#QGIS IRC Log - 2009-04-07

For logs after Feb 3, 2007, all times are GMT-8. Prior logs are GMT-9.
Back to Logs
00:05:17 CIA-56: mhugent * r10492 /trunk/qgis/src/core/ ( 8 files ): Added notes about new classes and methods in 1.1
00:06:28 sigq: tracfeed: Changeset [10492]: Added notes about new classes and methods in 1.1 <https://trac.osgeo.org/qgis/changeset/10492>
00:08:29 sigq: tracfeed: Ticket #1635 ( enhancement created ): Save bookmarks with project <https://trac.osgeo.org/qgis/ticket/1635>
00:10:30 sigq: tracfeed: Ticket #1636 ( enhancement created ): Make bookmarks dockable <https://trac.osgeo.org/qgis/ticket/1636>
00:21:18 CIA-56: jef * r10493 /trunk/qgis/ ( python/core/qgsdatasourceuri.sip src/core/qgsdatasourceuri.h ): added notes about new optional sslmode parameter in 1.1
00:22:34 sigq: tracfeed: Changeset [10493]: added notes about new optional sslmode parameter in 1.1 <https://trac.osgeo.org/qgis/changeset/10493>
00:50:42 sigq: tracfeed: Ticket #1637 ( enhancement created ): Legend Window; for drag-and-drop-actions, expand layer-group only after short delay <https://trac.osgeo.org/qgis/ticket/1637>
01:14:20 gabrimonfa: hi, I've built qgis_unstable rev 10491 and it segfaults
01:14:46 gabrimonfa: I'm on Linux x86_64 Ubuntu 2.6.27-14-generic
01:15:24 gabrimonfa: and gcc 4.3.2
01:24:20 CIA-56: mhugent * r10494 /trunk/qgis/src/plugins/ ( interpolation/qgstininterpolator.cpp qgisplugin.h ): remove debug code from triangle interpolation class
01:24:48 sigq: tracfeed: Changeset [10494]: remove debug code from triangle interpolation class <https://trac.osgeo.org/qgis/changeset/10494>
03:57:03 CIA-56: jef * r10495 /trunk/qgis/CMakeLists.txt: apply #1566
03:57:05 sigq: tracfeed: Changeset [10495]: apply #1566 <https://trac.osgeo.org/qgis/changeset/10495>
03:58:18 sigq: tracfeed: Ticket #1566 ( bug closed ): Fix install paths on 64 bit linux <https://trac.osgeo.org/qgis/ticket/1566#comment:3>
04:09:05 sigq: tracfeed: Ticket #952 ( enhancement closed ): Support PostgreSQL uuid type <https://trac.osgeo.org/qgis/ticket/952#comment:5>
04:35:31 CIA-56: jef * r10496 /trunk/qgis/src/plugins/grass/qgsgrasstools.cpp: open grass module tree collapsed ( implements #459 )
04:35:32 sigq: tracfeed: Changeset [10496]: open grass module tree collapsed ( implements #459 ) <https://trac.osgeo.org/qgis/changeset/10496>
04:36:53 sigq: tracfeed: Ticket #459 ( enhancement closed ): Collapse GRASS tool list <https://trac.osgeo.org/qgis/ticket/459#comment:12>
04:44:14 CIA-56: jef * r10497 /trunk/qgis/src/plugins/grass/config/default.qgc: remove empty sections from grass default.qgc
04:45:26 markusN: jef: thanks for r10496!
04:45:27 sigq: tracfeed: Changeset [10497]: remove empty sections from grass default.qgc <https://trac.osgeo.org/qgis/changeset/10497>
04:45:55 jef: markusN: what about r10497?
04:46:13 markusN: I meant "open grass module tree collapsed ( implements #459 ) "
04:46:37 jef: markusN: understood. but what about r10497?
04:46:42 markusN: r10497 I dunno...
04:47:02 jef: markusN: there were empty lines in the tree.
04:47:05 markusN: not sure what this label thing was
04:47:19 markusN: leo or pcav or .. are the authors
04:48:19 pcav: jef: going to look
04:49:22 pcav: jef: these were to keep various sections apart
04:49:45 pcav: so it was easier to locate the relevant command
04:50:13 pcav: now with collapsed command and the search list they shouldn't be necessary
04:50:27 pcav: I'm compiling, I'll let you know
04:50:32 pcav: ok jef?
04:50:51 jef: pcav: ok, just what I thought. removed them.
04:51:09 pcav: good
04:51:39 markusN: jef, is r10496 only trunk compatible?
04:53:04 pcav: what about https://trac.osgeo.org/qgis/ticket/1552 ?
04:53:06 sigq: Title: #1552 ( plug-in installer does not work behind proxy ) - Quantum GIS - Trac ( at trac.osgeo.org )
04:53:10 pcav: is this fixed in trunk?
05:08:12 jef: pcav: I think for wms/wfs it is ( r9939/r10009/r10031 ). not sure if the proxy settings also apply to python
05:09:24 pcav: jef: thanks, perhaps better to add this comment to the trac?
05:09:31 pcav: anybody can test?
05:18:06 pcav: jef: just compiled, but I do not see collapsed tools
05:19:01 jef: markusN: no, but it probably is a gui change
05:21:22 markusN: ok I see
05:24:44 jef: pcav: hm, I do. are you sure you're running the right binaries?
05:30:01 pcav: jef: I try again
05:40:31 CIA-56: mhugent * r10498 /trunk/qgis/src/plugins/qgsvectoroverlayplugin.cpp: remove qgsvectoroverlayplugin.cpp that not needed anymore
05:42:20 sigq: tracfeed: Changeset [10498]: remove qgsvectoroverlayplugin.cpp that not needed anymore <https://trac.osgeo.org/qgis/changeset/10498>
06:02:46 pcav: jef: now got it
06:02:54 pcav: looks nice, thanks!
06:02:58 jef: pcav: np
06:03:20 jef: timlinux: should r10496-7 be applied to 1.0?
06:03:48 jef: timlinux: seem there are quite a few people very enthusiastic about it ; )
06:06:40 jef: hm
06:56:29 jctull: jef: was qgsvectoroverlayplugin.cpp likely used in plugins to load vector layers?
06:58:12 jef: jctull: I don't think so. it was merged from the diagram ( aka vector overlay ) branch. guess that started as a plugin and was meanwhile migrated into app
06:58:52 jctull: ok, I wondered because I tried using one of the osm loader plugins and nothing was loaded to the legend/canvas after parsing. thx
07:04:47 jctull: I get a "Warning: Failed to load" message after parsing.
07:06:42 jctull: I think it is the plugin. It doesn't work in the older analysis branch build either.
07:10:47 jef: timlinux: should r10496-7 be applied to 1.0?
07:10:51 jef: timlinux: seem there are quite a few people very enthusiastic about it ; )
07:11:03 : * timlinux goes to look
07:11:15 timlinux: @revision 10496
07:11:30 timlinux: @changeset 10496
07:11:30 sigq: timlinux: http://trac.osgeo.org/qgis/changeset/10496
07:12:12 epifanio: guys ther is a little problem about grass-gui modules window in osx
07:12:41 timlinux: jef: sounds good
07:12:43 epifanio: maybe lookin at it now when you are working on .. is the right moment
07:12:59 epifanio: let me explain the problem
07:13:35 epifanio: when we start the grass tools, and we open a large number of modules
07:14:07 epifanio: the gui expand itself "horizzontal"
07:15:09 epifanio: and quit qgis, reopen the grass modules-gui .. it start with a large horizzontal size ... and is no longer resizable
07:16:10 epifanio: whle on linux, we have a nice arrow - button to scroll the menu-list on the top of the main-window( grass modules )
07:16:27 epifanio: these arrorow are not present on osx
07:22:23 epifanio: jctull: can you confir what i described ?
07:22:40 jctull: sorry, been elsewhere. let me review
07:23:52 jctull: epifanio: can you point to a particular module that is causing the expansion?
07:24:09 jctull: mine does not appear to auto-expand
07:24:44 epifanio: no it isn't related to one module, if you open more than one module .. add, add, ... it enlarge the window
07:25:04 epifanio: and then if you reopen it , it is not longer resizable
07:25:08 jctull: oh, I see what you mean.
07:25:30 epifanio: adding module you can have a window more larger than the screen :-(
07:26:07 epifanio: while on linux we have arrows on the top, to scroll between opened modules
07:26:36 jctull: I have to open something like 12 or more modules. Not exactly a practical issue, but an aesthetic problem nonetheless. I am guessing it will happen on other systems also.
07:27:17 jctull: I can resize mine after I close grass tools and repoen
07:28:01 jctull: At least to the minimum width or greater for the number of open modules
07:28:42 epifanio: ah .. i mean close the open module ( without close the window ) sorry, it stop its resize ont he max-size you have at the moment
07:31:32 jctull: epifanio: I see. Yes, it no longer can resize below the extent the window auto expanded to accommodate all the open modules. It gets stuck at the larger window size.
07:32:20 epifanio: yes, that's what i tried to describe ( thanks to give a better english description ;- ) )
07:32:46 jctull: np, don't ask me to do it in any other languages and we should be ok ; )
07:32:58 epifanio: hehe
07:34:36 cgs_bob: hello all. GRASS related question here : ) I have point feature that does not cover the whole region. I was to create a raster mask where there are no points that I would use with v.surf.rst. any suggestions?
07:35:23 jctull: cgs_bob: in grass ( or with grass commands ), you could set the region to your point layer with g.region vector=point_layer
07:35:34 jctull: then run your v.surf.rst command
07:35:56 jctull: Make sure your resolution is where you want it: g.region res=10 ( e.g., for 10m, if metric )
07:36:19 jctull: Is that what you wanted?
07:36:41 jctull: You don't need to mask it.
07:36:45 cgs_bob: jctull: I've done that. I'm just wishing that there was a concave hull
07:36:55 jctull: hmmm
07:37:41 epifanio: if you need to interpolate point in a precise ( limited ) zone that differ from the region ( recangular-size ) but it is an irregular zone ( like a poligon ) , you can create a mask of the zone you want interp and use it in v.surf.rset
07:38:19 CIA-56: jef * r10499 /branches/Version-1_0/src/plugins/grass/ ( config/default.qgc qgsgrasstools.cpp ): merge r10496 and r10497 to 1.0
07:38:38 epifanio: so the mask==zone where you need interp
07:38:47 sigq: tracfeed: Changeset [10499]: merge r10496 and r10497 to 1.0 <https://trac.osgeo.org/qgis/changeset/10499>
07:39:38 cgs_bob: epifanio: I could do that. thanks for the suggestion
07:41:01 epifanio: cgs_bob: are your point dense? are they elevation data, or other ?
07:41:10 cgs_bob: just wish there was as way to wrap a polygon around all the points.
07:41:53 cgs_bob: epifanio: yes they are. I have over 10000 points in a small area
07:43:01 epifanio: ah ok, so v.surf.rst is good, if you have instead point not with a omogeneus density you can also try : http://grass.itc.it/grass62/manuals/html62_user/v.surf.bspline.html .. but it is designed not for elevation data
07:43:02 sigq: Title: GRASS GIS: v.surf.bspline ( at grass.itc.it )
07:45:53 CIA-56: kyngchaos * r10500 /trunk/qgis/mac/xcode/ ( Qgis.xcodeproj/project.pbxproj qgis_settings.xcconfig ): vector overlay added, misc updates ( r10481, r10484 )
07:45:56 sigq: tracfeed: Changeset [10500]: vector overlay added, misc updates ( r10481, r10484 ) <https://trac.osgeo.org/qgis/changeset/10500>
07:47:21 jctull: cgs_bob: you want a minimum convex polygon of the points then
07:50:36 jctull: look at v.adehabitat.mcp in grass-addons
07:50:36 jctull: http://svn.osgeo.org/grass/grass-addons/vector/adehabitat/
07:50:37 sigq: Title: Revision 36609: /grass-addons/vector/adehabitat ( at svn.osgeo.org )
07:50:52 cgs_bob: epifanio: if I create a polygon around my points, convert it to a mask and use that with v.surf.rst, I might get stairstep caused by areas that don't have points. so I was hoping that I can do something like this: http://ubicomp.algoritmi.uminho.pt/local/concavehull.html
07:50:54 sigq: Title: Concave Hull vs. Convex Hull ( at ubicomp.algoritmi.uminho.pt )
07:51:24 cgs_bob: jctull: I'll take a look...thanks!
07:51:43 jctull: it's the same thing, basically. Just do an mcp with 100% of your points
07:52:03 jctull: maybe, you'll have to test
07:52:14 epifanio: the mask must be a raster, you have to convert the vector area with v.to.rast
07:52:24 epifanio: and use it as mask
07:52:30 cgs_bob: jctull: super! are there any dependancies?
07:52:32 jctull: an mcp does not do concave, only convex, so you will end up with some extra space in your mask
07:53:04 jctull: minimum convex polygon = mcp
07:53:34 cgs_bob: jctull: that's too bad :( I was hoping to see cfarmer to see if he can add concave hull to fTools : )
07:53:40 jctull: what about v.hull?
07:53:57 jctull: that exists
07:54:10 jctull: http://grass.itc.it/grass64/manuals/html64_user/v.hull.html
07:54:12 sigq: Title: GRASS GIS manual: v.hull ( at grass.itc.it )
07:54:40 jctull: no muss, no fuss as it should be already in your grass distro
07:54:56 cgs_bob: I've tried that. doesn't work well
07:56:08 cgs_bob: well, I have to run now. bbl
08:04:49 jctull: cgs_bob: maybe you need to buffer the result ( make it go out some distance from the v.hull vector ). Not sure... good luck
08:43:38 cgs_bob_: cfarmer: hi there! just wondering if you can add this to fTools: http://www.cs.duke.edu/~edels/Surveys/AlphaSurvey.pdf : )
08:44:25 : <_-Slash-_> \o/
08:44:26 jctull: cgs_bob: so, why does convex hull not work for you?
08:44:53 jctull: you need some additional buffering I am guessing
08:45:12 cgs_bob_: jctull: I'll answer you in a minutes...almost at work : ) brb
08:45:26 jctull: caltrain?
08:45:40 cgs_bob_: car : )
08:45:51 jctull: drive, man.. quit typing! ; )
08:45:55 cgs_bob_: brb
08:48:46 cfarmer: cgs_bob: when you are safely *not* moving, try using the delaunay triangulation tool in fTools... this should give you nearly what you are looking for, then the only remaining issue is to remove all polygons that share a boundary with the convex hull of the entire point dataset, and then dissolve the final output...
08:49:13 cfarmer: wait, not quite that simple, but you get the idea...
09:06:33 cgs_bob: jctull: I made it safely to work. just be glad you weren't in front of me : )
09:08:29 cgs_bob: cfarmer: just saw your note. I'll give it a try
09:09:13 cfarmer: cgs_bob: coolio.. I have to head out for a bit, but I'll check back to see how you make out...
09:09:59 cgs_bob: cfarmer: see ya soon
09:38:08 jctull: jef: FYI, selectplus is the plugin that is not playing well with postgis
10:01:42 macho: mrng
10:06:04 cfarmer: cgs_bob: how'd that work for you?
10:07:12 cgs_bob: cfarmer: if you remove all of the interior polygons, it looks like a convex hull : )
10:08:02 cfarmer: yeh, but if you remove the exterior polygons it makes what you want doesn't it?
10:09:43 cgs_bob: cfarmer: imagine a cluster of points shaped like a "C". a convex hull of that would give you an approximate semicircle...I want a "C" : )
10:11:49 jctull: cgs_bob: why don't you buffer your points, extract the nodes, then do a convex hull on that? Will that do it?
10:13:27 cfarmer: ok, no convex hull... let me think about this... but I think the delaunay triangulation is what you will need... modified slightly... will be back later...
10:13:45 : * cfarmer wbrb
10:14:43 jctull: jef: on OS X, when I save a project, it places it in a directory not selected in the file dialog.
10:15:15 jctull: So, I want to save to my desktop, I browse to the desktop, but the file is saved in some other directory ( the same directory every time, btw )
10:17:31 jctull: I will file a bug on this. epifanio: is this occurring for you also?
10:18:14 epifanio: let me check it now
10:20:09 epifanio: jctull: here it works as aspected
10:20:32 epifanio: i compiled qgis 1 hour ago using xcode
10:20:46 epifanio: but i never had these problem
10:20:58 dcope: hi all. looking to enable .ecw support in qgis under windows. do i need to recompile, or just copy dlls from somewhere? I have gdal running under fwtools with ecw support.
10:21:34 jctull: epifanio: you can save a project and it ends up where you browse to in the file dialog?
10:22:14 epifanio: yes, i select desktop to save project and is saved on desktop
10:22:57 jctull: ok, thanks
10:23:46 jctull: probably a qt bug. I'm still on rc1
10:23:50 jctull: 4.5-rc1
10:28:33 jctull: cgs_bob: I see, you need a concave hull of sorts. My mind is operating slowly today.
10:29:52 jctull: cgs_bob: http://lists.osgeo.org/pipermail/grass-user/2008-August/046080.html
10:29:53 sigq: Title: [GRASS-user] Concave hull of a set of points ( alphashapes ) ( at lists.osgeo.org )
10:30:17 cgs_bob: jctull: maybe we can get cfarmer to do something like this in fTools: http://www.cs.duke.edu/~edels/Surveys/AlphaSurvey.pdf
10:30:38 jctull: yes, but look at the above post for a grass solution
10:31:00 jctull: I suppose hand digitizing is not really a preferred option.
10:32:55 wildintellect: dcope, are you using the OSGeo4w installer?
10:33:41 cgs_bob: jctull: thanks a bunch...that is what I'm looking for!
10:34:09 dcope: wildintellect: no, i'm still on 0.11.0-Metis at work, since I am 'locked down' ( no admin rights ) and can't install OSGeo4W without an admin headache ; )
10:36:08 epifanio: OT. : new spatial lite release :- ) .. added : Routing - EXIF GPS - VirtualFDO - AMALGAMATION - GUI-tool enanchments
10:36:12 wildintellect: dcope, you could try copying all of the gdal stuff from fwtools over their conterparts in the qgis install folder( backup first )
10:36:57 dcope: wildintellect: which folders would i need to copy?
10:37:36 wildintellect: I don't have a copy in front of me but probably lib
10:38:14 wildintellect: this is all a guess though
10:48:53 dcope: wildintellect: tried copying 'lib' to 'lib' but no success
10:49:53 wildintellect: it's unlikely the ecw will show up on the list and I can't remember if 0.11 had the generic any gdal format or all files option
10:50:55 dcope: yes, the file names show up under "all formats" but i get "*.ecw is not a supported raster data source"
10:51:43 wildintellect: hmm, are there any other folders in your fwtools that have .dll files
10:56:23 markusN: hi jef - you there?
10:56:28 markusN: error: ‘mModules’ was not declared in this scope: http://rafb.net/p/1mv7nF75.html
10:56:29 sigq: Title: Nopaste - No description ( at rafb.net )
10:56:46 markusN: I assume that this is a typo in the recent fix/backport
11:06:47 markusN: indeed
11:07:05 markusN: fixed: http://rafb.net/p/Dl0x8D14.html
11:07:06 sigq: Title: Nopaste - No description ( at rafb.net )
11:09:32 dcope: hmm /bin has lots of .dll files
11:17:22 wildintellect: do any of them have the same names as files in the QGis folder
11:19:54 dcope: no. none of them seem to be the same names at all..
11:28:10 wildintellect: dcope, well backup qgis if you haven't already and try copying some over
11:28:59 dcope: wildintellect, yes, tried but with no success. Think i've burned enough time on this for now. Will mess with it at home and see what i come up with. Thanks for your help
11:29:23 wildintellect: dcope, email the qgis-user list someone may have a solution for you
11:29:41 dcope: thanks.
11:31:27 CIA-56: jef * r10501 /branches/Version-1_0/src/plugins/grass/qgsgrasstools.cpp: fix r10499 - ouch
11:31:49 sigq: tracfeed: Changeset [10501]: fix r10499 - ouch <https://trac.osgeo.org/qgis/changeset/10501>
11:33:04 CIA-56: jef * r10502 /branches/Version-1_0/ ( 11 files in 5 dirs ): update debian packaging
11:33:05 sigq: tracfeed: Changeset [10502]: update debian packaging <https://trac.osgeo.org/qgis/changeset/10502>
15:03:32 CIA-56: cfarmer * r10503 /trunk/qgis/python/plugins/fTools/tools/doGeometry.py: udpates to centroid function, should work with any and all projections
15:04:11 sigq: tracfeed: Changeset [10503]: udpates to centroid function, should work with any and all projections <https://trac.osgeo.org/qgis/changeset/10503>
15:24:32 CIA-56: jef * r10504 /trunk/qgis/src/core/ ( 16 files in 10 dirs ): upgrade internal spatialite from beta to amalgamated 2.3.0 release
15:24:33 sigq: tracfeed: Changeset [10504]: upgrade internal spatialite from beta to amalgamated 2.3.0 release <https://trac.osgeo.org/qgis/changeset/10504>
22:52:10 macho: mrng
23:04:45 jef: moin werner
23:15:42 macho: hi jef
23:15:47 macho: any news in Norden?