https://bugs.kde.org/show_bug.cgi?id=473529
Waqar Ahmed <waqar....@gmail.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |waqar....@gmail.com
--- Comment #3 from Waqar Ahmed <waqar....@gmail.com> ---
No, this is a real crash so no closing it please until we have some fix for it.
We have unchecked nullptrs, in this case QTextLayout*, basically everywhere.
The code used to be different before, but in master I changed the internal
line(int) function so that it *never* returns a nullptr. It will return an
KateLineLayout but never null. However, what I did miss was
1. Adapting all of the code to this new reality. Previously a lot of code
didn't bother that line(...) could return null, some did..
2. Making sure that the KateLineLayout::invalid(...) has an empty QTextLayout
and a non-nulll Kate::TextLine so that accessing it doesn't crash us. (not
even sure why layout stuff needs to store the Kate::TextLine instance).
If we don't crash, it will probably result in some weird behaviour. The weird
behaviour can often be easily noticeable and might actually be reproducible by
the user which can then be fixed. If we crash quickly, the user will most
likely not remember what kind of steps they were doing that triggered the
crash.
--
You are receiving this mail because:
You are watching all bug changes.
