gvSIG CodeSprint
Tuesday, November 30th, a day before the 6th edition of the International gvSIG conference, the CodeSprint event was held, where gvSIG developers gathered to discuss and solve some coding problems. This CodeSprint took place after the positive experience of the last gvSIG CodeSprint during the FOSS4G2010 event.
This event was sponsored by the Instituto ai2 of the Polytechnic University and was held in the Innova room of the Polytechnical City of Innovation of this University.
This document is a summary about how it was organized. The results are commented here:
When? Where?
Date: November 30, 2010
Time: 09:00 a.m. until 8:00 p.m.
¡It began at 10 a.m. and the last ones left at 7:30 p.m! ;-P
Arrival by Bus: Tram: Line 4 and 6, Serrería station; Bus: Line 1 and 41
Attending
26 developers from different companies and organizations:
Name | Company | Interests | |
1 | Jorge Piera | gvSIG | Coordinator of Code-Sprint |
2 | Nacho Varela | Xunta de Galicia | Working in Communities, small changes, bugs’ corrections. |
3 | Jorge Sanz | gvSIG | Not sure if attending, but if attends will help in the same areas as Nacho, providing ideas, etc. |
4 | Víctor Olaya | SEXTANTE | Improving Access to the raster layers for analysis in order to allow large layers to be processed in Sextant. I will try to take advantage of the code sprint to work with the part of the gvSIG team that handles these aspects, providing ideas and possible solutions. |
5 | Rafael Gaitán | gvSIG3D (ai2) | Improvements in gvSIG for better 3D integration. Correction of bugs. |
6 | Jordi Torres | gvSIG3D (ai2) | Improvements in gvSIG for better 3D integration. Correction of bugs. |
7 | César Ordiñana | gvSIG (DiSiD) |
Architecture and Design issues, gvSIG 2.0 and helping in other areas (bugs, improvements, etc.) |
8 | Francisco Puga | Cartolab | PostGIS |
9 | Andrés Maneiro | Cartolab | PostGIS |
10 | Javier Estévez | Cartolab | PostGIS |
11 | Pablo Sanxiao | iCarto | PostGIS |
12 | César Martínez | Universidad Autónoma Barcelona | OpenStreetMap Tile View. Create a complete editor analysis of OSM. Fix the SDE driver |
13 | José Carlos Martínez | Universidad Politécnica de Valencia | Possible improvements and bug correction in the Post GIS driver. Creation of a new driver for Jasp especially for H2. |
14 | José Javier Badía Castelló | Software Colaborativo | Correction of Bugs |
15 | José Vicente Higón | Software Colaborativo | WMS and WCS Services |
16 | Fran Peñarrubia | SCOLAB | Bugs PostgreSQL + Legends |
17 | Joaquin Jose del Cerro | gvSIG | Architecture and Design issues, gvSIG 2.0 and helping in other areas (bugs, improvements, etc.) |
18 | Nacho Brodin | gvSIG (Prodevelop) | Resolution of some WMS bug and raster. Collaboration with SEXTANTE in large layer and with AI2. |
19 | Jose Manuel Vivó | DiSiD | Help in various areas (Bugs, improvements, etc.) |
20 | Jorge Bellido | IKT | Help in various areas. Special interest in graphic edition. |
21 | Fernando S. Peregrino | Universidad de Alicante | Development of new raster algorythms on fussy entities. |
22 | María Arias de Reyna | Emergya | Community involvement |
23 | Dani González Moratona | Geodata Sistemas | Help in correcting errors and meeting people from the gvSIG community. |
24 | Romel Vázquez Rodríguez | Laboratorio de Computación Gráfica, UCLV, Cuba | Learn about SEXTANTE and raster analysis with gvSIG. |
25 | Francisco Díaz | gvSIG | Architecture and Design issues, gvSIG 2.0 and helping in other areas (bugs, improvements, etc.). |
26 | Daniel López Sánchez | Regsete | Postgis |
Features/Bugs addressed in CodeSprint
A wiki was created where each person could add the bugs and the features that he/she wanted to correct during CodeSprint. The list was edited the same day of Code Sprint, so that as the bug was corrected, it was updated. The definitive list, grouped by feature is as follows:
Regarding the installer:
- Feature: backport of the installer of gvSIG 2.0 plugins in the 1.11 line [jjdelcerro]
- Delete the copy of the libraries and archives of the license of the jai in the installer, as this dependence has been deleted.[jtorres]FIXED.
- Correctly install the direct access icons to gvSIG Linus.[jtorres]ALMOST FIXED, the only function left to finish is to write the script that detects the desktop[jjdelcerro].
- Change the lbz version that is provided for Linux, it is obsolete. [jtorres]FIXED. Update the library to the 1.2.3.3 version.
Regarding the ToC:
- Small Change: Delete the extension (.shp, .tif, .... ) when a layer is loaded. [nachouve]FIXED.
- Revise and integrate the existing patches in the OSR[amaneiro]:
- http://forge.osor.eu/tracker/?func=detail&aid=14605&group_id=89&atid=732
Regarding the Shapefiles:
- Implement the writing of M values (for 4D shapefiles) [fla83tn]
- Manage the editing of 4D shapefiles correctly (add vertex should initialize Z to 0 and M as the weighted average between previous and next vertex, explode should not remove the feature from the layer, insert multiline/line/arc etc with 4D if the layer allow it,etc) [fla83tn]
Regarding the Legend:
- Copy the styles quickly between the layers (this requires speaking about how this can be done internally, to be able to copy styles between different geometries, etc.. [nachouve] (This requires rethinking the architecture.)
- Stop re-coloring of the vectorial layers when the cancel button in the legend is pushed, and review all of the items that have been re-colored to see if anything has been colored unnecessarily. [jtorres] FIXED. The error was in: extDockingSkin.
- Stop creating a new query in the data base when changing the legend. (I think this was occurring). [nachouve]
Regarding WMS:
- SmallChange: automatically select the view projection at WMS projection selector (and probably also a preferred image format) [xurxosanz]
- Use of the scale hint parameter [jvhigon] [cmartinez] FIXED. Using MinScaleDenominator and MaxScaleDenominator
- http://forge.osor.eu/tracker/index.php?func=detail&aid=14856&group_id=89&atid=732
- GetCapabilities requests in a thread. Extensible to WCS [nbrodin]
- Every WMS download by libRemoteServices launches two threads but only one dies. Extensible to WCS. [nbrodin] FIXED by [rgaitan].
- Caching in the zones that are continually being requested?.
- At least what we have on the screen should be cached, as currently when a layer is activated or deactivated the WMS layers that are active are downloaded again (without changing anything in these layers).
- Revise and integrate the existing patches in the OSOR [amaneiro]:
- http://forge.osor.eu/tracker/?func=detail&aid=14736&group_id=89&atid=732
Regarding the WCS:
- Bug: the latest version of the geoserver isn´t working. [oscarfonts] FIXED by [jvhigon] ticket #14861
- Request of images with fixed resolution, that is, allow for omitting of WIDTH and HEIGHT parameters [fla83tn]
Regarding the WFS:
- Bug: do not replace underscores ('_') with spaces (' ') in attributes name of the layer [fla83tn]
Regarding the Catálogue:
- Feature: Development of OpenSearch client. [oscarfonts]
Regarding the gvSIG structure:
- There are many errors with large quantities of raster data. An example is to generate in sextant layers in large sizes (20000x20000 pixels, example). Error in the "java heap space" is very common. The developers of sextant say that it is a problem with gvSIG not with sextante. [Carlos Almeida]
Regarding the Topology:
- It is not possible to save the results of the topology in the shapefile. Only in the DXF, and GML [Carlos Almeida]
- The thiessem polygons create a lot of bugs that do not work with the majority of point layers that I have to test. [Carlos Almeida]
Regarding Oracle Spatial:
- gvSIG is restricted in some Oracle connections (bank configurations). When the geometries are edited, an error is saved (the units are changed, for example for 6000000, for 6000000000) [Carlos Almeida] >>this error is generated by the convections of the units defined in the operating system of which Oracle is the server. But, other software is not restrictive, for adding and saving new geometries in Oracle, independent of the configuration of the Unit. [Carlos Almeida]
Regarding generating PDFs
- gvSIG uses a lot of Ram memory (when there are defined labels by user and satellite images, or it has a Bug when users define the ZOOM by approximation. [Carlos Almeida]
- Exporting SVG to be able to finish laying out the maps in a design program.
Advanced Labeling
- gvSIG uses a lot of RAM memory to show all of the labeled elements in the screen. The functionality of joining the labels in polygons isn´t working. [Carlos Almeida]
- Allows for converting entries from the labels in the“advanced labels” or editing them individually in some form.
Regarding Postgis:
- Testing/Bugs: testing the postgis driver and correcting the possible bugs. [josecarlosmartínez]
- Re-connecting the connection after an error.
- Intending to fix the connection so that it doesn´t close so frequently. [nachouve]
- An example: When the user doesn´t have permission to edit the layer, an error is produced and only when closing gvSIG and loading the layer again, can the connection be repaired. [Carlos Almeida]
- When producing an exception in the SQL sentence (for example a violation of the restriction) the driver was not recovered from the exception cutting the connection. [josecarlosmartínez]
- Corrected and pending to be uploaded by OSOR; avoid closing the connection (and gvSIG getting stuck); when a user makes an error when writing a restriction during the process. [amaneiro]
- Lack of synchronization between the attribute table and the display when the geometry is updated from an external source [josecarlosmartínez]
- A provisonal solution is for the option that recharges the layer to always be active, when the origin is FLayerVectorialDB. I checked that when recharging the layer it is synchronized correctly. [fpuga]. Recharge always active. [#14857].
- Currently layers cannot be created or updated in 3D layers. Using the functions ST_FORCE_3D. [josecarlosmartínez].
- Writing geometries is carried out using WKT that can cause small displacements as well as slowness. The proposal is to use writers using EWKB the same as used in Reading. [josecarlosmartínez]
- A connection manager, mainly in tables displays the tables available. etc [Carlos Almeida]
- Ability to see the status of the connections [Read, Read-Write, Close, Reconnect], to indicate the permissions that the user has. [nachouve].
- The dialogue in alphanumeric tables of data bases should be the same and or similar to the geometric layers and should use the existing connections. Moreover, it should only list the table using alphanumeric numbers. [fpuga]
- FIXED. Allow for the name of the geometric layer to be in capital letters. [fpuga]. [144485]
- Allow connections to be made for a determined scheme, so that only the tables of this scheme are listed, or the tables can be filtered that appear in relation to the scheme in the layer selection dialogue. [fpuga]
- Automatic detection of the GID layer. It will begin to detect if there is any auto-incremental field and pre-select this is the combo box. [fpuga]
- When the type of data is numeric and has a value less than 1, for example 0.5 is introduced in the data base a 50000 is saved. [fpuga]
- [FIXED]Use of the cursor in postgis. Message in the list. [fpuga]
- Revision of errors reported in the OSOR by the users:
- Allow for the for parameter to be edited for the connection: http://forge.osor.eu/tracker/index.php?func=detail&aid=14739&group_id=89&atid=732
- When a user tries to load a layer a PostGIS without a previous connection, the following error is given gives an error: : http://forge.osor.eu/tracker/index.php?func=detail&aid=14599&group_id=89&atid=732
- Revision and integration of existing patches in the OSOR [amaneiro]:
- Patches prepared and pending to be uploaded to the OSOR: http://forge.osor.eu/tracker/index.php?func=detail&aid=14547&group_id=89&atid=732
SEXTANTE
- All that is related with the SEXTANTE, to improve its integration with gvSIG. I would like to listen to your ideas and the problems that the gvSIG team might have in this regard to know how to help you. [Victor Olaya]
Others:
- When you copy and paste groups of layers between views, the group are copied with the layer but the layers are copied again in the root. [jbellido]
- The “localization map” (below and to the left) should be a normal view (that is to say that this in the Project management can load layers, and comfortably change symbols, etc.)
- To set the localization map it should be selected in one of the created views.. [nachouve]This has been addressed and an interesting alternative has been searched for that can handle this in the future: copying layers from one view in the locater.
- The localization map should appear only when it is set and not by defect.
- Integration of Jaspa in gvSIG, specially in JASPA with H2 as the driver Jaspa/PostgreSQL driver is very similar to the PostGIS [josecarlosmartínez].
- Add a paginator to tables (e.g. do not fetch more than 500 records at a time) [fla83tn]
- The selection with the hyperlink tool should have a buffer when the points or a line are selected as sometimes is too difficult to get to the center of the point or the line.
- Idea: port the Load them all extension of QGis to gvSIG [xurxosanz]
- Idea: add a startup variable to select where to place the gvSIG configuration folder (from this post and older messages on mailing list) [xurxosanz]
Bugs and some features detected in the BugSprint:
- Comboscale empty, https://forge.osor.eu/tracker/?func=detail&aid=14793&group_id=89&atid=732[jbellido]
- Conversion from string to number on the field calculator,
- http://forge.osor.eu/tracker/?func=detail&atid=732&aid=13394&group_id=89
- Persistence error with WFS layer filtered by area, https://forge.osor.eu/tracker/?func=detail&aid=14789&group_id=89&atid=732
- No English translation for "capas", https://forge.osor.eu/tracker/index.php?func=detail&aid=14794&group_id=89&atid=732
- WFS error when re-projecting layer, https://forge.osor.eu/tracker/?func=detail&aid=14792&group_id=89&atid=732
- User defined labelling. gvSIG crashes, https://forge.osor.eu/tracker/?func=detail&aid=14795&group_id=89&atid=732
- Table name of joined tables do not update on the project manager, https://forge.osor.eu/tracker/?func=detail&group_id=89&aid=14798&atid=732
- Saving pies and bars symbology on .sld format, https://forge.osor.eu/tracker/?func=detail&aid=14802&group_id=89&atid=732
- Different name to the locator, https://forge.osor.eu/tracker/?func=detail&aid=14797&group_id=89&atid=790
- Language of gvSIG after installing, https://forge.osor.eu/tracker/?func=detail&aid=14791&group_id=89&atid=790
- Error in Symbology by expressions, https://forge.osor.eu/tracker/?func=detail&atid=732&aid=13381&group_id=89
- Wizard for labeling creation, https://forge.osor.eu/tracker/?func=detail&aid=14788&group_id=89&atid=790
Organization
Depending on the different interests a list of groups were created with a specific topic. In each group, a person was assigned to be responsible for the group’s coordination.
The list of groups was as follows:
Group | Leader | Participants |
Postgis | Fran Peñarrubia | Francisco Puga, Javier Estévez, Pablo Sanxiao, Andrés Maneiro, José Carlos Martínez, Daniel López Sánchez, María Arias de Reyna |
Raster and Sextante | Nacho Brodin | Fernando S. Peregrino, Romel Vázquez, V. Olaya |
Usability and user interface | Cèsar Ordiñana | Nacho Varela, Jose Manuel Vivó |
WMS and WCS services | José Vicente Higón | César Martínez, José Badía |
Bugs in general and what comes out of BugSprint | Jorge Piera | Dani González, Jorge Bellido |
Usability and user interface group
Author: César Ordiñana
The purpose is to discuss, design, and to develop, to test and experiment to improve issues related with the usability of gvSIG and its interface. Some things that occur to me for example are:
- When a task is carried out that requires a certain process, the gvSIG user usually does not receive any feedback. It would be good to at least show an hourglass cursor. An action could be attempted such as is presented in the article: Automate the hourglass cursor.
- Update the Swing theme. Currently, the Plastic is used, although maybe there is a something with a newer interface. At least it should be personalized, in linux, for example, the windows have a border that is so fine that it takes a lot of time to change the size.
- The docking Skin included in the installation of the 3D: to see its level of integration, its advantages and disadvantages compared with skin by defect, etc.
- The Project window, access it and propose improvements: perhaps use a tree with the documents by type, include more information (what can be seen in the properties of a view), etc.
- Revision and integration of the OSOR parches. [amaneiro]:
- http://forge.osor.eu/tracker/index.php?func=detail&aid=14342&group_id=89&atid=790
Summary of the conference
Start up
Author: Jorge Piera
The conference began late. It took awhile to find the room and the instructions that were sent were not that exact. Thankfully, Joaquín del Cerro came down to look for everyone, or we would have continued to go in circles.
At 9:45 a.m. we were all there except for those that arrived a little late. The first thing we did was present ourselves. I think we were brief but it served to get to know each other better.
The “expert” Jorge Sanz made a presentation as responsible for the technical collaboration of the project. He explained how to collaborate and he pressed us to read the documentation in the moment (this was nice of him.)
At 10 a.m. we began to work at an optimal level. Before the presentation we were already seated in groups, therefore it wasn´t difficult to get started.
"Postgis" group
Author: Fran Peñarrubia
I don´t know about everybody else, but the truth is, I really enjoyed code sprint. In our group we began to review the bugs that had been pointed out in this page, as they are the ones that bother us the most, and that we think can be fixed in only one day.
Once they were reviewed and we could see what could be handled we divided the tasks, and I think we were very productive. Jose Carlos Martínez fought all day with the 3D support in the geometries of the PostGIS (and with the M) and did a fantastic job, and finally almost finished the parch. I have his address in my mail and hope that when I have time we can finish this and load the 1.11 (we will have to hurry).
Daniel López began with a bug for small cuts in the PostGIS driver which is very bothersome, difficult to catch on to, and provokes breakage in the PostGIS. We believe that the problems have been solved, at least partly, we have to try to check it more thoroughly, to see if with this parch, the ones that the people from CartoLab (Javier Estévez) have tried and the one that I have been working on, we can leave the Postgis layer that is much more stable than it was. Andrés Maneiro has been working on finding cuts in the PostGIS connections that leave it in a invalid state and gvSIG has to be shut down to re-connect: 2 bugs were corrected that allow for reconnection to gvSIG without having to close it. Afterwards, he helped Daniel.
Pablo Sanxiao has been fixing some bugs that were being pointed out in OSOR. Francisco Puga has fixed the PostGIS driver so that it stops using the Geometry field in capitals, later I think that they improved the interface by adding layers so that they only show the layers that really have geometries and finally has been thinking about how to filter the layers by schema.
María Arias arrived later and joined the CartoLab group.
For my part, besides speaking about topics that I would not usually speak about with “normal people” by announcing The Big Bang Theory and listening to Siniestro Total (that didn´t work because of the speakers were too noisy) I was able to correct 2 bugs. The first one is something that usually happens to me. If you make a mistake when entering in the key when loading a Project, or in this moment you cannot connect to the database, errors begin to occur and the only thing that you can do is to close gvSIG. With the parch the Project is loaded with the layers in a disconnected mode (red cross) and the user can recover the connection and re-load the layers. The second bug is the one that is shown above. The PostGIS driver does not read the numeric fields well and an error is caused by the binary reading that we get from the recordset. When solving this issue I was lucky and I found codes on internet that read these types of fields. I added it to the driver and I think it is solved.
For our part, we still have to check the parches that have been generated, and if everything is working well, put them in the repository so that they appear in the next version.
I would like to say thank you to all of the participants for their work and the motivation they showed. We also should thank the AI2 personal for finding us a relaxed and fully equipped work place for the Code Sprint.
To conclude, I just want to say that this was my first code sprint and I enjoyed it, especially because of the people who participated, I hope that next year we can do it again, and from the gvSIG association we can work together to make it even better than this year.
"Ráster y Sextante" Group
Author: Nacho Brodin
There were four of us in the raster group we were able to solve some things. Victor Olaya brought up a pending issue regarding the big layers because he wanted to cache the outputs of the algorithms to avoid saturating the memory. It seems that after talking with him he was clear what was involved and will develop it outside Code Sprint, therefore the rest of the time was dedicated to revising some bugs that seemed to have been there for awhile and are bothersome in the sextante.
On the one hand, there is a problem with loading the single band images that are bytes, now that an enhancement filter is not being applied and they cannot be seen well. A solution is adopted but it needs to be checked on.
From the Cartolab group a parch was given to me to solve a problem with one of the scripts from before. It seems like the extRasterTools-SE it is not being distributed well. After explaining it we can discuss it.
On the other hand, I included in the WMS extension so that when a thread is launched when a petition is getCapabilities at the beginning of the session with the server. It has been solved and uploaded but a problem still persists and the downloading the files is never canceled and although you keep pressing cancel the xml is downloaded.
We talked with the 3D personnel about a problem with the threads when downloading files in remote services. It seems that it keeps running eternally and a solution needs to be applied so that it stops. Finally, I gave some code to Rafa but he has a better solution and keeps applying it and trying this solution.
Usability and User Interface Grupo
Author: César Ordiñana
After the presentation a small group of us, Nacho Varela, José Manuel Vivó, Francisco Díaz and a server, got together. After preparing the workspace of the trunk for awhile, Nacho Varela commented on some ideas that affected using gvSIG:
- To hide the extension of the files that is shown by defect in the TOC layers. In some cases the information is redundant, when there are a lot of layers loaded there is too much information in the TOC and it also affects the functioning of the map calculator of the Sextante. We decided that it could be an option for the user’s preference and Nacho began to develop it. It can be found at: feature request 14854.
- The type of layer it distinguished in the TOC using an icon, similar to the previous one, for example, this causes visual confusion with the information in the legend in the layers. After discussing this for awhile we were not able to make any solid conclusion, because we had to look for an alternative form to distinguish between the layers that would be more compact. It can be found at: klaus.schaefer4@giz.de
- Improvement in the locater
- In the set up, change the way of opening the window from model to tool.
- Substitute the layers for a TOC.
- Add an “import view” button that imports the structure of layers from an existing view in the current locater.
- José Manuel Vivó began with these changes and when he has more time he will finish them and sign up the corresponding feature request.
- A very useful function, when there are many layers with the same or similar symbols, would be to copy and paste between the same ones. Between the layers with the same structure of data and types of geometries would be simple, and we would have to see if we are able to copy.
On the other hand the following topics were revised:
- The feature request 14342: to activate the pan when pushing the center button of the mouse. We were performing tests in this regard but the parch forced the pan when pushing the center button within the MapControl code. As an alternative, José Manuel Vivó, preformed a quick test modifying the ToolListener that the pan makes to indicate which button should be pushed, registering it in all of the behaviors. Nevertheless, the majority of listeners perform its action with any button, which activated various options at the same time. As a result, it was decided that an update would be required in the API of the Behavior/Listener, in such a way that the specific buttons could be assigned to the Behavior. It can be found at: feature request 14869.
- The look and feel Nimbus. Available by defect from the 1.6 version update 10 of Java, it presents a new interface that is more modern and that can be rendered in any solution. gvSIG is started up with the latest versions of Java 1.6. We can activate it by the users preferences (General > Appearance). We activated and performed some tests; however, some panels with fixed sizes were moved out of position. We have to look in to whether gvSIG 2.0 is activated by defect, although for this purpose these panels will have to be checked and some basic recommendations given so that the panels can be seen in any look and feel.
- The bars are able to be re-positioned in swing, as they could be moved to the left, right, below or above the window. It is easy to activate this, but the option of being able to re-position the groups of buttons inside the button bar.
Nacho Varela uploaded the tracker of the Feature request the tickets:
[#14854] File extensions on ToC (with parche)
[#14855] Remove vector/raster ToC icons
"WMS and WCS service group"
Author: Jose Vicente Higón
Without any doubt, a magnificent initiative. I hope that next year we can do it again.
Regarding the work done by César Martínez, Jose Badía and I, we fixed some bug and added some interesting functions. Specifically, César has improved the treatment of the parameter that indicates the maximum and minimum scale of a WMS layer, also I added a position: “Zoom to the visible scale. The parches can be seen in: ticket 14856.
Jose Badía has been investigating the WMS services and solving a problem with the visualization of metadatos of the WMS service ticket 14849
For my part, I fixed the WCS parser so that it could be connected to the WCS-1.0.0 service that serves Geoserver, ticket 14861
"Bugs en general group and what will come out of Bug Sprint"
Author: Jorge Piera
The group composed of Jorge Bellido and a server, Daniel González joined afterwards. The first thing we did was present ourselves and speak about the development from our companies. You should see the amount of interesting extensions that people perform on gvSIG:
Afterwards, we got to work. First, we went over the list of bugs BugSprint and we tried to understand the bug that the others had detected the day before. Those testers are good.
Jorge Bellido explained to us about a bug that they show to their clients to demonstrate to them that gvSIG also has errors: when copying a group of layers between two views, the group of layers is copied as well as the individual layers that form the group. We got to work and the first bug was fixed! Ticket in OSOR and patch is attached.
Daniel González joined the group with a lot of strength and it didn´t take him very long to explain to us his first bug: when the zoom is selected from a single point in a layer of points the viewing scale is set to zero. Does this make sense? We discussed this for awhile but we didn´t arrive to any conclusion. We have to go over in more depth.
Afterwards we began with the list of the Bug Sprint. We took the first of the list that gave us an error when placing an invalid scale in the combination of scales that made this tool not usable. During the middle of the discussion Jorge Piera had had to leave to pick up María Arias that had just arrived and when he returned, que Jorge Bellido and Joaquín del Cerro had fixed the bug! Another ticket + parch en OSOR!
Daniel González and Jorge Piera spent the whole afternoon looking for a bug to fix. After a few attempts we addressed the bug fields of the field calculations of the " toNumber" method, that was produced when it attempts to convert a string to a whole.
We weren´t sure of the solution, so we asked the experts; Mario Carrera and Fran Peñarrubia were passing by and we kidnapped them to ask for their opinion on the value that would need to be returned in case the number to be converted was a string. The two of them agreed that the best solution would be to not do anything.
We spent awhile looking for a way to solve the error and after reaching the error limit, we consulted with Joaquín del Cerro as it was an error in the Jython. This error that seems trivial was more complex than we thought and we had to address the mission. Francisco Díaz, prepare to fix it soon.
Conclusions
Author: Nacho Varela
gvSIG is a FLOSS Project (Free Libre Open Source Software) that is opening up to the community more and more each day. The code-sprint conference held on November 30, 2010 in Valencia is the living proof that there is a great interest for this type of SIG on all levels. On a technical level, the knowledge and the ability to improve the program are clearly expanding, as many people work with this product that is distributed inside and outside of Valencia.
The code-sprint has been a meeting of developers of gvSIG that was composed of 26 participants that for one day were able share a common work space, experiences, ideas, advice, etc. But, above all the participants shared an objective: to build a gvSIG together, that is getting better every day.
Errors were corrected, by implementing new features, and issues were discussed and explained in a pleasant and participative environment. Nevertheless, the most important result of the code-sprint weren´t the lines of code source written that day; the most important were the relationships and communication mechanisms that were established between official developers of the core of gvSIG and external developers.
The network of participation in the project is growing, coordinated, sustained, solid and integrative. gvSIG is growing strong as it is supported by a busy, professional community that is willing to keep growing. The biggest challenge of the gvSIG Association now is to be able to collect the contributions of the users and to develop all of the current and future potential in the community. It is for this reason that the CDT meeting (Committee of Technical Direction of gvSIG) the 3rd of December with the participation of the new people who have been part of the code-sprint is a clear indicator of the intention of creating effective collaboration mechanism based on real contributions and according to the FLOSS dynamics.
Code-sprint has been a great icing for the 6th gvSIG conference: a collective, profitable and very motivating success. Thank you to all of the participants and organizers of the code-sprint and for your efforts so that everything went well. We look forward to seeing you in the next gvSIG Code-Sprint!