If I start kdevelop on my external screen (with 120x120dpi) it starts normal and I can work with it. If I try to start it on my internal laptop monitor (with 192x193dpi) it hangs and uses one cpu at 100%. As a test I started kdevelop with Xephyr using different dpi values: Doesnt work: Xephyr :1 -ac -dpi 192 -screen 1480x1680 &; DISPLAY=:1 kdevelop Does work: Xephyr :1 -ac -dpi 96 -screen 1480x1680 &; DISPLAY=:1 kdevelop I tried a few resolutions, it seems that the highest working value is 164 DPI. Any start of kdevelop with a setting above 164 DPI will cause kdevelop to hang at start. When I test it here in Xephyr I can see it instantly. If it hangs, I can not select any Text in its Editor or klick somewhere. Reproducible: Always Steps to Reproduce: 1. Xephyr :1 -ac -dpi 192 -screen 1480x1680 & 2. DISPLAY=:1 kdevelop Actual Results: Kdevelop hangs and cant receive input Expected Results: It does not hang
Can you try the following: export KDEV_DISABLE_WELCOMEPAGE=1 kdevelop Does it still hang then for the 164 DPI setting?
With: KDEV_DISABLE_WELCOMEPAGE=1 kdevelop it does not hang with 192DPI or 180DPI (I didnt try so many) so it seems to solve the problem. The problem (without your fix) did show when I started Xephyr with > 164 DPI, so I had to test it with values > 164 (just for clarification) Does a welcome screen (which I never saw again in this kdev session) prevent kdevelop to start properly?
Yes, we had a few problems with the Welcome Page in the 5.0.0 release. This should be fixed for the upcoming 5.0.1, though. Related patches: https://phabricator.kde.org/D2581 Another commit: commit 29d302ec5955ab0aec240ced6302247d903dc81a Author: Kevin Funk <kfunk@kde.org> Date: Sat Sep 3 01:50:51 2016 +0200 WelcomePage: Fix another buggy binding
Thanks a lot for your very (and I mean really fast) responses :-) I never saw a commercial support as good as you. Thanks for your great work. I use kdevelop for python and php development and I know exactly why I not use PHPStorm or PyCharm ;-)
*** Bug 368731 has been marked as a duplicate of this bug. ***