#OSGEO IRC Log - 2010-01-15

For logs after Feb 3, 2007, all times are GMT-8. Prior logs are GMT-9.
Back to Logs
03:09:07 sigq: osgeofeed: Changeset [2923]: avoid repeats in the package hog list <http://trac.osgeo.org/osgeo/changeset/2923>
03:09:29 CIA-14: osgeo: hamish * r2923 /livedvd/gisvm/trunk/bin/main.sh: avoid repeats in the package hog list
03:43:21 sigq: osgeofeed: Changeset [2924]: git package moved to gnuit <http://trac.osgeo.org/osgeo/changeset/2924>
03:43:37 CIA-14: osgeo: hamish * r2924 /livedvd/gisvm/trunk/bin/setup.sh: git package moved to gnuit
04:18:10 CIA-14: osgeo: hamish * r2925 /livedvd/gisvm/trunk/bin/install_gpsdrive.sh: updates for karmic pkg versions
04:18:12 sigq: osgeofeed: Changeset [2925]: updates for karmic pkg versions <http://trac.osgeo.org/osgeo/changeset/2925>
07:34:06 lambda: hi, currently I am testing OSGeo4W on MS Windows Server 2008 ( 64bit ) and I have significant problem with MSys, when I start MSys it's open 129times and then fails, has anyone similar experience here?
08:15:05 sigq: osgeofeed: Changeset [2926]: call4workshop and workshop/tutorial page added <http://trac.osgeo.org/osgeo/changeset/2926>
08:15:33 CIA-14: osgeo: lbecchi * r2926 /foss4g/2010/website/ ( index.php navigation.php ): call4workshop and workshop/tutorial page added
08:16:20 CIA-14: osgeo: lbecchi * r2927 /foss4g/2010/website/workshop.php: workshop/tutorial page now added
08:16:26 sigq: osgeofeed: Changeset [2927]: workshop/tutorial page now added <http://trac.osgeo.org/osgeo/changeset/2927>
14:27:35 dandy: question from the inexperienced for the experienced, are there any projections with different datums but using the same elipsoid?
14:35:32 jgarnett: thinking; not sure I am that smart in the morning
14:36:41 jgarnett: I am pretty sure there are ... but am drawing a blank at how to check.
14:39:29 dandy: another question I have
14:39:35 dandy: why does MultiPolygon exist?
14:40:55 dandy: as far as I can see the only difference is that MultiPolygon allows for multiple external fill rings
14:41:09 dandy: but I don't see why Polygon is restricted from that
14:41:16 dandy: is there some historical or legacy reason for this?
14:42:25 jgarnett: um a polygon is a single thing; ie from geometry class
14:43:08 jgarnett: when they start doing GIS to describe real world things
14:43:21 jgarnett: they run into the situtation where the same "idea" has several physical representations ( ie polygons )
14:43:21 dandy: Polygon allows for multiple rings
14:43:31 jgarnett: they would like to treat the group as a single name; and thus made up "multi polygon" for the purpose
14:43:32 jgarnett: an example would be the base supports of a large bridge
14:43:46 jgarnett: they are all "the bridge" but each support takes up some area on the ground that roads need to go around
14:46:26 jgarnett: thought it was a "bound area" of some sort
14:46:26 jgarnett: ie I don't think they allow for multiple rings
14:46:34 jgarnett: it could be that the kind of polygon used in gis is restricted as all get out; ie "simple" "not self intersecting" and so on ...
14:46:34 dandy: yeah I get the nomenclature thing but in GIS terms, Polygon and MultiPolygon differ only in the number of a certain type of ring.
14:46:34 dandy: Polygon does allow for multiple rings
14:46:35 dandy: maybe
14:46:35 jgarnett: so I expect this is an accident of history and math; and multi polygon is the work around...
14:46:36 dandy: do MultiPolygons allow the Polygons within them to intersect?
14:47:16 dandy: yeah it seems like it
14:50:49 dandy: i think I am going to ditch the concept of multipolygon internally in stuff I write and instead write something to help interoperate with it
14:51:08 jgarnett: hrm; you would be a bit on your own there
14:51:08 jgarnett: JTS, SDO_GEOMETRY and so on all take pains to seperate out the two cases
14:51:08 dandy: ?
14:51:08 jgarnett: for polygon having multiple rings; usually it is defined as a single outer ring; and many internal ones
14:51:09 jgarnett: ( some systems even force the rings to be counter clockwise for oter or clockwise for interior )
14:51:09 jgarnett: JTS Topology Suite; ported to C++ for GEOS etc...
14:51:09 jgarnett: SDO_GEOMETRY is the oracle spatial representation of geometry
14:51:09 pramsey: MulitiPolygons do not allow ring interactions.
14:51:09 jgarnett: when you say polygon does allow for multiple rings; what software are you refering to?
14:51:09 pramsey: However, dispensing with MultiPolygon is not the asnwer.
14:51:09 dandy: right
14:51:09 dandy: so internally for my algorithms I would have a polygon that supports multiple outter rings
14:51:09 dandy: but when I need to interoperate I can quickly convert that polygon or collection of polygons into something more conformant
14:51:51 pramsey: Dispensing with Polygon is. Since a Polygon can be easily represented as a one-part multipolygon.
14:51:53 jgarnett: dandy perhaps you are confusing what can be done with what is valid?
14:52:06 pramsey: ( in fact, that's how shape files model things. all the objects are assumed to be multipart )
14:52:28 dandy: well I suppose I am looking at it wrong, I am instead going to use a multipolygon and then later have to downconvert it to polygons
14:52:41 dandy: polygon winding is not an issue as long as it is consistent and can be easily corrected
14:54:35 : * pramsey thinks being explicit about what's an exterior ring and what's an interior one is more robust.
14:54:35 pramsey: that relying on any kind of vertex ordering rule
14:54:35 dandy: pramsey, I think that is how I should look at it, thanks
14:55:01 dandy: jgarnett, examples are shapefile, OGC, etc...
14:55:11 dandy: yeah I have a optional flag for hole or fill
14:55:15 dandy: but also allow for winding
14:55:26 dandy: I totally agree with the explicit flag
14:55:33 dandy: especially when dealing with multiple data sources
14:55:41 dandy: shp,sde,ogc, etc..
14:56:04 dandy: then bringing in data from non GIS sources tends to be CCW while SHP is CW if I remember correctly
14:56:05 jgarnett: cool
14:57:11 pramsey: WRT your projection question: yes
14:57:36 dandy: thanks
14:57:41 pramsey: lots of projections that are the same but for datum and lots of modern projections that differ in datum
14:59:13 pramsey: eg EPSG: 26910, EPSG:26710
14:59:42 dandy: perfect, love examples
15:01:41 dandy: thanks a bunch
19:04:07 FrankW: hobu, crschmidt: is something up with spatialreference.org? I don't seem to be able to reach it.
19:38:38 crschmidt: Working on it.
19:43:20 crschmidt: FrankW: fixed.
19:44:01 FrankW: thanks
19:44:11 FrankW: I must confess my need has passed.
19:44:38 crschmidt: Understood.
19:44:49 crschmidt: I apologize for not having instant reaction at 11:45 at night : )
19:44:53 crschmidt: ( or 11:15 )
19:45:06 FrankW: no problem, of course.
19:45:18 FrankW: I just consulted the libgeotiff epsg csv files.
19:45:30 FrankW: But some folks might not have mastered that fallback
19:46:05 : * crschmidt nods.