pilif

It's not just about two independent panes. It's about two panes in a single context. With this, there can be shortcuts that act in files and which now don't just have a source (the selected files in the active pane) but also a destination (the inactive pane), so copying, symlinking and moving all can be done with a single key without the need to provide a destination.

Two explorer windows are independent of each other. Any move or copy requires two keyboard actions or the use of the mouse

Looks like a different version of WSL2 with a newer kernel and newer VM host software

The Ultra 2 was sold to me as the watch that can do Siri without network connectivity. That was very useful to me to leave reminders for stuff I think about while on a run without the phone.

3 years later, Apple releases the OS 27 versions and Siri got improved to the point where it can’t work on-device any more on said Ultra 2

So to get the feature back which I paid for and which was specifically used to sell me the device and which now has been taken away from me, I just have to… buy the next hardware generation?

And how long will that last before they decide to remove the feature again?

Thankfully, the Ultra 4 provides some other compelling features making this less painful, but I would have expected at least a toggle to keep old Siri on the old watch. New Siri over the network is impossibly slow and buggy (I made the mistake of installing the betas, so there’s no way back now)

Once this was announced a few years back, I was very skeptical that a few dots on the screen would be all that's needed to help me.

And yet, apparently it was.

It's not perfect, but the dots help me to go from instantly feeling sick to lasting 30-40 minutes before I have to stop looking at the screen.

That’s not true for Postgres however: due to its usage of a shared memory pool, whenever a subprocess is terminated unexpectedly, Postgres will kill all other processes and enter recovery mode, replaying the WAL, during which time it will not accept connection requests.

It does this because it can’t possibly know whether the dying process did bad things to the shared memory pool.

it's also very tricky to do given the current architecture on the server side where one single-threaded process handles the connection and uses (for all intents and purposes) sync io.

In such a scenario, listening (and acting) on cancellation requests on the same connection becomes very hard, so fixing this goes way beyond "just".

Firefox is not in a position where it is the only browser allowed to run on a platform.

On iOS, you’re either doing a native app, sharing 30% of your income with Apple, or you’re restricted to Safari’s feature set. No browser in iOS can use anything but WebKit

On the other hand, by erroneously treating a SHOULD as a MUST, I would say that Google is the one who's not RFC-compliant

It’s open source and runs on all kinds of platforms. Original HL1 runs on old Windows and IIRC DOS. Nowhere else

I wouldn’t post some random vulnerability report, but the disclosure timeline at the end was very interesting to me and not putting the vendor in a great light