Le 17/12/2024 à 23:07, Jean-Marc Lasgouttes a écrit :
Le 17/12/2024 à 23:01, Pavel Sanda a écrit :
On Tue, Dec 17, 2024 at 10:38:08PM +0100, Pavel Sanda wrote:
Hi,
this happened couple times today in the particular document, which
loads ton of pdf images,
somewhere in the middle of initial loading and scrolling (at the same
time).
I'll try to come up with some recipe, but it seems very hard to
reproduce on demand,
I need to load it like 20x to get it again with some lucky random
scrolling pattern.
Combination of page up/downs seems to be enough to triggers this. And
cache
needs to be clean before running so the convertor mechanisms take time
to load.
Thanks for the backtrace. So is it only a thing in branch? Note that the
assertion in ParagraphMetrics::position() is new, so this might be an
old problem that we detect only now.
The patch below should fix the symptom, and I think there is no other
cause to cure. I am not sure, though, since I do not understand this
beforeDisplayPosX stuff.
Pavel, could you try it out?
Could you also confirm whether this is branch-only ? This would be
surprising and point to a missing update somewhere, requiring a
different approach.
JMarc
diff --git a/src/Cursor.cpp b/src/Cursor.cpp
index fa844b2946..58647f7c49 100644
--- a/src/Cursor.cpp
+++ b/src/Cursor.cpp
@@ -772,7 +772,8 @@ bool Cursor::getStatus(FuncRequest const & cmd, FuncStatus & status) const
void Cursor::saveBeforeDispatchPosXY()
{
- getPos(beforeDispatchPosX_, beforeDispatchPosY_);
+ if (bv().paragraphVisible(*this))
+ getPos(beforeDispatchPosX_, beforeDispatchPosY_);
}
--
lyx-devel mailing list
lyx-devel@lists.lyx.org
https://lists.lyx.org/mailman/listinfo/lyx-devel
