[Kst] [Bug 134987] Skipping frames sometimes misses the last point

[Nicolas Brisset]

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.
         
http://bugs.kde.org/show_bug.cgi?id=134987         




------- Additional Comments From nicolas.brisset eurocopter com  2007-03-29 17:50 -------
I am somewhat dubious here. We rarely (or never) use average/filtering on load, and I insist that it is very confusing for a user to indicate "read 1 sample every 25 frames", have 3501 frames and only 140 samples. I think the mental image the typical user has of skipping is that it reads the first data point as indicated in the UI, then skips (skip-1) points, then reads the next sample, and so on. If there is a sample #3501 that lies just after 24 skipped points, it should be read. Anything else will cause confusion ! Of course, there may be performance issues or consistency between the behavior with/without averaging on, and all sorts of reasons. But in the end it IS confusing.
I am aware that taking into account all aspects (averaging on/off, read from start/end, skip or not, updates paused or not, performance, ...) makes this pretty tricky. I am just not sure that there isn't a way to remove this confusion while preserving a sensible behavior for the rest. That said, I haven't spent hours thinking about it... and maybe there is no better solution ?