#QGIS IRC Log - 2009-06-26

For logs after Feb 3, 2007, all times are GMT-8. Prior logs are GMT-9.
Back to Logs
00:28:03 timlinux: hi all
00:28:06 timlinux: jef you about?
00:28:19 macho: hi timlinux
00:28:23 timlinux: hi macho
00:28:31 macho: any news?
00:28:31 timlinux: anyone here using osgeo4w
00:28:32 timlinux: ?
00:28:37 macho: not me ..
00:28:55 timlinux: macho, I'm still trying to get grass to work properly in my 'all in one' qgis windows installer
00:29:14 timlinux: == pushing splinters under my nails
00:31:13 timlinux: anyway as a backstop measure, I am updating the vmware image to have qgis 1.0
00:31:19 timlinux: 1.0.2
00:32:33 macho: 1.0.2 - the rockstable QGIS ; )
00:32:49 timlinux: oops I wrote that to wrong channel :- )
00:33:07 timlinux: we are getting ready to do a training course on qgis over on #osgeo-africa channel
00:33:42 timlinux: which is why I am also trying to make the win build work nicely
00:33:59 timlinux: macho did you see the vietnamese translation mail?
00:36:13 timlinux: epifanio, hi
00:36:27 timlinux: epifanio, was it you that was involved in makeing livecds once?
00:36:56 macho: timlinux: already commited yesterday
00:37:02 timlinux: macho, cool
00:37:06 macho: timlinux: forgot to write email - > busy time right now
00:37:11 : * timlinux should read his commit logs
00:37:18 timlinux: I wrote him a little reply
00:37:26 timlinux: so you prolly dont need to worry now
00:40:04 jachym: guys: what OGC standard and which versions does qgis understand ?
00:48:30 timlinux: jachym, wms and wfs
00:48:43 timlinux: jachym, ( and hi :- ) )
00:48:53 timlinux: jachym, epifanio is working on wps
00:55:18 jachym: timlinux: great!
00:56:07 timlinux: I guess we also support simple features via gdal / postgis
00:56:16 jachym: timlinux, I know, it is wfs and wms - but what about wcs? and which versions ? for wms it is 1.1, 1.1.1 and 1.3 ( according to documentation )
00:56:39 jachym: and there is some csw implementation as well?
00:56:43 timlinux: and some folks are working on metadata which I think will be iso19115 part 1&2 based
00:57:09 timlinux: which I think matches ogc metadata standards ( speaking under correction here )
00:57:31 jachym: timlinux: afaik iso 19115 is how the metadata should look like
00:57:33 timlinux: wcs not yet except via gdal ( guru users only would be able to use it )
00:57:47 jachym: 19110 ( or what ) how they should be stored as XML
00:57:55 jachym: and csw is , how to get them
00:58:02 timlinux: right
00:58:12 timlinux: there is work on csw happening I believe
00:58:22 jachym: yes, I saw something
00:58:30 timlinux: right now the only stable integrated ones are wfcs wms and simple features
00:58:39 timlinux: for versions I am not sure
00:58:51 timlinux: I believe wms is 1.3
00:58:59 jachym: I will have to ask the ml then ;- ) thanks anyway
00:58:59 timlinux: but wed need to check with marco
00:59:03 timlinux: np
01:06:25 jachym: some little demonstration ?
01:06:41 timlinux: jachym, ?
01:06:55 timlinux: jachym, how do you mean?
01:10:27 jachym: foss-workshop
01:20:40 pcav: so now qgis is OGC registered : )
01:32:17 timlinux: pcav, just saw you message
01:32:33 timlinux: pcav, what is the procedure to request travel costs for the hackfest
01:32:48 timlinux: I need to book a ticket soon as its cheaper the earlier you book
01:34:13 pcav: for you, no procedure : )
01:34:32 pcav: I think we can make a call on the dev ML
01:34:34 pcav: I'm going to do it now
01:34:37 timlinux: have you had any other requests?
01:34:41 timlinux: ok
01:34:42 timlinux: great
01:34:49 timlinux: pcav, are you going to be able to attend?
01:34:59 timlinux: pcav, do you have any windows qgis users there?
01:35:22 timlinux: I've been trying to resolve the grass issues in my standalone qgis installer
01:35:41 timlinux: in particular it crashes out when trying to make a new mapset
01:36:04 timlinux: I just discovered that my osgeo4w installed qgis does the same thing
01:40:37 pcav: timlinux: yes, I'll be there
01:40:41 pcav: no win here, sorry
01:40:41 pcav: yes, I know
01:40:42 pcav: how do you compile grass for this?
01:40:51 timlinux: no I'm using all osgeo4w deps
01:41:44 timlinux: my new build system ( what I worked on at the last hackfest a bit too ) lets you build qgis easily under win using osgeo4w for dependencies but then packages it as a standalone installer
01:41:58 : * timlinux off for a toddle with the dog for a few mins - back just now
01:42:13 pcav: but grass is compiled with mingw?
01:42:22 pcav: and qgis with msvc?
03:01:19 jef: pcav: yes
03:19:27 pcav: I know _nothing_ about win compilation
03:26:14 jef: timlinux: do we have a stable 1.0.x api or a stable 1.x API
03:53:24 timlinux: jef, we have a stable 1.0.x api and 1.x api should retain backwards compatibiliti ( ie. its addititive only )
03:53:45 timlinux: jef you should mark methods with doxygen if they should be deprecated
03:53:49 jef: grmp
03:54:14 timlinux: pcav, to answer your question above, yes
03:54:23 timlinux: jef, what you wanting to change?
03:54:34 jef: timlinux: nothing. that's the problem
03:54:39 timlinux: :- )
03:54:54 jef: past tense. ; )
03:54:56 timlinux: thats very confuscious-alistic
03:55:08 timlinux: did the api get changed already?
03:55:42 jef: yes. QgsVectorLayer::addAttribute( ), QgsVectorDataProvider::addAttributes( )
03:55:47 jef: and the providers...
03:55:58 timlinux: urf
03:56:02 jef: @changeset 10863
03:56:02 sigq: jef: http://trac.osgeo.org/qgis/changeset/10863
03:56:08 pcav: jef: timlinux: according to previous packager
03:56:11 timlinux: recently? I never saw that happen
03:56:28 jef: if 4 weeks is still recent...
03:56:29 pcav: mixing mingw and msvc packages leads to crashes
03:56:52 pcav: I think this is one of the reasons why he stopped contributing btw
03:57:05 jef: who?
03:57:39 jef: marcopx?
03:57:48 timlinux: pcav, the mingw grass packages are called in their own thread
03:57:58 timlinux: so should not cause crashes
03:58:02 tackat: hi timlinux
03:58:06 timlinux: grass apps I should say
03:58:31 timlinux: pcav, and jef cleverly built enough of the grass libs natively that we need to interface with other grass stuff
03:58:34 timlinux: tackat, hi
03:58:41 tackat: how are you doing? : )
03:58:52 tackat: We are about to release Marble 0.8 soon
03:58:58 timlinux: jef, from the looks of that commit it could have been fixed easily by doing:
03:59:23 jef: pcav: and I never found a crash yet, that wasn't caused by unclean memory management on the grass end.
03:59:29 timlinux: a wrapper call around bool addAttribute( const QgsField &field );
03:59:43 timlinux: so that it internally just calls the new method
03:59:52 timlinux: with default values
03:59:58 timlinux: same for addAttributes
04:00:32 timlinux: and then virtual bool addAttributes( const QList<QgsField> & attributes ); and bool addAttribute( const QgsField &field ); could be marked as deprecated in doxy
04:00:36 timlinux: if that makes sense
04:00:43 timlinux: tackat, cool thanks and you? :- )
04:00:57 timlinux: well done - marble is catching up to qgis in version numbers :- )
04:01:11 timlinux: I havent tried marble for quite some months now :-(
04:02:16 jef: timlinux: do you know if python can have method with the same name, but different signatures?
04:02:28 timlinux: jef um
04:02:41 timlinux: ---> python idiot <------
04:02:51 timlinux: ( timlinux is a )
04:03:10 timlinux: I only know enough to fake my way along in geodjango
04:04:42 spacedman: jef: python doesn't have signatures.... you probably need to fake them by using **kwargs args that match all, and then write the despatching code yourself
04:36:55 aghisla: good to know
04:37:12 : * aghisla is an _actual_ python idiot
07:25:09 CIA-55: jef * r10981 /trunk/qgis/src/ ( 43 files in 9 dirs ): automatic indentation update ( r10918-r10980 )
07:25:52 sigq: tracfeed: Changeset [10981]: automatic indentation update ( r10918-r10980 ) <https://trac.osgeo.org/qgis/changeset/10981>
07:27:30 timlinux: go jef go :- )
07:27:43 : * timlinux hasnt committed anything in years :-(
08:59:53 CIA-55: jef * r10982 /trunk/qgis/ ( 6 files in 2 dirs ): re-add vector attribute method removed in r10863
09:00:50 sigq: tracfeed: Changeset [10982]: re-add vector attribute method removed in r10863 <https://trac.osgeo.org/qgis/changeset/10982>
09:33:22 CIA-55: jef * r10983 /trunk/qgis/python/core/qgsvectordataprovider.sip: sync QgsVectorDataProvider sip interface
09:33:22 sigq: tracfeed: Changeset [10983]: sync QgsVectorDataProvider? sip interface <https://trac.osgeo.org/qgis/changeset/10983>
11:08:46 timlinux: can anyone try to make a new location using the new mapset icon in grass plugin?
11:09:06 timlinux: I get a g_make_location( ) failed for both lin qgis 1.0.2 and win
11:10:34 timlinux: ah I see why
11:28:22 ttestt: hello. is here someone who use python to made own plugins?
11:29:58 jef: timlinux: what do you see?
11:38:57 jef: timlinux: did you also have a directory you couldn't write to?
11:48:58 ttestt: do you know, is it possible to 'hijack' data changes from the qgis' table? ( using python ). it is to change point coordinates when it's values changed in a table
11:51:14 jef: ttestt: QgsVectorLayer emits a attributeValueChanged signal
11:52:08 ttestt: thanx, jef !
12:02:20 ttestt: jef, you know, when standard Actions will be available to execute from python? and 2: is it possible to made personal popup menu for each layer ( modify layer's popup menu from python )?
12:10:18 jef: ttestt: 1. still no idea. 2. no
12:13:30 CIA-55: jef * r10984 /trunk/qgis/src/app/qgsmaptoolnodetool.cpp: fix warning
12:14:14 CIA-55: mhugent * r10985 /trunk/qgis/src/app/composer/ ( qgscomposer.cpp qgscomposer.h ): Replaced activated signal with triggered signal in composer to get rid of qt3 support issues
13:51:23 ttestt: on the page http://qgis.org/en/download/unstable-software.html there is a link to md5 of 1.1.0. this link is not correct
13:51:25 sigq: Title: Unstable software packages ( at qgis.org )
13:52:29 ttestt: there is a link to http://download.osgeo.org/qgis/win32/QGIS-1.1.0-Setup.exe.md5, but needs to be an http://download.osgeo.org/qgis/win32/QGIS-1.1.0-0-Setup.exe.md5
21:51:54 sigq: tracfeed: Ticket #761 ( bug closed ): GRASS shell <https://trac.osgeo.org/qgis/ticket/761#comment:10>
21:53:54 sigq: tracfeed: Ticket #1748 ( bug created ): GRASS shell non functional on Windows <https://trac.osgeo.org/qgis/ticket/1748>
21:55:54 sigq: tracfeed: Ticket #1458 ( bug closed ): GRASS shell: corrupted chars in error message <https://trac.osgeo.org/qgis/ticket/1458#comment:5> || Ticket #1456 ( bug closed ): QGIS GRASS shell doesn't allow moving cursor to the right <https://trac.osgeo.org/qgis/ticket/1456#comment:4> || Ticket #1217 ( bug closed ): some GRASS commands block the GRASS shell <https://trac.osgeo.org/qgis/ticket/1217#comment:3>
21:59:38 sigq: tracfeed: Ticket #1457 ( enhancement closed ): key combos, mouse actions and commands for GRASS shell <https://trac.osgeo.org/qgis/ticket/1457#comment:2>
23:04:16 sigq: tracfeed: Ticket #1749 ( enhancement created ): add WMS tile caching support <https://trac.osgeo.org/qgis/ticket/1749>