[Sidefx-houdini-list] Weird cooking problems
craigleehoffman at aol.com
craigleehoffman at aol.com
Fri Feb 13 20:34:42 EST 2009
I just figured this out about an hour ago.? My stamp expression had bad syntax like this:
with one period instead of two like this
Strangely enough it seemed to work fine when only one was displayed (which is why I didn't think the expression was bad), but when two objects with the same copy/stamping were displayed things weren't happy.
When using one period, is that perhaps pointing to some global variable that was colliding when two were trying to access it perhaps?
From: Louis Dunlevy <louisdunlevy at gmail.com>
To: sidefx-houdini-list at sidefx.com
Sent: Fri, 13 Feb 2009 4:28 pm
Subject: Re: [Sidefx-houdini-list] Weird cooking problems
hate to ask obvious questions but your stamp function does say something
The first thing I can think of is that the objects are cross referencing
each other because you haven't set the path to the copy sop explicitly
enough to make sure it only looks in the current network..
or can you try the foreach sop?
craigleehoffman at aol.com wrote:
> I have 3 Objects that each have CopyStamping going on inside of them.? Most of
the complex stuff has been cached out to bgeo files or locked in the networks.?
Each one by itself (only it is turned on) is lightning fast to move around in
the Viewport at the OBJ level or to dive in and out of with no re-cooking (as is
expected due to no external references and no updating really being needed), but
when two or more of them are turned on, every selection of one of them, or
interactively moving a transform handle at the Object level, or diving into
SOPs, or jumping up from SOPs causes a very slow re-cook, with major time hits
on the Copy SOPs (from the Performance Monitor), even though nothing is being
changed- just selecting them or grabbing a transform handle, etc.? There is
nothing affecting these to cause them to re-cook, but yet they do.? They do
> I tried upping the COP Cache and the OBJ Cache to the max in the preferences
thinking that it must be losing the memory due to them all being on (and they
aren't that big or crazy), but this didn't help.? Again, one at a time they are
lightning fast, but with two everything grinds to a halt.?
> Anyone have any ideas?? The second two were each copied from the first, but I
made sure there are no references where one relies on anything in another- could
something I am not aware of be colliding??
> Sidefx-houdini-list mailing list
> Sidefx-houdini-list at sidefx.com
Sidefx-houdini-list mailing list
Sidefx-houdini-list at sidefx.com
More information about the Sidefx-houdini-list