Humboldt University Berlin had officially licensed a UNIX System III as a research institute. They run it on a K1600 series Robotron system, which was a PDP-11 clone. I had a few sessions on it ca. 1985 as a 16-year old kid as a member of the Mathematical Student Society of Humboldt University.
I remember being challenged to learn about the file system. All I was told was, use the man command. I knew CP/M, or better the East-German clone SCP, but that OS didn't know directories. I had to learn the concept from the man pages. There were no UNIX books in libraries or book stores. But it was fun, I managed to write a simple compression program doing run-length encoding on that system.
Theoretically you can use channels to simulate a mutex, but I agree with you there are use cases where a mutex makes more sense. They are even used in the standard library, for instance to implement sync.Once.
But generally I would agree that if you need to code parallel execution, channels are a good way to do it, because you can avoid race conditions if you share data only over channels. The biggest problem is that a lot of people don't understand, that channels with a buffer larger than 1 are a sign of problems in the architecture.
There is a type of parallel programming with workers for specific functions, that always leads to performance issues. The problem is you need to right-guess the distribution of work, when you have to define the amount of workers for a specific function. At least one go routine for one request is a much better approach than function-specific workers.
The RST (Reset) is sent to inform the client that the data it sent was not read by the server. The RST avoids here the 4-way handshake for the TCP connection closure and the long wait times, if the client doesn't behave normal.
For the case here the server should call shutdown with SHUT_WR after sending the data and then drain the incoming data before closing the socket.
The interesting aspect of the Cloudflare support, which is not clarified, is how they came to the risk assessment that it is ok to roll out a change non-gradual globally without testing the procedure first. The only justification I can see is that the React/Next.js remote command execution vulnerabilities are actively exploited. But if this is the case they should say so.
This thing crashes on Ubuntu LTS 24.04 during start. Apparently all these agents are not able to ensure that a desktop app starts on a popular Linux distribution.
If Google has forgotten how to do Software, than the future doesn't look bright.
In my org nobody has admin rights with the exception of emergencies, but we are ending up with a directory full of Github workflows and nobody knows, which of them are currently supposed to work.
Nothing beats people knowing what they are doing and cleaning up behind them.
The statement "An interface type A is assignable to interface type B if A’s methods are a subset of B’s." is wrong. It is not in the current language specification of Go. The author misunderstood the term type set from the Go language specification. The type set of an interface is the set of types implementing the interface and not the set of methods of an interface. If you use the right meaning of type set, the subset makes sense again.
The density of black hole decrease by the inverse square of the mass of the black hole. That means massive black holes have a much lower density that small black holes. So they are more likely to form than small black holes. Dark matter will have played an important role in the creation of those early black holes. If there is no dark matter and some form of MOND theory of gravity is correct, the Schwarzschild formula will require a modification for large black holes. In that case galaxy centers will not require large masses to see the same effects.
There are two scenarios: First: Microsoft uses the JWT signing keys in memory and the attacker were able to get access to it by injecting code or get access to the memory image of such a process. Second: Microsoft actually uses HSMs but has to distribute the keys geographically and the attackers were able to get access to the key this way.
The first scenario is more likely, but you cannot exclude the second as well.
The comments are full of statements regarding security capabilities for passkeys. But there is no public specification that even defines requirements for the exchange of passkeys between devices. Google and Apple make statements on their websites regarding the security, but all of it is practically unverifiable. Please note that end-to-end encryption is useless, if you are not controlling all the endpoints.
Sites of course could use the device public key extension of the WebAuthn protocol, to rely on more than a private key copied intransparently between devices, but I wonder, who will even know about it and actually use it. Google has stated they support the extension, but I cannot find a statement by Apple. A question whether DevPubKey is supported by Apple is unanswered on the Apple Developer Forums.
"Der kluge Hans" is not directly a Brother Grimm title. There is a story "Der kluge Knecht", which translates into "The clever farmhand". The hero of the story is called in the story kluger Hans. There is another story called "Der gescheite Hans", which means exactly the same. Both stories are not fairy tales but droll stories about a simple-minded person.
At first I was on the fence too. I'm not use it all the time, but when I need it, it works as expected and it is much less a hassle than I remember writing C++ templates.
I have not seen a lot of comments that complained about the slower compile times. In my own experience it didn't really had an impact. But I agree the compiler should not become slower over time, so I appreciate the effort of the Go team to bring the compiler speed back.
I don't think that code readability is so much impacted. The square brackets work well. I find the angle brackets from C++ harder to read and there is the problem that >> is a token and cannot be used for two closing template angle brackets.
The increased complexity of the compiler is an issue, but cannot be avoided if you want to support Generics. But they took the time to make it right and as I stated it works for me.
I don't think that there is low adaption. Using type parameters visibly in a public API, breaks the API, which is the reason there are not a lot of uses in the standard library and with popular packages now. But this will change when maps and slices will be integrated in the standard library, which provide completely new APIs. Yesterday I found a library writing and reading parquet files, which used it quite extensively. But since I simply checked what libraries existed to assess how well the file format is supported, I cannot say much whether the use of type parameters by the library is useful.
I actually use my German ID card to communicate with the Elster service of the German tax offices. My old USB signing stick would need to be replaced next year, but using my ID card was the cheaper option.
The article contains a number of problematic statements. One is for instance the statement that Linux destroyed value. It may have for the proprietary UNIX vendors. But this should be balanced by the value Linux and OSS has created overall. Cloud Computing would never have scaled the way it has, if there would be operating system licenses and cloud computing providers would not be able to modify the OS according to their needs. I doubt that the profits of the proprietary UNIX firms over their lifetime are larger to what Amazon Web Services makes in a month.
OSS is basically exploiting a network effect that aggregates a lot of smaller contributions to a large one. Learning how to use a specific open-source software is the contribution the corporate free riders make.
The question is, how can the maintenance work be decentralized and distributed. Kubernetes changes the release manager for every version, so distributes the work over time. As long we rely on the single maintainer the results will not satisfy expectations. BTW a commercial company providing the support needs to solve the same problem.
It was not so easy back then as well. They had draft riots in New York City, which had to be suppressed by regular Union regiments. There were also deserters and bounty jumpers.
The South had trouble to support their armies. Lee basically planned the Gettysburg campaign to live of the land. Soldiers were happy because they had enough to eat. A lot of soldiers lacked shoes.
Somebody told me a few years back that the life time of a CISO in a larger organisation is not larger than 24 months. In my organisation that proved to be true so far. Here the rule applies as well.
People should have a look as k-sortable unique identifiers (KSUID). Binary they are represented by 20 bytes and their string representation has 27 characters, which is shorter than UUIDs since the use a base62 encoding. They are sortable since the 20 bytes start with a 32 bit UNIX timestamp followed by random 128 bits. They should be very efficient for clustered indexes / B+-Trees.
Note also that as long as you have a single central database you don't need UUIDs. They are only needed if you have several processes creating objects without coordination.
I remember being challenged to learn about the file system. All I was told was, use the man command. I knew CP/M, or better the East-German clone SCP, but that OS didn't know directories. I had to learn the concept from the man pages. There were no UNIX books in libraries or book stores. But it was fun, I managed to write a simple compression program doing run-length encoding on that system.