I would like to express a different point of view. One of the nice things about GRASS in its current form is that one download will get you hundreds of modules without having to worry about installing add-ons.
What's the difference between having 300 or 500 modules in the base distribution? GRASS has the file layout required to scale up to this size easily, and each individual module is small and lean. Even if we were to triple the number of modules in the base distribution, the download size would not increase significantly (and vice versa). On the other hand: - Having to compile add-ons written in C/C++ is a big obstace, g.extension or not, for many users, especially on Windows. - It is hard to ensure consistent quality across external add-ons. - Documentation for such add-ons cannot be included in the main online index. - There is always a risk that useful add-ons maintained by external developers will be abandoned. - Putting modules into extension groups will never be free of overlap and always leave people wondering where the wanted module is. The basic r.*, v.* etc. and some good keywords are a much better system IMHO. So I am not convinced about the merits of breaking the GRASS distribution into smaller pieces and add-ons. Instead, I would say the base distribution should grow as required and include new, useful modules such as r.modis. Cheers, Ben On 10/30/2012 11:43 AM, Yann Chemin wrote:
Hi Martin, I have no worries, since I do load modules directly. I am worried about the learning curve for Windows-bound students in need of MODIS images, but not comfortable with GRASS GIS yet. But this is an old story coming back, I am all for reducing significantly the number of modules in GRASS, to have a lean download. We can indeed have groups of plugins to get through g.extension: - Hydrologically related - Imagery - Temporal - 3D - DB (though I am not a user, so cannot be sure of how much needed) This would keep import/export, some of the basic raster and vector ones. Neat small download, use g.extension for anything else, by group or on demand. Yann On 30 October 2012 15:08, Martin Landa <landa.mar...@gmail.com <mailto:landa.mar...@gmail.com>> wrote: Hi, 2012/10/30 Yann Chemin <yann.che...@gmail.com <mailto:yann.che...@gmail.com>>: > as a regular user of MODIS, I would like to call other MODIS users to > express interest to include r.modis into GRASS 7 SVN. we cannot extend number of modules in trunk forever. Probably some modules should be reviewed and moved to addons. Do you have any problem with installing r.modis using g.extension? Martin -- Martin Landa <landa.martin gmail.com <http://gmail.com>> * http://geo.fsv.cvut.cz/~landa <http://geo.fsv.cvut.cz/%7Elanda> -- Yann Chemin _______________________________________________ grass-dev mailing list grass-dev@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/grass-dev
-- Benjamin Ducke {*} Geospatial Consultant {*} GIS Developer bendu...@fastmail.fm _______________________________________________ grass-dev mailing list grass-dev@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/grass-dev