| romangg | |
| mart |
| Plasma |
Because of the nature of floating point math and various Qt and X11 bugs, limiting the
scale factor increment to 0.0625 (6.25% in percentage form) will improve the display in
many apps.
For more information, see the discussions in https://bugreports.qt.io/browse/QTBUG-66036 and
the following bug reports:
CCBUG: 412447
CCBUG: 390451
CCBUG: 373232
| No Linters Available |
| No Unit Test Coverage |
| Buildable 17756 | |
| Build 17774: arc lint + arc unit |
I would just reformulate
"Scale factors are limited to multiples of 6.25% on X11 to minimize visual glitches in applications."
to
"Scale factors are limited to multiples of 6.25% to minimize visual glitches in applications."
I think adding X11 only makes this harder to understand for the users.
Actually I would like to expose the fact that we are doing that only on X11. People not knowing what this is can ignore it. Also it is a single factor, not multiple as with multi output DPI scaling on Wayland.
How about:
The global scale factor is limited to multiples of 6.25% to minimize visual glitches in applications on the X11 windowing system.
@cullmann from what I understand, it is related to what you wrote in https://cullmann.io/posts/kde-qt-highdpi-scaling/ ? How is it only related to X11?
Just a small note: we will probably see some platform that do not use IEEE 754 floating points appear in the next years (unums like posit are well positioned to make an entry on the market, together with RISC-V CPUs). Those will have different tradeofs.
On Wayland one seems to always scale up to the next full factor (e.g. 2) and then scale down the rendered stuff via gl to the selected factor (e.g 1.5).
That leads to other issues but doesn't let rounding stuff creep into the rendering via Qt.
For non IEEE stuff: perhaps that will be available in the future, but at the moment all really available RISC-V ISAs provide just IEEE, too.
And staying in some 1/(2^x) range won't hurt any alternative approach I assume.