| 00:47:17 | CIA-70: | gjm * r10583 /trunk/qgis/src/app/qgisapp.cpp: Fix a typo in the about dialog box text |
| 00:47:18 | sigq: | tracfeed: Changeset [10583]: Fix a typo in the about dialog box text <https://trac.osgeo.org/qgis/changeset/10583> |
| 01:03:19 | sigq: | tracfeed: Ticket #1639 ( enhancement closed ): New vector layer SRS <https://trac.osgeo.org/qgis/ticket/1639#comment:1> |
| 01:08:05 | CIA-70: | jef * r10584 /trunk/qgis/i18n/qgis_de.ts: update german translation |
| 01:08:06 | sigq: | tracfeed: Changeset [10584]: update german translation <https://trac.osgeo.org/qgis/changeset/10584> |
| 01:12:06 | alan_s: | howdy |
| 01:12:21 | alan_s: | is anyone familiar with the print composer? |
| 01:16:11 | doktoreas: | hello folks |
| 01:16:28 | doktoreas: | jef sorry for feature request.. |
| 01:17:09 | doktoreas: | So save as shapefile is not related to that setting |
| 01:17:09 | jef: | doktoreas: np. that's what you wanted, isn't it? |
| 01:17:29 | doktoreas: | sure, now I enabled always ask for CRS |
| 01:17:52 | doktoreas: | don't you think that shape as shapefile should follow the same way? |
| 01:18:14 | doktoreas: | basically once the layer is opened the user has already set a projection |
| 01:18:36 | doktoreas: | but don't know..just my 2c |
| 01:23:09 | jef: | doktoreas: probably yes. it's just that the current behaviour is bound to layer insertion. Is the saved layer also inserted? |
| 01:23:50 | doktoreas: | yep |
| 01:24:06 | doktoreas: | damn gotta run..sorry brb |
| 02:29:06 | schnuddelbuh: | hi all, on qgis.org the download links don't seems to work ( german page, firefox 3.0.8, ubuntu 8.10 )? |
| 02:52:54 | jef: | schnuddelbuh: even worse - there are no download links |
| 02:53:24 | jef: | schnuddelbuh: qgis.org |
| 02:53:37 | jef: | schnuddelbuh: um, http://download.qgis.org/downloads.rhtml |
| 02:53:38 | sigq: | Title: Quantum GIS Downloads ( at download.qgis.org ) |
| 02:55:05 | jef: | macho: huhu |
| 04:23:34 | macho: | jef: jep? |
| 04:24:10 | jef: | macho: schnuddelbuh's comment... |
| 04:24:10 | macho: | jef: german qgis page is not ready now .. there seems to be a problem with submenus |
| 04:24:28 | macho: | but english website should work |
| 04:25:03 | macho: | jef: cant find the bug .. |
| 04:26:23 | macho: | its surely somehow joomla related - but .. in german version submenues doesnt "display" |
| 04:26:38 | macho: | while in english they work as expected |
| 04:30:55 | doktoreas: | hello folks.. |
| 06:05:59 | CIA-70: | wonder * r10585 /trunk/qgis/src/plugins/oracle_raster/qgsselectgeoraster_ui.cpp: Fixed compilation error with CSLCount not being declared. |
| 06:06:52 | sigq: | tracfeed: Changeset [10585]: Fixed compilation error with CSLCount not being declared. <https://trac.osgeo.org/qgis/changeset/10585> |
| 08:27:46 | sigq: | tracfeed: Ticket #1639 ( enhancement reopened ): New vector layer SRS <https://trac.osgeo.org/qgis/ticket/1639#comment:2> |
| 08:30:30 | CIA-70: | kyngchaos * r10586 /trunk/qgis/mac/xcode/Qgis.xcodeproj/project.pbxproj: fix theme copy paths |
| 08:30:31 | sigq: | tracfeed: Changeset [10586]: fix theme copy paths <https://trac.osgeo.org/qgis/changeset/10586> |
| 08:36:37 | jctull: | gsherman: It would be nice to see that zoom patch included in 1.1 |
| 08:36:47 | gsherman: | agreed |
| 08:36:59 | gsherman: | but it may be too late : ) |
| 08:37:09 | jctull: | so, +1 from me fwiw ( not much ) |
| 08:37:19 | jctull: | yeah, maybe so |
| 08:37:36 | gsherman: | have to wait for the release managers opinion...if he gets back from the bush |
| 08:38:01 | jctull: | right, trekking in SA right now |
| 08:38:16 | jctull: | I wonder which is worse, a hungry grizzly or a hungry lion |
| 08:38:24 | jctull: | you two should sort that out |
| 08:38:30 | gsherman: | i'd call it a draw |
| 08:38:40 | gsherman: | although a bear won't always eat you... |
| 08:38:46 | jctull: | true |
| 08:39:09 | jctull: | cats, on the other hand... even a house cat will eat you |
| 08:39:17 | gsherman: | course that doesn't matter much if you are dead... |
| 08:39:39 | jctull: | wouldn't bother me much |
| 11:32:25 | TylerM: | hi all |
| 11:32:58 | TylerM: | i'm telling someone about running GRASS in QGIS.. should I be pointing them to a specific version of GRASS, or will it work with various ones? |
| 11:38:03 | wildintellect: | current qgis is bundled with Grass 6.4RC |
| 11:38:22 | TylerM: | bundled in windows i guess? |
| 11:38:35 | wildintellect: | well actually on most platforms |
| 11:38:49 | wildintellect: | the qgis-grass plugin is compiled against specific versions |
| 11:38:52 | TylerM: | hmm didn't see it on mac, but i might not have looked close enough : ) |
| 11:38:56 | TylerM: | okay, thanks alex |
| 11:39:07 | wildintellect: | if the want to use another version they'd have to compile the plugin themselves |
| 11:39:27 | wildintellect: | ah I'm not sure what comes on the mac build |
| 11:39:58 | TylerM: | i just had to download it separately tis all |
| 11:42:15 | wildintellect: | yes, I guess now on all platforms it's a separate download |
| 12:26:01 | timlinux: | morning all |
| 12:26:28 | timlinux: | springmeyer: how goes the qgis <-> mapnick hack? |
| 12:31:12 | springmeyer: | hey timlinux! |
| 12:31:29 | springmeyer: | well, going good |
| 12:32:13 | springmeyer: | gotta figure out how to render onto the map canvas though |
| 12:33:15 | timlinux: | springmeyer: how do you mean? |
| 12:33:25 | timlinux: | you draw rasters in mapnick? |
| 12:33:32 | timlinux: | and you want to place onto canvas? |
| 12:33:38 | springmeyer: | I mean can I hypothetically create my own QPainter object? |
| 12:33:46 | springmeyer: | and then do: mapCanvas.mapRenderer.render( MyPainter ) |
| 12:33:57 | timlinux: | yes |
| 12:34:01 | springmeyer: | assuming I have an image buffer from mapnik |
| 12:34:07 | springmeyer: | how to place that... |
| 12:34:20 | timlinux: | or you can paste your buffer into the mapcanvas' painter |
| 12:34:34 | springmeyer: | paste? interesting |
| 12:34:52 | timlinux: | using qpainter->drawImage( ) |
| 12:35:11 | timlinux: | or |
| 12:35:48 | : | * timlinux goes to poke in the source |
| 12:35:56 | springmeyer: | ewww, cool : ) |
| 12:39:01 | springmeyer: | last I was working on it |
| 12:39:19 | springmeyer: | I has created a QImage from the mapnik buffer |
| 12:39:27 | springmeyer: | stuck that in a QPixmap |
| 12:39:41 | springmeyer: | then created and new painter from that |
| 12:39:52 | springmeyer: | and tried to pass that off as an argument to the renderer |
| 12:40:11 | timlinux: | do you render exactly the same extents as the current aoi in the mapcanvas? |
| 12:40:17 | springmeyer: | but I must need to pass the pixmap image off to the existing painter? |
| 12:40:18 | springmeyer: | ya |
| 12:40:46 | springmeyer: | hmmm, maybe? self.canvas.paintEngine( ).drawPixmap( ) |
| 12:40:54 | : | * springmeyer boots up |
| 12:43:16 | timlinux: | springmeyer: I would think so yes |
| 12:43:33 | timlinux: | springmeyer: there just isn't really an accessor for the paintdevice itself |
| 12:43:43 | timlinux: | so Im just looking for the best way to get it |
| 12:43:58 | springmeyer: | cool, thank you |
| 12:44:01 | timlinux: | you kinda want to prevent the normal vector rendering process |
| 12:44:08 | : | * springmeyer qgis currently crashing python... |
| 12:44:32 | timlinux: | so Im wondering if creating a subclass of QgsMapCanvas or QgsMapRenderer ( probably the latter ) wouldn't be the way to go |
| 12:45:42 | timlinux: | and then just overload the render( ) routing |
| 12:45:45 | timlinux: | routine |
| 12:48:30 | timlinux: | springmeyer: otherwise the quick way you are playing with could get the qpainter like this: |
| 12:49:32 | timlinux: | QPainter * mypPainter = qGisInterface->mapCanvas( )->renderContext( )->painter( ); |
| 12:50:20 | timlinux: | QPainter * mypPainter = qGisInterface->mapCanvas( )->rendererContext( )->painter( ); |
| 12:50:36 | timlinux: | ( sorry mistake in first line above so use second ) |
| 12:50:44 | timlinux: | that assumes you are getting it via a plugin |
| 12:51:26 | timlinux: | which it probably a less tidy way to do it since you can't really suppress vectors drawing with the current renderers too |
| 12:53:39 | springmeyer: | hmm, is the renderercontext exposed in Python? |
| 12:53:47 | springmeyer: | trying via the python console... |
| 12:53:48 | springmeyer: | >>> iface.mapCanvas( ).rendererContext( ) |
| 12:53:48 | springmeyer: | exceptions.AttributeError: rendererContext |
| 12:53:58 | timlinux: | ah you are using python I forgot sorry |
| 12:54:10 | timlinux: | let me check |
| 12:55:18 | timlinux: | you would have to do: |
| 12:55:25 | timlinux: | ( scuse my crap python ) |
| 12:55:34 | springmeyer: | : ) |
| 12:55:43 | timlinux: | iface.mapCanvas( ).mapRenderer( ).rendererContext( ) |
| 12:56:00 | timlinux: | just checking it is exposed to python like that |
| 12:56:26 | timlinux: | yup it is |
| 12:56:44 | timlinux: | springmeyer: btw Im basing discussion off QGIS trunk |
| 12:57:04 | springmeyer: | : ) |
| 12:57:24 | wildintellect: | I hope that's what's being used since jef commited a fix just for springmeyer |
| 12:57:42 | springmeyer: | surely not : ) |
| 12:57:57 | springmeyer: | I'm using 1.1. from william's binaries |
| 12:58:17 | : | * springmeyer wouldn't know the first thing about compiling qgis |
| 12:58:47 | wildintellect: | ah on mac, I can't really help you, but yes the .database( ) you wanted is in trunk now |
| 12:59:02 | springmeyer: | yes, I saw those commits |
| 12:59:14 | springmeyer: | but not anything about exposing the renderContext( )? did you? |
| 12:59:36 | : | * springmeyer heads to the timeline |
| 13:00:25 | springmeyer: | ah ha! |
| 13:00:25 | springmeyer: | https://trac.osgeo.org/qgis/changeset/10467 |
| 13:00:26 | sigq: | Title: Changeset 10467 - Quantum GIS - Trac ( at trac.osgeo.org ) |
| 13:00:42 | timlinux: | springmeyer: btw 2: if you want to know what is really in the python api you can look in python/core/*.sip |
| 13:00:42 | timlinux: | and python/gui/*.sip |
| 13:00:42 | timlinux: | whatever is in those files should be in the compiled py bindings |
| 13:00:42 | timlinux: | sometimes people forget to add new methods to the sip binding files when they change classes |
| 13:01:05 | springmeyer: | okay, thanks |
| 13:01:54 | timlinux: | yup that has what you need |
| 13:32:48 | timlinux: | springmeyer: its pretty easy to compile on mac if you have williams frameworks |
| 13:33:04 | timlinux: | qgis cmake should pretty much pick up all deps |
| 13:33:12 | timlinux: | he's also added an xcode project into trunk |
| 13:33:26 | timlinux: | which I believe is pretty much automatic if you have his deps |
| 13:37:51 | springmeyer: | 'compiling 33 of 133 source files..' = ) |
| 13:38:30 | : | * TylerM waves |
| 13:38:33 | TylerM: | hi guys |
| 13:39:03 | TylerM: | springmeyer: you going to make a mapnik rendered in qgis? |
| 13:39:40 | racicot: | TylerM: he is working on it... |
| 13:39:47 | TylerM: | hi aaron - sweet |
| 13:40:09 | racicot: | I have a version of it working in its current state and am trying to help get it integrated in |
| 13:40:32 | TylerM: | I remember first time I rendered a large vector map in mapserver.. it took like 3 seconds.. but in arcmap/arcview took like 30.. |
| 13:40:38 | TylerM: | it was a mindbender to think of : ) |
| 13:40:53 | timlinux: | hi TylerM |
| 13:41:06 | springmeyer: | hey guys |
| 13:41:13 | TylerM: | happening place! : ) |
| 13:41:19 | springmeyer: | : ) |
| 13:41:20 | TylerM: | sounds good racicot |
| 13:41:43 | : | * racicot is working to try to get the output into a tab currently |
| 13:41:53 | springmeyer: | oh la la la la |
| 13:42:32 | timlinux: | quick off topic django question - is Book.objects.annotate( Count( 'authors' ) ) style summarising only in trunk? can I do something similar in 1.0? |
| 13:42:35 | TylerM: | i'm interested to learn sometime how mapnik reads in the qgis project info! |
| 13:43:04 | springmeyer: | nope, only trunk or upcoming 1.1 release |
| 13:43:09 | TylerM: | springmeyer: we talked about this a while back right? http://wiki.osgeo.org/wiki/OSGeo_Cartographic_Library |
| 13:43:10 | sigq: | Title: OSGeo Cartographic Library - OSGeo Wiki ( at wiki.osgeo.org ) |
| 13:43:12 | timlinux: | TylerM: springmeyer used microsoft .net universal anything to anything translator to do it |
| 13:43:29 | TylerM: | timlinux: ohhh sillly me.. Does it run on Mac? :D |
| 13:43:42 | springmeyer: | timlinux: huh? |
| 13:43:44 | : | * TylerM searches for his Silverlight implant... |
| 13:43:49 | timlinux: | TylerM: sure under win 3.1 for workgroups emulation mode |
| 13:44:08 | springmeyer: | : ) |
| 13:44:14 | timlinux: | springmeyer: sorry I was tapping into my innate ability to talk crap :- ) |
| 13:44:22 | TylerM: | !lart timlinux |
| 13:44:28 | timlinux: | @lart TylerM |
| 13:44:28 | : | * sigq pokes TylerM with a dog eared copy of 'The international guide to Morris Dancing' |
| 13:44:29 | TylerM: | come on sigq... |
| 13:44:31 | TylerM: | ouch |
| 13:44:35 | TylerM: | @lart sigq |
| 13:44:36 | : | * sigq pokes TylerM with a dog eared copy of 'The international guide to Morris Dancing' for trying to dis sigq |
| 13:44:36 | timlinux: | gotcha first :- ) |
| 13:44:43 | TylerM: | :o |
| 13:44:44 | TylerM: | biab |
| 13:45:02 | wildintellect: | I'm having connection issues can someone confirm they see this message? |
| 13:45:15 | timlinux: | wildintellect: yup you are here |
| 13:45:40 | timlinux: | springmeyer: is 1.1 officially released yet? |
| 13:45:48 | springmeyer: | no, delayed |
| 13:46:02 | timlinux: | ok |
| 13:46:05 | springmeyer: | I've been using it for a while though, that aggregation is awesome |
| 13:46:16 | timlinux: | yeah its just what I need :-( |
| 13:47:01 | springmeyer: | does QGis need sqlite compiled with RTree support? |
| 13:47:18 | : | * timlinux wants to distill this down to a single rec for each ip address showing latest visit from that ip: wget http://linfiniti.com/users/visitorkml |
| 13:47:36 | wildintellect: | the new spatialite included should have rtree already included |
| 13:47:42 | timlinux: | springmeyer: I believe so yes if you want spatiallite support ( new to upcoming 1.1 ) |
| 13:47:48 | springmeyer: | okay |
| 13:47:50 | TylerM: | I broke something |
| 13:47:56 | TylerM: | Linking CXX shared library libqgis_core.dylib |
| 13:48:02 | TylerM: | /usr/libexec/gcc/i686-apple-darwin8/4.0.1/libtool: can't locate file for: -lgcc_s.10.4 |
| 13:48:15 | timlinux: | TylerM: cmake build or xcode? |
| 13:48:20 | TylerM: | cmake |
| 13:48:29 | TylerM: | where do i slap it? |
| 13:48:57 | timlinux: | TylerM: did you just update for the first time in long time? |
| 13:49:13 | TylerM: | timlinux: not really, but most of my builds have been failing for last week or so |
| 13:49:25 | timlinux: | did you do any recent system upgrades? |
| 13:49:44 | TylerM: | hmm nothing that looked serious |
| 13:50:01 | TylerM: | but maybe did a week or two ago |
| 13:50:26 | TylerM: | my darwin ports hasn't been happy lately either :P |
| 13:50:32 | TylerM: | I'm pretty sure its a conspiracy |
| 13:50:33 | timlinux: | TylerM: try doing a make clean and rerunning ccmake .. |
| 13:50:39 | timlinux: | and then build again |
| 13:50:53 | timlinux: | its not really because I know what your problem is caused by |
| 13:51:13 | timlinux: | but it looks like some of your libs are moved / changed / gone |
| 13:53:53 | TylerM: | same.. of course : ) |
| 13:54:07 | TylerM: | i just did an svn update before trying today |
| 13:54:38 | TylerM: | I have to run, but if you know where the secret sauce is, I'd love to build this : ) but I'm sure it's alll my fault anyway : ) |
| 13:54:45 | TylerM: | good to see you timlinux |
| 13:55:15 | springmeyer: | TylerM: do you need stuff from macports for building qgis on your system? |
| 13:59:27 | timlinux: | cya TylerM |
| 13:59:28 | timlinux: | springmeyer: no you don't need it |
| 13:59:28 | timlinux: | ah on his specific system |
| 13:59:37 | timlinux: | maybe but in general its not needed to have any macports stuff to build qgis on mac |
| 14:00:09 | springmeyer: | right, okay, ya macports just seems like trouble to me |
| 14:01:02 | springmeyer: | hmm, I'm getting close here |
| 14:01:09 | timlinux: | ya me too I avoid them |
| 14:01:23 | springmeyer: | but the XCode project is failing to find my basic python headers |
| 14:01:27 | timlinux: | springmeyer: close to building qgis or close to making so mapnick in qgis magic? |
| 14:01:28 | timlinux: | ah |
| 14:01:52 | timlinux: | there is a settings file for the build in the xcode dir you can try to tweak |
| 14:01:55 | springmeyer: | error: Python.h : No such file or directory |
| 14:02:00 | springmeyer: | ya, been doing that... |
| 14:02:13 | springmeyer: | uncommented line about using leopard system py |
| 14:02:16 | springmeyer: | hmmm |
| 14:14:39 | springmeyer: | yes! |
| 14:14:40 | springmeyer: | >>> iface.mapCanvas( ).mapRenderer( ).rendererContext( ) |
| 14:14:40 | springmeyer: | <qgis.core.QgsRenderContext object at 0x12e51078> |
| 14:34:21 | springmeyer: | timlinux: still around? |
| 14:34:50 | springmeyer: | I'm built but doing nothing but crashing qgis now as I hack at the plugin again |
| 14:36:51 | springmeyer: | running in GDB via xcode gives "EXC_BAD_ACCESS" |
| 14:37:17 | gsherman: | springmeyer, that's no accomplishment, even i can crash qgis |
| 14:37:49 | springmeyer: | whew |
| 14:38:26 | springmeyer: | in the gdb term i get: |
| 14:38:27 | springmeyer: | Program received signal: “EXC_BAD_ACCESS”. |
| 14:38:27 | springmeyer: | warning: .o file "/Developer/SDKs/MacOSX10.4u.sdk/usr/lib/gcc/i686-apple-darwin9/4.0.1/libgcc.a( _eprintf.o )" more recent than executable timestamp |
| 14:41:37 | gsherman: | odd |
| 14:44:42 | timlinux: | springmeyer: yup still here |
| 14:45:33 | timlinux: | springmeyer: sounds like some clock skew |
| 14:45:52 | timlinux: | springmeyer: what if you do : sudo touch /Developer/SDKs/MacOSX10.4u.sdk/usr/lib/gcc/i686-apple-darwin9/4.0.1/libgcc.a |
| 14:46:14 | timlinux: | ( /me accepts no responsibility for breaking your system with that :-( ) |
| 14:46:24 | timlinux: | s/me/timlinux |
| 14:47:50 | springmeyer: | hmm, ya same error |
| 14:48:06 | springmeyer: | here, I'll paste my python code, see if I am doing anything fishy? |
| 14:48:16 | timlinux: | springmeyer: sure |
| 14:48:40 | timlinux: | bearing in mind my previous comments about python deficiencies |
| 14:48:45 | timlinux: | ( my own that is ) |
| 14:48:58 | gsherman: | @faq paste |
| 14:48:58 | sigq: | gsherman: Avoid flooding - paste your long clipboard contents at http://rafb.net/paste or http://pastebin.com |
| 14:49:02 | springmeyer: | http://dpaste.de/44K0/ |
| 14:49:03 | sigq: | Title: dpaste.de: Snippet #939 ( at dpaste.de ) |
| 14:49:07 | gsherman: | very good |
| 14:49:09 | wildintellect: | I'll see if I notice anything amok |
| 14:49:43 | gsherman: | first problem -- using mapnik |
| 14:50:12 | springmeyer: | hmm: second is that I uncommented lines 8 and 9 both |
| 14:50:18 | springmeyer: | I've been trying on or the other |
| 14:53:17 | wildintellect: | I'm a little confused by the mapnik_map which seems to go unused since you're calling it from self.mapnik_map instead |
| 14:54:12 | springmeyer: | ah, ya confusing |
| 14:54:30 | springmeyer: | one sec |
| 14:54:45 | springmeyer: | getting no crashes now... |
| 14:54:50 | springmeyer: | inching forward |
| 14:55:05 | timlinux: | springmeyer: Ill also need to check what canvas refresh does - i.e. if it wipes out the paint device before redrawing |
| 14:55:05 | gsherman: | wildintellect, good catch |
| 14:55:12 | timlinux: | or did you check that |
| 14:56:09 | timlinux: | springmeyer: 365 clear( ); gets called in qgsmapcanvas.cpp |
| 14:56:13 | timlinux: | during refresh |
| 14:56:26 | springmeyer: | okay, no crashes now with this: |
| 14:56:26 | springmeyer: | http://dpaste.de/FeOz/ |
| 14:56:27 | sigq: | Title: dpaste.de: Snippet #940 ( at dpaste.de ) |
| 14:56:32 | timlinux: | which may well result in you not seeing any mapnick goodness |
| 14:56:40 | springmeyer: | but also no effective drawing on the canvas |
| 14:56:44 | springmeyer: | right |
| 14:58:13 | timlinux: | try self.canvas.updateMap( ) |
| 14:59:05 | springmeyer: | okay, did so |
| 14:59:13 | springmeyer: | nothing changes |
| 14:59:14 | timlinux: | no that prollly didn't do it |
| 14:59:33 | timlinux: | ya its just a signal sorry |
| 14:59:50 | timlinux: | so is the map blank or drawn but wihtout your pic? |
| 15:00:02 | springmeyer: | unchanged |
| 15:00:12 | springmeyer: | existing layer stays put, no visible refresh |
| 15:00:51 | timlinux: | springmeyer: try this: |
| 15:01:34 | timlinux: | self.canvas.mapRenderer( ).render( self.canvas.mapRenderer( ).rendererContext( ).painter( ) ) |
| 15:01:54 | timlinux: | wait let me check if it clears if first there |
| 15:01:58 | springmeyer: | : ) |
| 15:02:07 | springmeyer: | actually just did that in the console before you pasted |
| 15:02:13 | springmeyer: | crash! : ) |
| 15:02:16 | timlinux: | ouch |
| 15:02:18 | springmeyer: | will try from plugin... |
| 15:02:29 | timlinux: | it doesn't clear the map that way so should in theory work |
| 15:02:50 | timlinux: | not sure how python stuff deals with pointers though |
| 15:03:01 | timlinux: | render( ) expects a QPainter * |
| 15:04:43 | springmeyer: | no change with that, should I call updateMap( ) after render? |
| 15:05:13 | timlinux: | you could try - it just sends a signal so I need to track what is listening for the signal |
| 15:05:44 | timlinux: | you could try doing: void saveAsImage( QString theFileName, QPixmap * QPixmap = 0, QString = "PNG" ); |
| 15:05:45 | timlinux: | on the canvas |
| 15:05:56 | timlinux: | and seeing if its rendered out properly in the file |
| 15:06:03 | springmeyer: | hmm, that's what prompts the crash, adding that line ( updateMap( ) ) |
| 15:06:04 | TylerM: | Do any of you know if, on Mac, we could create a bundle of libs and apps that all just reside in a folder instead of handling frameworks and the built in .app issues, etc? I'm thinking specifically for distributing demo packages. |
| 15:06:05 | timlinux: | ( just as a debugging measure ) |
| 15:06:20 | TylerM: | sorry, didn't mean to interrupt your non-english discussion ; ) |
| 15:06:30 | timlinux: | TylerM: Williams xcode build is supposed to do that for you |
| 15:06:38 | TylerM: | and I'm not just talking about QGIS. |
| 15:06:49 | TylerM: | I'm not quite understanding the benefit of xcode unfortunately :( |
| 15:07:14 | timlinux: | TylerM: you get to build qgis in a flash gui interface rather than a flash bash shell |
| 15:07:15 | gsherman: | TylerM, yes, you can create a bundle with *all* deps, no frameworks |
| 15:07:36 | timlinux: | TylerM: I actually have / had one that I made once |
| 15:07:45 | timlinux: | and they were built as universal |
| 15:07:51 | TylerM: | are you familiar with ye 'olde FGS for linux? |
| 15:07:55 | timlinux: | I still have the docs explaining the procedure more or less |
| 15:08:15 | TylerM: | similar to the concept of the OSGeo4W platform too |
| 15:08:17 | TylerM: | but for mac |
| 15:08:32 | TylerM: | unzip a package.. run an app, move on with life ; ) |
| 15:08:48 | TylerM: | http://www.maptools.org/fgs/ was one approach using linux |
| 15:08:49 | sigq: | Title: FGS.MapTools.org ( at www.maptools.org ) |
| 15:09:11 | TylerM: | self containing all needed libs etc and configured to work together under it's own tree from the root. |
| 15:09:21 | gsherman: | TylerM, piece of cake - let me know when you're done so i can test it |
| 15:09:33 | TylerM: | ok cool... |
| 15:09:39 | TylerM: | sending now via PSK31 |
| 15:16:52 | timlinux: | springmeyer: what's happening there in the trenches? |
| 15:20:02 | springmeyer: | back to crashing : ) |
| 15:20:43 | springmeyer: | I'm thinking I've got a bogus build |
| 15:21:32 | jctull: | springmeyer: if you are up to date on william's frameworks, the build is pretty simple |
| 15:21:47 | jctull: | if you want some tips... |
| 15:22:49 | jctull: | do you have cmake installed? |
| 15:23:24 | springmeyer: | ya, I'm up to date, just for this, and installed fresh cmake |
| 15:23:35 | jctull: | my cmake command is thus: |
| 15:23:36 | jctull: | cmake -D CMAKE_INSTALL_PREFIX=/Applications -D CMAKE_BUILD_TYPE=Release -D GRASS_PREFIX=/Applications/GRASS-6.4.app/Contents/MacOS ..' |
| 15:24:07 | springmeyer: | I used the Xcode project |
| 15:24:11 | jctull: | everything else gets picked up automatically, although I use the stock python framework, so not sure if you have upgraded |
| 15:24:17 | springmeyer: | here is an example crash: |
| 15:24:19 | springmeyer: | http://dpaste.de/BzQ6/ |
| 15:24:20 | sigq: | Title: dpaste.de: Snippet #941 ( at dpaste.de ) |
| 15:24:50 | jctull: | did you edit the qgis_user.xcconfig? |
| 15:24:59 | jctull: | and are you using the stock python? |
| 15:25:00 | springmeyer: | yes, to use the system python |
| 15:25:33 | jctull: | hmmm, well, you should try cmake. There may be a change to the xcode project that William needs to make based on today's updates in trunk |
| 15:25:35 | timlinux: | springmeyer: I wonder if its receiving a pointer to a qpainter |
| 15:25:41 | timlinux: | 0x0027664c QgsMapRenderer::render( QPainter* ) + 162 <-- there |
| 15:25:56 | timlinux: | springmeyer: did you build with Debug? |
| 15:26:05 | springmeyer: | not sure |
| 15:26:06 | jctull: | yeah, that too. I was just getting to that. ; ) |
| 15:27:42 | jctull: | let me try to update with the xcode project. It worked last night for me. |
| 15:28:41 | jctull: | This will take about 9 mins ( almost twice as long as cmake ) |
| 15:29:12 | timlinux: | springmeyer: change jctulls options above to: -D CMAKE_BUILD_TYPE=Debug |
| 15:30:22 | jctull: | ...if you use cmake to build |
| 15:30:47 | timlinux: | yup |
| 15:31:10 | timlinux: | ok I have to make like springmeyer's app....crash |
| 15:31:15 | timlinux: | cya tomorrow folks |
| 15:31:19 | jctull: | timlinux: take care |
| 15:31:25 | springmeyer: | thanks for the help |
| 15:31:31 | springmeyer: | I'm about to roll too |
| 15:31:50 | timlinux: | np |
| 15:31:59 | jctull: | springmeyer: do you get the crash when you start Qgis.app or are you doing something else? |
| 15:32:56 | springmeyer: | well, I'm trying to override the standard renderContext( ) from python |
| 15:33:08 | springmeyer: | very 'liminal' proposition |
| 15:33:24 | jctull: | are there any commands for me to try and duplicate your crash once my build finished? |
| 15:33:32 | jctull: | s/finshed/finishes |
| 15:33:45 | springmeyer: | ya, one... |
| 15:35:24 | springmeyer: | load up a shapefile |
| 15:35:27 | springmeyer: | open the python console |
| 15:35:31 | springmeyer: | paste: |
| 15:35:32 | springmeyer: | p = iface.mapCanvas( ).mapRenderer( ).rendererContext( ).painter( ) |
| 15:35:53 | jctull: | ok, seemed to work on last nights build for me |
| 15:36:08 | springmeyer: | iface.mapCanvas( ).mapRenderer( ).render( p ) |
| 15:36:26 | jctull: | so far so good |
| 15:36:38 | springmeyer: | that's odd syntax but crashes me every time |
| 15:36:40 | jctull: | ...with a blank canvas anyway |
| 15:36:51 | springmeyer: | and is similar to the objects I'm working with |
| 15:37:00 | springmeyer: | jctull: what about with some data? |
| 15:37:07 | jctull: | trying now |
| 15:37:16 | jctull: | spinning beachball, then crash |
| 15:37:30 | CIA-70: | borysiasty * r10587 /trunk/qgis/python/plugins/plugin_installer/ ( 5 files ): Plugin Installer update - cleaning and refactoring |
| 15:37:31 | sigq: | tracfeed: Changeset [10587]: Plugin Installer update - cleaning and refactoring <https://trac.osgeo.org/qgis/changeset/10587> |
| 15:38:29 | jctull: | let me try a debug build with cmake... |
| 15:39:16 | jctull: | give me 5 minutes |
| 15:39:19 | jctull: | if you can |
| 15:42:44 | TylerM: | i think i've broken my c++ compiler on this mac.. fun fun : ) |
| 15:42:49 | TylerM: | probably deleted it when i shouldn't have hee |
| 15:43:51 | gsherman: | deleting it is one way to break it |
| 15:44:41 | jef: | springmeyer: hm, what are you trying to do? render( ) actually uses the painter to setup a QgsRenderContext. |
| 15:45:46 | springmeyer: | hey jeff, trying to do something I probably shouldn;t via python : ) |
| 15:46:03 | springmeyer: | trying to creater a Painter and render it |
| 15:46:24 | springmeyer: | er, more specifically trying to draw on the existing painter |
| 15:46:29 | springmeyer: | but I;ve tried both |
| 15:50:08 | jctull: | springmeyer: cmake build, unsurprisingly, crashes also |
| 15:51:19 | jctull: | http://rafb.net/p/FfyGvm62.html |
| 15:51:20 | sigq: | Title: Nopaste - No description ( at rafb.net ) |
| 15:52:23 | jef: | springmeyer: looks like there is no existing painter. painter is created before calling render( ), passed in, copied into the context, used for rendering and when render( ) returns, the caller can clean it up. so painter( ) is invalid in the context you call it - just residue from the last render run. |
| 15:53:44 | springmeyer: | okay, sounds like a 'crash course' : ) |
| 15:53:56 | springmeyer: | so jef: big picture question then... |
| 15:54:58 | springmeyer: | if I have an image and I want to put it in the map canvas how can I do it ( in a way that will re-request the image each time the map update signal is called ) ? |
| 15:56:35 | jctull: | springmeyer: mapnik needs cairo? that's why you have macports... |
| 15:56:57 | springmeyer: | fully optional |
| 15:57:02 | TylerM: | macports is pretty sweet usually |
| 15:57:10 | springmeyer: | but ya, macports helps with cairo |
| 15:57:20 | gsherman: | incoming... |
| 15:57:23 | springmeyer: | use CAIRO=False |
| 15:57:23 | jctull: | I quit it long ago... too many problems were arising for me |
| 15:57:31 | springmeyer: | to skip it |
| 15:57:44 | jctull: | the... zoom ...patch... |
| 15:57:45 | jctull: | ? |
| 15:58:01 | TylerM: | you quite cairo? or macports? |
| 15:58:11 | jctull: | macports |
| 15:58:19 | jctull: | I do not have cairo currently installed |
| 15:58:35 | TylerM: | something toasted mine.. now I can even install the macports base upgrade |
| 15:58:36 | jctull: | I think I have not had luck with it in the past ( cairo ) |
| 15:58:38 | CIA-70: | gsherman * r10588 /trunk/qgis/ ( 13 files in 7 dirs ): |
| 15:58:38 | CIA-70: | Applied patch from ticket #1522 for Zoom Next functionality. |
| 15:58:38 | CIA-70: | Includes icons for zoom next tool ( same icon for all themes ). |
| 15:59:00 | CIA-70: | gsherman * r10589 /trunk/qgis/images/themes/gis/mActionZoomNext.png: |
| 15:59:00 | CIA-70: | Applied patch from ticket #1522 for Zoom Next functionality. |
| 15:59:00 | CIA-70: | Includes icons for zoom next tool. |
| 15:59:08 | jef: | springmeyer: hm, that's just what I got from backtracing the crash. I didn't dig deep into rendering yet. |
| 15:59:32 | : | * gsherman completes his commit for the year |
| 15:59:55 | TylerM: | \o/ |
| 16:00:36 | : | * gsherman seems to boggled up the commit message in the transision from git to svn... |
| 16:00:40 | jctull: | That's infinitely more commits than you will see from me |
| 16:00:44 | gsherman: | heh |
| 16:00:58 | gsherman: | that's like two or three for me this year, including a bug fix |
| 16:01:42 | jctull: | see, you finish your book and you have all kinds of free time |
| 16:01:43 | TylerM: | .. yeah time to back off a bit |
| 16:01:49 | TylerM: | heh |
| 16:02:04 | gsherman: | hmm, is that what you call it? |
| 16:02:05 | gsherman: | : ) |
| 16:03:06 | jctull: | Seriously, if we could just get you off your 'American Idol' obsession... ; ) |
| 16:03:55 | gsherman: | more like if 'i could remember how to write code...' |
| 16:04:19 | gsherman: | last commit was a patch, i didn't write it, just applied and tested it |
| 16:04:28 | jctull: | it fades quickly... I used to be good with Pascal, several decades ago... |
| 16:04:35 | gsherman: | btw, that commit was "blessed" by the release manager |
| 16:04:36 | jctull: | well, maybe not good |
| 16:04:52 | jctull: | I saw the email and was waiting with baited breath for the commit! |
| 16:05:01 | gsherman: | i know about 15-18 programming languages, but i can only remember one at a time. |
| 16:05:04 | jctull: | let me build it now |
| 16:05:29 | gsherman: | oh, you expect it to build? |
| 16:05:34 | jctull: | : ) |
| 16:06:19 | jctull: | Made it to 10%, 15%... |
| 16:11:11 | gsherman: | oh, well, that's better than 5% |
| 16:12:36 | jctull: | 83%... |
| 16:14:35 | jef: | springmeyer: maybe like the copyright plugin does: connect to the renderComplete( QPainter * ) signal of mapCanvas( ) and drawPixmap( ) your image to that painter. |
| 16:18:20 | jctull: | gsherman: now I just have to figure out what to do with this zoom to next feature : ) |
| 16:18:38 | gsherman: | jctull, zoom in a few times and then start clicking zoom last/next |
| 16:20:16 | jctull: | That's great. What does refresh do, or is that old? |
| 16:22:31 | gsherman: | been there since the dawn of time |
| 16:22:38 | gsherman: | down of time == 2003 |
| 16:23:40 | jctull: | ok, well it still seems to do nothing, but the others work. |
| 16:23:57 | gsherman: | refresh just redraws the canvas |
| 16:23:57 | jctull: | thank you for this. It is something that I greatly need, and everyone else I'm sure |
| 16:24:33 | gsherman: | it may be as useful and selling ice cubes in the arctic at this point |
| 16:24:35 | jctull: | positive karma coming your way ( fresh out of cash ) |
| 16:24:54 | gsherman: | cash is overrated |
| 16:25:03 | gsherman: | i need more ohloh karma |
| 16:25:20 | jctull: | ok, bbl |
| 16:39:22 | CIA-70: | jef * r10590 /trunk/qgis/i18n/qgis_de.ts: german translation update |
| 16:39:49 | sigq: | tracfeed: Changeset [10590]: german translation update <https://trac.osgeo.org/qgis/changeset/10590> |
| 18:22:00 | springmeyer: | WOOOOT! |
| 18:22:30 | gsherman: | springmeyer, find a quarter on a ground? |
| 18:23:04 | springmeyer: | just came back and saw jef's hint |
| 18:23:06 | jctull: | solution, methinks |
| 18:23:13 | springmeyer: | we've got a solution : ) |
| 18:23:26 | jctull: | springmeyer: I've had to give up on mapnik. The dependencies are killing me. |
| 18:23:35 | springmeyer: | on-the-fly rendering now working |
| 18:23:49 | springmeyer: | jctull: what did you try? |
| 18:24:48 | jctull: | well, I ended up digging up the gtk framework hoping that would work, but no cairomm lib. Running autogen.sh on the mapnik source kicks an error on missing libtool, despite having that. |
| 18:25:01 | springmeyer: | ack |
| 18:25:13 | springmeyer: | ya, the gtk framework cairo is a dead end |
| 18:25:22 | springmeyer: | and DONT use autotools |
| 18:25:32 | springmeyer: | use the official scons build |
| 18:25:33 | jctull: | ok, how can I configure then? |
| 18:25:38 | jctull: | ok, scons |
| 18:25:49 | springmeyer: | http://trac.mapnik.org/wiki/UsingScons |
| 18:25:53 | springmeyer: | sorry about that |
| 18:26:00 | jctull: | you need gtk for cairo right? |
| 18:26:04 | springmeyer: | we forgot to strip autotools from the 0.6.0 release |
| 18:26:12 | springmeyer: | ( should only be in trunk for a few devs that like it ) |
| 18:26:18 | springmeyer: | but it don't really work |
| 18:26:32 | springmeyer: | jctull, worry about cairo once you've got mapnik built |
| 18:26:41 | springmeyer: | rebuilding with cairo support is easy |
| 18:26:53 | jctull: | ok |
| 18:27:04 | springmeyer: | ( once cairo is installed - which I usually do via macports ) |
| 18:34:11 | springmeyer: | http://dbsgeo.com/tmp/mapnik_backend.mov ( 6mb ) |
| 18:36:42 | racicot: | springmeyer: nice! |
| 18:36:54 | springmeyer: | so, in the end just having the 'painter' passed as an argument from a canvas signal did it |
| 18:36:55 | springmeyer: | http://dpaste.de/WZHO/ |
| 18:36:56 | sigq: | Title: dpaste.de: Snippet #944 ( at dpaste.de ) |
| 18:37:33 | jctull: | very nice springmeyer! |
| 18:37:46 | springmeyer: | now I just need to figure out how to turn off the QT rendering automatically when mapnik rendering starts |
| 18:38:05 | springmeyer: | otherwise I'm pulling from the OSM database via two connections |
| 18:38:07 | jctull: | so qgis is rendering the vector layer, then you pass the mapnik render to the canvas and paint over the qgis render? |
| 18:38:21 | jctull: | you answered my question above. |
| 18:38:23 | springmeyer: | ya, exactly |
| 18:38:29 | springmeyer: | just painting over the top right now |
| 18:38:35 | springmeyer: | unless the layers are turned off |
| 18:38:40 | jctull: | killing the primary render would help the peformance, obviously |
| 18:38:43 | springmeyer: | then things speed up quite a bit |
| 18:38:46 | springmeyer: | ya |
| 18:38:59 | springmeyer: | better way that just turning off the layers? |
| 18:39:42 | springmeyer: | so jef is some kinda of wizard huh? |
| 18:40:05 | racicot: | yes he is |
| 18:40:16 | springmeyer: | he is the most amazing guy |
| 18:40:39 | jctull: | and he never sleeps! |
| 18:40:42 | racicot: | so, do you think you want to keep just painting on the main canvas like that, or do you want to go to a tab? |
| 18:40:45 | jctull: | robot, I tell you |
| 18:41:03 | springmeyer: | would be sweet to be able to go to a tab |
| 18:41:21 | springmeyer: | but switching between both should be do-able |
| 18:41:33 | racicot: | then just create a new canvas and do the same thing you are... but would not have to worry about the repaint of the other layers |
| 18:41:47 | springmeyer: | good point |
| 18:41:53 | : | * racicot goes to get dinner |
| 18:42:05 | springmeyer: | me too in a sec, here |
| 18:42:13 | springmeyer: | thanks =racicot |
| 18:44:21 | springmeyer: | hey, thanks all : ) |
| 18:44:29 | springmeyer: | nothing like getting a bit sidetracked |
| 18:44:44 | springmeyer: | ( all I planned on starting this week was a mapnik xml generator ) |
| 18:49:28 | springmeyer: | jctull: if you pick up the mapnik install again... I recommend boost from source |
| 18:49:37 | springmeyer: | you can find instructions on the mapnik trac site |
| 18:49:53 | springmeyer: | later guys, thanks again |
| 18:59:06 | jctull: | gsherman: zoom to last just used in real-world case. |
| 18:59:45 | jctull: | For some reason, it zoomed way out and not to my original extent on what should have been the last zoom extent ( the original extent ) |
| 18:59:48 | gsherman: | now if only i could take credit for it |
| 19:00:02 | gsherman: | odd |
| 19:00:18 | jctull: | I'm doing another zoom to last and it is churning right now |
| 19:00:40 | gsherman: | don't tell i'm going to have to rip it out |
| 19:00:56 | jctull: | will it pick a layer and zoom to its extent? |
| 19:00:59 | gsherman: | it worked in my exhaustive 2 zoom test |
| 19:01:26 | jctull: | It appears to have backed out to the extent of one of my layers, but it was not a zoom level I had already been at. |
| 19:01:49 | gsherman: | no it's just a stack of zoom levels and not per layer |
| 19:01:51 | jctull: | I.e., does it simply use a history of zoom levels or will it look at the next largest extent possible based on all the available layers? |
| 19:01:57 | jctull: | nevermind, answered |
| 19:02:24 | jctull: | well, I still like it... But it does seem to have a bug, at least for me |
| 19:02:50 | jctull: | That's why I made a bookmark first : ) |
| 19:04:03 | jctull: | FWIW, I have many layers in my project, did a couple zooms and a few canvas drags before I tried to back them out |
| 19:21:42 | gsherman: | jctull, not sure if it handles pans does it? |
| 19:22:20 | jctull: | maybe that is what messed it up? I don't know either. |
| 19:22:34 | gsherman: | I guess more testing and maybe a bug report is in order |
| 19:22:55 | jctull: | I'm going to pass on the report so I can get this last map finished. Beer awaits. |
| 19:23:05 | jctull: | Wait, why should beer await! Silly me. |
| 19:23:08 | gsherman: | what kind of maps? |
| 19:23:44 | jctull: | Some field maps for southern nevada with usgs drg underlays, land ownership, a utm grid, hillshade, roads, wildlife corridors, etc. |
| 19:23:55 | jctull: | Like the ones I was making during the hackfest |
| 19:24:01 | jctull: | updated and more of them... |
| 19:24:31 | gsherman: | ah |
| 19:24:42 | jctull: | I'm pretty comfortable with the composer for pretty good maps. |
| 19:24:49 | jctull: | Only we need collision detection in labels. |
| 19:24:54 | jctull: | And better labeling in general. |
| 20:10:33 | cgs_bob_: | gsherman: just tried Zoom Next on Ubuntu Hardy Heron. haven't noticed any problems so far |
| 20:11:24 | jctull: | cgs_bob_: did you mix some panning into the zooming, then try to do multiple 'zoom to last' clicks? |
| 20:13:54 | cgs_bob_: | jctull: just tried it again...panned around in a circle then zoomed in. Zoom Last went in reverse and Zoom Next replicated all of my pan/zooms |
| 20:14:18 | gsherman: | as usual, os x is the problem child : ) |
| 20:14:40 | jctull: | cool. I'll play some more later this week. If it can be replicated I'll try to file an "OSX-only" bug. Not too many of those. :D |
| 20:15:57 | jctull: | phew, rendering my last map for the night. Hopefully for the week. And still no beer in hand. I have to remedy that and say 'night to everyone. |
| 20:16:01 | cgs_bob_: | what would be nice if there was a way to zoom/unzoom that uses the left click/right click. |
| 20:16:39 | cgs_bob_: | gnite jctull. go ahead and get a beer : ) |
| 20:18:37 | jctull: | beer in hand. checking out now... |
| 22:09:58 | sigq: | tracfeed: Ticket #1522 ( patch closed ): Zoom to Next functionality <https://trac.osgeo.org/qgis/ticket/1522#comment:5> |