FOSDEM: more boring shit
Let's take a look at my annotated copy of the FOSDEM 2019 main talk schedule, shall we?
The sign-in sheet for this session doubles as a census of people nobody wants to talk to.
Some bureaucrats make excuses for the conference attendees' hypocrisy. In place of productive information or actual technical content, the speakers will whine about which programs nobody's writing for free. The audience will take notes on Macbooks, iPads, and other Apple products.
A primordial webshit travels to a conference devoted to serving a specific copyright cult, and then holds a lecture about why the copyright cult is really important. Nobody in the room will learn anything new, and the speaker will not reach any audience not already in the cult. While this is arguably a massive failure of advocacy, it's also about on par with copyright cultists' track record to date.
Another bureaucrat speaks to a very specific confluence of misapprehensions, to wit:
- Any functioning human gives a shit about bitcoin,
- Bitcoin has any effect on human society at all,
- Anyone has found a valid use case for blockchain technology,
- Blockchain dunces are ever given any position of responsibility,
- Anyone cares about the ethical value judgments of a professional copyright cultist
Nobody attending the talk will have the heart to point any of this out.
The speaker is very angry that people are using computers in a manner that renders specific copyright licenses irrelevant. The audience will be instructed to care even harder, even though no solutions are available. The speaker works for a company whose profitability depends on a customer base who gives a shit about copyright licenses.
Jon Hall observes the 50th anniversary of the creation of UNIX, and generously includes Linux in the observance. Linux does not deserve this, and the FOSDEM audience does not deserve Jon Hall, but on the bright side this talk will ruin the "100% useless dipshittery" streak in this year's FOSDEM lineup. That's right: they can't even fuck things up correctly.
if they'd just started with this talk we'd all be spared a lot of noise
Why, you might ask, would "open source" layered extensibility be different from any other extensibility approach (layered or otherwise) in any other software? It would not, of course, but the CEO of the company needed a reason to fly to Brussels for the weekend, and this is as good a reason as any to write it off as a business expense. Spoiler alert: "layered extensibility" here is a code for "everything is webshit."
A webshit will try to convince the audience that replacing other protocols with webshit is healthy and good. "Why" will be answered with "because we want to track you more closely," "how" will be answered with "extremely poorly," "when" will be answered with "as soon as the Chrome team tells you to" and "who gets to control how names are resolved" will of course be "Google."
A corporate drone explains that using non-Linux operating system makes it easier to get your company's shitware committed upstream, spreading out the maintenance burden and magically converting some of your operating expenses into externalities you can ignore. Sorry, I mean "engaging with a close-knit developer community benefits your product's ROI."
A database management program has not yet ceased development. The speaker will read the version control commit log for one hour.
A Postgres will mumble about RAM management features that everyone has been using for approximately twenty years. This talk belongs at some kind of professional database symposium, but it's happening at FOSDEM, so the audience will consist entirely of the speaker's co-workers and the people who are going to use the room next.
Another Postgres regrets to inform us that they've been fucking up basic storage primitives for at least the entirety of the 21st century, and continue to fuck them up today. This talk will comprise an extended apology, followed by a lot of excuses and whining, wrapping up with wholly unsubstantiated claims regarding future improvements. The only people who care already know.
The programmer behind an also-ran database program will discuss the only interesting aspect of that program: someone else's algorithm. The program in question is a port from Java to C++ of a database management program so shitty that Facebook, who created it, almost immediately discontinued all use and development thereof and donated the software to the Office of Software Terminal Care, also known as the Apache Foundation. The speaker chooses to work on this disgraceful travesty full-time.
Some corporate programmers are released from the mines long enough to pretend anyone cares about the things they work on. The anxiety and excitement induced by being allowed out of the cave caused them to paste the talk description into the web form twice. Nobody noticed.
This talk is an hour-long sales pitch for the speaker's employer, which sells a product that other companies can use to provision TLS certificates and spyware updates to touchscreen refrigerators. In order to lure idiots into attending the sales talk, the speaker will point out that the license of the software conforms to the thematic goals of the conference.
This talk is from a VA Linux alum, which means the talk will almost certainly be a story about how the speaker spent way too much money on poorly-performing hardware, then made it worse by crowbarring a half-working Linux installation onto it, resulting in a very expensive disaster that works about a third of the time. This will be followed by an invitation to others to follow suit in purchasing and then ruining a hundred-thousand dollar car, leading to a brand-new interpretation of the well-worn phrase "malfunctioning Linux driver."
Some rando is excited because you can now check if err != nil on smaller processors than ever before. Like literally every other language, the entire project is just transpiling to llvm intermediate code and then letting the existing toolsets handle the rest, completely removing any advantage of using a specific programming language.
A bureaucrat will ramble incoherently about seeking life advice from copyright law cultists.
Some Googles will present propaganda in order to sell low-quality laptops, on the principle that they can sometimes be tricked into functioning like actual laptops. With any luck, this will give Adsense valuable insights into the behavior patterns of gullible idiots, which sounds like a pretty juicy demographic to advertise in front of.
The speaker would like anyone at all please to use software maintained by the speaker. To achieve this goal, the speaker will try to convince a bunch of random strangers to choose graphical design tools not on their suitability for a given task, but based on the copyright licenses under which they are available. This conference is ridiculous.
A Microsoft tries to convince everyone to install a shitload of .dll files onto their Linux systems in order to use expensive IDEs to produce the same shit everyone else already does. Along the way, the Microsoft will brag about tricking many rubes into working for free on corporate platform code.
The rich and lengthy tradition of the Free Software community making inferior copies of other, better-engineered systems continues, in this instance enabling underemployed nerds the world over to shout things at their computers. Shouting at computers while some shoddy software desperately attempts to parse and respond to this input is apparently preferable to using any of the well-supported existing input devices that come with every single computer on earth, so this talk will be well-attended and extremely beneficial to anyone who has nothing better to do on Sunday afternoon.
This talk, from a Facebook lawyer apparently being punished with public relations duty, has nothing to do with privacy at all. The entire point of the talk is to make the audience believe that Facebook gives a single shit about their opinions regarding intellectual property law. The actual product whose weaponized license caused a shitstorm in nerd circles is not mentioned even once: misdirection, or idiocy? That's a trick question; nobody can tell the difference.
An academic, enrolled in the Tim Berners-Lee fan club, will engage in a performance-art piece hypothesizing about a world where anyone gives a shit about what Tim Berners-Lee wants. In accordance with the colorful tradition of web-reinvention nutcases, bold claims and broad promises will rain down upon a rapt audience. After forty minutes, the speaker will ask for questions from the audience, who will respond by spending ten minutes' sober contemplation of profound questions like "what if we COULD do things better?," "I wonder what room I was supposed to have been in?," and "if I leave now, can I get something to eat before the next unhinged rant?"
The United States Defense Department's most successful honeypot sends its apex bureaucrat to reassure paranoid Europeans that they can still totally trust all this stuff, you guys. Everything's fine. We're on your side. Route all your traffic through us. It's for your own good.
A researcher would like the audience to consider what happens when government oppression is executed in a mode that allows blame to be shifted to ostensibly-unbiased computer programs, but the talk description suffered a transporter malfunction and was merged with every edition of Roget's Thesaurus of English Words and Phrases. Those audience members with a sufficiently enhanced vocabulary will have to undertake the task of translating the meaning of this talk into emoji for wider dissemination.
A search engine programmer will deliver a sales talk.
The audience for this talk, whose topic is low-level systems diagnosis, will consist entirely of webshits who are under the delusion that "distributed" means it uses http as a core protocol. None of them will understand it, but understanding is not required to paste the relevant keywords into the end of their resumes.
Because Kubernetes, a clustered-application execution platform, was invented by Google, there is no way to usefully inspect any aspect of it. The speaker is one of a large class of poor sods who have to resort to injecting code at runtime into the kernel to make up for the absense of fundamental operational functionality. The talk will be well-attended, because injecting code at runtime into the operating system kernel is simpler than debugging Kubernetes.
A programmer will attempt to Fix All The Problems with perl by reimplementing all of the other projects created to Fix All The Problems with perl. This talk was primarily scheduled to keep the room available for people who couldn't get into the ChromeOS talk in the other room.
A Red Hat drone thinks that the problem with local filesystems is they don't have enough of the network stack involved. This talk will explain how to rectify this omission, as well as some tips on how to make block storage as unreliable as everything else Red Hat pays to develop.
A FreeBSD developer thinks that five-year-olds are interested in why storage software caches things into RAM and how that cache is managed. The only five-year-olds the speaker has ever met are bugs in the software being discussed. The word 'works' in the talk title is to be interpreted as an aspiration. We can reevaluate things after the FreeBSD project deletes all this shit and gets in line behind the Linux weenies.
Another sales talk designed to turn this weekend into a business expense. No apologies to this software's victims are promised. The sales talk highlights using the software as an abstraction layer for AWS, Azure, and Google Cloud Storage, without mentioning what a high-maintenance treadmill nightmare the solution is.
An OpenStack refugee disguises a CV as a talk, using numbers that seem large to attract an audience. Most of the performance of the namedropped system comes from add-on 'accelerator' cards which are so unusably terrible that Intel discontinued the entire product line out of shame. If the OpenStack project had any dignity, it would follow suit. The talk description does not contain a reason the speaker would admit in public being involved with any of this.