Good catch on the index. I've fixed the testcase and now display the indexes starting from 0 similar to the fullscreen test code. The updated testcase is attached.
This is a screenshot of my KDE Display Configuration page:
Here's what I get when I run three different versions of test/fullscreen:
From left to right, the first window shows the output from the fullscreen test code in FLTK 1.3.6. (Gentoo doesn't install the "config.h" file with the library so I commented that include out, hence the "This requires GL" message. Since FLTK 1.3 doesn't scale, this window is also smaller at 100% resolution on the HDPI monitors.)
The second "FLTK<2>" window uses a local git fetch of FLTK with the following message at the top of "git log": commit 85af0036f9fc733b2e8ad3ae62fd00bc783db738 (HEAD -> master, origin/master, origin/HEAD)
Author: Albrecht Schlosser <scrub email addr>
Date: Sat Nov 12 20:14:44 2022 +0100
Reformat Cairo support Fl_Cairo.cxx for CMP compliance
Only formatting, no code changes.
So this local git fetch is from a few weeks ago. (I have a few different projects using FLTK that I sporadically update.)
The last window "FLTK<3>" is test/fullscreen from my latest FLTK fetch, with the following git log: commit bcb739f17f2036ce30ed5cd58627841b59190ea3 (HEAD -> master, origin/master, origin/HEAD)
Author: Albrecht Schlosser <scrub email addr>
Date: Sat Nov 26 21:59:03 2022 +0100
Update dependencies
My default KDE scaling is 150%, and both of these 1.4.0 windows pick that up and scale appropriately. (Curiously enough, one of my FLTK projects stopped automatically scaling with 1.4.0 a few weeks ago; I'm still trying to sort that out as well.) As you can see, sometime in the last few weeks the FLTK 1.4 code went from reporting three screens to one screen. In both cases the FLTK 1.4 code also displays incorrect monitor resolutions. I believe the FLTK 1.4 code used to match the FLTK 1.3.6 code not that long ago, but I only recently noticed this regression. The test/fullscreen output matches the output from my testcase code.
My WM is KDE running on Xorg, not Wayland. (xwininfo, xdpyinfo etc all confirm):
$ xdpyinfo
name of display: :0
version number: 11.0
vendor string: The X.Org Foundation
vendor release number: 12101004
X.Org version: 21.1.4
etc. etc.
Unless you have any quick idea of what's going on, I can try to git-bisect the issue back to where it began. Thanks!
On Monday, November 28, 2022 at 12:45:24 AM UTC-8 Manolo wrote:
A major problem in your testcase.cpp is that the 5th argument of the call to Fl::screen_work_area() is set to 1, 2, 3 when it should be 0, 1, 2 .
Please, change that and compare what you get with the corrected testcase.cpp and with test/fullscreen app that gives in its bottom panel the sizes of all screens of of the work areas of each screen.
Also, what do you get with the "System Settings" application in its "Display Configuration" item? Please, report the Resolution and Scale values shown there for each of your 3 screens.
--
You received this message because you are subscribed to the Google Groups "fltk.general" group.
To unsubscribe from this group and stop receiving emails from it, send an email to fltkgeneral+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/fltkgeneral/2bc5ed97-afdf-4278-9f42-a1a5ee90d9abn%40googlegroups.com.
[ Direct Link to Message ] |