[Kst] kst2 issues....

[Barth Netterfield]

On Monday 01 February 2010 17:35:25 Matthew D Truch wrote:
> A few more things, this time mostly small (and/or large) feature
> requests).  Please let me know if I should be posting bug reports to the
> bugzilla rather than sending messages like this to the list.

This is easier, but Bugzilla lets there be a discussion.  This way, I may just 
say "no" (eg, items 1 & 2 below) or ignore it pending more info (eg, #4).

If you want a discussion, or think we might need one, use bugzilla.  
Otherwise, email is fine.
 
> 1. Double click on a tab should open the "rename tab" dialog instead of
> doing nothing.  kst 1.x does this.

There is an rmb context menu, which is standard for tabs in other apps.  I've 
never heard of double clicking on a tab.... so... "no".

> 2. If you choose a symlink as a datasource, and the symlink changes to
> point to something else, kst should catch this change.  If the user
> didn't want the symlink, they would have chosen the actual data file.

No.  The current implementation respects the behavior of the underlying OS, 
making it standard with the behavior of other apps.

Of course the specific issue you are concerned about is the behavior when we 
open defile.lnk, and defile gets restarted.  I think the current behavior is 
better than auto-reload, since this way, of you were looking at old data, you 
won't suddenly have it disappear if defile gets restarted.  The cost is you 
have to hit reload to move on.  Steve likes the current behavior, and so do I.  
I've been bitten by the old behavior on kst 1.x

> 3. The data manager needs purge functionality.  It is very easy to build
> many plots with many points using up mucho memory, and when you delete
> them from all views, its nice to reclaim that memory as well.  kst 1.x
> has this functionality.

Agreed.  Its already on my list.

> 4. When live data is scrolling by (at least with dirfile datasources),
> sometimes (rarely) the most recent ~0.5 seconds of data is drawn as a
> flat line instead of showing it's true values.  This is usually
> (always?) corrected on the next redraw.  Slightly annoying and certainly
> false.

Hmmm.. haven't seen that.  Have you ever managed to catch it on a pause?

> 5. There doesn't seem to be a pause keyboard shortcut.  In kst 1.x, I
> think this was 'p'.  That would be acceptable, as would the space bar
> (which is pause on virtually all multimedia players, which at least the
> pause button on kst is modeled after).

OK.  Fixed.