[kepler-dev] Problem with ImageReader actor

[Christopher Hylands Brooks]

ImageReader could be updated to transfer ImageTokens, but it might
play havoc with backward compatibility, we should update all the
classes in actor/lib/image to use ImageToken. 

I'm not sure how to make ImageReader backward compatible so that
an actor that was expecting an ObjectToken and that gets an ImageToken
does not barf.

The classes in actor/lib/image were initially written well before
ImageToken came around.

It looks like ImageDisplay already expects ImageTokens.

ImageDisplay has some FIXMEs that would be nice to address.

If someone wants to hack around with the image actors and have them
use ImageToken, that would be ok by me.  The image actor code
is in the "Red" category, which means it can change.  It would
be nice if the changes were backward compatible, but it is not
necessary.  I'm the primary author of the first pass at these actors,
and James Yeh made some changes when he added ImageToken and the
JAI and JMF actors.  James graduated and left, so this code could use
a champion to clean it up.

The easiest way to get changes in to the Ptolemy II tree is to send
them to me.  I'm more than willing to give Ptolemy II CVS read/write
access to Kepler developers.  The reason we don't do this in the
default is because it takes a certain amount of fooling around with
the departmental user database to create CVS read/write accounts.
Creating CVS read only accounts is much easier.

When folding changes in, my concerns are:
- Breaking the build
- Copyright violations
- Coding style
- Backward compatibility 

-Christopher





--------

    On Wed, 21 Jul 2004, Daniel Higgins wrote:
    
    > to the 'fire' method in ptolemy/actor/lib/image/ImageReader after the
    
    By the way, should we change ImageReader (et al) to injest/produce
    ImageToken rather than ObjectToken(java.awt.Image) ?
    
    Also, is there a procedure for submitting changes to the ptolemy tree back
    to the Ptolemy folks?
    
    --tobin
    
    _______________________________________________
    kepler-dev mailing list
    kepler-dev at ecoinformatics.org
    http://www.ecoinformatics.org/mailman/listinfo/kepler-dev
--------