OSEDUCONF-2015, the tenth conference “Free Software in Higher Education” - videotapes and overview

A couple of weeks ago in Pereslavl-Zalessky, the tenth anniversary conference “Free software in higher education”, briefly OSEDUCONF-2015, was held.

As the name of the conference suggests, reports all revolve around open-sourcing Linux and education, sometimes with a bias towards purely scientific-technical things, sometimes pure soft, sometimes into politics. There are also incendiary reports, and frankly boring, deep-philosophical and earth-realistic, concrete-practical and academic reports with the title in three paragraphs, of which all the same nothing is clear, complaints about life and success stories. Open-source developers and maintainers, rectors and university professors, industry practitioners, officials from ministries, scientists of various degrees of madness, students who earn points for a diploma.

Now the reports are looking for the audience - I recorded and assembled them qualitatively (montage from several cameras, screen and voice recorder, Animation technology ™, etc.), so, as an illustrative picture, a minute video is almost a cock, well suited to instantly transmit and general impression and quickly flip through all the reports - maybe someone, even without reading the review annotations, will immediately catch the eye on the interesting.

')
And under the cat a brief overview and classification of reports from both the last conference and retrospective classification for the last three years - do not be afraid to click, there is only a compact text, and all videos are links.



This year, perhaps the most interesting report for a wide audience was the “ UNIX way crisis and fundamental IT education ” ^{supplementary materials} . Religious trend report on the crisis of faith. Beliefs in UNIX-way — everything that was taught turned out to be wrong — the Unix-rights system, interprocess communication, text logging, the graphic system, and ... where you don’t spit in Linux is everything, revising all architectures. Is there something invariably axiomatic in this crazy changing world that can be taught to students for decades without regaining consciousness? It seems only the generalized principles of good architecture, and it must be admitted that the “classic unixway” often did not satisfy these principles. In a sense, the report continues the pessimistic line of the report on the open-source crisis , which is also recommended for viewing. The report is expressive, and the speaker’s image of a mad professor from the Blood Stain comic book ... will be boring to watch.
SPO for effective teaching

• " Free and free software for creating mathematical sites " ^{additional materials} - how to effectively make formulas and graphs for web content? Tried different things, from GIF-generators on LaTeX to MathML and JavaScript / HTML5-draggers. In my opinion, the author missed one of the most effective solutions that I use - LaTeX SVG formulas.
• “ MediaWikiQuizzer or WikiExams - tests that are convenient for teacher and student ” ^{additional materials} - my story about the system of test questions MediawikiQuizzer and generally ease of use of the modified MediaWiki for convenient maintenance of educational content, and the tests are also “first class citizens”, and it’s also convenient to start and reuse, quickly correct erroneous or unsuccessful questions.
• “ Web-control on programming with automatic generation of tasks ” ^dopmaterialy - here, too, about the tests, with an emphasis on the importance of generating the tasks themselves, so that lazy students could not stupidly burn / deflate / zabombi / code all possible questions and answers.

SPO in teaching IT disciplines

• “ Building Embedded Systems ^{Programming Workshops} ” ^dopmaterials - history, as they set the appropriate course for two Belarusian universities, with the desire to work on real ARM hardware, but still got a Kalashnikov machine program for core programming, prog and all that and on emulators. But still, a good experience, all the materials on the githaba.
• Several reports related to Kumir-Piktomir - a distinctive, perhaps somewhat ahead of time system of teaching children to programming, and its inventor himself spoke.
  • “ Algorithms and programming for preschoolers and undergraduates - global trends and domestic experience ” ^{additional materials} - Experiments with PiktMir, i.e. "Pretext" programming systems for kindergarten. However, now probably everyone has already seen Blockly Games - so, ours have invented this for a long time, only popularity is low ... And through official channels, with techniques and retraining of informatics teachers, with hundreds of page manuals everything goes slow.
  • “ Five workshops of K. Yu. Polyakova on programming with automated testing in the CuMir system. The results of the retraining of informatics teachers " ^{additional materials} - This is just about the training of teachers, and there it is perhaps worse than with the training of kindergarten. In general, as it turned out, not every informatics teacher will master independent learning, although they have found life hacking — you can watch at what speed he solves a couple of tasks in order to understand whether he has a chance to learn independently or need in-person training.
  • “ IDE for learning Python ” ^dopmaterials - you shouldn't think that all this Idol is some kind of “alternative special way”, moreover, based on IDE for Idol they try to make a good IDE training for Python. There is a need - there are no good free IDEs, the project started before the emergence of academic licenses for PyCharm, and in this case, there are difficulties and fuss with asking for licenses and overloading the IDE itself. I immediately asked the author for the implementation of the REPL-cycle in debug mode - a thing that is wildly helpful in learning, and which is only in expensive paid IDEs. Soar or not - see in the coming year. I would like to, yes.
• “ On one experience of conducting a special course on Python with automatic checking of homework ” ^{dop.materialy} - Remote report from half of a screencast, on a half-distance course of study at VMK-MSU, through live screening on python with screencast viewing, and tournament solving of problems on the ejudge Olympiad system. “The higher the name in the tournament table under which you can log in, the higher your score is ©.” There are a lot of dubious decisions, such as automatic scoring by points, and risky advice to tame a difficult ejudge ... but the experience is interesting anyway.
• “ Developing information systems using the MongoDB database software NoSQL ” ^dopmaterials - in fact, this is not about development, but about teaching development, so here people with experience are unlikely to find something new, on the contrary, they can lash out at a bunch of questionable statements ("NoSQL is without transactions and is not suitable for financial applications, etc."). But from the point of view of teaching, yes, it’s time to start teaching only the classical RDBMS device, based on the System R theory and practice on phpmyadmin, it’s time to show the full diversity of the DBMS and teach you to make the right choice.
• “ Learning to process management on free software ” ^dopmaterials - “Business processes, managers with pathos people draw arrows in BPMN-notations and then model the Corporation” - you can play all this on the domestic sources. And for free, and invented various tasks, there are educational materials and all that.

SPO and organizational process in universities

• “ The learning management system and the unified educational environment of the MGIU ” are ^{additional materials} . At MGIU, students are taught Ruby / rails, and therefore they can do everything from the ERP system for the university, to the CMS portals and even the anti-plagiarism system. Of course, crossing Linux, GitHub, Ruby on Rails, PostgreSQL, MongoDB, CAS, Ace, Pandoc, Markdown, LanguageTool, TeXLive, MathJax, jQuery, Reveal.js ... are all fashionable and tasty from the open source world. It is sad that all this is in the spirit of the announcement “I’m not looking for anyone just bragging ©”, i.e. the decision itself is not published ...

Open-source in scientific and near-scientific research
Yes, it's not about education at all; rather, let's say, Original Scientific Research.

Static code analysis

Reports with titles longer than reasonable (> 256): “ Building an Equivalent Representation of Dependencies in the Source Text of Programs Using the Universal Intermediate Representation, ” ^{additional materials} and “ Experimental study of quantitative patterns in analyzing the source code of open source software using equivalent representations. ” ^{Additional materials} are all static analysis, with compiling code in different languages ​​in Abstract Syntax Tree (native parsers), and translating this AST into some more abstract and generalized "SCP". Well, with the help of established frameworks we load different open source code (where can I get a lot of code from ...), and we consider some very strange metrics. To me, here are graphs of the type “Dependence of the proportion of typed duck fields on the threshold” that the brain actually carried out (“Agronomist, but got drunk as a PhysTech ™”).

Datamining and searching

• “ BigData analysis with Apache Spark and ontological engineering methods ” ^{additional materials} - By the way, here is an example that the conference is not strictly prolinux - and a speaker with mail in the “outlook.com” domain, and the very first phrase, “like this, in your Linux for the full display screen? "(by the way - F5). But to keep the good ones - the simplest data mining (scrapping Wikipedia-index-counting something simple on the graph) is already taught in the most recent - Apache Spark + Scala.
• “ Improving the quality of searching in large volumes of text documents using a genetic algorithm as a way to support scientific research ” ^dopmaterials - uff ... another attempt to make a Semantic Search Engine with Ontologies. In the world of fashion, this has already subsided, and good full-text engines have killed the mantra “no semantic search is impossible”, but experiments continue. However, it does not matter, because the speaker is a beautiful girl. Everything is also on Apache Spark + Scala (“Scala is a fashionable language in our circles”). I really didn’t understand why a genetic algorithm is needed there, if it is used only for strange selection with maximization of fitness by randomly taken parameters ... (falling into ontology, weights of found terms in different sections of the document ...) i.e. it seems that the main thing for genetic algorithms does not occur - “crossing” solutions.

Chemistry

• " ROSA Desktop Fresh in the chemical laboratory " ^dopmaterialy - alas, here we are not talking about automation of the Breaking Bad open source, but rather some review of analytical programs useful to the chemist. Directly devices (mass spectrometers and others) still run Windows with closed drivers. However, it is not so thick with the near-chemical programs, if you throw out dozens of handicrafts, such as variants of the "periodic table". Alas, a good objective open-source software can only be born from a person with an understanding of the subject area, i.e. one must either wait when the chemist learns to program, or when he marries a programmer. Or vice versa, as in this case.

Linguistics

• “ Practical Sanskrit morpho-analyzer - Morpheus ” ^{additional materials} - the audience was filled with spells in dead languages ​​when the author demonstrated his latest Firefox plugin - for now this is only the inverse task of morphoanalysis, the direct one is a generator of word forms in the near future. And of course, down with Ruby, only javascript, only hardcore.

“ What is expected in the Eighth Alt Linux ^Platform? ” ^{Additional materials} are probably a well-deserved sponsorship report, for ALT Linux is the permanent organizer of this conference . Well, we, ROSA , regularly help with the technical part, video recording and publication.

By the way, these are still beta versions of the video, with “debugging information” embedded in them — for example, red time markers of true time. So if suddenly you find some critical bug in the video (the sound is gone, you need to make the screen bigger in order to read something, or vice versa, remove the screen to see the laser pointer) - write me , or comment on the video, or PM , in general, as you like - by specifying this very range of time and problem - as long as (a couple of weeks), you can still fix it.

Yes, there was another report, even with a very cool result (something like “P = NP !! 111”, but it’s not even clear to me at once), but actually the speaker and the rector of that university decided to keep secrecy and not to write it down (speech talked about some patents and other games in priority). Of course, it wouldn’t stop me from recording, but “Homeostatic Universe TAK © really happened”, snowfall, I was late by stupidity on a non-driven car ... but even having reached the beginning, I stumbled 20 meters before the venue (generator → battery → rode without headlights → and still stalled at the last corner) and was busy just before the end of this report, which was the first to go. So it goes.

Well, once you've read this far, we offer you a close-packed review-classification of the conference for the last three years, with links, videos and other materials, because even pompous names and bloated annotations can be tricky, everything is as brief as possible, in twitter-style.

---

Follow the links - there will be annotations, and abstracts, and a detailed infographics of slides, thematic classification and information about the authors. By the way, it was often the most difficult moment to publish - to understand who spoke: for some reason the speakers did not present themselves, in the program of the report according to the Soviet habit ten dozen co-authors were declared, it is not clear who of them spoke, and according to the foolish tradition, in the program only surnames with initials, foolish names, by which not only figs that you google, but even the author’s gender cannot be understood. We had to spend hours searching for sanitized information about the authors. Speakers, I call once again - read my memo to the speaker , you will make the world a better place.

, , , , . , — , . , - ( ).

— . — :

• :
  • - , , 19 … , … .
  • .
  • — , .
  • ( — , , , ), , .
  • — - — , .
• — «» , — , , Linux- — , … , , .

, , , .