The first QGIS bughunting event on December, 5 2009 was held at the Institut of Ecology and Evolution (Russian Academy of Sciences), Biodiversity Conservation Lab. The institute provided the venue, audiovisual equipment, kitchen facilities and internet access. We are extremely grateful for helping us hold our event.

We would also like to thank GIS-Lab.info for organizing and bringing together people from very different backgrounds, fields of expertise and ideologies.

The event was rather small, so no financial arrangements were made and funding (except personal for beers ;) was not sought after.

Participants

Bughunt was organized as 2 simultaneous events, with 14 people present in person in Moscow and around 10 on IRC channel. We were searching for bugs for about 3 hours and then had about 3 hours more of relaxed discussion of everything GIS related with lots of talks about QGIS and world-domination plans.

Summary

The event was planned to fill the gap of extensive and well organized user testing of QGIS with careful registration of all inconsistencies these being bugs or suggestions for improving the interface. The approach was the following:

• Bring your own data  (though our testing dataset - Geosample) was also provided
• Level of QGIS experince doesn't matter
• Choose area of work: printing, geoprocessing, OSM, labeling etc.

Though we planned to split the area of work by topics, this didn't work out in the end, just because people were overwhelmingly new to QGIS. The work was coordinated in GIS-Lab.info forum acting as a buffer and refined results were then transfered to main QGIS trac by more experienced users.

Results

Here are some of the results of our work:

• Around 90 reproducible bugs were found during period right before the bughunt and during the event;
• More than a half of them were considered valid and reported to QGIS trac.

Experiences

Here are some of our experiences and lessons learned from this event in no particular order.

• It works great for a group of enthusiasts to start working on bugs before the bughunt to get feeling of how it goes.
• Be prepared to work a lot after the bughunt, to sort bugs, remove duplicates, translate, put in QGIS trac.
• 3 hours are not a lot, 6 would be much better. Most newcomers started to get a flavor of rather specific mode of work closer to the end of the meeting.
• For some people, it would work better if more specific guidelines are provided, such as tutorial that shows how to do specific tasks so they can look for bugs along the excercise;
• Though we worked on trunk, QGIS was very stable with people expressing concerns that everything "just works" and it is hard to break it;
• Lots of concern was expressed regarding performance of QGIS with big vector datasets, this need to be addressed in the future.

Further links

• Bughunt summary on GIS-Lab.info blog with more photos (in Russian)