Sure thing, I'll simplify it out and send it on to you. A semi related head's up; I can't reproduce this, but describing it may help to mention in case it ever comes up - yesterday I had a scenario where an iteration count on an iterator, when connected with an input splitter that was a number splitter, yielded one less visual iteration than using an insert splitter that was an index splitter (nothing to do with lfo's or square waves inside the iterator, I was using interpolation for placement). I double checked that the splitters where delivering the same count at the output of the splitter, as well as the input of the iterator, regardless of splitter type used.
It's as if data conversion was going awry at the input port of the iteration count on the iterator. After closing and reopening QC, this stopped happening (and I had no exceptions or other weirdness in that session, all stock patches, etc.) -George Toledo On Fri, Jul 2, 2010 at 5:17 PM, Christopher Wright < christopher_wri...@apple.com> wrote: > > Your composition and result does make sense while it's running, and is > what I would expect. > > *phew* glad that went well, because ... > > > I guess what is still "weird" to me, is how I can coax 2 rendered objects > out of this setup at two iterations, but only if I manually commit, and > choose one iteration first. In your composition, if I stop it at two > iterations, the placement of the iterations shifts from bottom, and top > right corner, to both at top after a few seconds. Everything considered, I > can't understand why the state of the composition would shift like that, > when all value are consistent. > > ... this sounds like a bug. > > > So, I do understand what you are conveying about this generally being > not_a_good_idea, and the reasons why. However, the random shifting (in > render count or placement) still seems flaky even considering your points, > and like placement/rendering should lock in one way or another (thankfully, > I've not been bitten by thinking this worked predictably). > > Can you make a simplified example (this is already pretty simple, so feel > free to reuse it), and provide some easy-to-follow instructions with > screenshots at each step (shots of the viewer's output, not the editor ;)? > I ask because there are some subtle bugs on 10.6 with iterators and > changing counts that might be causing the strangeness you're observing. If > it's what I think it is, there's already a radar tracking it, but I want to > see what you're seeing just to make sure they're the same problem. > (Additionally, testing it on 10.5 might be interesting, to see if it's a > 10.6 regression) > > (you can PM me those steps+pics if you'd like, since they'll likely be > largish and the list might frown upon that)
_______________________________________________ Do not post admin requests to the list. They will be ignored. Quartzcomposer-dev mailing list (Quartzcomposer-dev@lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/quartzcomposer-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
