funny thing is I have not really been having these problems here - cache away 
on the farm here and all attributes we wanted cached have worked so far [touch 
wood and hope that I did not just put a curse on it].

S.


Sandy Sutherland<mailto:sandy.sutherl...@triggerfish.co.za> | Technical 
Supervisor
[http://triggerfish.co.za/en/wp-content/uploads/udf_foundry/images/logo.png] 
<http://triggerfish.co.za/en>
[http://static.ak.fbcdn.net/rsrc.php/v2/ym/x/lFV-lsMcC_0.png] 
<http://www.facebook.com/triggerfishanimation>

[https://si0.twimg.com/a/1349296073/images/resources/twitter-bird-white-on-blue.png]
 <http://www.twitter.com/triggerfishza>
________________________________
From: softimage-boun...@listproc.autodesk.com 
[softimage-boun...@listproc.autodesk.com] on behalf of Martin Chatterjee 
[martin.chatterjee.li...@googlemail.com]
Sent: 06 February 2013 19:24
To: softimage@listproc.autodesk.com
Subject: Re: [Mini rant-let] Caching attributes - again.

I'd second that.

I normally use all attributes I want to force evaluation for to construct a 
zero length 3D vector.  Then I do    GetPointPosition --> AddMyZeroVector --> 
SetPointPosition  as the last Execute in my tree.

Has been working this way for me for years now.

Cheers, Martin


--
       Martin Chatterjee

[ Freelance Technical Director ]
[   http://www.chatterjee.de<http://www.chatterjee.de/>   ]
[ https://vimeo.com/chatterjee ]


On Wed, Feb 6, 2013 at 5:35 PM, Alok 
<alok.gan...@modusfx.com<mailto:alok.gan...@modusfx.com>> wrote:
I would still advise using the forcing to be done through some arithmetic nodes 
- like adding a zero vector to point positions etc.

Alok
Lead TD
Modusfx
On 06/02/2013 11:14 AM, Vladimir Jankijevic wrote:
the problem with this is that with the log node present in the tree, the 
icetree will evaluate on every frame even if you have two different evaluation 
stacks. Like if you have a tree in the modeling stack and additionally to that 
a tree in the sim stack. If you now put a log node into the tree in the 
modeling stack, it will force the tree to evaluate on every frame even though 
the sim stack would make the stacks below only evaluate on the first frame. 
Hope this makes sense. This case makes it even more clear that we need a way of 
disabling the automatic optimization of ICE values.

Vladimir


On Wed, Feb 6, 2013 at 5:00 PM, Christian Keller 
<chris3...@me.com<mailto:chris3...@me.com>> wrote:
yep, that´s the most reliable way to do it. just make a small compound with the 
log node and deactivate the log checkbox.
then you can drop it onto your connection to the set data or whatever node ...

--
christian keller
visual effects|direction

+49 179 69 36 248<tel:%2B49%20179%2069%2036%20248>
chris3...@me.com<mailto:chris3...@me.com>

http://vimeo.com/channels/96149

Am 06. Februar 2013 um 16:05 schrieb Vladimir Jankijevic 
<vladi...@elefantstudios.ch<mailto:vladi...@elefantstudios.ch>>:


one way of doing it is to put a "Log Values" between the data and the setData 
node. Even if the log node is muted, Log is unchecked, this node forces the 
attributes to be evaluated.

Cheers
Vladimir


On Wed, Feb 6, 2013 at 3:51 PM, Rob Chapman 
<tekano....@gmail.com<mailto:tekano....@gmail.com>> wrote:
I know I know, I used to have various hacky ways to 'force' an
attribute to be stored in a cache. daisychaining them or storing
userdata in the color or whatever ;)   all of them have failed me at
some point until I happened upon a demo ICE topo scene provided by
Ciaran Moloney and saw that every single attribute used in the tree
was diligently applied as a separate Custom attribute display.  Even
some built in attributes that you would expect to cache. when trying
to recreate the exact setup, if any of those attributes were NOT in a
custom display attribute then the topo caching / read /write system
would fail.

so, like I said, to me, this is the ONLY way to reliably store user
stuff in a cache that I have found so far.  I like to think when I am
setting up each of these custom display attributes (after a failed
attribute cache) that each button press and menu select in the process
is like hammering in a nail .  that *hit*  goddam *hit* attribute
*hit* had better *hit* store *hit* this time!!!  *Hit* *Hit* *Hit*


On 6 February 2013 14:28, Sebastian Kowalski 
<l...@sekow.com<mailto:l...@sekow.com>> wrote:
> ...or multiply by 1 and key that param, or put an fcurve in between.
> still i would prefer a "do what i told you to" switch ;)
>
> Am 06.02.2013<tel:06.02.2013> um 15:21 schrieb Rob Chapman 
> <tekano....@gmail.com<mailto:tekano....@gmail.com>>:
>
>> long way around but definitely a brute force approach. make a custom
>> attribute display for each attribute you need storing.
>>
>> On 6 February 2013 14:14, Sebastian Kowalski 
>> <l...@sekow.com<mailto:l...@sekow.com>> wrote:
>>> +1 for brute cache !
>>>
>>> Am 06.02.2013<tel:06.02.2013> um 15:06 schrieb Dan Yargici 
>>> <danyarg...@gmail.com<mailto:danyarg...@gmail.com>>:
>>>
>>>> While I appreciate the software's efforts to cache only the ICE attributes 
>>>> that it deems necessary (unless I list all the ones I want - in this case, 
>>>> many); can we please, please, get a "Just bastard cache everything, I 
>>>> don't care how big the cache files are" button?
>>>>
>>>> I've lost hours today trying cheap and dirty tricks to get a reliable 
>>>> cache.... grrrrr.
>>>>
>>>> DAN
>>>>
>>>
>>>
>>
>
>




--
---------------------------------------
Vladimir Jankijevic
Technical Direction

Elefant Studios AG
Lessingstrasse 15
CH-8002 Zürich

+41 44 500 48 20<tel:%2B41%2044%20500%2048%2020>

www.elefantstudios.ch<http://www.elefantstudios.ch>
---------------------------------------



--
---------------------------------------
Vladimir Jankijevic
Technical Direction

Elefant Studios AG
Lessingstrasse 15
CH-8002 Zürich

+41 44 500 48 20<tel:%2B41%2044%20500%2048%2020>

www.elefantstudios.ch<http://www.elefantstudios.ch>
---------------------------------------

No virus found in this message.
Checked by AVG - www.avg.com<http://www.avg.com>
Version: 2012.0.2238 / Virus Database: 2639/5583 - Release Date: 02/05/13


Reply via email to