It's like complaining that the biggest problem with books is that you can read them and/or (god forbid) copy them.

Since the dawn of times, the single thing that protects against this "biggest weakness" is the ability to inspect the copy/fork, and then, by using your brain, decide if the copy/fork is any good (inspect the commits).

In general, the best solution to show authority, is to selfhost the git repository on your own domain/url/server, and not use something like Github.

Finally a clear analysis.

Search for "crossing english channel" "world record", and you should find Rob and Morgan Wylie.

One problem is the mold for the case. It costs something like $2000 to make one. If you have only 10 people wanting that layout, it's $200 extra cost for that keyboard. Another problem was that the person producing these keyboards (yes, it's a one-man show) had to pick what to offer at least four years prior to having anything concrete. So if you embark on this mission without prior experience in manufacturing (abroad), you want to keep the offerings simpler at first. He started it around 2016, it wasn't until May 2020 that he got the quality of all parts to match or exceed IBM's original quality. He now has the experience to make more concrete offerings and possibly more variations because he knows what can be (easily) manufactured.

One of the problems here is reliability. These keyboards are built to last 40 years. Using cheaper technology for some keys will render those keys inoperable earlier than after those 40 years.

A neat trick which dampens the spring-ringing-sound is using Oral B Superfloss: just insert them in the center of the springs (30 to 60 minutes of work for the whole keyboard).

Try the Matias Quiet Pro keyboards. Best compromise between tactile keys and silence.

I'm typing on a new model F77 as we speak. Built as a tank, easily serviceable, touch like it was 1981.
The lack of function keys can be alleviated by using the so-called tap-dance QMK extension, it allows for triggering the function keys using either:
- Holding down the key for longer than 250ms will produce a function key.
- Double tapping the key produces the function key right away.

As for the backtick/tilde and escape, there are other solutions for that, as in (not necessarily exclusive):
- Ctrl key tapped produces ESC.
- Shift Esc, gives a tilde (unchanged muscle memory).
- Ctrl Esc, gives a backtick.

Incidentally, I mapped Caps-Lock to Ctrl, of course.

As for the numerical keypad, the F77 gives you a 3 x 5 keypad to play with. It usually allows for most incantations of numpad use.

With regard to next year's model, that is NOT going to be buckling spring (model F). If you still want the buckling spring, you should order right now. The window for new orders is likely to close soon (end of june perhaps), after that the manufacturing will close down permanently.

Next year's model is going to be a beam-spring keyboard. And yes, an option then is a full model-M like layout. However, the price is going to be higher. The buckling springs are for less than $500, the beam-spring keyboards are likely to cost around $1000 next year.

Procmail as such, as far as I can recall, didn't have any bugs that could be directly exploited through a specially crafted E-mail.
However, obviously people could create procmailrc files which acted on incoming E-mails, and yes, you can then create something which perhaps takes unvalidated information from the E-mail and then executes a command based on that. Actually creating procmailrc scripts that are vulnerable in this way is hard though (none of the examples do this).

Sendmail, OTOH, had an infamous remotely exploitable vulnerability which allowed for the successful distribution of the Morris worm.

Quite. I still vividly remember the days when every hacker worth their salt was releasing actual software on USENET (for others to compile), all the while when ESR focused on "compiling" and releasing The Hacker's Dictionary on USENET.

Real hackers compile code, not dictionaries.

I couldn't agree more.

It is long overdue that all social media platforms are forced to pick a side:
a. Either willfully mess with the order of content, with the (in)visibility of content (boost or suppress due to some vague machine-learning algorithm) and *accept* the fact that you are considered a publisher. And thus are held accountable for *everything* visible through your medium (just like all other publishers).
b. Or be protected as a medium, but then there cannot be any hidden algorithms, or reordering, or extra (in)visibility.

Actually, the corollary objective here is to minimise on-the-ground time of the plane. The plane earns money when in flight, so making sure that the boarding process goes faster will boost the airline's bottom line by a few percent because the airplane can shave a few minutes off of its time on the ground.

You mention "yet" as in "I know that it is scheduled to be open sourced" ?

Googling around I find OceanBase 2.2 to be the latest, perhaps. It is supposed to support Oracle/PostgreSQL dialects now too. But if it's not open source (anymore), there is no point in looking into it.

It says OceanBase is open source; but the copy on github is ancient.
Does anyone know if they went closed-source or where the actual current repository is?

This is close to my own solution, which is going frameworkless (no jquery, no react, etc.), but still use npm, but only pick packages that have *zero* dependencies (one way to find them is to look for the tag "zerodeps").

Going about it this way avoids dependency-hell and sort of guarantees a baseline quality level. Because, let's face it, 99% of all packages on npm are not worthy of the term "library".
Traditionally, a library has no dependencies, and involves *at least* a thousand lines of code (not counting empty lines or comments).