#QGIS IRC Log - 2009-04-05

For logs after Feb 3, 2007, all times are GMT-8. Prior logs are GMT-9.
Back to Logs
00:18:45 sigq: tracfeed: Ticket #1623 ( enhancement created ): add r.external module <https://trac.osgeo.org/qgis/ticket/1623>
00:20:46 sigq: tracfeed: Ticket #1624 ( enhancement created ): show GRASS region settings <https://trac.osgeo.org/qgis/ticket/1624>
00:30:49 sigq: tracfeed: Ticket #1618 ( bug closed ): color table for GRASS rasters <https://trac.osgeo.org/qgis/ticket/1618#comment:3>
00:36:51 sigq: tracfeed: Ticket #1625 ( enhancement created ): document how to write GRASS modules <https://trac.osgeo.org/qgis/ticket/1625>
00:38:52 sigq: tracfeed: Ticket #1626 ( enhancement created ): r.los, r.cost: take input points from the canvas <https://trac.osgeo.org/qgis/ticket/1626>
00:42:52 sigq: tracfeed: Ticket #1627 ( enhancement created ): v.distance, v.to.db: add data to vector table as new column, not replacing an existing one <https://trac.osgeo.org/qgis/ticket/1627>
00:44:51 sigq: tracfeed: Ticket #1628 ( enhancement created ): r.out.tiff: does not allow selection of destination directory <https://trac.osgeo.org/qgis/ticket/1628>
00:56:29 pcav: hi all
00:56:35 pcav: just here on a course
00:56:38 pcav: in Evora
00:56:46 pcav: wonderful weather : )
01:20:33 epiway: hi all
01:37:51 macho: mng
01:38:11 macho: pcav: which course?
01:52:20 timlinux: hi all
02:00:47 macho: hi timlinux
02:07:53 timlinux: hi macho
02:15:06 pcav: macho: http://faunalia.pt/curso
02:15:09 sigq: Title: Curso integrado de GRASS e Quantum GIS ( QGIS ) Software SIG Open Source | faunalia.pt ( at faunalia.pt )
02:15:19 pcav: coffee break now...
02:24:55 jef: timlinux: ok, spatialite working on windows - integrated in core.
02:25:36 timlinux: jef: oooer nice!
02:25:51 : * timlinux fires up his vm
02:26:30 timlinux: jef: so I can chuck away my hand built spatiallite libs on linux too?
02:26:40 jef: timlinux: yes.
02:26:59 : * timlinux *grins*
02:27:07 timlinux: \o/
02:27:19 pcav: jef: good!
02:28:52 CIA-56: jef * r10468 /trunk/qgis/src/app/attributetable/BeataView.cpp: fix a warning
02:29:23 sigq: tracfeed: Changeset [10468]: fix a warning <https://trac.osgeo.org/qgis/changeset/10468>
02:29:44 CIA-56: jef * r10469 /trunk/qgis/src/app/qgspluginmanager.cpp: include error string when loading plugins
02:29:59 CIA-56: jef * r10470 /trunk/qgis/src/app/qgsserversourceselect.cpp: fix warning
02:31:24 sigq: tracfeed: Changeset [10470]: fix warning <https://trac.osgeo.org/qgis/changeset/10470> || Changeset [10469]: include error string when loading plugins <https://trac.osgeo.org/qgis/changeset/10469>
02:36:16 CIA-56: jef * r10471 /trunk/qgis/ ( 13 files in 4 dirs ): update debian packaging
02:37:26 sigq: tracfeed: Changeset [10471]: update debian packaging <https://trac.osgeo.org/qgis/changeset/10471>
02:40:00 CIA-56: jef * r10472 /trunk/qgis/src/ ( 2 files in 2 dirs ): make spatialite provider work on windows
02:40:30 sigq: tracfeed: Changeset [10472]: make spatialite provider work on windows <https://trac.osgeo.org/qgis/changeset/10472>
02:42:34 CIA-56: jef * r10473 /trunk/qgis/ ( 46 files in 18 dirs ): add internal spatialite support
02:42:35 sigq: tracfeed: Changeset [10473]: add internal spatialite support <https://trac.osgeo.org/qgis/changeset/10473>
02:48:00 CIA-56: jef * r10474 /trunk/qgis/src/providers/postgres/ ( qgspostgresprovider.cpp qgspostgresprovider.h ): use ctid as temporary feature id in postgres provider if there is no other primary key available
02:48:31 jef: timlinux: now my tree id in sync again ; )
02:52:21 timlinux: jef you mean you have no more uncommitted changes?
02:52:48 timlinux: jef did you change the libnames in the debian packaging to make debianistas happy again?
02:53:20 timlinux: jef oh - was that support for keyless tables / views you committed there too?
02:53:29 jef: timlinux: I think.
02:53:41 timlinux: cool
02:53:47 jef: timlinux: no, just for tables. view don't have a ctid
02:53:49 timlinux: lots of nice goodies :- )
02:54:00 timlinux: ok
02:54:17 timlinux: I guess people will ask you to support htat next :-P
02:54:37 timlinux: what is ctid? pg internal identifier?
02:55:54 jef: timlinux: yes. represents the physical location of the row in block number/offset. block is 32 and offset is 16 bit. so that's only used when the block numbers are <0x10000
02:56:48 timlinux: jef by the way did you ever have a problem where a table was too big for pgdump to dump?
02:57:44 jef: timlinux: no
03:00:00 jef: timlinux: hm, we should remove internal sqlite support - we have sqlite3 twice now.
03:00:46 timlinux: jef yes
03:01:07 timlinux: jef so how do I enable internal spatial lite support? it happens automatically?
03:01:48 jef: timlinux: did you ever use that on windows? I didn't notice any dllexport stuff in internal sqlite3
03:01:57 timlinux: do I need to trash my build and configure from scratch?
03:02:11 timlinux: jef I think I did
03:02:19 timlinux: can't recall now
03:02:27 timlinux: maybe it was on mac
03:02:37 timlinux: my brain is mush :- )
03:03:11 timlinux: ah ok I see the WITH_INTERNAL_SPATIALITE option now
03:03:34 : * timlinux starts building
03:03:58 timlinux: is spatiallite built into the provider or core?
03:04:09 jef: timlinux: core
03:04:20 jef: timlinux: just like sqlite was
03:04:51 timlinux: ok
03:07:14 epifanio: timlinux: jef .. so now spatialite is in the core :- ) ??? and we can set it in the cmake, right ?
03:07:44 CIA-56: jef * r10475 /trunk/qgis/cmake/FindIconv.cmake: missed a file in last commit
03:08:33 jef: epifanio: right
03:08:34 sigq: tracfeed: Changeset [10475]: missed a file in last commit <https://trac.osgeo.org/qgis/changeset/10475>
03:08:54 epifanio: cool !
06:11:45 : <_-Slash-_> yop
07:02:09 jef: moin horst
07:12:37 CIA-56: jef * r10476 /trunk/qgis/src/core/CMakeLists.txt: actually use iconv include path
07:12:38 sigq: tracfeed: Changeset [10476]: actually use iconv include path <https://trac.osgeo.org/qgis/changeset/10476>
07:14:28 macho: moinmoin
07:14:42 macho: lots of jef commits ...
07:14:51 macho: jef: busy weekend? ; )
07:22:12 hdus: jef: moin jef
07:23:25 hdus: jef: I saw that you committed today a patch to connect to non primary key postgis layers. Am I correct?
07:35:47 jef: hdus: yes
07:46:50 CIA-56: jrm * r10477 /docs/branches/1.0.0/french/user_guide/ ( 10 files ): Fichiers avec accentuation correcte
07:46:51 sigq: tracfeed: Changeset [10477]: Fichiers avec accentuation correcte <https://trac.osgeo.org/qgis/changeset/10477>
08:42:38 CIA-56: wonder * r10478 /trunk/qgis/ ( 16 files in 5 dirs ): Added possibility to express tolerances for digitising also in pixels in addition to map units. Patch contributed by Richard Kostecky.
09:24:43 epifanio: hi guys .. i'm a bit embarassed .... can i ask a pyqt help :- ) ?
09:28:36 cgs_bob: epifanio: don't have much experience with pyqt, but I thought you would find this book useful: http://www.qtrac.eu/pyqtbook.html : )
09:28:37 sigq: Title: Rapid GUI Programming with Python and Qt ( at www.qtrac.eu )
09:29:54 epifanio: ohh yes it will give me a good help ... need to buy it!
09:31:59 cgs_bob: I just got it. so far it is a good read.
09:45:31 sigq: tracfeed: Ticket #1629 ( bug created ): replace v.buffer with v.buffer2 for GRASS 6.4 <https://trac.osgeo.org/qgis/ticket/1629>
09:49:32 sigq: tracfeed: Ticket #1630 ( enhancement created ): v.db.addcol: checkbox for data type <https://trac.osgeo.org/qgis/ticket/1630>
09:52:25 CIA-56: kyngchaos * r10479 /trunk/qgis/mac/xcode/Qgis.xcodeproj/project.pbxproj: tolerance ( r10478 ), spatialite ( r10473 ) updates ( spatialite sources added to project, but not integrated in build yet )
09:54:40 sigq: tracfeed: Ticket #1632 ( bug created ): v.db.addcol: fails if path contains spaces <https://trac.osgeo.org/qgis/ticket/1632>
10:15:40 sigq: tracfeed: Ticket #1633 ( enhancement created ): editing GRASS tables <https://trac.osgeo.org/qgis/ticket/1633>
10:18:25 timlinux: jef I have all kinds of problems since your sqlite patch
10:18:32 jef: timlinux: like?
10:19:11 timlinux: I did make uninstall on spatiallite and if *very* kindly removed /usr/lib/libsqlite-3.0 which broke all kinds of things in my system
10:19:17 timlinux: ( not your fault )
10:19:40 timlinux: - I had installed spatiallite into /usr/local
10:20:28 timlinux: if I have both sqlite3 and spatial lite set to internal then try to build I get: /home/timlinux/dev/cpp/qgis/src/core/spatialite/virtualtext/sqlite3/sqlite3.c:94656: multiple definition of `sqlite3_global_recover'
10:20:28 timlinux: which is logical
10:20:38 timlinux: but if I try to disable sqlite3 internal I get
10:20:45 timlinux: CMake Error at CMakeLists.txt:119 ( MESSAGE ):
10:20:45 timlinux: sqlite3 dependency was not found ( try compiling with sqlite3 internal )!
10:20:57 timlinux: so it puts me in a kinda no win situation
10:22:08 timlinux: I guess I should try with a clean build dir
10:25:37 jef: timlinux: and you don't have sqlite3 installed externally, I suppose.
10:34:27 jef: timlinux: ok, removing WITH_INTERNAL_SQLITE3
10:47:39 timlinux: jef good
10:55:49 sigq: tracfeed: Ticket #1634 ( bug created ): snapping tolerance in pixels does not work if on-the-fly reprojection is enabled <https://trac.osgeo.org/qgis/ticket/1634>
11:03:19 CIA-56: jef * r10480 /trunk/qgis/ ( 6 files in 6 dirs ): internal sqlite3 was replaced with spatialite
11:03:20 sigq: tracfeed: Changeset [10480]: internal sqlite3 was replaced with spatialite <https://trac.osgeo.org/qgis/changeset/10480>
11:07:56 cgs_bob: jef: hi there! with your latest change, will I still be able to use the spatialite that I installed in /usr/local or will I have to uninstall it?
11:11:59 jef: cgs_bob: should work with WITH_SPATIALITE=ON and WITH_INTERNAL_SPATIALITE=OFF
11:12:04 pcav1: beware jef
11:12:19 pcav1: there is a new version who fixes several issues
11:12:25 pcav1: I wrote Sandro
11:12:34 pcav1: to keep thngs updated
11:13:00 pcav1: macho: you were asking about the course yesterday?
11:15:03 jef: pcav1: sandro doesn't have a repository, does he?
11:17:29 pcav1: not yet
11:17:39 pcav1: I'm convincing him to set up one
11:17:48 pcav1: that's the problem
11:17:51 pcav1: a lone wolf
11:19:21 pcav1: bye all
11:31:56 timlinux: doesn't sandro have write access to qgis now?
11:32:20 timlinux: or who is the new guy that wrote the spatial lite provider?
11:34:51 cgs_bob: jef: thanks! now I'll test you changes
11:45:48 : * timlinux is building with jefs changes now too
11:57:28 timlinux: jef itworks
11:57:53 timlinux: jef how do I create a new sqlite spatial db? do I need to do anything special?
11:58:01 timlinux: or just create a normal sqlite db
11:58:02 timlinux: ?
11:59:12 jef: timlinux: no idea - I just tested shortly with the 2.2 dataset
11:59:43 timlinux: :- )
11:59:59 : * timlinux will go and rtfm
12:04:07 epifanio: timlinux: i used the gui from sandro's site, then i loaded some shape ... it will create a spatiallite.sqlite db
12:42:11 wildintellect: when timlinux gets back tell him there's an sql script to initialize the spatial metadata tables
12:42:37 wildintellect: http://www.gaia-gis.it/spatialite/init_spatialite-2.2.sql.gz
12:45:32 jef: hm, that could probably be taken from srs.db
12:46:43 wildintellect: well sorta
12:46:48 wildintellect: it inserts some triggers
12:47:00 wildintellect: and a table to keep track of which columns in what table are spatial data
12:47:10 wildintellect: in addition to the srs stuff
12:48:03 jef: probably even better to include a template database and just copy it.
12:48:42 wildintellect: I agree that's a good idea, but maybe also just include the sql script and have a command to trigger it
12:48:55 wildintellect: that way you can convert existing sqlite databases
13:15:55 : * doktoreas tries the Metadata plugin
13:16:07 doktoreas: is the suthor here :D ?
13:17:04 doktoreas: *author
13:22:45 jef: wildintellect: still there?
13:22:58 wildintellect: yes
13:23:15 jef: wildintellect: timlinux too ; )
13:23:19 wildintellect: doktoreas, I'm one of the authors
13:23:34 wildintellect: I'm aware the plugin is incomplete
13:23:41 doktoreas: hi wildintellect and nice to meet you : )
13:23:51 doktoreas: is it normal with 0.1
13:24:14 doktoreas: wildintellect, I tried with the italian schema
13:24:16 wildintellect: oh yes, I just wanted to get a public release out so I can invite more developers
13:24:18 doktoreas: are workin fine
13:24:38 wildintellect: cool, keep in mind it's only a reader right now
13:25:12 wildintellect: timlinux, to answer your question earlier you need to run http://www.gaia-gis.it/spatialite/init_spatialite-2.2.sql.gz
13:25:14 doktoreas: I was thinking about a way for uploading data to Geonetwork
13:25:24 timlinux: wildintellect: ok thanks
13:25:29 doktoreas: hi timlinux
13:25:47 wildintellect: doktoreas, that would be a good idea, want to add a ticket for it we have a trac site
13:26:04 timlinux: wildintellect: I guess I'm wanting to kinda go 'new vector layer' and create a new layer in the sqlite ( or postgis for that matter ) backend
13:26:06 doktoreas: I can give a hand if needed..
13:26:18 doktoreas: wildintellect, nice..trac URL?
13:26:44 wildintellect: http://apps.sourceforge.net/trac/qgismetaedit/
13:26:45 sigq: Title: qgismetaedit ( at apps.sourceforge.net )
13:27:04 wildintellect: if you have a sourceforge account I can add you as a developer so you have commit access too
13:28:17 wildintellect: timlinux, that makes sense in the case of sqlite it should check if the spatial metadata tables exist and if not run the init sql before creating the new table
13:28:49 doktoreas: wildintellect, probably I have one..but I don't remember : )
13:29:08 wildintellect: you can always sign up for a new one
13:29:42 doktoreas: wildintellect, is doktoreas
13:31:03 timlinux: wildintellect: any screenies / docs of your metadata editor doing its thing?
13:31:41 wildintellect: we have an algorithm image, but that's it for now aside from tooltips on everything
13:32:27 wildintellect: doktoreas, you should have access now
13:33:30 wildintellect: I'm actually surprised we got anything working, we only met 10 times for class
13:33:59 timlinux: wildintellect: heh
13:34:01 doktoreas: wildintellect, thank you I'll talk with Ticheler about this.
13:34:29 wildintellect: timlinux, the plugin was a class project
13:34:52 wildintellect: we met once a week for 2 hours and did a few hours each on the side
13:35:35 wildintellect: none of use had ever written a qgis plugin, and I was the only one with python experience
13:35:58 wildintellect: so we spent most of the time just trying to understand qt
13:36:09 timlinux: wildintellect: yes I remember you announcing the project before you started coding
13:36:29 wildintellect: ya, a new announcement is going out today to invite more coders in
13:36:33 timlinux: qt is thankful well documented
13:37:02 wildintellect: well sorta, there's one really good howto
13:37:17 wildintellect: and once I learned to read the qt docs I think I get them now
13:40:23 timlinux: wildintellect: yeah the little read more.... link at the top of each class doc is the first thing I click on
13:40:33 timlinux: as they often have little code example
13:40:34 timlinux: s
13:41:50 wildintellect: ah yes, but mostly c++ examples
13:42:12 timlinux: just s/->/./g :- )
13:43:42 wildintellect: what really got me was, not realizing that the QPlainTextEdit was new in qt4.4
13:44:03 doktoreas: wildintellect, going to bed
13:44:05 wildintellect: so right now qt4.3 people will get an error
13:44:08 doktoreas: see you tomorrow
13:44:10 doktoreas: bye folks
14:03:16 timlinux: http://www.flickr.com/photos/timlinux/3415349301/ <-- pet sandwich
14:03:17 sigq: Title: Sandwich on Flickr - Photo Sharing! ( at www.flickr.com )
14:04:44 Kuril: Haha. Cute!!
14:04:52 wildintellect: looks like your cat is ignoring the existence of the dog
14:05:16 timlinux: :- ) they are still lying like that only the dog has gone to sleep now too :- )
14:19:43 wildintellect: Kuril, I just made an OSGeo facebook group - check you notifications
14:49:44 timlinux: jef its strange on my windows trunk qgis I don't get the spatiallite option in cmakesetup
14:50:01 timlinux: its at rev 10480
14:50:42 timlinux: and I deleted my cmakecache and configured everything from scratch
15:33:08 Kuril: wildintellect: okie dokie
15:45:00 jctull: wildintellect: metaedit will not allow xml files to be selected as templates in the browse for xml file dialog no OS X
15:45:14 jctull: xml files are greyed out and unselectable
15:45:22 wildintellect: interesting
15:45:49 jctull: If I choose an existing template, such as ISO19139_elements_en.xml, I get this error
15:45:56 jctull: loadDoc:: Failed to load: /Users/jctull/.qgis/python/plugins/metaedit/Templates/xml/ISO19139_elements_en.xml 'ascii' codec can't encode character u'\xe7' in position 15: ordinal not in range( 128 )
15:46:15 wildintellect: oh we know some of them don't work - still trying to decipher why
15:46:16 jctull: My guess is there are some qt-mac encoding issues
15:46:48 wildintellect: why they are greyed out is something I haven't seen
15:46:57 wildintellect: but then again none of us had a mac to try on
15:47:32 jctull: anyway, good to see this start. This is sorely needed.
15:47:58 wildintellect: if you have a sourceforge account I can give you svn access
15:48:13 wildintellect: we are inviting contributors
15:48:15 jctull: your assuming I know how to do anything besides complain ; )
15:48:20 jctull: Unfortunately, I don't
15:48:26 wildintellect: someone needs to file bugs in trac
15:48:38 jctull: I have an account, so can do that
15:48:56 wildintellect: people don't realize this but one of our team members isn't a coder at all
15:49:12 wildintellect: she worked on the gui and docs primarily
15:50:14 wildintellect: if you could post a screenshot of your issue on OS X with the dialog that would help
15:53:52 wildintellect: jctull, can you check if you can file tickets - I'm not sure if you need to join the project or just have an sf account
15:54:26 jctull: I cannot login to the apps page
15:57:14 jctull: screenshot of xml issue: http://www.flickr.com/photos/67875808@N00/3416482710/sizes/l/
15:57:15 sigq: Title: Sign in to Yahoo! ( at www.flickr.com )
15:58:36 jctull: wildintellect: xml files show up if I select browse from the 'browse' button next to 'open template', but it fails if I use the browse button from the open template dialog window
15:58:57 wildintellect: ah ok that's a good hint
15:59:53 jctull: It also looks like the loadDoc issue fails partway through an xml load.
16:00:27 wildintellect: depends on the xml file
16:00:32 wildintellect: some load, some don't
16:00:33 jctull: It appears to possibly be due to a character that is not being handled
16:01:09 wildintellect: I see that from the error you posted earlier
16:01:55 wildintellect: jctull, that screenshot link is bunk
16:02:35 jctull: not for me
16:02:45 jctull: try http://farm4.static.flickr.com/3642/3416482710_2fe53433b2_b.jpg
16:03:12 jctull: I think a ≥ character is killing the load on one of my own files
16:04:00 jctull: yep, verified it by changing that to 'gt'
16:04:40 jctull: then the xml loaded nicely
16:04:46 wildintellect: so I'm looking at that pic, what's the issue - nothing appears grey
16:05:08 wildintellect: I guess we'll need to add a python sanitizer to the xml parsing
16:05:32 jctull: see all those files... They are not selectable. They are greyed out ( look at the black text in the sidebar for context )
16:06:14 wildintellect: interesting, it shows the files but won't let you select
16:06:26 jctull: Is there no editing capability yet? If I change anything, the save button remains greyed out.
16:06:33 wildintellect: correct
16:06:36 jctull: wildintellect: yes to your comment
16:06:44 jctull: ok, good to know. Thanks.
16:06:48 wildintellect: you can monkey to your hearts content but no save
16:06:52 jctull: This is a nice start
16:06:55 wildintellect: the undo/redo kinda works
16:11:03 wildintellect: jctull, so the browse on the initial page works fine for you
18:47:28 jctull: wildintellect: been away, as you probably guessed. I'm not sure what you are asking.
22:38:46 CIA-56: jef * r10481 /trunk/qgis/src/core/CMakeLists.txt: also install qgstolerance.h
22:38:48 sigq: tracfeed: Changeset [10481]: also install qgstolerance.h <https://trac.osgeo.org/qgis/changeset/10481>
23:33:01 SmokeyD: hey eveyone, is it possible to change the gradient when setting a graduated color when classifying a layer? QGis always uses green to blue. Can't I make it use red to green or any other gradient?
23:33:29 SmokeyD: I can manually change the color of each class when I classified the layer, but it would be so much quicker to set the gradient before hand