It would seem your baseless assumptions are wrong.

Sure, that was 3 years ago. But you seem to be ignoring the following facts:

• AMD's proprietary driver shares most of its code between Windows and Linux.
• NVIDIA's proprietary driver shares code between Windows, Linux, Solaris, OSX and assorted BSDs.
• The majority of a device driver has to do with the device itself, and all kernels are trying to get the device to do the same thing and so will expect similar hooks. It is annoying to support multiple targets, but certainly not difficult.

You have failed to understand the analogy.
Deutsche Bank = Microsoft
Branch = servers in Ireland
New York Times = EU citizens using servers in EU
 

I appreciate the criticism. Did another crack at research, managed to find an xorg.conf that (supposedly) does multiple sessions with one graphics card with the Nvidia driver. I think I'll try to redo the setup this weekend, if I can find the time.

I wish all my side projects were scrutinized by persons as knowledgeable as yourself. Seriously, thanks.

I setup such a system a year and a half ago (4 stations off one computer, totaling > 10mpx of desktop space).

Allowing me to be abundantly clear: I wish Wayland was production ready a year ago. Multiseat-X sucks for these reasons:

• You either need a separate graphics card per session, or use framebuffered Xephyr windows to run your desktop (I chose the latter, as performance wasn't a concern, but adding more graphics cards was)
• Dynamically assigning keyboards/mice to different screens based on a common USB host... sucks. SystemD manages sessions and integrates with udev so a hotplugged USB devices can be limited to it's associated session. Seriously awesome.
• The above is doubly a problem as when a usb peripheral disconnects momentarily, an X session forgets about it and never sees it again... Fixed by following what was happening with GDB, putting together an awful hack, and nuking X updates from the package system.
• All users sessions run on the same virtual terminal. I'd often ssh into the system to maintain it rather than my usual ctrl+alt+F1
• Documentation is erratic and frequently wrong.
• What I have works, but I would not deploy it to someone else's work environment

All that badness aside, I do appreciate only having to maintain one machine for software updates etc, not having to setup network shares, or worry about users logging in with NFS + LDAP. It made it quite easy for 4 people to share a powerful machine that they only needed the speed from on occasion, and collaboration was super simple.

I look forward to rebuilding this setup with pure SystemD + Wayland in the next 18-24 months.

I apologize, I didn't make myself clear enough to justify that paradox, that apparent contradiction. My point about responsibility being a zero sum game was more about how even though A did actions that enabled B to do C does not mean B is not wholly responsible for C (Though A may also be partially, or wholly responsible. Or not).

Let me give you some other examples of non-zero sum responsibility. One can:

• Prevent theft by being poor.
• Never be cheated on by always being single.
• Not die in a plane crash by never flying.
• Avoid criticism by never doing anything noteworthy.

  • In the first three situations it would be common to blame the thief, the unfaithful, and the airline. But one can't deny the theoretical control the first party has. Just because one can be responsible for an end result (here, rubbing people the wrong way) does not imply one is accountable for the negative consequences of same (reciprocation with hateful, and harmful actions).

    Really I just found how the structure of this conversation parallels conversations I've been involved in recently regarding gender and IT, and was trying to explore the "Maybe sexism would be a step in the right direction," angle, as it at least implies acceptance of an external group of people. I think often people of our ilk tend to be dismissive and derogatory toward those who reach a different conclusion based on values we disagree with, such as the creators of systemd and their opponents who clearly value different things at this point, to the point where they can't even agree on what those different things are).

So, victim blaming? I see your point, but the abuse he describes experiencing (as a young white male) reminds me of a trendy workplace topic. Maybe women aren't in IT because of sexist assholes, but because of unqualified assholes.

I agree, he has some power to control the response he gets from the community. But responsibility isn't a zero-sum game, and should he have to?

I really doubt SpaceX is going to stop work on a vehicle they were developing before they were awarded the contract.

Boeing, on the other hand...

That's what __attribute__((cleanup(function))) is for. Thank you, GNU C extensions!

Actually, I'm Canadian.

My point was that regional blocking of Canadians by .com's predated the updated copyright legislation, they are entirely separate issues. damien_kane below provides a nice explanation why.

As an aside, the updates to the Copyright Act were to satisfy obligations to WIPO treaties more than a decade old. And it beat the hell out of C-60 that the liberals proposed in 2005, which was just "Let's add the bad parts of the DMCA to Canadian copyright law and call it a day." (Don't get me wrong, if I had my druthers I would have put Charlie Angus (NDP) in charge of this bill, but that's not going to happen any time soon. Lazy youth voters, grr).

Those limits were imposed by modifications to the copyright act 2 years ago. What was there excuse before then?

The line immediately above:

EnvironmentFile=/etc/sysconfig/irqbalance

If you can't figure out that an "Environment FIle" might contain these things known as "Environement Variables," I question your claim to have been using Unix since 1978.

The environment file sets all the environment variables the service sees. It sees no more or less than what is in that file by default, which is both more secure and maintainable.

No, he said there exists N such that managing > N linux servers is easier than managing > N windows servers. Then he went on to compare them in single server environments, and discussed why windows is often used in small businesses as their only server.

As to installing proprietary software: honestly, my experience says that this tends to suck equally on all platforms. Had to maintain a small CS2 deployment years back, standard operating procedure when it fucked up was to reinstall the OS. I've never had to resort to such levels to clean reinstall proprietary software on linux, but I've also had difficulties with some proprietary software on Linux. I really think that tends to be a quality of the vendor thing more so than the underlying OS.

Uhm... cp has the exact same behaviour as systemd. Examples from my system:
: cp -ar bin bin2
: cp -ar bin /bin2
cp: cannot create directory ‘/bin2’: Permission denied
: echo $?
1
: systemctl start dmraid.service
: systemctl start imaginary.service
Failed to issue method call: Unit imaginary.service failed to load: No such file or directory.
: echo $?
6
:

So, systemctl behaves exactly the same as cp - error message and non-zero return on failure. Perhaps you were thinking of the verbose option for cp, which some people alias in permanently but is very non-unixy?

Obviously you've never had to switch out a library or framework before, or done anything cross-platform. The author talked about writing libraries that sit on top of the API, not simple thin wrappers.