NRK

That change was intentional. But since it ignores unknown specifier in normal cases (which I didn't consider at the time), I can see how this can be confusing. We can...

> I'd like to hard error on unknown specifiers. For specifiers, sure. I don't mind that. But what about backslashes? The current behavior on backslashes is _extremely confusing_ (and borderline...

It *should* fix the issue, but some actual real-world testing is still needed.

> The opacity setting doesn't seem to work in general: Are you running under a compositor that supports `_NET_WM_WINDOW_OPACITY`? https://github.com/resurrecting-open-source-projects/scrot/blob/6e2b5221bc25bb073552a2b0707ee99803f1a44f/man/scrot.txt#L163-L164 > since nobody complained about it not working in all...

> picom (a fork of xcompmgr) also doesn't it seems What WM were you running? Perhaps this flag makes it work? https://github.com/yshui/picom/blob/3aed5599c3f73cbfa53b0249795e76ab07cf9ecd/man/picom.1.asciidoc?plain=1#L130C1-L131

If no one can test this, I'll just merge it in a day or two, since the math adds up.

> Edit: `poll()` doesn't have subsecond precision. `poll`'s timeout is in miliseconds. And there are much bigger issues with that code. I'm at the end of my awake hours, so...

> And there are much bigger issues with that code. 1. `scrotXEventVisibility` tries to check if the window becomes visible. However, that event will *never* come. Because we never subscribed...

> I'd much rather figure out what exact problem that code was trying to solve Related: https://github.com/resurrecting-open-source-projects/scrot/pull/67 My wm doesn't seem to respect the raise request. So it'll be difficult...

One more bug, if the selected window was already raised, then there won't be any visibility event (as nothing changed) and so scrot will just sleep for that entire duration.