[Kst] Vector skip patch ?

[Adam Treat]

On Wednesday 18 October 2006 9:09 am, Barth Netterfield wrote:
> On Tuesday 17 October 2006 10:59, Adam Treat wrote:
> > What I am supposing (and the patch is doing) is that when in
> > count-from-end mode that we count back from the end the requested length
> > and the frame we arrive at is then f0.  We then divide up the length
> > according to skip.
>
> Imagine what then happens if one more frame gets added to the file...  f0
> increments by 1, and the entire vector changes.  So as data comes in, you
> get very unnerving flicker, and you take a big performance hit.

Gotcha.

> The solution we came to was to insist that new f0 = n*skip where n is an
> integer ("skip boundries").  This is what the old code did.  This is
> required for count from end, and so, for consistency, we did it everywhere.

Ah, that is why the f0 requested is not always what the user gets.  I'll 
modify the testcases.

> The only change in behavior from the old code we need to make is to ensure
> that the last skip-boundry point is read.

Ok.

Adam