What the duck Microsoft bullshit is this?
There is no concept of locked files in extfs, much less inside the kernel. Resource locks and unkillable processes is some windows bullshit that no sane operating system would touch with a ten foot pole.
What the duck Microsoft bullshit is this?
There is no concept of locked files in extfs, much less inside the kernel. Resource locks and unkillable processes is some windows bullshit that no sane operating system would touch with a ten foot pole.
There’s a fine line between disrespecting a fallible opinion and disrespecting the person. In writing, it’s easy to cross that line. It’s ok to disagree with people, but it’s important to sometimes take the step back and remember that the person is larger than any singular decision. There’s likely context you’re missing that lead them to that decision.
The biggest thing to remember is that more likely than not, if you really and truly fuck up your job, chances are the worst you do is create extra work for your team. They probably won’t even be in danger of losing their jobs if you truly screw up. It’s not likely that people will die. The blast radius of most software engineering jobs is incredibly small.
Good luck with it!
It is “fun”, and there are lessons that Ruby has that should be taken elsewhere, like the principle of least surprise. The most important one is MINASWAN: Matz is nice and so we are nice.
Matz described Ruby as Lisp with C semantics, and Perl convenience.
There’s a reason the CLI flags can put it in “perl mode”
The actual SEC report is relatively short - and surprisingly accessible.
Nope. Bitkeeper used it in the master-slave pairing and the term was carried forward. Gitlab did a whole writeup about it.
$previous_job allowed us to pick. One of my coworkers had to replace his laptop, and I convinced him to try out Linux this time. I handed him the bootstrap script and he was back to working by the afternoon.
Our CEO got wind of this and said as a matter of policy everyone is switching to Linux unless they have a good reason (needing excel for financial reports is a good reason). The two new hires who had been setting up their dev environment for over a week at that point were the trigger for this.
TCP Selective Ack is very much a thing, but it does take extra memory so lots of TCP stacks exclude it or disable it by default.
TCP was never designed with wifi in mind. TCP retransmission was only ever meant to handle drops due to congestion, not lossy links.
Tmux is a wonderful complement to mosh. Together you get persistence even when your local client loses power (speaking from experience)
I worked with mosh for years to connect to servers on other continents. It was impossible to work otherwise. It only has two small warts: forwarding, and jump hosts.
The second is fixable/ish with an overlay network, but that isn’t always an option if you don’t control the network. I tried to solve this with socat but wasn’t able to configure it correctly - something about the socket reuse flag was very unhappy.
First turned on in 2010.
Git-svn was always the blessed path for converting. GitHub supporting svn was more about getting heterogeneous orgs to buy enterprise subscriptions.
Also that in order to exploit this it requires an active man in the middle. Which requires any of the following:
Almost all of those have decent mitigations like 801.x and BGP monitoring. The best mitigation is that you can just change your client config to disable those ciphersuites though.
I was not expecting the drama around it. Is the issue truly a different orthography or is more like a different font/ligature issue?
EDIT: forgot the article I found on it: https://restofworld.org/2021/tulu-unicode-script/
Locks are only held during system calls. Process termination is handled on the system call boundary.
You’re projecting windows kernel insanity where it doesn’t belong.