PROTON_DUMP_DEBUG_COMMANDS=1 - outputs a launch script you can use outside of Steam in /tmp

True. I knew I should have left that as “NFS 4” because someone would comment this. From what I’ve read (never used it), NFS 3 is very different to 4 and also just kind of not worth using, especially just for Windows, since it has no security at all.

Please just use Kerberos instead of fiddling with uids. It’s the only sane way to get NFS access controls and user mapping. Works on both Linux and macOS (but there’s no NFS on Windows anyway).

I’d say you can run the Kerberos KDC on the NAS but if Synology has some locked down special OS you’ll need another machine for that (edit: but you say you have other servers already so that shouldn’t be a problem).

Unfortunately SMB is so screwed that you can’t reuse ordinary Kerberos for authentication there, which is unfortunate if you want to have both that and NFS. I’ve yet to look into whether Samba AD can be used for both.

That would make sense if they were overlapping. They aren’t. There’s no need to “focus” the window.

Window focus is important for things like determining where keyboard input goes. If you want to type text into another window that isn’t focused, you need to switch focus before continuing to type so your text goes into the right window.

The need to focus on the window before clicking?

It doesn’t delay the click action for a double click because it already does it for a single click, so it would be pointless to do the same for a double click. If you’re double clicking, it’s pretty much always because you actually want to double click on something specific in the UI.

I don’t understand the question. All of them.

Skill issue.

No. It doesn’t. I’m beginning to think you’ve never used a Mac.

I use a Mac almost as much as I use Linux, which is almost daily, right now exclusively even since I’m not at home where my Linux computer is.

I’m beginning to think you’ve never used any computer since you don’t even know what window focus is for.

Well, theoretically yes. On a Mac, no.

Yes, even on a Mac. Necessarily so since it strictly places windows on one monitor. You’re always switching to a window on another monitor.

Can you give an example of what you’re talking about?

Why would I want to do that? Why does double-clicking suddenly remove that need?

So you can activate a window without first having to find a free space in the UI to click on (especially if it partially overlaps). It much increases the surface to click on to focus a window and therefore makes it faster since you can be more inaccurate in where you move the mouse.

What need?

No you can’t. It just minimizes them. Just like the yellow button.

It does not.

Can you give an example of a window that gets minimized by clicking the red button?

Like I said, sometimes you can, sometimes you can’t. Apple does not give any fucks about consistency or intuitive design.

No, drag and drop tile actions always work, even if it doesn’t entirely make sense (e.g. windows that can’t be resized).

Can you give an example of a window that it does not work with?

You have to click to switch monitors but if you do it twice it registers as a double click so you have to click…wait…then click again.

You don’t switch monitors, you switch windows. That is how it works for all windows. It’s like that so you can click anywhere in a window to focus it without activating something in the window by accident.

You can’t close anything from the window buttons and the red and yellow buttons do the same thing. You have to go into the taskbar and right click to close them.

You can close windows with the red window button, and the yellow button minimizes a window. Absolutely not the same thing. The whole application you can close via the dock, or the menu bar, or cmd+q. Two different things.

Some apps have a single main window though which will reopen when the dock icon is clicked (e.g. Mail), but that is still different to what the yellow minimize button does. The distinction is much more useful for document style apps like TextEdit which can have multiple windows (or none, if no file is open). There is also Hide which hides the entire application and all its windows until it’s activated again.

Then they took the time in Sequoia to add window tiling but it’s just such an awful experience. You have to hover over the green dot and wait for the prompt to popup and choose from a drop-down menu. WHY CAN’T YOU JUST DRAG AND DROP!?

You can absolutely drag and drop to tile windows, and there are also keyboard shortcuts for it. Check the Window -> Move & Resize menu for that.

On macOS, the correct thing to use is usually NSUserDefaults API, or - [NSFileManager URLsForDirectory:inDomains:] with NSApplicationSupportDirectory (gives a list of paths to find your config file in, like XDG_CONFIG_DIRS)

This seems super overcomplicated. What I would do is put all the subdomains on the public DNS, let HTTP(S) through the firewall for the respective hosts, deny everything from outside of your local network on the http server that isn’t under the HTTP challenge path and then run the HTTP challenge as you would for a public site.

Then you can get certs, everyone outside trying to access will get 403, and inside the network you can access as normal.

Of course you’ll have to trust your http server’s ACL for that, but I’m just going to assume servers like nginx (which I use) have a reliable implementation.

That’s a killer feature. Sucks that it got removed.