Re: feedback on today's teleconf

On 21 mei 2009, at 11:51, Silvia Pfeiffer wrote:

> Tracker, this is ISSUE-10.
>
> There is another test case to consider,  which I forgot yesterday: we
> have to take into account whether a time frame is within the duration
> of a file or outside.
>
> So, for #t=x,y request, with x,y integer and y>x , we have several  
> cases:
> * x,y <= duration which is the good case,
> * x,y > duration, in which case it should be "416 Requested Range Not
> Satisfiable" IMHO (same as the other problem cases)
> * x,y with x < duration and y > duration, in which case I am not
> really sure what to return (probably 206).


My suggestion for the latter would be to clip it, i.e. treat it as  
"t=x,".

Whatever we do, we should make sure that we do the same for spatial.  
In other words, if we decide that xywh=0,0,200,200 on a picture of  
100x100 will return xywh=0,0,100,100 then the same should happen for  
temporal clips. And if either returns an error then so should the other.

And to pick a nit: I think the "trouble" cases are for x >= duration  
and y > duration.
--
Jack Jansen, <Jack.Jansen@cwi.nl>, http://www.cwi.nl/~jack
If I can't dance I don't want to be part of your revolution -- Emma  
Goldman