It is interesting to me that ULTRIX 4.5 -- from 1995 -- still has a bunch of executable binaries in /etc. It also has no /sbin, and most of what I'd expect to be in /sbin is in /etc, much like Unix v6/v7.
I remember that Linux machines had MAKEDEV and a couple of other scripts in /etc in that time frame, but I think by 1995 /etc on Linux was almost exclusively data.
An aperture card! Used as a very early image tagging/sorting system, you'd have microfilm mounted in a punch card, so you could have a regular punch card organizer/sorter find your images. The DoD loved them because they could store tons and tons of engineering blueprints in an organized, computer-sortable and -searchable way.
Clearly WD has top notch QC in place.
17 days in, maybe I should have used more than one core?
This surely looks extremely dangerous. #electronics
@freakazoid @jens @jauntywunderkind420 @rubenerd We need to remember that TCP is an attempt to emulate a serial byte-pipe over an unreliable network; that is, it's whole reason for existence is to preserve the semantics of the leased lines that the Internet was first built upon.
UDP was an encapsulation that basically exposed raw IP to clients, with the addition of a port to allow for application multiplexing over a single network cable.
It seems to me, as with people who saw fit to route SSH over HTTP (not even joking) because some network operators are too lazy to manage ports on behalf of their users, we have yet another network-stack-on-a-network-stack solution with the datagram extension. You're just re-inventing UDP ... on top of (ultimately) TCP.
You're not alone, though; the BEEP (nee BXXP) protocol attempted to do the same exact thing. It was summarily dismissed by the community as needlessly redundant.
When will people just bite the bullet and tell network administrators, "Look, you have to manage ports. Do your F'in job."
Protocols like QUIC and BEEP are solving non-problems; or, more precisely, problems of their own creation. I wholeheartedly believe that it's actually easier to push and adopt protocols like QUIC and BEEP than it is to fix the root cause. But, it only offers asymptotic levels of benefits. Eventually, it will cease to be economical to do so. But, by then, it will be too late. Infrastructure will have become immutable by then.
In my opinion, it is our duty as technologists, programmers, and systems architects to push back on this unchecked re-encapsulation HARD.
A review and close look at the LEGO Ideas 21327 Typewriter. Looks amazing.
"I appreciate SDF but it's a general-purpose server and the name doesn't make it obvious that it's about art." - Eugen Rochko