Recently, the sixteenth conference of free software developers was held in the glorious city of Kaluga. The conference was organized by Bazalt-SPO . Embox , being a free project, also spoke at the conference. I, traditionally, will not review all the reports, but will talk about those issues raised at the conference that were of interest to me personally.
Before considering the reports, I’ll say a little about the organization of the conference. She, as I said, was held in Kaluga. Many of those who did not attend the conference are interested in why not in Moscow, where there would be more visitors. Indeed, getting from St. Petersburg to Kaluga was less convenient than to Moscow. I got from Moscow by train in two hours. Muscovites said that from Moscow two hours by car along a good highway. Both one and the other option, in my opinion, are not very tiring.
The benefits, in my opinion, are greater. Firstly, Kaluga is a beautiful city, there is where to walk, what to see. For example, the space museum . Secondly, this is an occasion to see something other than Moscow, where the majority happens often enough, and many just live. Well, such an unobvious moment as a party in the evenings. In Moscow, people quickly creep into their homes, and then people talked in bars, cafes, or just walked around the city. To communicate, not only was a telegram chat started , from it I not only found out where the people were hanging out, but, for example, why the conference had such an emblem. It turns out that the emblem is a modification of the universal emblem of the community of all hackers , which is a glider of the “Game of Life” .
Some netizens who learned about the conference were also worried about mandatory registration. The conference itself is free, but you had to register by leaving an e-mail. Registration was needed only so that the organizers could evaluate how many people will attend the conference, because the number of seats, you know, is limited. As a result, at the conference no one controlled the registration and you could go in to sign a badge for yourself (or even be present without a badge) and calmly listen to reports, chat with people and eat pies. So everything was more than free!
Reports
The first report was devoted to a very important topic - the interaction of the media and a free project. The report was entitled “How to Organize Media Work with a Free Project,” by Sergey Golubev. The report suggested, as an alternative to publications in traditional specialized media, creating their own blogs to talk about the project.
The importance of the topic was well demonstrated by the participant’s question after a report by Ivan Panchenko, “The development model of PostgreSQL as an international free product and community . ” It sounded like this: here you are so good and free, but why do you talk so little about yourself? I hope this conference allowed her to at least get better at the ideas of open source software.
As authors of the blog, it was proposed to attract students for credit. The author’s idea is as follows. Not everyone knows how to write complex software, and who knows how, often does not want to be distracted by publications and documentation. But everyone can describe complex software. This is at least easier than developing. Accordingly, a student can, on the one hand, bring noticeable benefits to the project, and on the other hand, may not be as qualified as in the case of development.
Personally, I was a little touched by a few points. Firstly, the work is set off. I already told in an article on the hub , in a report at the TechTrain conference, and after the report I also inserted a comment about the work in the standings. I don’t know how the others do, but in Embox, offset is a very weak motivation. I’m not talking about the moral side of using shareware, I’m talking about the low quality of the result obtained with the help of an unmotivated participant in the project. Therefore, in our project, even if possible, the test (diploma, exam, ...) was never the main motive. Yes, we write diplomas, term papers and other scientific works, but at the same time the goal is to make something interesting for yourself, and, of course, useful for the project.
The second thesis, which caused me some skepticism, is that publishing is easier than developing. Actually, I would agree that writing publications is easier, but we are talking about technical publication, and not about posting somewhere, “oh look what a cool project, it still has boring wallpapers.” And it seems to me that there is some easy snobbery in speaking - look, we are such cool developers, well, at least you write about us.
I will quote the thesis about simple documentation on Embox. We have the problem of publications and documentation is very acute. And to be honest, we tried to correct the situation with the help of an approach through credit to students. However, this did not work, in search of a solution, I wrote an article “How we develop documentation in an open Embox project” . What I agree with the author is that if you want to help a free project, then this can be done not only through code development! The users who use the project are very valuable, those who leave feedback are even more valuable, and the value of those who make documentation and publications can hardly be overestimated. But still, the motive should not be a set-off, but a kind of “karma”. That is, by publishing competent technical articles, you create credibility in this area, improve your resume, so to speak. In addition, this may be the way to develop code, because in order to describe something, you yourself need to figure out what you are talking about.
With the main idea of ​​the report that students should be involved in participating in open source software projects, be it publications or source code, I certainly agree. This is a mutually beneficial cooperation, because students gain experience in an area of ​​interest to them, practical skills in team work, moreover distributed, and, of course, raise their credibility, and therefore their status in the labor market. In this regard, a report by MIET students “Solving the problem of organizing quick communication between performers when working with business processes for implementation in the free RunaWFE system” was indicative. The content of the report is not so important, it is important that both the report and the development were made by students. In addition to this report, there were also reports from other universities: TSPU im. Tolstoy, MTUCI, Brest State Technical University.
The problem of attracting students (participants) to open source software projects was voiced in the already mentioned report “PostgreSQL development model as an international free product and community”. The author expressed the idea that there is a danger for the development of open communities in that new participants are not well attracted. After all, it used to be cool to be a contributor in open source projects, and the developers strove for this. But now, students are interested in quickly creating some kind of startup, because it is at startups that the halo of success is now. And to create a startup, you do not need to deeply understand technologies, you need to be able to use them. But probably, there were always those who created the technologies and who used them. The structure may have changed a little, now the development of open projects is often carried out with commercial companies.
The importance of users for a free project was discussed in Denis Silakov’s report “OpenVZ Customer Experience Program, or user data collection in OpenVZ 7” . The essence of the report is that it is very important to get feedback from users in an open project, but at the same time, of course, not violate the laws of countries where the product and the privacy of user data are used.
A series of reports was devoted to processor architectures. Perhaps this is due to the fact that Basalt-SPO pays great attention to porting its distribution to non-Intel architectures, as I already noted in the article about OSDay . At this conference, there were also many different "non-Intel" glands. Elbrus, RISC-V, RaPi4. I, of course, also did not stand aside and showed Qt on STM32F747i-discovery .
The most frequent entertainment at the stand with various pieces of iron on which Alt-Linux is installed, besides toys, was a performance comparison. Of course, modern Malinki are inferior to modern intellects, but only several times, respectively, are quite suitable for use as office, and not only office, desktops. During the presentation by Anton Midyukov “Mkimage-profiles is a flexible distribution build tool for many platforms.” NVIDIA Jetson Nano was connected to the projector, naturally in Alt, and everything worked perfectly. And Alexey Novodvorsky said that he ordered such a piece of iron from China for use as a home desktop. The report itself was devoted to the automation of creating distributions for various platforms and architectures.
Naturally, the most common non-Intel architecture was ARM. Roman Stavtsev from the company BAIKAL ELECTRONICS made a presentation at the conference with a report “Processor Baikal-M. Software and hardware environment. “ And spoke about their new Baikal-M processor based on ARM Cortex-A57. Unfortunately, the boards did not manage to be brought to the conference, but according to my information, they were presented at the Microelectronics 2019 conference already next week. And since the architecture is standard, then the software package is standard and, of course, open source software. Gcc compiler, qemu emulator, U-boot loader, Linux kernel, ...
Employees of Basalt-SPO devoted two reports to domestic processors with Elbrus architecture . As you know, the architecture is original, but unfortunately quite closed, hence its advantages and disadvantages. More of course flaws. Of the advantages, I only note that in the report of Andrei Savchenko “Features of porting open source software to the Elbrus platform.” It was said that when porting various software to this architecture, they found problems in this software that rarely appeared on common architectures and even patches were accepted in some projects. Igor Chudov's report “Problems porting SBCL to new hardware platforms.” Was more pessimistic. It was about the same Elbrus, but in the name instead of “features” sounds already “problems”. The author of the report complained that he could not even find a clear description of ABI, I could not resist and invited him to read our article “Climbing Mount Elbrus - Combat Intelligence. Technical Part 1. Registers, stacks and other technical details . ” True, the author of the first report also noted that some information about the registers of Elbrus is on the wiki Alta . In general, everyone agreed that the architecture is certainly interesting, but the closeness strongly hinders the development and complicates the development. Everyone hoped that over time the situation will change, and the ICST will make its products more open, and we (the community) will render all possible assistance in this.
Another very fast-growing architecture presented at the conference was RISC-V. A fairly quick prototype was presented at the demo stand. The report was presented by Nikita Ermakov, it was called “RISC-V Architecture” . And was dedicated to the ecosystem around this open processor architecture.
In conclusion, I will give a link to the conference program and a link to all the videos from this conference. After all, as I said, I did not review the conference, and did not dedicate many points. Of course, there were much more interesting reports. In general, the annual conference is sometimes consecrated on a hub , unfortunately, quite a few recently. There are a lot of interesting reports on it, here is a video from conferences for all years . And the atmosphere at the conference is traditionally very pleasant, including lively discussion, and just communication with smart, qualified specialists. Come to the conference, it’s not so far from Moscow.