C. Barth Netterfield
netterfield at physics.utoronto.ca
Mon Jul 26 21:07:23 CEST 2004
I'm fairly certain this is not a good idea.
Here are two common cases:
What if the user is plotting something whose Y axis goes to +-10^-10? All is
well with the old behavior, and not at all well with the new behavior.
What if the user is plotting something whose X axis goes from 10^9 to 10^9 +
100 (eg, time_t seconds)? It doesn't take much zooming before the range
labels saturate, but the change here does not fix it.
The actual solution will have to involve new label formats (like a left side
base, and then tick mark offsets if the label dynamic range gets saturated).
But I think this change is bad. Please revert (or discuss why it is ok as it
On Monday 26 July 2004 02:09 pm, Andrew Walker wrote:
> CVS commit by arwalker:
> Prevent the user from zooming in too much, which ends up with an axis'
> labels all displaying the same value. The user is now limited to zooming in
> to range of 0.000001 on both the x and y axes. This can be modified by
> changing the values of X_MIN_DIFFERENCE and Y_MIN_DIFFERENCE in
> kst2dplot.cpp. Set these values to 0.0 to revert to the previous behaviour.
> M +123 -104 kst2dplot.cpp 1.178
> M +28 -11 kst2dplot.h 1.74
> Kst mailing list
> Kst at kde.org
More information about the Kst