO ? |
| 17:54.05 | cr2 | kiozen: will test n560 gps now. |
| 17:54.23 | kiozen | cr2: mostly historic ones like Potsdam et al don't know and need local fixing |
| 17:54.29 | cr2 | was dumb enough to remove the toolchain today. |
| 17:54.38 | cr2 | ok. |
| 17:55.01 | cr2 | +datum=potsdam does not know ? |
| 17:55.02 | kiozen | yes GeoGridViewer is the medicro stuff you get with the maps. It's as good as Mapsource |
| 17:55.17 | cr2 | lol |
| 17:55.56 | kiozen | yes for Potsdam you need the BETA2007 correction factors to get good results |
| 17:56.18 | kiozen | without I would live at the wrong side of the river in front of my house |
| 17:56.24 | BabelO | + |
| 17:57.10 | kiozen | GeoGridViewer looks like win 3.1 and has no mouse wheel zoom |
| 17:57.14 | tsdogs | goxboxlive: yes, finally got Uni to boot qtopia again :) |
| 17:58.08 | tsdogs | new kernel and replaced tslib (from fic toolchain) |
| 17:58.22 | kiozen | can't wait to get the Loox back and to start, that OE / kernel /Qtopia compiling business again :/ |
| 18:01.14 | dcordes | bye |
| 18:06.39 | cr2 | kiozen: soon :) |
| 18:07.24 | kiozen | cr2: no rush, hope that all these current problems are fixed when I have to do it :P |
| 18:07.51 | cr2 | pH5: i don't see the max15xx on n560, but some pxa gpio sequence is used when the cpu speed is changed. is it possible to add this to the cpufreq infrastructure ? |
| 18:09.04 | cr2 | kiozen: it seems that many people using ozi don't have a clue when it comes to projections. |
| 18:10.13 | pH5 | cr2: you'd have to plug it in similar to what pxa27x_voltage does (register a cpufreq frequency transition notifier) |
| 18:11.05 | cr2 | kiozen: the only thing i may do reliably is to convert the datum to epsg code. the rest may need some manual changes. |
| 18:11.16 | cr2 | pH5: is it already done somewhere ? |
| 18:11.41 | kiozen | cr2: epsg is would be fine |
| 18:12.20 | cr2 | kiozen: but only for the 'gcs', not 'pcs' |
| 18:12.24 | pH5 | cr2: what? execute some code to change voltage before/after frequency changes? yes, pxa27x_voltage + pxa27x_voltage_max158xx do that for max158xx connected to pxa27x's power_i2c bus. |
| 18:12.27 | kiozen | I think I should clean up code, update the homepage and release GT |
| 18:12.44 | cr2 | pH5: ok, i'll have a look. |
| 18:13.30 | cr2 | kiozen: qlandkarte.sf.net still says 'GPSmap 76Cx (usb, untested)' |
| 18:13.49 | *** join/#htc-linux dcordes (n=snoopdog@unaffiliated/dcordes) |
| 18:13.54 | kiozen | yes, you have to wait for the next release ;) |
| 18:14.23 | cr2 | ok. then the Tbl10 too :) |
| 18:54.29 | *** join/#htc-linux dcordes (n=snoopdog@unaffiliated/dcordes) |
| 18:57.24 | kiozen | cr2: fixed Tbl10 |
| 19:03.52 | tsdogs | BabelO: smplayer is working, but vo of mplayer is fb :( |
| 19:04.17 | tsdogs | also need to fix size issues with app |
| 19:12.24 | *** join/#htc-linux CVirus (n=GoD@62.135.96.245) |
| 19:14.05 | cr2 | kiozen: will check now :) |
| 19:14.21 | cr2 | kiozen: GT does not accept epsg:54004 for projection ? |
| 19:16.01 | cr2 | -gcp c r x y z |
| 19:16.32 | cr2 | is also not accepted, but it's a legitimate gdal_translate syntax. |
| 19:16.55 | cr2 | c and r are ok, but x y z is put into the label |
| 19:22.20 | cr2 | kiozen: nice. but it spews some 2byte hexcodes to stdout. |
| 19:23.14 | cr2 | kiozen: the labels are ucs-2 ? |
| 19:27.43 | *** join/#htc-linux patoche (n=patoche@lns-bzn-56-82-255-254-232.adsl.proxad.net) |
| 19:43.52 | kiozen | cr2: add "#V1.0" to the first line of the gcp |
| 19:44.28 | kiozen | but probably won't work, too. |
| 19:51.58 | *** join/#htc-linux AlGe (n=alge@chello080109231226.4.uni-klu.teleweb.at) |
| 20:03.43 | dcordes | cr2, what's module name of blueangel keyboad driver? |
| 20:05.04 | *** join/#htc-linux patoche_ (n=patoche@lns-bzn-56-82-255-254-232.adsl.proxad.net) |
| 20:05.19 | AlGe | dcordes: blueangel_kp, but it is built-in not module |
| 20:05.44 | dcordes | AlGe, hey I just managed to get phone with your image working |
| 20:06.15 | AlGe | Nice, still the audio issue? |
| 20:06.39 | dcordes | yes right. I have to set otuput muxer to mixer everytime after boot. also mic is silent when set to internal |
| 20:07.44 | AlGe | I thougt I would find some time to incorporate an according alsa call into gomunicator, but still didnt find the time to think about it ... |
| 20:08.33 | dcordes | isn't there a way to save this alsamixer setting? |
| 20:08.46 | AlGe | yes, but it doesnt help. |
| 20:09.00 | AlGe | it gets reset to the wrong value every now and then |
| 20:09.08 | dcordes | ok anyway one time after boot setting it is ok. |
| 20:09.27 | AlGe | not for me, if I remember correctly |
| 20:11.02 | dcordes | AlGe, anyway thanks for the images. I spent much time trying to get phone working in openmoko but didn't work yet |
| 20:15.27 | dcordes | AlGe, did you ever get problems with the keyboard? |
| 20:16.04 | AlGe | yes, that infamous never ending repeat problem |
| 20:17.46 | AlGe | and sometimes the keypad dies silently and doesnt work until reboot. but this error occurs not so frequently |
| 20:21.08 | dcordes | AlGe, repeat problem here all time |
| 20:21.43 | *** join/#htc-linux TimRiker (n=timr@pdpc/supporter/bronze/TimRiker) |
| 20:22.26 | AlGe | we should ask BabelO if he has some time to look at this problem once again :-) |
| 20:23.54 | *** join/#htc-linux patoche (n=patoche@lns-bzn-56-82-255-254-232.adsl.proxad.net) |
| 20:24.28 | dcordes | AlGe, if one builds the keyboard driver as module, would it help to reload this module when repeat happens? |
| 20:25.43 | AlGe | I think so, but it is not too simple to configure all the platform devices as (re)loadable module. |
| 20:26.16 | AlGe | I tried this when I started to experiment with the BA and ended up with a non booting kernel |
| 20:31.04 | dcordes | won't try it then |
| 20:31.35 | dcordes | BabelO, how is openmoko going on BA? |
| 20:32.12 | dcordes | AlGe, does gommunicator directly speak to /dev/ttyS0? |
| 20:32.28 | AlGe | yes, |
| 20:32.40 | dcordes | and ttyS0 is provided by gsmd? |
| 20:33.01 | AlGe | no, it needs my small settihtc utility running before |
| 20:34.18 | AlGe | this only switches the line displine and keeps the device open, so that the line displine stays there |
| 20:35.36 | dcordes | and the rest is done by gsmd? |
| 20:36.02 | AlGe | no, there is no gsmd |
| 20:36.36 | AlGe | gomunicator does the rest (or it should, as I broke some sms functionality in the current version) |
| 20:38.08 | AlGe | suspend/resume stuff is handled by a script that I added to the settihtc package (http://www.algepop.net/users/alge/angstrom-ba/settihtc-0.1/) |
| 20:39.19 | *** join/#htc-linux tsdogs (n=tsdogs@62.123.180.130) |
| 20:39.25 | dcordes | AlGe, I would like to disable resume (boardid6 here) how would I go about this? |
| 20:41.03 | dcordes | AlGe, I disabled it in gpe-settings but when I plugin the power supply, sometimes it goes standby |
| 20:53.13 | BabelO | dcordes: nobetter screen with openmoko |
| 21:01.40 | dcordes | BabelO, do you have big icons? |
| 21:12.08 | BabelO | dcordes: no |
| 21:20.43 | dcordes | BabelO, can you make screenshot of mainscreen? |
| 21:22.10 | AlGe | dcordes: that "suspend on plugging into power" which I also can observe. maybe I did something wrong in http://www.algepop.net/users/alge/angstrom-ba/kernel-patches/ba-resume.patch |
| 21:23.29 | dcordes | AlGe, I don't understand c |
| 21:23.38 | BabelO | dcordes: screenshot with what ? |
| 21:24.19 | dcordes | BabelO, openmoko on your blueangel |
| 21:25.43 | dcordes | BabelO, it looks like this here http://bugzilla.openmoko.org/cgi-bin/bugzilla/attachment.cgi?id=558&action=view |
| 21:25.52 | dcordes | on my ba |
| 21:29.46 | dcordes | AlGe, would it be possible to adpat openmoko gsm system on top of your programs? |
| 21:32.55 | AlGe | I think so, when I first thougt of this I couldnt build it in oe last year, but this may have changed now. |
| 21:33.17 | AlGe | gsmd at least runs inmy image if you install it. |
| 21:36.05 | dcordes | AlGe, I just tried it http://pastebin.com/m4276b84d |
| 21:36.12 | dcordes | installed task-openmoko-phone |
| 21:43.37 | AlGe | dcordes: you should then at least stop settihtc: /etc/init.d/settihtc stop |
| 21:43.51 | AlGe | before starting gsmd |
| 21:48.51 | Raven-coda | Can someone give me some pointers on where to find info on low level kernel debugging? |
| 21:49.24 | Raven-coda | I got a kernel that won't boot :-( |
| 21:50.52 | dcordes | AlGe, I get the same output when I start gsmd after stopping settihtc |
| 21:51.55 | dcordes | libgsmd-tool says Can't connect to gsmd |
| 21:52.15 | dcordes | Raven-coda, what's your machine? |
| 21:52.25 | Raven-coda | htc wizard |
| 21:52.38 | dcordes | which kernel do you try to boot? |
| 21:52.40 | Raven-coda | working with the linwizard group |
| 21:52.56 | Raven-coda | 2.6.23 |
| 21:53.06 | dcordes | it's omap850 right? |
| 21:53.10 | Raven-coda | right |
| 21:53.50 | *** join/#htc-linux pH5|N810 (n=user@p5485E39E.dip.t-dialin.net) |
| 21:54.53 | dcordes | Raven-coda, where did you get it / how did you build it? |
| 21:55.26 | Raven-coda | https://sourceforge.net/projects/linwizard/ |
| 21:55.52 | Raven-coda | used a custom cross compiler |
| 21:56.22 | Raven-coda | I was able to build a run the 2.6.16 kernel from the site so I know my tool chainworks |
| 21:57.11 | dcordes | Raven-coda, yes 2.6.16..... doesn't mean 23 will work for you |
| 21:57.28 | dcordes | it won't make your phone more functional even if it would build |
| 21:59.28 | Raven-coda | k, my hope is to run android on the phone more as a proof of concept |
| 21:59.44 | Raven-coda | and it was my understanding that I need atleast 2.6.23 |
| 22:00.04 | dcordes | no matter what you will run on it you should wait for the device drivers to be ready. |
| 22:00.59 | Raven-coda | you mean the device drivers for omap850 |
| 22:01.09 | AlGe | dcordes: I'll try maybe tomorrow, rather sunday gsmd on a new sdcard image ... I think I got it working already ... |
| 22:01.35 | Raven-coda | ? |
| 22:02.20 | dcordes | AlGe, that would be nice |
| 22:02.40 | dcordes | additional sd is a good idea. atm I keep a dd of a working alge image :) |
| 22:09.52 | Raven-coda | so I shouldn't try to get newest kernel working? |
| 22:09.59 | dcordes | no |
| 22:10.13 | dcordes | it won't make your phone more functional even if it would build |
| 22:18.11 | Raven-coda | I just thought since the 2.6.23 has support for 850 why not use it rather then the 730 in 2.6.16 |
| 22:19.28 | dcordes | I didn't hear that sorry. can't help you much though |
| 22:20.02 | dcordes | you could state your experiences at xda wiki or linwizard sourceforge |
| 22:22.13 | Raven-coda | thx :( |
| 22:23.49 | dcordes | Raven-coda, well where exactly does it fail to build? |
| 22:23.58 | Raven-coda | it builds fine |
| 22:24.18 | Raven-coda | it just doesn't boot |
| 22:25.39 | dcordes | oh I see. how did you create your defconfig? |
| 22:26.01 | BabelO | pH5 pH5|N810: still around |
| 22:26.26 | Raven-coda | It's a striped down version of the defconfig used on the 2.6.16 |
| 22:26.40 | Raven-coda | I used xconfig and compaired the two |
| 22:27.13 | dcordes | Raven-coda, maybe you have missed something important. are you booting using haret? |
| 22:27.34 | Raven-coda | yeah, I used haret's logging but didn't see anything funny |
| 22:27.40 | dcordes | do you have a correct default.txt for haret? |
| 22:28.00 | cr2 | 0x34/2 |
| 22:28.25 | Raven-coda | angain just using a modified version of the 2.6.16 version that worked |
| 22:28.40 | Raven-coda | *again |
| 22:30.04 | Raven-coda | I prob am just missing something important I'm just trying to figure out troubleshooting steps |
| 22:31.21 | dcordes | Raven-coda, do you see something after haret go go go? |
| 22:31.27 | dcordes | does it just freeze? |
| 22:31.47 | Raven-coda | screen slowly fades to whit then soft resets |
| 22:32.07 | Raven-coda | no go go go last line |
| 22:32.26 | dcordes | Raven-coda, sounds liek a framebuffer problem to me |
| 22:33.02 | dcordes | are you sure everything provided in the 2.6.19 kernel from linwiz is also provided in your build? |
| 22:34.36 | Raven-coda | k, I think so. I git the omap repo first then add the linwiz server to the checkout |
| 22:50.33 | dcordes | BabelO, are you able to make calls in openmoko? |
| 22:50.39 | pH5 | yay, the images finished |
| 22:50.59 | pH5 | i guess testing is for tomorrow, then |
| 22:51.22 | BabelO | pH5: gsmd does not work with be :( |
| 22:51.37 | BabelO | pH5 gsmd -p /dev/ttyS0 -F -s 115200 |
| 22:51.56 | BabelO | pH5 and libgsm-tool -m shell or atcmd |
| 22:52.02 | BabelO | never get return |
| 22:52.23 | pH5 | anything interesting in /tmp/gsm.log? |
| 22:54.01 | dcordes | I think there is the basis to make it work. I installed gsmd on top of alge image and got this http://pastebin.com/m4276b84d pH5 |
| 22:54.08 | cr2 | hmm. new problems |
| 22:54.23 | dcordes | pH5, only installed openmoko-task-phone |
| 22:54.49 | dcordes | also in the OM images there was some kind of activity in the gsm panel icon |
| 22:55.07 | BabelO | pH5: nothing :( |
| 22:55.28 | BabelO | pH5 it is like if line disciplinei s not set or bad eol |
| 22:56.16 | cr2 | pH5: the magician_phone is huge... |
| 22:56.44 | dcordes | pH5, do you have your images available online? |
| 22:58.44 | pH5 | yuck, looks like my initramfs was borked. |
| 22:59.13 | pH5 | dcordes: I'll upload what I have when it at least boots, I only have 128kbit upstream here |
| 22:59.28 | dcordes | pH5, ok nevermind then |
| 22:59.48 | pH5 | cr2: never mind the suspend/resume parts. |
| 23:01.12 | cr2 | pH5: what do you think about arch/arm/mach-pxa/mioa701/mioa701_gps.c |
| 23:02.25 | cr2 | pH5: i'd like to stop the cloning of the code for applying power to XXUART. |
| 23:06.23 | pH5 | cr2: so you have sirf3, too? |
| 23:07.01 | pH5 | looks good to me. if you can just put the gpio numbers in platform_data you should use that, too. |
| 23:07.20 | cr2 | yes, sirf3. |
| 23:07.23 | cr2 | the same on athena |
| 23:08.07 | cr2 | pH5: can you create a small template ? |
| 23:08.28 | cr2 | it's late and my brain stops thinking logically ;) |
| 23:11.30 | pH5 | yeah, same here. |
| 23:11.38 | pH5 | let's see if we can come up with something working tomorrow |
| 23:11.42 | *** join/#htc-linux CVirus (n=GoD@62.135.96.245) |
| 23:12.49 | cr2 | ok. |
| 23:15.40 | pH5|N810 | good night |
| 23:34.22 | dcordes | is there a ctrl key on ba keyboard? |
| 23:38.50 | AlGe | dcordes: yes, "sym" key right to operator key (which is ESC) |
| 23:49.49 | cr2 | <1>[ 165.430000] Unable to handle kernel NULL pointer dereference at virtual address 00000004 |
| 23:50.01 | cr2 | <4>[ 165.450000] PC is at pxa_ir_configure+0x74/0xc0 |
| 23:50.04 | cr2 | oops |
| 23:54.06 | cr2 | BabelO: how did you disable IR on stuart ? |
| 23:57.09 | BabelO | cr2: STISR=1 |
| 23:57.31 | cr2 | hx4700 does the opposite, and i'm lazy to read the pxa263 dtaasheet |
| 23:59.37 | cr2 | btw, can you boot BA with linload.exe ? after the changes made by psokolovsky and fixes by Kevin2 i cannot anymore. |