Talk:GRASS Community Sprint Prague 2013
- 1 Participants' reports
- 1.1 Joint efforts
- 1.2 Individual Reports
- 1.2.1 Martin Landa
- 1.2.2 Helena Mitasova
- 1.2.3 Markus Neteler
- 1.2.4 Yann Chemin
- 1.2.5 Margherita Di Leo
- 1.2.6 Markus Metz
- 1.2.7 Anne Ghisla aka aghisla
- 1.2.8 Sören Gebbert aka huhabla
- 1.2.9 Stepan Turek
- 1.2.10 Thomas Leppelt
- 1.2.11 Milena Nowotarska
- 1.2.12 Anna Petrasova
- 1.2.13 Vaclav Petras
- 1.2.14 Eva Stopková
- 1.2.15 Tereza Fiedlerová
- 2 Options to discuss with Sprint participants
Major achievements have been done! See below for an incomplete list...
... and: 30 YEARS OF GRASS GIS! http://tinyurl.com/30-years-of-GRASS-GIS
(see here for participants' names)
- Creating 3D vector test data (for 3D interpolation) - ES, SG, MN
- Added General introduction (based on manual of G7/i.vi by Yann Chemin) - MN, MN, YC
- Discussions on Bundle Block Adjustments - YC, ST
- wxGUI refactoring - AP, VP
- Presentation of state of image processing in GRASS 7 - YC, MN
- Presentation of state of Temporal GIS Algebra in GRASS 7 - SG, TL
- ISIS-GRASS bridge port to GRASS7 - YC, VP, ML, MN
- Discussions on potential integration of Rasdaman software as raster data backend - SG, YC, and Jachym (virtually)
- Presentation on vector conflation - TF with discussion
(participants in random order)
- Community Sprint organization
- OSGeo4W GRASS 6.4.3RC4 package published
- Fix on MS Windows,
- Fix on MS Windows,
- Update OSGeo4W package 'msys' to 1.0.18, http://trac.osgeo.org/osgeo4w/wiki/pkg-msys
- Preparation of 'mingw' OSGeo4W package
- Discuss compiles issues on Mac OSX - with AP, VP
- Discuss Temporal GIS Algebra implementation - with SG
- Community Sprint Budget management (available on request)
- r.wf Addon fixed: and
- r.ipso Addon fixed:
- Draft Doxygen'iation of the CDHC library for testing normality & exponentiality
- Added General introduction to the image processing wiki/manual page (based on manual of G7/i.vi by Yann Chemin) - MN, MN, YC
- Backport of i.latlong to GRASS 6 (Addon) in .
- screenshots for example in new module v.surf.mass (see Markus Metz entry)
- Merging in translations (i18N) done by Milena Nowotarska and Andreea Marin
- ISIS-GRASS bridge port to GRASS7 (with Vaclav, Martin and MarkusN)
If ISIS (http://isis.astrogeology.usgs.gov) is installed, GRASS will load it in the background, the banner and prompt will change to ISIS-GRASS. Both ISIS and GRASS commands are available at the ISIS-GRASS prompt.
- g.isis3mt exports a ISIS3 template from current location, ported into GRASS 7
- Tested direct access of ISIS3 data into G7, r.in.gdal and r.external both work (tested Mars MOLA and Dawn test_datasets)
- Discussed with Soeren, Vaclav and Jachym about Rasdaman, Voxel lib and Raster handling in Grass 8, we agreed on investigating handling, interfaces and performances.
- Went through the Bundle Block Adjustment files with Stepan
- Landsat 7 chain processing using pyGRASS goes in testing mode with MarkusN and Soeren
- Worked more on r.crater
Margherita Di Leo
- r.basin ported to G7 ,
- r.basin html fixed
- r.stream.* modules: clean up before moving from addons to trunk ,
- GRASS wiki image processing update (MM, MN, YC)
- standardizing options for v.db.* and db.* modules, updating respective manuals
- new module v.surf.mass for Tobler's pycnophylactic area interpolation (manual page)
- new module v.area.weigh for area interpolation with known cell weights, preserving area properties like Tobler's pycnophylactic area interpolation
- i.segment: object recognition fixed for special cases, tested with 2 billion cells
- v.kcv optimized (thanks to Jan Ruzicka and Jan Vandrol)
Anne Ghisla aka aghisla
- Google Summer of Code live discussions
- v.krige fixed in G7 and following
Sören Gebbert aka huhabla
- Google Summer of Code Temporal Algebra presentation and implementation support
- Discussion of RASDAMAN GRASS integration
- Discussion and implementation of the new temporal module t.rast.accumulate and the use of a modified version of r.gdd (implemented by Markus Metz) instead of r.mapcalc
- Writing and updating of many temporal modules manual pages
- Implementing a prototype that uses Python multiprocessing and a new display C-library function derived from g.pnmcomp to speedup the image composition in the wx GUI
- Working on Google Summer of Code project
- Went through the Bundle Block Adjustment files with Yann
- Released source code of scatter plot backend to backend and turns backend
- Discussed GSoC with Martin and Soeren
- Working on GSoC Project: Temporal Algebra
- Discussing the algebra concept.
- Implemented functions for if-statements in the temporal algebra.
- Added several temporal functions like start_date(), end_time(), start_datetime().
- Implemented topological relationships for if-statements.
- Added topological relations for comparison operators in conditional statements.
- Finished to implement the hash-operator (#) for map counting.
- New module t.select alpha version in repository ready for testing. GSOC-wiki
- GUI translation to Polish
- i18n fixes
- winGRASS testing
- attempts to compile wxPython with wxGTK3/Broadway (HTML5) backend, so far not successful, see discussion
- wxGUI refactoring
- wxGUI design
- toolboxes: discussion, adding <addons> tag
- compiling wxPython with wxGTK3/Broadway with Anna
- discussion of Rasdaman GRASS integration
- wxGUI refactoring
- wxGUI design
- different source file for module search/tree and main menu
- ISIS-GRASS bridge port to GRASS7 with Yann
- discussing wxGUI programming with Stepan
- handling of translations (calls of gettext underscore function), fixes bug in wxGUI (parts remained untranslated)
- debugged plotting variogram output in module v.kriging (not published yet)
- improved and debugged other functions of module v.kriging
- discussion about kriging methods in GRASS GIS with Anne G.
- discussion about optimalization and testing modules (not published yet) v.kriging (2D/3D kriging) and v.nna (2D/3D Nearest Neighbour Analysis) with Markus N. and Soeren G.
- discussion about 3D kriging, 3D Nearest Neighbour Analysis and possibilities of future work on the modules with Helena M.
- numerical testing of v.kriging:
- prepared inputs
- fixing the bugs
- learning something about GRASS GIS vector modules
- discussion about own (yet unpublished) QGIS plugin and C++ library for vector conflation and its possible implementation to GRASS
- beginning to write C API for vector conflation library (needed for GRASS implementation)
Options to discuss with Sprint participants
- IRC - chat
- Google Hangout (see below)
- come to Prague :)
Google Hangout option:
Q: do you have to be signed up with a Google+ social media account?
- in theory no, but in practice they strongly funnel you that way
Q: is Jitsi on a desktop computer compatible with G+?
- at the text chat level, yes. video and voice via Google Talk, reported to work with Voice, not sure about video
- It is compatible with GoogleTalk instant messaging (and most other instant messaging including the XMPP RFC like Facebook; Skype is the notable one that it doesn't work with), not sure about GoogleTalk's voice calls, I'd give that a 80% chance of working.
- An interesting side-project Jisti offers its own work-alike free multi-person video chat for XMPP(Jabber compatibles) which is not limited to 10 people: VideoBridge. Given enough bandwidth and a modern desktop CPU acting as the bridge server, you can have someone in a G+ hangout share their desktop to pass through the other members of the video hangout.
- Also there's a functional alpha-release Android client.
- For Jitsi <-> Jitsi video you may need a valid XMPP account. Many (federated in a single grid) providers for that, so it doesn't matter much who. Jitsi run a simple registry at http://jit.si
- All quite interesting possibilities and the development is quite active -- but is it there yet? I'm not sure, probably a not far away "almost"; if so, perhaps we could run the bridge on Adhoc for a few days (can run as an unprivileged user, needs Java 6, chat "host" user needs shell access to set magic pass word and get the assigned port numbers to plug into their client), or perhaps host it at fsv.cvut.cz? (i.e. locate it at the shortest hop to max number of participants using it) --HB