FOSDEM: more boring shit (2017/01/05)
Let's take a look at my annotated copy of the FOSDEM 2017 main talk schedule, shall we?
The speech is 25 minutes so it will fit into your Burger King lunch break!
"GIFEE" stands for "Google Infrastructure For Everyone." Presumably your computers are too fast and you need some more abstractions to slow them back down. No, I don't know what the second 'E' stands for.
"The Free Software Commons" is just some bureaucrat's idea to put a shitload of free software on a hard drive somewhere. The talk description does not explain why archive.org is insufficient for this task. I think it's because archive.org will not establish a salary for the bureaucrat.
An FSF bureaucrat spreads FSF propaganda.
Someone is trying to use linux to do something.
This one is only ten minutes so you can attend it and get back before the fries are done.
The talk description claims RISC-V is in use in both Internet of Things and supercomputing, which is amazing since nobody's produced any working hardware yet.
Hewlett-Packard came up with a bespoke machine. They could not afford any marketing experts, so this machine is called "The Machine." The big innovation is that you just shit data onto the disk instead of bothering with a proper filesystem. The speaker is the same Keith Packard who made the X Window System even worse than it was before.
Google, typically, open-sourced some trash software that only serves their needs. This guy is here to explain to you why you should ever, ever give a shit, and also you should probably get involved with the project and give Google free labor.
In addition to "30.000" being a strange grade of precision to apply to the number thirty, the second the audience realizes this idiot is talking about Ubuntu everyone will know this "never break" bullshit is a lie.
The talk description is in English and the talk will be given in English. You can find out more on the FOSDEM web page, which is in English. This shows you how important localization is.
Someone realized that it's much easier to secure software if you throw half of it out and ignore the other half. He has started a business with this as the planned profit source.
Someone realized that "containerized applications" -- the practice of just tarring up all your shit with its shared library dependencies and treating that as a unit -- is part of what made Windows suck. The speaker's other talk involves advocating for 'rolling release' distributions. That's right: even the rolling-release idiots think AppImage sucks.
An idiot is afraid to say fuck. He works for Red Hat and Apache at the same time so obviously he is the world's leading authority on what documentation should look like: either contained entirely within runtime error messages or hidden behind a paywall.
Another Red Hat idiot wants people to stop writing documentation and write stories instead.
The talk description spouts meaningless shit about storytelling, but the real common factor between Rowling and Linux is their presence in the homes and hearts of millions of children and/or mouth-breathing retards who are impressed only because they haven't seen what competence looks like.
Finally, someone comes up with open-source gaming technology even less interesting or popular than VR.
An academic has a radio.
An idiot calls multidimensional arrays "datacubes." Nobody knows why, or wants to find out.
Daniel Stenberg is here to tell you how important curl is, because telling people how important curl is happens to be his full-time job.
All I Want To Know Is Why The Woody Allen References*
*(to a movie that came out before you were born)
An absolute asshole tries to justify the massive crowdsourced surveillance program they've built.
Organizational advice from two idiots who have never organized anything.
The idiot from last year is back, trying to sell you laptops preloaded with a version of Coreboot that has all the useful shit stripped out.
It is kind of gross to make this some kind of scheduled event. Have some class and do an informal meeting after hours for this, people.
Yet another webshit trying to make his webshit faster.
Red Hat would like the ability to significantly increase the amount of data collected for the bug reports it ignores.
Security and Encryption
Yeah, it's still just SHA-3, man.
An Exercise in Full-Stack Irrelevance!
Apparently nobody's told Red Hat that if something can be automatically decrypted you may as well never have encrypted it to start with.
Recognizing that IPSec is an overcomplicated monster that nobody uses, an idiot sets about writing a simplified monster nobody uses.
A moron thinks someone wants to use third-party servers with Open Whisper Systems' Signal protocol. Don't bother to learn his name; you won't need it.
Listening to a Russian talk about quantum computing is akin to asking your dog how to replace a circuit breaker.
FOSDEM 2017: more boring shit (2016/12/29)
Things are a little different this year, because FOSDEM has their shit together even less than usual. As of 29 December 2016, just about a month out from the conference, they still have not got any keynotes, main talks, or lightning talks scheduled. If we're very lucky, they never will.
Meanwhile, we'll make fun of the devrooms.
All of them.
Two guys think they're going to "focus on the weakness" of their respective projects without anyone trolling. NetBSD is not participating because there is nothing that NetBSD does not suck less than.
Disk encryption was difficult to implement on FreeBSD because the programmer did not know what he was doing.
This has nothing to do with BSD.
That guy from Nuxi shills his poor-man's-selinux project.
No content, but at least it has the same title twice.
A hobbyist backup project that "requires no maintenance." It creates bootable ISOs that put your computer back the way it was. No, I don't want that either. It only works with a few linux distros, none of which you want to use.
A shitload of bash scripts to control apache, tftp, and dhcpd to do all the maintenance shit that the previous product claims not to need. The guy seems to think someone will pay him to work on this. God help anyone who does.
The first actual viable backup project under discussion -- it's a fork of Bacula. Turns out all they've done since the fork is make a web interface.
A Red Hat employee thinks QEMU is the right program to handle incremental backups. Nobody is surprised.
SuSE is back shilling the idiotic ISO maker again.
"requires no maintenance"
All the people who gave the earlier talks gang up to ask an empty room what features it wants.
A nerd does not know when internet conversations end because he lacks the typical real-world signals to which he is accustomed (primarly the other party rolling their eyes and walking away). Having described the task as "hard," he relieves us all by being just the man to teach us.
Truly, half an hour is not long enough. This program is a shit festival of stupendous proportion. It's a python thing ... installed by downloading a VM ... from an OS repository ... which they tell you not to check the certificate of. The python thing uses mysql to store its data, but it does everything over http... except the user interface, which is java. It's like someone threw a grenade into a room full of bad decisions.
'TheForeman' is a webshit systems administration interface, and this talk is about how to make it show the output of a real configuration management tool. The speaker is the "Stop Talking" expert from the lonely Community devroom, so with any luck the audience will get to witness his mastery at "stop talking now."
The talk description contains a typo and refers to Docker as 'Dicker'. This is a much more valuable contribution than anything else the Ansible team has done.
If that title is hard to read, it's only partly because it was written by a person whose first language is not English. The other reason it is hard to read is that it's a terrible idea attempting to solve a problem caused by incompetence.
This is a pretty common theme in modern systems administration (primarily among the people who call themselves "devops" and mean it): take a handful of tools that almost everyone on earth uses, make incorrect assumptions about how other people are using them, and then put it all on Github and give it a name. Congratulations, all those things are now Your Product, you genius! Please note: none of the tools this person is presenting are a spellchecker.
This speaker has never actually done anything except give a lot of talks and write a lot of books about everything Smelling -- code, design, projects, everything. I think he might just need to clean his laptop.
This better not actually take five entire minutes.
This is a talk about porting things to Mir, the Canonical-sponsored Wayland clone which absolutely nobody on Earth likes or cares about. The speaker attempts to remain relevant by being pretty sure some of this might apply to Wayland, maybe?
This guy's primary hobby is 'trying to get KDE people to care about snap and flatpak.' Since snap and flatpak solve problems nobody has, and solve them in ways nobody likes, I imagine his hobby is a challenging one.
Haiku, an OS that is entirely a reimplementation of another OS sends a representative to talk shit about "stealing ideas." The asshole they sent thinks that being multithreaded will help a program run well on older machines, despite older machines being more likely to have one core.
This is a handy talk to attend if you want to know which European administrations are running out of money.
Namedropping aside, this talk is actually about some guy using QML.
This has nothing to do with Desktop.
SPOILER ALERT: This program does not exist.
Kolab (the company that sells the imap shitware) is making noise about producing a Qt client for their product, and they have dispatched a drone to fluff it.
A clueless retard, who happens to work for SuSE, shills SuSE's new product. Tumbleweed is an attempt to offend the gods by crossbreeding an overengineered bureaucratic fuckpile like SuSE with an undisciplined usually-broken teenage jerkoff festival like Arch. The retard in question has no idea what drawbacks could possibly be had by changing major versions of basic system components. That's why he works in QA.
Homebrew is like pkgsrc but not implemented as well and full of macshit. Literally nobody cares what version of your package manager you're using. The speaker can't even be bothered to describe the talk, because he knows this.
Unsurprisingly, SuSE has shit out tons of repeated-work packages to format documentation. They send an employee to talk about it, in case anyone wants to hear how great it is.
OBS, the one genuinely nice thing SuSE provides to the world, can apparently get you 'a new Linux appliance with every commit', which is both hazily defined and doesn't seem like all that good an idea. The speaker mentions this functionality powers Tizen, which explains why Tizen is completely absent from the Western world.
Fedora is ever-so-slowly figuring out why the BSDs have the concept of 'base system' and 'ports' as separate entities that do not meaningfully interact. Two Red Hat employees are trying to nail down sufficiently devopsy terms to replace shitty utilitarian drab terms like 'base system' and 'package group'.
A genius wants to reimplement Solaris boot environments with btrfs. Yes, this is already trivially easy with zfs. No, btrfs has not fixed their data-loss bugs. Yes, this guy is serious.
The only competent person in the building, in an act of pure kindness, will display a fully-functioning thought process to the audience. None of them will understand any aspect of it and the best-case scenario is mild confusion.
To cleanse the audience palate after the preceeding scary-ass display of intellect, the devroom organizers present the rantings of a mentally disabled lunatic on the topic of functional packaging of webshit atrocities. Relieved beyond words, the audience will take careful notes and rush home to throw themselves into implementation of this garbage -- an effort to drive out the painful memories of Natanael Copa making rational decisions and producing working software.
A lifetime inmate of the MySQL ecosystem/prison gets super pissy about packagers, because his project's bullshit is apparently their problem.
Some asshole wants you to GPL your makefiles.
They added a plugin interface. No, I don't know why that's worth an entire talk.
Hahaha, not integrated gnucap simulation with Kicad.
Attempt #10,467 to replace Verilog and VHDL with some hipper shit; this time it's Scala, functional programming, and OOP. Don't settle in; it will be gone before you can finish asking why the fuck anyone would use this.
A carefully-titled talk, because it is not possible to create anything *but* simple FPGA designs with OSS tools. The first step in creating complex FPGA designs is always "buy the toolkit from the FPGA manufacturer."
Not. Everything definitely goes into the "not" category here.
The talk description is empty, which at least accurately reflects the availability of FLOSS tools for FPGA work.
Scheduled for twenty-five minutes, with four participants, which should leave about twenty minutes of uncomfortable silence.
The guy who wrote it attempts to justify its existence. This should be easy, since it doesn't actually do anything.
This definitely needs to be a presentation instead of a page on their website. I'm glad to see FOSDEM enabling important work like this. The world would be a darker, more efficient place without it.
A brave man will attempt to impart meaningful programming guidance in a 25-minute window. Subtracting the traditional fifteen minutes to get his laptop to recognize the projector and the ancient five-minute "is my mic working yet" ceremony, and he'll have just enough time to pagedown past the license header before he has to stop.
Someone reverted all the way back to the circuit design methods of the mid-1980s, except this time with lisp. Upon this base, they are reimplementing AutoCAD.
The 'EDA tool' under discussion here is a web shopping cart.
Nothing has changed; no changes are forthcoming. But did you guys see the new web shopping cart?
Chromebooks are crippled laptops designed to corral people into Google's surveillance platform. A man offers to teach you how to adapt some of their controller software to produce your own crippled devices.
Remember that news story where some guys demonstrated the ability to remotely fuck over a Jeep? Red Hat is here to teach you how to design security vulnerabilities for your car company!
Raspberry Pi is a series of shitty computers designed to save parents money in the process of pretending their children are geniuses. An important feature of the Raspberry Pi is that it has hardware built in that allows you to blink LEDs. This guy sells the LEDs.
SPI is a light-blinking interconnect for Linux.
Autohell clone #4,566 for the entire operating system.
"Because GNOME can't just use working shit that already exists" and "poorly", respectively.
Turns out hardware vendors don't give a shit about Linux programmers. The speaker presents a series of tactics for working with people who don't care about you and are not interested in your bullshit.
The guy who tried to fix random-number generation is back to tell us that testing things is worthless and you should just use statistics to predict your failures.
GPS is not good enough for this guy's autonomous indoor flying drone. I eagerly await the video of the talk so I can see if he's figured out how to make his drone localize and avoid a baseball bat.
The speaker teaches us how to make a Wayland compositor "from scratch":
- have Qt do it
His Turing Award is in the mail.
Jenkins plugins for testing your security vulnerability platforms.
A Tizen developer wants you to watch him install some software. It's going to take half an hour. The last step in all Tizen-related test suites is "ensure this hardware is never sold in any country where anyone might speak English or have any money."
The most important part of designing an embedded security vulnerability platform is the ability of your hardware to connect to the internet. Without an internet connection, Brazilian funboys can't root your skateboard/refrigerator/other stupid thing you're selling. The speaker presents technology allowing Linux-based devices to share software attacks among a localized mesh network.
Someone wants to replace CANBUS and QNX with this shit and linux.
Exception in thread "main" java.lang.NullPointerException at devroom.Welcome.main(Welcome.java:1)
Still everyone's favorite half-assed dodge to avoid paying Oracle any money!
IBM has written so much shit in Java that they can't go back. They also won't be paying any money to Oracle, so they need something that works better than OpenJDK. They plan to open-source all of their JVM, which means you will be able to download it as an Eclipse extension sometime in 2035.
You can use Java to hot-patch Java into your Java so that the Java can monitor the Java!
I wouldn't take this literally. This guy's employment contract hasn't been optimized away yet.
25 minutes to describe every available OpenJDK garbage collector and also introduce a new one, to a room full of Java programmers who are willfully ignorant of memory management to start with. This is another error that should have been optimized away.
What a shame this talk comes right after the Shenanodah garbage collector talk! After that tour-de-force, why would anyone want to work on a competing project?
Oh, right. Because this one is coming from Oracle and is therefore going to actually be used.
The speaker claims they'll "improve Java tool performance by a factor by 1000"[sic], which raises the question of what the hell is wrong with the idiots who wrote the old Java tool.
I'm sorry. I was going to say something funny here but the talk title is making me physically nauseous. Let's just move along.
The production environment is Twitter.
The talk description is just a condensed CV for the speaker. He seems to divide his time between molesting Lucene and posting his CV to conference websites.
No talk description, but the speaker is a Red Hat employee, so I'll just assume this talk is about systemtap.
If I understand the badly-formatted talk description correctly, Oracle is excited that they can get people to give them feedback. Presumably "The Awesome Parts" is a kinder title than "Look At All These Emails We Ignored."
A. lol I dunno
Small panel mockery
The only relationship this talk has to GNU Guile is that Guix is written in it. It's like presenting a how-to-install-Windows-XP talk in the C++ devroom, only nobody uses Guix or Guile.
The speaker is careful to point out in the talk description that Guile is turing complete. In order to convince people that Guile is suitable for UI programming, the speaker holds a goddamn ncurses program up as an example.
Oh thank christ someone wrote a colorized REPL for Guile. This will fix all the adoption problems!
In a rare flash of clarity, some GNU programmers stop what they're doing and realize they have no idea what the fuck anything is. Sadly, instead of learning anything, they built a shitshow to "help users and sysadmins reason about the whole system," because clearly only GNU programmers are capable of reason unassisted.
In typical GNU fashion, the speaker assumed 'subtitle' meant 'subset of the title,' and so we have the above nonsense.
GNU still thinks they can reimplement Twitter. Maybe they can, but nobody will ever know, because nobody will ever care.
In this context, 'script writing' is text-based role playing game scripts. Please allow a couple minutes for the last couple of people to file out of the room before you begin this ridiculous display.
Turns out it's a lot of hard work to make scheme fast! Who knew! Oh, right: everybody knew that. The speaker doesn't have a solution, but instead wrote software that throws shit out and rewrites your code with the absolute minimum amount of information retained.
Lisp guys are still mad that computer hardware doesn't speak lisp.
At long last, someone is tackling the important work: porting software nobody uses, written in a language nobody uses, to an operating system nobody uses, to enable a package manager nobody uses.
A key-value store for your scheme programs, because it's 1987.
As opposed to hydrospatial.
Some guy wrote a wget equivalent to download shit from the government. He named most of the functions after himself.
Jesus, I *really* hoped we were past this 'semantic html' bullshit, but here we are again.
By "best place," they mean "best place to live," but the example site defaults to Milan, which is not within a thousand miles of the best place to live.
They had over twenty kids cranking out software all summer! Come and see the wondrous benefits of hordes of undereducated programmers!
An OSM guy is slightly pissy about the filthy casuals who invaded his nerd kingdom.
Guys, it's soooo easy! All you need is python, flask, sqlite, sqlalchemy, osmalchemy, angular.js, a live network connecti-- wait where are you going
SPOILER ALERT: This software doesn't exist.
Some dude wants to crowdsource awareness of legal public notices. I think he's serious.
Stupid programmer, you can't fly a hangar.
Hiring a surveyor is a waste of time, when you can just hope some dude flies a drone over the land in question! As long as you know what camera lens he's using!
A fly-by-night shitware developer talks about reimplementing Uber in third-world countries.
A career bureaucrat attempts to justify the overengineered garbage he helped ram through a standards committee or two. With luck he'll divulge the real motivation to his audience: it has generated an unending stream of grant funding for his university.
The same failure to execute we got from every other MySQL version! The speaker talks about GIS-specific functionality being planned, without bothering to explain why you'd use MySQL for GIS to start with.
Incredibly, Boost is still a thing that people actually consider using. Unsurprisingly, Oracle is funding its development.
I'd say an Oracle-employed MySQL coder giving a Boost talk already has three strikes against him.
A thirty-minute overview of the entire history of graph theory and its applications in computing is surely worthwhile -- especially since this talk is likely a Neo4J sales pitch.
Shoving a graph database into Drupal is definitely worth your time! Just ask this guy, who makes money off selling Drupal!
Someone has noticed that running queries on huge-ass databases can be slow. Instead of reconsidering their approach to data analysis, they applied several caching bandaids and are very proud of the resulting corpse.
The ability to visualize and interact with very large datasets is an important and fast-moving sector in information technology. These people have put together a system that enables a lot of interesting funtionality, and then glued it to the worst, least-interesting possible application: Twitter spam.
Oh jesus, not these assholes again
It should be obvious by now, but anyone who looks at that syntax and is not immediately repulsed is a monster and should not be trusted with any important task.
Ah yes, let's use graph databases to implement the "customers also bought:" upsell advertisements in shopping carts. We'll sleep better tonight knowing we've enhanced the human experience.
SAP sends an employee to shill its products. No edition of SAP HANA is open-source. Some parts of SAP's UI toolkit are. It's not clear what the hell this talk is doing here.
MPP here is code for "we use SQL and we are resistant to actual parallel processing." This talk is basically advertising for Apache MADlib, which is a bundle of third-party packages shoved into a database interface.
More Neo4J spam, this time about how to work around its crippling performance problems. SPOILER ALERT: throw shitloads of hardware at it.
These people are working on figuring out when someone is using your Hadoop cluster to do bad things. Or something. They mumble about cybersecurity and network analysis but never make any actual points.
Some dudes are desperate to make containers relevant to HPC. They brag that you won't need to have any applications installed on your cluster (making the incorrect assumption that all users are going to play with this container toy) and that users' programs will run on diverse supercomputers across the TOP500 (making the incorrect assumption that users commonly have accounts on a wide subset of TOP500 supercomputers).
Like everything else in Cuba, other countries paid for everything.
Two GNU programmers with no visible HPC experience set out to solve problems that don't exist in ways nobody wants.
Someone ported a Python app installer to Cray PE because they got sick of waiting for Cray support to get around to it.
Someone wrote some software to figure out which programs are being used on a computer. Hope you like Elasticsearch.
Someone wrote a scheduler and implemented it as a python library. This way, you don't have to learn to write real code, and your code won't run right on clusters with real schedulers.
Someone noticed that programming for GPU accelerators is a pain in the ass. They responded by inventing a whole new language with its own compiler, for masochists, I guess.
Another guy trying to solve a lack of systems administration skills with containers and/or openstack.
An Apache programmer thinks anyone in HPC wants to use Bigtop or Spark. He is mistaken, but is willing to set that aside to mislead the audience about the usefulness of trying to use third-party tools to make your software scalable.
The only one of these talks to acknowledge that Spark is trapped in the JVM. Pleads with users of other languages to take it seriously anyway.
Someone wrote some software to determine which half-assed 'big data' framework is faster. Unfortunately the software does not seem to automatically insult the user for nerding out about benchmarking bad code instead of writing better code.
Someone decided to use natural-language processing for the most important task humanity has ever underaken: policing the Fedora wiki. Why does Fedora have a wiki? Nobody knows. Search stackoverflow and see if someone explains it there.
The speaker informs us that "little previous knowledge" is required, but declines to specify whether he's talking about himself or the audience. My prediction: both.
Have you ever wanted to use a web browser to draw pictures of where your data should be stored? Have we got the product for you! Stop laughing, this is serious, people use this for serious things, STOP LAUGHING
Not answered: WHY
More benchmarking noise for terrible software.
Some CS professor made a faster version of Hadoop to attract grant funding. It's still not as fast as getting your data structures right.
This one's got a PDF attached! Let's take a look!
Oh, for fuck's sake. They're monitoring hotel prices to write travelblog spam. GET A JOB, ASSHOLE
Someone wrote checkpointing software that only works with one specific Apache product. NEXT
You should absolutely not care about SQL for big data and nothing anyone does to 'help' is going to make it any better. "SQL" is the prototypical "I don't understand data structures" technology and "big data" is code for the current overpaid charlatans of the "I refuse to learn data structures" ilk.
Since all of the actual developers are engaged in other devrooms, it makes perfect sense to concentrate all the legal shit in once place while they're busy.
A bureaucrat pontificates about bureaucracy.
A Bloomberg employee seems to think that binding legal agreements should prioritize ease of use. This problem is real, but whining about it in Europe is probably not going to fix it.
Turns out there are consequences to playing fuck-fuck games with your project's licensing when you try to monetize it. Who knew? (Everyone knew)
SPDX is a license classification scheme which has had the living shit trademarked out of it by the Linux Foundation. The speaker would like basically everyone to use it, generating a lot more work for themselves with no measurable gain.
You should start a trade association to maintain a vicelike grip on your source code. Eclipse did it, and everyone loves Eclipse, right? Right? [SFX: lone IBM employee cheering]
A team of academics crowdsources the topics for their next round of grant applications.
A Red Hat bureaucrat rambles about GPL enforcement, which has never actually happened.
This would be an interesting talk, because GPLv3 is a god-awful license written by facists. However, the core assumption that Samba is relevant in any marketplace is fatally flawed.
Some Intel employees waste everyone's time talking about Open Hardware, which does not exist.
The Fedora Legal Chair, who is not a lawyer, takes time out of his work (as the leader of Red Hat's Education Outreach team, whatever the fuck that is) to describe his hobby in great detail.
Some Europeans are concerned that they will have to abide by radio transmission regulations. They are attempting to derail the legislation by pretending it affects anyone but them. An FSF bureaucrat takes the lead.
"GPL enforcement" is a technical term that means "writing angry letters to people who use your source code." Someone from the FSF shows up to whine about how people aren't paying sufficient respect to the Founding Fathers of Shitty Software: the FSF.
This guy started a business doing this; presumably if anyone actually cared they'd have hired him by now.
Yes. What's the other 24 minutes going to be about? Nobody knows.
Some idiots decided to open their own patent office. Nobody cared.
The answer is "nobody cares", because it's never come up in court, and it never will. That's not going to stop bureaucrats fantasizing about what might happen, or playing wargames in their heads about how they'd handle the case, but it doesn't mean anyone will ever have to give a shit.
An idiot thinks it's possible to determine the efficacy of a Code of Conduct by grepping for swear words.
Some dipshit registered a nonprofit and wants someone to write him software to run a certificate authority. Does FOSDEM not screen speakers at all? This is the conference equivalent of a dude posting 'someone write me a better Facebook' on rentacoder.
At long last, a Mozilla employee shows up to explain why they refused to ship a root cert from Honest Achmed's Used Cars and Certificates.
An FSF shithead shows up to screech about the dangers of having different opinions on software licensing. The incontrovertible and widely-reported fact that actually-free licenses are being chosen at a higher rate than GPL facism is completely wrong and bad and we'll all be sorry and it's not really happening anyway so everything is fine. The subtext is that anyone who doesn't like or use the GPL is an ignorant asshole destined for an ignominious fate. It is not clear whether such a fate is preferable to being associated with the FSF in any way.
The leaders of several bureaucracies are here to tell you why you need bureaucrats.
Nothing in this talk matters until well into 2018.
However the fuck we want, Deb.
A post-mortem on the dangers of doing any business whatsoever with a multinational corporation. This guy suggests getting a lawyer; it's probably better just to delete every email from anyone working on any Apache project.
...but it absolutely will be if you use Kubernetes, which can't do ipv6 right, still hasn't nailed down nebulous concepts like 'storage', and is so otherwise full of warts that I think it might actually be the software version of herpes.
Also could be titled "deliberately sticking your hand into the oven."
Google's in-house REST replacement for HTTP2. Useless to adults.
Kubernetes is so great at orchestration they require third-party software to actually orchestrate anything. This talk is about etcd, but they call it "an Operator" to avoid recognizing that literally nothing else even almost works.
A Canonical employee describes trying to shove UID hacks into the linux kernel.
"Get hired by Red Hat." Without that step, you'd first have to answer the question "why the fuck would you want to develop with Openshift?"
SPOILER ALERT: everyone loses.
"Cloud native" means "runs in Kubernetes." This talk describes how to design a Java program to run in Kubernetes, despite Kubernetes' promises that shit can Just Work with it, and also despite Java's "write once run anywhere" promises.
User-level containerization to make high-performance applications even harder to debug. Thanks, kid!
The linux people are rewriting cgroups, presumably because nobody is allowed to touch them except systemd.
You can save a lot of time by just Shooting Kubernetes, but the speaker would like to teach you how to manually capture a core dump from a running container and figure out which DNS problem is breaking your software.
Pointless containerization of virtualization, or pointless containerization of virtualization administration? The possibilities are endless, just like the stupidity of either of those plans!
Oh, they're serious about this Lua/webshit mashup. Gross.
Amazing both that there was a previous generation of the LuaRocks test suite and that they could trick some poor sod into wasting time improving it. They never did get it working correctly on Windows, but like all open-source projects, they just marked the problems "known issues" and ignored them forever.
You know those people who beg for attention on Reddit and Youtube by programming an arduino to flash some lights, glue the whole thing to a pair of shoes or something, and then declare themselves Makers?
If your scripting language is slow, shove some JIT into it! Enjoy this display, since nobody will ever look at any of this code again.
Fun fact: not all Lua programs will work on these ports. Success!
BEHOLD: OUR IRRELEVANCY VORTEX
Write a whole bunch of software, then use it to implement OS primitives. Not sure why this talk exists, or who will attend, but then again that sentiment is the theme of this entire track!
Paper-writing factory for European grad students.
Someone describes stack-and-register fuckery which is of use to nobody and does not work on currently-available computers.
The programming language that Jesus Christ personally delivered unto Man is being used to write an operating system with all the sharp edges filed off so stupid users can't hurt themselves. Also impossible with this kernel: everything else.
Hooray, now we can just use the linux kernel and run everything there! Seems like the next logical step is just installing linux and using that instead -- that's what the Genode developers do, after all.
The guy who wrote the TCP/IP stack for MINIX finally admits it's shit.
The Genode developers edge closer to Plan-9-style namespaces, but stop well short of implementing anything useful.
Porting virtualization tools to a virtual operating system seems crazy. Hurd has virtually existed for many years now and it virtually works fine.
The attached report is, hilariously, hosted on Google Drive. It is also utterly useless in every way, because it was poorly designed to ask the wrong questions of the wrong people. At least the presentation is short.
Mozilla has successfully crowdsourced localization work, and would now like people to handle development builds for free as well.
The talk description starts out "No, Firefox DevTools are not Firebug" without stopping to apologize for that fact or explain why we got such a shitty alternative built in instead.
Mozilla is also outsourcing the extension API. To their biggest competitor.
Sales pitch for Selenium with apologetics from the guy who wrote the plugin to make it work with Firefox.
Poorly, slowly, and loudly. The dashboard in question is not written in Rust.
Homeboy shows up to explain why he wrote a Rust development tool in Haskell. SPOILER ALERT: it's because Rust fucking sucks.
Nobody uses any of this shit.
Covered earlier, nothing added: shitty not-VR implemented in a browser.
Of course it's not actually important to care about users, regardless of platform -- nonetheless, these losers have created an entire infrastructure around their desire to bitch that some website didn't work on their browser. They've truly leveraged a tremendous amount of ingenuity to apply industrial processes and tooling to the art of being ignored by webmasters.
Nobody really gives a shit, until the day inevitably comes that Google decides to shove QUIC down everyone's throat.
Some nobody shat out some PHP to implement a buggy and insecure OAuth2 setup after Mozilla left him out to dry.
Ruby webforum shit. Disregard.
Mozilla finds it crucial to support women in technology, as long as they're not queers.
Mozilla's corporate political lobbying efforts.
Dropbox goes into great detail describing all the ridiculous lengths they go to -- including rebuilding MySQL with specific compilers -- all in an attempt to avoid learning to program computers properly.
Be like Dropbox! Profile MySQL so you can find out exactly which part of your shitty software should have been written by someone better than you.
Github, a company famous for capricious and unexplained downtime, would like you to run the software they wrote to molest your database on a fundamental level.
This talk contains the term 'split-brain,' which 100% of the time is an indicator that the person speaking has no fucking clue how anything works. Surprise: the speaker has no fucking clue how anything works.
Setting up notifications to go off if someone accesses a 'honey pot table' has to be the dumbest shit I've ever heard from the database world. What kind of bush-league amateur would ... oh, it's a MySQL employee. Imagine that.
The author helped found the MariaDB fork. I wonder which of the above he prefers?
Bash, Docker, Ansible, ProxySQL, MHA, and MySQL! Painless! Next talk: carving a pocket watch out of granite with your dick! Painless!
This is literally a regex substitution on your sql statements.
Sweet, more performance advice from people stupid enough to use SQL in performance-critical roles. I'll never get sick of this!
Hilariously, the methodology seems to be "take the data out of MySQL immediately, use literally anything else during computation, and then put your results back into MySQL."
Absolutely nobody uses this shit, because it doesn't work reliably. I can't tell if this is a series of excuses or an apology.
This is on the same topic, except it was written in a parallel dimension where the software works and is something to be proud of.
A master class on workarounds to your shitty RDBMS implementation, presented by a representative of the company that would like to charge you money for shitty RDBMS implementations.
RocksDB is a database engine for people who write a lot of data and then never read it (i.e. idiots). It is trivially smaller than InnoDB, slightly faster than InnoDB, and nowhere near as mature as InnoDB. Switch today!
MyRocks, being a shitty implementation of a bad database for a terrible RDBMS, is of course something Facebook badly needed to integrate into their platform immediately. The speaker brags about achieving during the transition the kind of stability that would be expected as a baseline from any other RDBMS.
In case you thought that LibreOffice developers were sad nerds with wannabe-enterprisey delusions, overcome with a desire to do literally anything to seem interesting to healthier human beings: you are correct.
More interface bikeshedding because people have grown used to MS Office's 'ribbon' thing and now get angry when it's not in LibreOffice.
Because there are so many buttons that it's hard to find the buttons, they're moving the buttons around again, but this time with fewer rules.
LibreOffice ports everything to Glade, which nobody has used for ten years.
A developer explains the phenomenal amount of work that goes into making sure LibreOffice slightly fucks up all your formatting at every opportunity.
As soon as someone decides to use LibreOffice webshit, it will immediately explode, but this guy's ready to figure out why.
Talk description not found, exactly like the point of extensions in LibreOffice.
Probably the absolute easiest part of any word processor, now available as a documentary.
If you spend your life writing Excel macros in an Excel clone, you might care about this crap. The rest of us can merely pause to enjoy the irony that someone went to all this work so they could use a text editor that didn't come with their word processor.
Why spend your time developing software when you can spend your time developing software development software?
Even the speaker found this talk too dull to describe.
We live in a world where people have to work on performance optimizations for word processors. This is the same industry developing autonomous cars. Good luck, world.
Again: your 2GHz processor is not fast enough to run this word processor. It needs more than one.
The only thing better than a webshit word processor nobody uses is a webshit word processor nobody uses EVERYWHERE!
If your office software needs a 'safe mode,' perhaps you should take a moment to re-examine your software choices. Perhaps you should immediately stop using that office software forever, and use something that won't destroy your computer instead?
Well, thank god for this! I'm tired of having to figure out which part of my office software is crashing my computer. If it's all one importer, it'll be much easier to assign CVEs to all of the buffer overruns.
Talk description not found, much like my interest in this entire devroom.
Here's our subtle warning: LibreOffice is going to start trying to track its users, in the name of harassing them into installing updates, whether the user cares or not. And they're getting the tech from Mozilla, who know user-tracking better than any other open-source project! Hooray?
In the spirit of fairness, LibreOffice is also tracking its developers. This is much more efficient than treating them like humans.
This is the same talk description, given by the same speaker, with a different title. They really take this OpenOffice/LibreOffice split to heart.
Sure, FOSDEM already has a lightning talks track. But fuck that NIH shit! We'll do it ourselves!
Resume-building fluff from a non-participant.
"Rapid prototyping" is only required when you forgot to have a design stage in your software.
An idiot still thinks people want to interact with computers by waving their hands around. He offers to teach you how to make things like that.
Artists working hard to raise awareness about their art projects.
Someone hooked up a motion sensor to a Raspberry-Pi-like light-flashing platform.
Yet another "game console" that consists of shitty hardware connected to shitty software, with its shittiness excused by being called "retro." No one is fooled; this just sucks.
Yes, this idiot would like you to electrocute yourself as a form of entertainment. No, I don't suggest you participate. Yes, I'm positive this talk will be overflowing with people eager to electrocute themselves. Yes, they are idiots.
That guy from before who wanted you to GPL your makefiles also wants you to GPL your pictures and sound.
This kind of shit is pretty revolting. Of course increased diversity is much to be desired, but setting up some kind of god damn money pit is not the way to do it. Rather than reaching into the industry and working with established parties to further their mission, they're striking out on their own and maximizing their buzzword potential. This enhances the grant flow, but it lessens the impact, and silos their work.
See, this is what I mean. These people had an idea and instead of founding a fucking Institute and spreading out their Impact Factor, they hired some fucking game developers and made a goddamn game. Take notes, previous talk.
That's great, but it doesn't appear to be open source. What the hell is it doing *here*?
Entry #5,193 in Educational Games Hell. No source code available. Open what, exactly?
Webshit to clone Scratch.
Brand-new project, unknown if it will even survive -- but props for using mercurial instead of fucking github.
Yeah, *that* Mono. We're still pretending C# is open source, I guess.
Just what we needed! Yet another finicky runtime that probably doesn't work on anything but Ubuntu!
HELLO WE DO NOT SOFTWARE! LOOK AT PICTURE PLEASE
This is a sales talk.
It would be nice if anyone relevant used this, but they don't, so this talk is mostly worthless, except for the link to the European Broadcasting Union's Timed Titles project, which they have named EBU-TT. Hahah.
No mention of the horrible security vulnerabilities that keep popping up as a result of gstreamer's whorish acceptance of any code spurted at its repository.
More wannabe-VR crap you can safely ignore forever.
This talk will be standing room only; every single broadcast professional in Europe will be here.
"Poorly," says MediaArea founder and technical lead Jérôme Martinez, "and with very low accuracy."
A hex editor on steroids.
A VP9 fork continues to be irrelevant.
Here's the traditional "oh my god what about jitter" crew, back to show you how they use RTP to stream audio.
Same shit but for switches and knobs.
Shame, retreat, denial.
SSL and Kerberos, just like literally everything else. NEXT
The above question, while trivial and uninteresting to most people, is actually just a representative sample of several similarly trivial and uninteresting questions which will be addressed in this trivial and uninteresting talk.
"Important information for anyone running replication or backup." Not important enough to link to anything. Just come to the talk.
This is simple. In the beginning, there was no fault tolerance. After many years and a lot of hard work, developers successfully implemented a new kind of fault tolerance: none. Today, there is still no fault tolerance, but there's a shitload of code you can shove in front of, behind, and all around PostgreSQL to fake it.
When all you have is a PostgreSQL, all your data looks like nails. Nails fired into a toilet door with a broken nailgun by a drunken idiot. But hey, at least you only have to learn one database!
Complaining that the PostgreSQL documentation is just too darn thorough, the speaker offers to helpfully obfuscate the topic and introduce uncertainty where there was none before. With luck, he will also leave out important details, hand out rank misinformation, and then tell the audience they should have just read the documentation.
GNU Radio, bored sitting alone in a corner totally unused, decides to experiment with machine learning. Nothing happens.
Everyone in this room is old.
This is a sales pitch for the author's embarassing garbage, which requires a version of Ruby newer than that installed on most computers in the world, and doesn't seem to actually do much for the user.
Ruby has had five different garbage collectors in the past few years. None of them cured it of its fatal flaw (being Ruby) and none of them will make this talk interesting at all to anyone.
The Rust Evangelism Strike Force deploys shock troops deep into Ruby territory, confusing the natives with offers of Rust-base API translation.
I don't know what kind of psychopath deliberately chooses to interact with Java, Ruby, and Rails at the same time, but it's probably not a coincidence that the speaker's name is listed as "Nutter."
Who cares? I can't even be bothered to read far enough into the talk description to see if he defines any of these ridiculous names.
Any challenge involving Ruby is easily surmountable: stop fucking using Ruby. There was never a reason to use Ruby; you're only doing it because you're an idiot. Stop it.
The trick reported by the speaker is "stop using Ruby." He says things get way faster when you use Java instead! Good news: actual scientists do not use this garbage anyway.
The speaker develops software for his light-flashing platform that will flash lights if his house catches fire.
A sales talk for the speaker's university grant-funding cash cow.
Reverse engineering hardware support that nobody asked for.
Intel. Presumably so their Omnipath shit doesn't suffer from Aries Shitty TCP Syndrome.
Intel again! Nobody else actually gives a shit about DPDK.
At this point in the devroom, corporate control is handed over from Intel to Cisco, who proceeds to talk about shit only Cisco cares about.
This is Go shit that smells especially Googly, even though the speaker's name is not divulged and Google seems curiously hands-off on the project itself. It probably violates some ridiculous patent that Google doesn't want to fight about.
An especially shitty idea: keeping all network packet state in an in-kernel bytecode program loaded from userspace. What could possibly go wrong?
Java shit for software-defined networking! Next year's talk will be a smashing success: "Getting the Fuck Away from OpenDaylight"
BGP is such a simple protocol that I can't imagine the fuckwittery required to deliberately involve a gigantic Java mess.
Shoving like four different organizational platforms into Openstack and hoping the Jenkins install doesn't fall over: the talk
Network analytics software sales pitch.
Ansible, a particularly bad configuration management tool, is singularly ill-suited for managing software-defined networks. Here's how to do it anyway!
Yeah, turns out Cisco has been doing this shit a long time. Not sure what this has to do with open source, but whatever.
This introductory talk is twice as long as any other introductory talk. I don't know why.
Why is DARPA hosting a hackfest? Why are they doing it in Brussels? What the fuck is going on over there?
I didn't know copyright law was one of the four fundamental forces of the universe. I guess the other three are emacs, C++, and fatness.
This goddamn talk description keeps saying 'ham' and 'ham blend' and now I want a sandwich
Yep, another GNU Radio sales pitch.
All three challenges are GNU Radio. I hope one day it can be overcome.
Actual technical content at a technical conference? The speaker will be ridden out of town on a rail and forced to live in exile.
Not the first FOSDEM talk with a corporate logo attached to it, but definitely the most useless FOSDEM talk with a corporate logo attached to it.
Interestingly, the entire content of this talk is just pasted into the talk description, complete with bibliography. Less interesting: the topic of the talk.
This dude is honestly concerned about the security of temporary traffic lights.
Some nerds use a light-flashing computer to flash radio signals.
A grad student is paid to develop software to detect signals. He forgets to mention why.
Same grant-money funnel as described in the opening talk of the SDN devroom.
You know how you can triangulate on a signal source if you take readings from a few different locations? This guy likes that plan, but he wants it to be way, way more complicated, because he is an academic.
Find a clock signal, receive it, process out original signal, calculate time of flight: measure altitude of ionosphere. Simple, yes, but it's also a hell of a lot more work than it needs to be!
Memcheck: the quickest way to feel like you're helping without having to read anyone's code.
This appears to be a 'how to use Valgrind' talk but the description is too incoherent to be sure.
Simple! Turn on FORTIFY_SOURCE and throw Valgrind in the trash. Done!
Who is this talk for? People interested in currently-ongoing performance ricing which will probably be upstreamed and forgotten about next month? Who sees this and says "hell yeah, lock me in a room with THAT for half an hour!"
After a couple-hour break from his Memcheck retrospective, this speaker returns to demand the audience tell him what to do to fix his old code.
In an awkward turn of events, this speaker describes the ways in which he forked and fixed the previous speaker's old code. I predict angry staring from the back of the room.
Red Hat says: keep shitting out that free code, citizens! We're counting on you!
A Red Hat employee pretends anyone wants to use Openshift.
Whatever the fuck VDEplug is, it can use different kinds of virtual networks now.
Citrix wants you to know that you can 'live patch' Xen. Why that is preferable to just moving the Xen instance to another host, patching, and moving it back is not explained.
Another Red Hat employee shows up, this time pretending anyone wants to use oVirt garbage.
More Red Hat, doing more stupid shit like trying to crossbreed oVirt and Kubernetes.
The Red Hat festival continues, this time trying to make QEMU development less scary to unpaid open source developers so that Red Hat doesn't have to pay for so much open source development.
NBD is an NIH alternative to AoE or iSCSI. The project maintainer is here to shill for it.
A "unikernel" is when you have just enough shit in your OS image to run your program. This is obviously completely retarded, which is presumably why this guy (who founded a business hosting such programs) has to go to conferences and give talks about how to do it. Left to their own devices, it's unlikely anyone would want to try.
Lago is to oVirt what Vagrant is to VirtualBox. Yep. Red Hat.
This Guy's Career: a Red Hat Retrospective
The Red Hat Tribute to oVirt continues, this time about solving bad systems administration with software! How refreshing!
A Red Hat engineer discusses how nonvolatile memory enables them to do all kinds of wonderful new shit that they could have been doing all along with regular hard drives.
A little after 6 pm, in a shocking twist, an Ubuntu developer wrests control away from the Red Hat mafia! Not only does the heretic not talk about oVirt, he suggests using Openstack to run containers! He is quickly escorted out in protective custody by Red Hat security guards.
All that distraction allows another interloper to grab the mic: a Xen developer! And the bastard works on FreeBSD, which Red Hat doesn't even sell! SEIZE HIM
FOSDEM 2016: more boring shit (2015/12/01)
Let's take a look at my annotated copy of the FOSDEM 2016 main talk schedule, shall we?
Probably a lie.
Places we're NOT taking it include "Interoperability City," "Lake Reliability," and "The People's Republic of Sanity".
Alternate subtitle: "we didn't give a shit about any of these people before we saw them on the news, but look what we can accomplish when we need to feel righteous about our shitty hobbies"
I think the printer's jammed, we need you guys back at the office
You know that huge pile of bad software the "devops" people wrote so that they wouldn't have to ever actually install their software? This guy wants to make that the norm. Everywhere.
The speaker confuses reproducible builds with package signing, then makes a ton of excuses about why his talk summary is meaningless noise.
Except it's not really "a linux session" as much as it is "onboard cpanel shit for people who should not be trusted with computers." Naturally, the strongest evidence that this is a mistake is its inclusion in Fedora by default.
A GNOME developer talking about security.
This isn't "open source" in the "use this how you want" sense. This is the "help us do work for no money" sense.
Every goddamn thing, to the author's surprise. Formalization of the process doesn't seem to improve anything, it's demonstrably irrelevant which formalized process your team adopts, and the author is "trying to make code review work as expected with my research," which will probably be too dull to be hilarious.
Setting aside the fact that "machine learning" has yet to prove worthwhile in basically any current implementation, and the fact that "big data" is almost always code for "shitty programmers who ran out of memory," this abhorrent pile of dogshit is intended to strap together a bunch of computers scalemp-style in an attempt to make single-threaded hobby projects relevant. Absolutely disgusting.
Everyone's favorite also-ran clustered filesystem continues thrashing around in irrelevancy, but at least they're spending a lot of time rewriting existing functionality.
This one came as something of a surprise, since I've been using linux with multi-petabyte filesystems for many years. Unfortunately, the actual speech seems to be based around a version of Plato's allegory of the cave, but instead of shadows on the wall the speaker sees webshit technology and assumes there is no other reality. Because webshit runs in userspace, POSIX is irrelevant and filesystem i/o should be moved out of the kernel. Yes, the author works on Gluster.
The speaker assumes that since FPGAs are getting cheap, all the problems will go away. The fact that all available FPGA development kits are either ruinously expensive or iredeemably shitty does not appear to factor into this.
Another entry in the "pretentious three-part FOSDEM title" history books. Seems like it could be an interesting talk if it were just about FM radio and less about GNU/crap.
"so much industry momentum" that nobody gives a shit at all! Incedentally, if you've noticed how bad your haswell video support is on linux, this speaker is part of the reason why. Cheers!
Seems like a risky move putting the entire one-hour presentation directly into the talk title, but maybe it'll work out. We'll have to spend less time getting excited that they've finally got around to supporting laptops made a mere six years ago, or that post-Ivy Bridge laptops can never be supported. At least you can install this on your servers, thereby destroying all the useful remote operations shit that might have otherwise made your life easier!
The answer to this has always been "yes, as long as you're willing to sign an NDA before getting audit access to the code." The talk, predictably, has nothing to do with this question, instead focusing on handwaving an open-source software that everybody already uses.
Bonus answer: why the fuck would anyone pay for the infrastructure under these circumstances
Facebook figured out InnoDB sucks. Unfortunately, they're so rich that writing another database backend seemed like a good idea. Oh well.
No, the speaker has never designed a kernel API. He just writes books about them, and asks for money in exchange for "training" about them. I really don't understand why there is a talk here aimed at kernel developers, since they already have a mailing list, and their own conferences, and firmly-entrenched opinions on the topic.
some webshit thinks you can solve social problems with git
Ah, this must be the sort of "momentum" that Vulkan experiences. In practice, an ODF document means one thing: the user downloaded OpenOffice and forgot to save in Word format. Good news, though! A quick email should be enough to get them to re-save it and send it to you.
In which it is revealed that their fake version of Office365 is wildly resource-hungry and subject to crippling security problems. At the end of the presentation, everyone who is surprised by this information will be given a Mac and a job at Facebook.
The Don Quixote of FOSDEM 2016. This speaker is attempting to speak to conference attendees about one of the lost arts of the internet: email that is not hosted by Google. He tries to make it interesting for the typical FOSDEM attendee by tossing terms like "http" and "json" into his slide deck, but then he ruins it all with technical discussion of hash algorithms and abstract syntax trees. If he just renamed the project "spam.ly" and gave the talk wearing jeans and a sport coat, Paul Graham would give him eighteen million dollars.
Last year I made fun of the systemtap team for their desperate attempts to seem relevant. This year they're back, and it's even worse. In 2015 there was a linux kernel vulnerability that had to do with a specific variable being signed instead of unsigned. Our team patched the kernel and redeployed and we were fine within hours. Red Hat's official "Band-aid" for several weeks was to use systemtap to bitwise-AND the value with INT_MAX, during runtime, on every boot of every computer. I tell this story to systems administrators as an example of how shitty Red Hat security responses are. Now they are giving a talk about how wonderful it is.
The second theoretical talk at FOSDEM: the speaker does not work in the telco field. Presumably this talk is a sales pitch for all those telecom executives at FOSDEM. Tons of those.
An introductory howto for people who actually want to know how to use the features their operating systems provide. Presumably scheduled after lunch on Sunday to give the engorged web crowd somewhere to sleep. I suppose it's possible that this talk is presented as a curiosity, but then it would have been titled "What people do in a real conference" or something.
In this talk, the speaker brags about Kubernetes fault tolerance by unplugging several shitty computers from a cluster of shitty computers. It is not mentioned that Kubernetes is sufficient for this task: it will generate its own faults regardless of what hardware you run it on.
Amit Shah finally gets around to explaining why QEMU live migration used to work fine and now it completely sucks.