[postgis-users] st_transform crashes
Hi there! I have a table containing unprojected footprints from observations on Mars, just polygons and multipolygons (no curved). For Mars, reference systems are defined by IAU2000 codes, similar to EPSG codes for earth data. This is the projection the data currently has: http://spatialreference.org/ref/iau2000/49900/ ...and I need it to be projected into this one: http://spatialreference.org/ref/iau2000/49910/ Therefore, I took the PostGIS insert statements on the respective pages, ingested them into the database and then tried to project the data using select st_transform(geom_shifted, 949910) from alldata The result is that the database crashes with the message *** Error *** and I need to restart the PostgreSQL service in order to use it again. I have no clue why the database crashes. This is what I tried to far: (1) set the srid to epsg 4326 just for testing and projections work fine, for example to 90093 (2) projected from IAU2000:49900 to some other common mars projections listed on http://spatialreference.org/ref/?search=mars . all of the projections worked except for the Equidistant Cylindrical ones (database crash as described), but those are exactly the ones I need! (3) exported the data to a shapefile and tried to project it using ogr2ogr. command: ogr2ogr -f "ESRI Shapefile" -s_srs "+proj=longlat +a=3396190 +b=3376200 +no_defs" -t_srs "+proj=eqc +lat_ts=0 +lat_0=0 +lon_0=0 +x_0=0 +y_0=0 +a=3396190 +b=3396190 +units=m +no_defs" "1.shp" "2.shp" worked fine! The proj4 definition I needed for the ogr2ogr command for IAU2000:49913 cannot be exported on spatialreference.org...Any clue why? The 'custom', non-EPSG codes can't be the problem because of (1) and (2). Why would ogr2ogr do the job and PostGIS would not? Is "+proj=eqc +lat_ts=0 +lat_0=0 +lon_0=0 +x_0=0 +y_0=0 +a=3396190 +b=3396190 +units=m +no_defs" even correct? I found that proj4 string in other data projected with an equidistant cylindrical projection. Is anyone able to translate that into a PostGIS insert statement? I was not able to find anything on problems with PostGIS and equidistant cylindrical projections... I uploaded a shapefile containing the data: http://www.file-upload.net/download-3621892/problem.zip.html Thanks in advance! Kind regards, Bryan ___ postgis-users mailing list postgis-users@postgis.refractions.net http://postgis.refractions.net/mailman/listinfo/postgis-users
Re: [postgis-users] PostGIS documentation license
+1 with Regina. This is exactly my experience. My clients have no issues contributing back because the cost of constantly merging their "private" code does not offset the value added by keeping it private. On the other hand they are VERY reluctant to integrate any GPL code base with their products because they do not want to risk polluting the code that is legitimately proprietary because it in inadvertently mixed in some GPL code. A good example is mapserver.org that uses an MIT-X license and a huge part of the development effort is funded by commercial companies contracting to developers to add this or that feature to the base product. I don't what to start a this or that is better war. These things tend to be very personal beliefs and there are many models and all seem to be working, I just think it is a shame the the ecosystem is broken into all these camps. -Steve On 7/27/2011 1:50 PM, Paragon Corporation wrote: Paolo, I think software is more valuable the more people are using it and stress testing it. That said, I'm more concerned about people not using PostGIS because they fear they will have to release their proprietary source code, more than I am worried about people not giving back to the community. Even people who don't give back, find bugs and complain which makes the software stronger and more robust when we fix it. There are plenty of people producing proprietary software that have no qualms about giving their enhancements back to PostGIS/GEOS etc and in fact beg us to take them so they don't have to cut in there changes with each release. We have several customers like that -- e.g. the tiger geocoder work we are doing, x3d export, the ability for shp2pgsql to be able to read raw dbf (with no shape), the ability to export .prj were all pieces of work we did as part of work we did for clients marketing closed source SaaS that relies on PostGIS. They're willingness to give back these changes wouldn't have been any different if PostGIS were GPL or BSD because to them PostGIS is just a wheel in their armor like any other database software would be. However the fact that PostGIS is GPL does give some a pause for concern as to how they distribute it etc and their willingness to even use it since it does bring up the question of where their software begins and PostGIS ends. If you use Oracle, SQL Server etc, the fact there is a clear payment and exchange of goods makes it in some cases a safer choice if you are worried about protecting your intellectual property. Thanks, Regina -Original Message- From: postgis-users-boun...@postgis.refractions.net [mailto:postgis-users-boun...@postgis.refractions.net] On Behalf Of Paolo Cavallini Sent: Wednesday, July 27, 2011 2:46 AM To: postgis-users@postgis.refractions.net Subject: Re: [postgis-users] PostGIS documentation license Il 26/07/2011 22:06, Paragon Corporation ha scritto: Regarding GPL. Mostly I just find the whole licensing confusing. I don't think I'm the only one who would be happier if PostGIS was under a BSD/MIT or some other licensing. Mostly just for the headache of arguing about what you can and can't do with it for commercial purposes. For what is worth: I'm against it, as this would make it easy to use the code we developed in proprietary projects, returning nothing to the project. All the best. ___ postgis-users mailing list postgis-users@postgis.refractions.net http://postgis.refractions.net/mailman/listinfo/postgis-users
Re: [postgis-users] PostGIS documentation license
I kind of agree with Regina here... On 28/07/2011, at 1:50 AM, Paragon Corporation wrote: > They're willingness to give back these changes wouldn't have been any > different if PostGIS were GPL or BSD because to them PostGIS > is just a wheel in their armor like any other database software would be. > However the fact that PostGIS is GPL does give some a pause for concern as to > how they distribute it etc and their willingness to even use it since it does > bring up the question of where their software begins and PostGIS ends. We certainly don't have the ability to give much back - except the bug reports and maybe some help on lists, but we do occasionally have clients (some of them very small local parts of very large multinational organisations) who want solutions without any of the complications of the licensing required by GPL. If it looks like getting confusing or complicated, they would rather just pay. While I personally might believe that this is misguided, I can't really afford to turn them away... although if they could give me the money they shelled out on proprietary database licences, I wouldn't have so many problems... Anyhow... ___ postgis-users mailing list postgis-users@postgis.refractions.net http://postgis.refractions.net/mailman/listinfo/postgis-users
Re: [postgis-users] PostGIS documentation license
On Wed, Jul 27, 2011 at 01:50:54PM -0400, Paragon Corporation wrote: > I think software is more valuable the more people are using it and stress > testing it. Only if such stress-testing results in actual fixes to the code... > That said, I'm more concerned about people not using PostGIS > because they fear they will have to release their proprietary source code, > more than I am worried about people not giving back to the community. Even > people who don't give back, find bugs and complain which makes the software > stronger and more robust when we fix it. If a company's strategy is to sell a proprietary portion of a system, there might be a tendency to invest more on workarounds than on fixes to the free software part. Doing so would increase the added value in the mixed proprietary/free solution. This is to say that stress-testing might result in proprietary improvements more than bugfixes, unless the users of the software are obliged to release the derivative work as free software as well. > However the fact that PostGIS is GPL does give some a pause for concern as > to how they distribute it etc and their willingness to even use it since it > does bring up the question of where their software begins and PostGIS ends. I don't see how is this bad. > If you use Oracle, SQL Server etc, the fact there is a clear payment and > exchange of goods makes it in some cases a safer choice if you are worried > about protecting your intellectual property. Excellent. This is really how copyleft should work. It _must_ be an expensive choice not to give back to the community. If you're willing to pay big bucks to be greedy, go ahead :) --strk; () Free GIS & Flash consultant/developer /\ http://strk.keybit.net/services.html ___ postgis-users mailing list postgis-users@postgis.refractions.net http://postgis.refractions.net/mailman/listinfo/postgis-users
