As most of you will already be aware of, it has again been an interesting and productive Google Summer of Code year for our community. As usual the GSoC year is concluded by a summit for the mentors organized by Google and we thought we’d share some bits and pieces as well as talk about the experiences of this year in general.
As we do every year, we sent out two of our mentors to represent our community at the summit. Last year it was Noam and Mattan, this year Assaf and Mattan (David was our first pick as he never attended so far, but it didn’t work out in his schedule).
The event is a two day conference held in Google’s offices in Mountain View, California. Hospitality was wonderful, nice hotel rooms, great food and lots of friendly people. The breakout sessions during the summit covered a variety of topics, all focusing on the summer of code and open source in general.
When looking back to this GSoC year, one sad point was when Assaf had to fail a student. This was a very hard decision for him and the whole team, as it is always hard to be the “bad guy”. The student possibly planed with the money and might be in trouble without it. Additionally, this whole endeavor has to be counted as a failure which reflects on him and his CV. But unfortunately, the student struggled from the beginning (mainly in the communication and community-involvement department) and it was a hard choice to even give him a passing grade in the mid-term.
The GSoC summit now offered a good chance to discuss the underlying question “When should we fail a student?” with other project mentors and more importantly Google people. Google summer of code program manager (Carol Smith) has given a definite answer for this dilemma: “Fail soon, fail often”. She stated that failing a student will never reflect badly on the project in Google’s eyes and if we are not sure what we should do – the default choice would be to fail over passing. The general opinion was that most (if not all) of the students that you are not sure about in the mid-term – will fail in the end – so they should fail the mid-team, harsh but logical. In our case this year – this was also true. It is never easy to fail a student, but we are too nice, we perhaps need to improve here 🙂 .
But we also want to highlight the bright sides of this year’s GSoC for Ogre3D: We had three other, very successful projects that contributed a lot and we again want to thank all the students for their efforts! We will also be releasing a short series of posts, each of which will quickly outline the results of the four projects and show where to get additional information. So check them out to get some insight into what transpired this year (in case you didn’t follow this area).
One other topic that was and is very important for us is “How to get more great students?”. The solution seems to be: “Do some groundwork”. Other projects suggested writing to previous year’s successful students and asking “How can we get more great students like you?”. Asking them to contact their friends, prof
essors, student association and such and ask them to send students our way. We are going to do that this year and even more than that, the core Ogre3D team will try to look up new contacts by contacting people that published paper in our field on computer graphic and asking them to be our students this year or recommend us to their co-students.
One of the GSoC summit sessions also focused on tools for code hardening. In that session we learned about a great tool that is given for free to open source project called “Coverity“. Coverity is a static code analysis tool for C, C++ and other programming languages, used to find defects and security vulnerabilities in source code. It was highly recommended by other projects, so we asked Coverity for a free license and got it. The next stage is for someone from the community to pit in the effort and test our code base using Coverity and perhaps even fix the issues found. Contact us if you want to volunteer and assist us in further improving our code base. Any help is welcome! And no need to do it all on your own. Also small steps can help a great deal…
One of the nicest things about the conference was seeing how well recognized and valued Ogre3D is! So many people are using it or plan to check it out that it really made us proud and makes all our hard efforts worthwhile!
We would also like to take the opportunity to thank Google for their hospitality and for the support they give us every year by organizing this great open-source project that “Google Summer of Code” is!
So all in all, this year was already really good, but next year, we will (try to be) be even better. We have just begun! This year brings lots of new opportunities and innovations to Ogre3D, so lay back, watch and be amazed…or even better jump on board and help in whatever capacity you can!