- From: Kari Pihkala <kpihkala@cc.hut.fi>
- Date: Mon, 18 Feb 2002 11:37:32 +0200 (EET)
- To: Patrick Schmitz <cogit@ludicrum.org>
- cc: www-smil@w3.org
Hi Patrick, I've noticed that the section 3.4.2 has other errors as well. (I have checked the errata doc) In the beginning, it defines i = floor((t*n)/d). This causes i to be n (i=n), when t = d (because 0 <= t <= d, defined in section 3.2.1 and the cumulative and frozen animation functions in section 3.2.5 call f(d) ) However, this will cause discrete animation function f(t) = value[i] to access value[n], which doesn't exist. (because there are n entries in values attribute, so only value[0]...value[n-1] exist) The same array access problem exists in linear and paced equations, as well. For instance, in paced anim, it explicitely says 0<i<= n, so i can be n, which is wrong! D(i) = dist(value[0],value[1]) + dist(value[1],value[2]) +...+ dist(value[i-1],value[i]), for integers i with 0<i<= n. I haven't checked the spline equation. regards, Kari On Wed, 2 Jan 2002, Patrick Schmitz wrote: > Hello Taras - > > Sorry for the long delay in responding - I was busy with some other work and > took a break over the holidays. > > In the equation you mention, the subscript i refers the current interval > from the set of n-1 intervals. As such, t(sub)i is the time associated with > the beginning of the 'i'th interval. > > As I review the section, I believe I see an erratum in this equation - I > will take this up with the SMIL WG. > > Thanks - Patrick > > > -----Original Message----- > > From: Taras Shymbra [mailto:terry@inobject.com] > > Sent: Tuesday, December 18, 2001 3:16 AM > > To: cogit@ludicrum.org > > Subject: SMIL Animation > > > > > > Dear Patrick Schmitz, > > > > In the SMIL 2.0 Specification (part 3 -> "The SMIL 2.0 Animation Modules") > > it is declared: > > > > "For linear animation, the duration is divided into n-1 equal periods, and > > the animation function is a linear interpolation between the > > values at the > > associated times: > > f(t) = value[i] + (value[i+1]-value[i]) * (t-ti)/d. " > > > > Would You please clarify what is the semantics of the "ti" in > > this animation > > function equation? > > > > Thank You in advance, > > Sincerely Yours, > > > > Taras Shymbra > > >
Received on Monday, 18 February 2002 04:37:40 UTC