The next trip report is from Gabor Pali:
This year the FreeBSD Foundation helped me with travelling to EuroBSDcon organized in Warsaw, Poland, in the beautiful building of the Warsaw University of Technology. It is an excellent opportunity to meet fellow developers in person and discuss some of the on-going issues or just to hang out somewhere in a nice European city while drinking some fine local beer.
Similar to the previous years, I volunteered to organize and lead the ssociated FreeBSD Developer Summit (overlapping with the tutorials) and chair the Developer Summit Track at the main conference. This was my third time as a "remote organizer", and thanks to Pawel Jakub Dawidek and the student volunteers at the Warsaw Univesity of Technology, everything went smooth, resulting in a productive summit. I tried to improve the format learned from the BSDCan developer summit organizers, and offer the visiting developers a way to exploit their time spent together well.
Besides taking care of the usual organization tasks, I also had some time to actively participate in some of the sessions during the summit. For example, I shared my experiences earned in teaching university courses from a practical aspect at the "Teaching FreeBSD as a University Course" group, lead by Benedict Reuschling. I think one could manage to find complete but not-that-complicated real-life examples in the source code that could serve as a demonstration tool to introduce problems and their potential solutions. This motivates students as they can see what causes headaches to the FreeBSD developers and how they try to resolve issues. As a special guest for this session, Pawel brought Andrzej Tobola, a lecturer from the University, who has been using FreeBSD for a long time and he uses it for teaching. He provided us with interesting feedback.
Again, similar to previous years, I managed to put a schedule together for the Developer Summit Track as part of EuroBSDcon 2012. It featured brief summaries of the working group leaders, so the conference visitors could see what is planned in the FreeBSD Project these days. But we also had many interesting talks on many of the work-in-progress projects: the XML transition in the documentation source tree, the plans for the new ports building infrastructure, and the presentations of some of our Google Summer of Code students from this year. I joined as well and gave a presentation on the project I did at the Cambridge University Computer Laboratory during the summer. This featured the FreeBSD port of the Mirage network stack, which is written in OCaml.
From the conference, Dan Langille's inspiring talk comes to my mind: he decided to reveal his secrets on how he organizes the premier BSD conference in Canada, BSDCan. I am glad that he did so and finally made it to Europe to see how his competitors are doing. I found his presentation very useful as it perfectly supported the experiences I have earned so far.
I really enjoyed the developer summit and the associated dinners, as well as the conference and the city itself. I think the fellow organizers did their best to bring the spirit of the European BSD Conferences to Warsaw -- and do not forget that this was also made possible with the help of its gold sponsor, the FreeBSD Foundation. Every donation to the Foundation helps us to maintain the tradition of such a great conference.
The FreeBSD Foundation is a 501(c)(3) non-profit organization dedicated to supporting the FreeBSD Project. The Foundation gratefully accepts donations from individuals and businesses, using them to fund projects which further the development of the FreeBSD operating system.
Friday, November 30, 2012
Wednesday, November 28, 2012
EuroBSDCon Trip Report: Gabor Kövesdán
The next trip report is from Gabor
Kövesdán:
It was my very first time at a FreeBSD Developer Summit and a EuroBSDCon and I was very excited about it. Recently, I have been actively working on two important projects, which I wanted to discuss and report about. One of them is the XML migration and clean up of the documentation and the second one is the regular expression code.
I arrived on October 17 in Warsaw and I was about to go to the dinner but it was really late when I arrived at the accommodation and got prepared for the dinner. But I met some of the other developers after dinner and we had some interesting conversations, but no one stayed too long at the hacking lounge because next day’s developer summit was scheduled to start at 8 am.
On the next day I arrived on time at the developer summit. There was a bit of delay since the majority of the people had problems finding the place on time. After the short opening talk, I participated in the Toolchain working group. I am not directly involved in the toolchain and compiler work in the source tree but I have been working on important command-line utilities like grep and sort, which are usually used in build processes and I work on some massive changes on the documentation infrastructure, which also has a build system, so I thought it was an interesting session for me. I learned about the ongoing work and bmake’s more advanced features. Someone raised performance questions and utilities so we also talked a bit about grep and similar things. I didn’t stay at the afternoon’s session because I still wanted to polish some stuff I was going to present the next day.
The second day of the developer summit was very productive. We started with the documentation session and we covered lots of important topics. The two most important I want to highlight was the XML and toolchain migration and the printed edition of the Handbook. I summarized the rationale of the changes and I presented rendered output made with the old and the suggested new tools and I also showed some other pdf documentations from open source projects. I highlighted in what sense I think we are behind and what we should do to have a modern and pleasant look. In fact, our HTML documentation is visually quite nice but if we compare it to the printable output, the latter is far behind. We could render fancy and modern output with much more advanced features with Apache FOP but it depends on Java, which raised some concerns. I emphasized that Java would only be required for printable formats, which are not built by default, not even on FreeBSD.org, so the impact of this dependency is not that serious. The second most viable solution would require LaTeX, which is also a heavy dependency and we do not even have good LaTeX support in FreeBSD since the TeXLive ports are not yet ready for production use. As for the Handbook work, we talked about different options to reorganize current content and other items that need to be done before publishing, like updating indexterms, glossary, and adding more conceptual figures to help understanding the textual explanations. We also mentioned some possibilities of an ebook format. We also talked a bit about tooling and the possibility of using UTF-8 as a unique global encoding. For this, we invited Ed Schouten, who gave us a bit of a status report and overview of what is ready and what needs to be done to fully support UTF-8 at the console. After the documentation session, I participated in the vendor summit where we heard useful industry experiences from third-party vendors. I then talked a bit to Ulrich Spörlein about some ideas concerning mdocml.
The first day of the conference, I participated in the developer summit track and presented my two projects. After the presentation I got some encouraging words from some participants and now I have a potential tester for this work. I spent the rest of the day at the developer summit track, listening to fellow developers and all of the talks were quite interesting.
The final day I listened to the BHyVe and the ZFS talks and then I was hanging around in the sponsors room thinking about and discussing the Handbook reorganization. I talked about it to other developers and I also interchanged some experiences with a NetBSD doc fellow. I am really glad that I participated at the conference since it was a really good experience both from a technical and a social viewpoint. I am very thankful to the FreeBSD Foundation that made this trip possible!
It was my very first time at a FreeBSD Developer Summit and a EuroBSDCon and I was very excited about it. Recently, I have been actively working on two important projects, which I wanted to discuss and report about. One of them is the XML migration and clean up of the documentation and the second one is the regular expression code.
I arrived on October 17 in Warsaw and I was about to go to the dinner but it was really late when I arrived at the accommodation and got prepared for the dinner. But I met some of the other developers after dinner and we had some interesting conversations, but no one stayed too long at the hacking lounge because next day’s developer summit was scheduled to start at 8 am.
On the next day I arrived on time at the developer summit. There was a bit of delay since the majority of the people had problems finding the place on time. After the short opening talk, I participated in the Toolchain working group. I am not directly involved in the toolchain and compiler work in the source tree but I have been working on important command-line utilities like grep and sort, which are usually used in build processes and I work on some massive changes on the documentation infrastructure, which also has a build system, so I thought it was an interesting session for me. I learned about the ongoing work and bmake’s more advanced features. Someone raised performance questions and utilities so we also talked a bit about grep and similar things. I didn’t stay at the afternoon’s session because I still wanted to polish some stuff I was going to present the next day.
The second day of the developer summit was very productive. We started with the documentation session and we covered lots of important topics. The two most important I want to highlight was the XML and toolchain migration and the printed edition of the Handbook. I summarized the rationale of the changes and I presented rendered output made with the old and the suggested new tools and I also showed some other pdf documentations from open source projects. I highlighted in what sense I think we are behind and what we should do to have a modern and pleasant look. In fact, our HTML documentation is visually quite nice but if we compare it to the printable output, the latter is far behind. We could render fancy and modern output with much more advanced features with Apache FOP but it depends on Java, which raised some concerns. I emphasized that Java would only be required for printable formats, which are not built by default, not even on FreeBSD.org, so the impact of this dependency is not that serious. The second most viable solution would require LaTeX, which is also a heavy dependency and we do not even have good LaTeX support in FreeBSD since the TeXLive ports are not yet ready for production use. As for the Handbook work, we talked about different options to reorganize current content and other items that need to be done before publishing, like updating indexterms, glossary, and adding more conceptual figures to help understanding the textual explanations. We also mentioned some possibilities of an ebook format. We also talked a bit about tooling and the possibility of using UTF-8 as a unique global encoding. For this, we invited Ed Schouten, who gave us a bit of a status report and overview of what is ready and what needs to be done to fully support UTF-8 at the console. After the documentation session, I participated in the vendor summit where we heard useful industry experiences from third-party vendors. I then talked a bit to Ulrich Spörlein about some ideas concerning mdocml.
The first day of the conference, I participated in the developer summit track and presented my two projects. After the presentation I got some encouraging words from some participants and now I have a potential tester for this work. I spent the rest of the day at the developer summit track, listening to fellow developers and all of the talks were quite interesting.
The final day I listened to the BHyVe and the ZFS talks and then I was hanging around in the sponsors room thinking about and discussing the Handbook reorganization. I talked about it to other developers and I also interchanged some experiences with a NetBSD doc fellow. I am really glad that I participated at the conference since it was a really good experience both from a technical and a social viewpoint. I am very thankful to the FreeBSD Foundation that made this trip possible!
Tuesday, November 27, 2012
EuroBSDCon Trip Report: Alexander Pronin
The Foundation sponsored several developers to attend the EuroBSDCon DevSummit and conference in October. Here is the trip report from Alexander Pronin, a GSoC student who worked on parallelization in the ports collection and pkgng utility.
I arrived in Warsaw on Wednesday. After dropping off my luggage at the hotel Gromada Centrum, I stayed for the rest of the day at the hotel, preparing for the coming devsummit. Unfortunately, I was late to join devsummit participants for dinner and discussion of the coming devsummit.
The next day the devsummit started. The first person I met was Pali Gabor Janos, the devsummit organizer. I also became acquainted with Aleksander Dutkowski, another Google Summer of Code student at The FreeBSD Project. During the briefing part of the devsummit, participants introduced themselves and went to the working group rooms. First, I attended the Ports working group. The working group covered the following topics. Beat described the current state of movement to SVN, and explained problems that he faced and the long period of movement to the SVN repository. Finally, utilities that still work with the CVS repository were covered. Then Baptiste Daroussin discussed the current state and further implementation of the pkgng utility, as well as topics related to build cluster, such as the new portbuild2 script and schedule of building of binary packages sets. Unfortunately, the problem of UNIQUENAME global variable is still open. Finally, the new options framework and default sets of options for various FreeBSD builds were discussed. Kris Moore also took part in this discussion in scope of default options for PC-BSD.
During lunch I met Gavin Atkinson. He kindly offered his help in case of any questions or if I need his help with introduction to somebody in the community. I also get acquainted with Ulrich Spoerlein. We discussed the current state of mirroring FreeBSD source code to GitHub, the importance and benefits of his work, and the importance of maintaining GitHub repositories for GSoC students at The FreeBSD Project. Then I met Aleksandr Rybalko who works on porting FreeBSD to embedded platforms in the Ukraine.
After lunch I attended the OS Course working group, where we discussed different techniques of teaching FreeBSD at Universities, such as a FreeBSD administration course and FreeBSD system programming course. During this discussion, the most important and interesting aspects of FreeBSD where outlined as well as potential student projects and student materials. I met Andrzej Tobola, with whom I discussed academic research in Poland. After the discussion group I met with Benedict Reuschling (leader of the group) and spoke about the possibilities of teaching FreeBSD at his university, and possible relations between FreeBSD and PhD academic research.
During dinner I talked to Hiroki Sato on usage of FreeBSD in students circles. I met with Eric Allman and Kirk McKusick, we talked a bit on my future talk in the scope of the Devsummit Track.
The second day I attended the Quality Assurance Testing group, where Hiroki Sato presented his Documentation testing framework. Necessary testing environments for ports commiters were also covered. Then went the vendor group, where Alistair Crooks gave a very interesting talk on Netflix and FreeBSD. Finally I visited the Desktop working group with Kris Moore as leader. After dinner I get acquainted with Baptiste Daroussin, with whom I discussed my GSoC project, and the state of my changes. He also encouraged me to start setting up a roadmap for commiting my changes to the source tree.
The third day of the conference started. Eric Allman gave a keynote talk on the OpenSource project life cycle in comparison with academic research projects and commercial projects. I stayed at the Devsummit Track, where I gave my talk after lunch. After my presentation, I talked to Kirk McKusick on the state of my project, and he also encouraged me to get up to speed and prepare my changes for the source tree. After the Devsummit Track I attended Marc Espie’s presentation dedicated to the packaging system in OpenBSD.
During the Social event Brooks Davis gave me a short talk regarding the current state and future changes in the ports collection and related utilities, and shared his experience with me regarding contribution to the ports source.
On the last day of conference I attended the following talks:
And finally, Kirk McKusick gave his overview of different types of locking in the FreeBSD kernel, and discussed where each type of locking is appropriate and why.
I arrived in Warsaw on Wednesday. After dropping off my luggage at the hotel Gromada Centrum, I stayed for the rest of the day at the hotel, preparing for the coming devsummit. Unfortunately, I was late to join devsummit participants for dinner and discussion of the coming devsummit.
The next day the devsummit started. The first person I met was Pali Gabor Janos, the devsummit organizer. I also became acquainted with Aleksander Dutkowski, another Google Summer of Code student at The FreeBSD Project. During the briefing part of the devsummit, participants introduced themselves and went to the working group rooms. First, I attended the Ports working group. The working group covered the following topics. Beat described the current state of movement to SVN, and explained problems that he faced and the long period of movement to the SVN repository. Finally, utilities that still work with the CVS repository were covered. Then Baptiste Daroussin discussed the current state and further implementation of the pkgng utility, as well as topics related to build cluster, such as the new portbuild2 script and schedule of building of binary packages sets. Unfortunately, the problem of UNIQUENAME global variable is still open. Finally, the new options framework and default sets of options for various FreeBSD builds were discussed. Kris Moore also took part in this discussion in scope of default options for PC-BSD.
During lunch I met Gavin Atkinson. He kindly offered his help in case of any questions or if I need his help with introduction to somebody in the community. I also get acquainted with Ulrich Spoerlein. We discussed the current state of mirroring FreeBSD source code to GitHub, the importance and benefits of his work, and the importance of maintaining GitHub repositories for GSoC students at The FreeBSD Project. Then I met Aleksandr Rybalko who works on porting FreeBSD to embedded platforms in the Ukraine.
After lunch I attended the OS Course working group, where we discussed different techniques of teaching FreeBSD at Universities, such as a FreeBSD administration course and FreeBSD system programming course. During this discussion, the most important and interesting aspects of FreeBSD where outlined as well as potential student projects and student materials. I met Andrzej Tobola, with whom I discussed academic research in Poland. After the discussion group I met with Benedict Reuschling (leader of the group) and spoke about the possibilities of teaching FreeBSD at his university, and possible relations between FreeBSD and PhD academic research.
During dinner I talked to Hiroki Sato on usage of FreeBSD in students circles. I met with Eric Allman and Kirk McKusick, we talked a bit on my future talk in the scope of the Devsummit Track.
The second day I attended the Quality Assurance Testing group, where Hiroki Sato presented his Documentation testing framework. Necessary testing environments for ports commiters were also covered. Then went the vendor group, where Alistair Crooks gave a very interesting talk on Netflix and FreeBSD. Finally I visited the Desktop working group with Kris Moore as leader. After dinner I get acquainted with Baptiste Daroussin, with whom I discussed my GSoC project, and the state of my changes. He also encouraged me to start setting up a roadmap for commiting my changes to the source tree.
The third day of the conference started. Eric Allman gave a keynote talk on the OpenSource project life cycle in comparison with academic research projects and commercial projects. I stayed at the Devsummit Track, where I gave my talk after lunch. After my presentation, I talked to Kirk McKusick on the state of my project, and he also encouraged me to get up to speed and prepare my changes for the source tree. After the Devsummit Track I attended Marc Espie’s presentation dedicated to the packaging system in OpenBSD.
During the Social event Brooks Davis gave me a short talk regarding the current state and future changes in the ports collection and related utilities, and shared his experience with me regarding contribution to the ports source.
On the last day of conference I attended the following talks:
- The BHyVe Hypervisor In Depth (Michael Dexter)
- Tips on running a conference for 250 people all by yourself (Dan Langille)
- OpenBSD and ‘real’ threads (Philip Guenther)
- Touch your NetBSD (Pierre Pronchery)
- A Fault Aware Global Server Load Balancer in DNS (Stefan D. Caunter, Allan C. Jude)
And finally, Kirk McKusick gave his overview of different types of locking in the FreeBSD kernel, and discussed where each type of locking is appropriate and why.
Tuesday, November 6, 2012
New Funded Project: Porting Efika
The FreeBSD Foundation is pleased to announce that Aleksandr Rybalko has been awarded a grant to port FreeBSD to the Genesi Efika MX SmartBook laptop and SmartTop nettop devices.
Both use the Freescale i.MX515, an ARM Cortex-A8 System-on-Chip (SoC). These low power devices will provide convenient reference platforms for FreeBSD on ARM, as they are low-cost complete systems. The Smartbook includes a 10" display, 3G connectivity and a battery life of 6 to 8 hours for $199.
When this project is completed, it will be possible to run X11 applications on FreeBSD on the Efika, with full support for sound and networking. It will also make it much easier to support other devices, such as some Android tablets, that ship with the i.MX515 SoC.
This project will be completed by the end of 2012.
Both use the Freescale i.MX515, an ARM Cortex-A8 System-on-Chip (SoC). These low power devices will provide convenient reference platforms for FreeBSD on ARM, as they are low-cost complete systems. The Smartbook includes a 10" display, 3G connectivity and a battery life of 6 to 8 hours for $199.
When this project is completed, it will be possible to run X11 applications on FreeBSD on the Efika, with full support for sound and networking. It will also make it much easier to support other devices, such as some Android tablets, that ship with the i.MX515 SoC.
This project will be completed by the end of 2012.