Hi Bruno, Thanks for following, the cache invalidation part is solved now as I was not aware that a param's "EvaluationOnChange=0" practically disabled the cache invalidation flag. Back to the plug-in, it is sort of an image generator based on a fluid sim, so it is very sequential and I am actually trying to solve the interactive UI issue, e.g. try to pause the simulation and post a persistent message then provide a means to resume the sim via a button and so on .... i know it's ugly.
mike ----- Original Message ----- From: Bruno Nicoletti <br...@thefoundry.co.uk> Date: Monday, April 4, 2011 10:48 pm Subject: Re: [Nuke-dev] Cache invalidation from OFX To: Mike Wong <smm...@cityu.edu.hk> Cc: Nuke plug-in development discussion <nuke-dev@support.thefoundry.co.uk> > Hi Mike, > > what kind of interframe dependency do you have? Do you actually need > the last frame processed or do you only need data about the previous > frame(s) (eg: a transform or luminance or some such)? > > If it is only data about the last frame you might consider doing a > separate analysis pass over the image sequence. In response to a > button press, analyse each frame in the sequence and write key frames > to some set of parameters. When you come to render, simply look up the > parameter value and use that instead. > > The Foundry's stabilisation tools in Furnace are OFX plugins that work > in this way. It removes any interframe dependency during rendering, > which is a complete pain for compositing software to deal with, as > they generally assume each frame can be computed independently of all > others. True interframe dependency is pretty evil in such a situation, > for example, you cant spread renders around render farms and so on. > > b > > > On 29 Mar 2011, at 08:55, Mike Wong wrote: > > > > > Hi there, > > > > I am working on an OFX plug-in on Nuke which requires sequential > rendering and I would like to find a way such that the plug-in may > invalidate the cache to maintain continuity or re-render if necessary > (e.g. the user drags the timemark in Nuke) . > > > > I have tried to set "kOfxParamPropCacheInvalidation" property on > parameters but with no success. > > > > any idea? > > > > Mike > > _______________________________________________ > > Nuke-dev mailing list > > Nuke-dev@support.thefoundry.co.uk > > http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-dev > > -- > Bruno Nicoletti, > Chief Technology Officer > The Foundry > 6th Floor, The Communications Building, > 48, Leicester Square, > London, WC2H 7LT, UK > Skype: brunonicoletti > Tel: +44 (0)20 7968 6828 - Fax: +44 (0)20 7930 8906 > Web: www.thefoundry.co.uk > Email: br...@thefoundry.co.uk > > The Foundry Visionmongers Ltd. > Registered in England and Wales No: 4642027 > _______________________________________________ Nuke-dev mailing list Nuke-dev@support.thefoundry.co.uk http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-dev
