Andy HJul 3
James Calligeros

Users just want their computer to work with minimal fuss. They get told to try a Linux distro because it’s safer and faster than the alternatives, and the Linux desktop is totally ready for primetime now!

“Why do my games launch on my secondary vertical monitor? Why can’t I just pick a default monitor?”

“We aren’t implementing that because it’s from Windows. If you want your applications to launch on your main monitor go and use Windows.”

How the fuck can we expect to retain users when a nontrivial minority of developers seems totally fine treating them with such contempt?

Jul 2 at 11:25pmWeb
Show threadSimo ✔️Jul 3
@chadmed maybe realize there is no collective we? There are people, some are decent, some are a*holes, in any generalized group; FOSS is no exception.
Show threadJames CalligerosJul 3

@simo5

Put 40 Keynesians, 30 Georgists, and 10 Chicago school economists in a room, and tell them to design a taxation policy. You set the rules such that no policy is to be ratified unless 30 of the Keynesians, 20 of the Georgists, and 7 of the Chicagoans sign off on it. Do you think the _primary_ source of friction here is going to be 3 of the Chicagoans simply being annoying dickheads?

Show threadSimo ✔️Jul 3
@chadmed I am not an economist so I do not understand the example, sorry.
Show threadJames CalligerosJul 3

@simo5 They are three schools of economic thought. Keynesianism and Georgism can be thought of as roughly aligned, whereas the Chicago School is essentially everything the former two stand opposed to.

The idea here is not about the economics but rather having one group of people, on who's approval you rely to progress things, be staunchly opposed to the direction the rest of the group wants to go in on purely ideological grounds. The personalities of the people involved are not the main source of contention in these situations.

Show threadEmil Jacobs - CollectifissionJul 3

@chadmed On the reverse however: users expect customer service level support from people who are generally overworked, doing this in their free time, and are upset when they're told to read the (often excellent) documentation.

People want a solution to their problems, but want it for free and fast.

Disclosure: I tried to setup a company supporting users in their Linux usage, but failed miserably because no one actually wanted to pay me for my time.

Show threadfcalvaJul 3
@chadmed Pretty sure that's a feature already though ?
Games in particular also usually have settings for the specific screen they are on that persist.
Edit : Removed most of it because i am pretty sure it's just how it felt to me, not how it happens.
Show threadLeelooJul 3

@chadmed
Is that an actual quote, or is it made up FUD?

It doesn't sound real to me, but you never know.

A more probable explanation is that it's controlled by the game and we can't force the games to do the right thing. Or the game leaves it up to the window manager, which is why it's good to have multiple different window managers to choose from.

Show threadLeelooJul 3

@chadmed
In fact I've seen the exact opposite claim - years ago, auto detection of screen dpi was removed, and hard coded to 96 dpi. At the time, it was claimed that this was because that's what windows did.

Later I managed to find the real reason. The api in question only supported one dpi, but with multiple monitors, there is't one dpi. A compromise had to be made and that compromise ended up being 96 dpi. There is a newer api which gives the dpi per monitor, and that api still works.

Conclusion: Anyone can make up reasons for why something works a certain way, so unless you find a quote from the developers who made the decision, you can't trust the claim.

Show threadJames CalligerosJul 3
@leeloo https://gitlab.freedesktop.org/wayland/wayland-protocols/-/issues/179
Output selection for games (#179) · Issues · wayland / wayland-protocols · GitLab

Game toolkits like SDL are in a bit of a difficult situation when it comes to selecting the output to place a fullscreen game on. The current way...

GitLab
Show threadLeelooJul 3

@chadmed
I do not see anything in that thread that matches your claim.

I see a lot of trying to figure out what the problem is, and several people arguing that it's SDL abusing the existing api, and as a result not getting the result they would have if they'd used the api correctly. Which of course a new API is not going to change.

Show threadJoshixJul 3

@chadmed i downloaded a python script that can move a window and configured keyboard shortcuts to move the active window between the monitors. If something like that were builtin into desktop environmemts it would be fine, but the current situation sucks.

(I'm using X11, no idea about wayland)

Show threadJames CalligerosJul 3
@joshix this is related to ongoing discussions about standardising Wayland's approach to this issue. DEs that support the concept currently do so in disparate and hacky ways which causes issues further up the stack, e.g. in SDL3.