#378 Matrix: Pixel snow in canvas after repeated scrolling
Please followup on rg-devel as well as here.
In the matrix editor, there is sometimes "pixel snow"
from the edges of notes in the matrix canvas after
having done several cycles of rapidly scrolling to the
right followed by rapidly scrolling to the left, or
vice versa. The pixel snow is always a 1 or
2-pixels-wide vertical line of the same colour, height
and position as notes that were on the display before
the scrolling. If the affected notes were in a chord
then you see a set of pixel snow lines stacked
vertically above one another.
A version of this bug was originally reported on
rosegarden-devel
on Wed, 10 Mar 2004 23:23:21 GMT
in Message-Id:
200403102323.XAA22445@liam2.demon.co.uk
Richard Bown confirmed the bug
on Thu, 11 Mar 2004 09:32:34 +0000
in Message-Id:
200403110932.34948.richard.bown@ferventsoftware.com
However from William's matrix bugs I can reproduce
these:
[...]
* Pixel snow in matrix editor after rapid scrolling
Logged In: YES
user_id=4816
This one is caused by the QCanvas sometimes not refreshing properly. It
happens too when moving the playback pointer at high zoom levels. I'm
not sure what we can do about it...
Logged In: YES
user_id=663564
I'm experiencing the same thing on my X terminal. It's a NVIDIA
RIVA-128 with the standard xfree86 "nv" driver. I'm getting snow
all over the place sometimes. Rosegarden and the Kicker are
the usual victims.
I ran that box for a good couple of months without experiencing
this. The difference is that when I set it up as an X terminal, I
installed the version of xfree86 I had on-hand after getting my big
box up, which was 4.3.x. I was running 4.2.x while I was using
the thing on a daily basis.
Everything else being equal, I have to wonder if it doesn't have
something to do with the new xfree doing something screwy.
Anyway, there's nothing we can do about it.
Logged In: YES
user_id=1004577
No, this particular pixel-snow bug doesn't depend on
XFree86; it occurs independently of both type of graphics
hardware and version of XFree86 X server.
There is, however, another pixel-snow bug which is related
to a known bug in the XFree86 drivers for nVidia graphics
hardware:
https://sourceforge.net/tracker/index.php?func=detail&aid=929488&group_id=4932&atid=104932
Logged In: YES
user_id=4816
Basically the same problem as 888431 (Trail from blue bar in
matrix view at low zooms). Will be fixed after matrix rewrite,
yada yada.
Logged In: YES
user_id=1004577
This bug is totally different from 888431 (Trail from blue
bar in matrix view at low zooms) -- this bug does not depend
on zoom, it occurs only during rapid scrolling and it
affects notes but not the blue playback pointer.
I'm re-opening this bug report, leaving it at priority 5
which in Chris' priority scheme means it won't be fixed
until after 1.0.
Logged In: YES
user_id=4816
William, I've been dealing with the QCanvas for more than 4
years, I have implemented the matrix view, and the playback
cursor. In short, I know what I'm talking about.
There is no reason to leave this bug opened, given that it's on
a part of the code we know we're going to throw away soon,
and there's another BR which I'm leaving open only as a
reminder.
Logged In: YES
user_id=1004577
If the reason you want to close this bug report is that the
canvas code will be changing soon, please would you explain
why you are not also closing the many other canvas-related
bug reports for the same reason. Secondly, given that this
bug report describes a different problem from your bug
report #888431, please would you explain why you think this
bug report is a duplicate.