Possible Solution for Color Definition Cleanup

[Fred Ridder]

Jim Pinkham wrote:

> However, part of my clean-up process also includes a "Wash All Files in
> Book via MIF" step. I did that first and then repeated my effort to
> delete the cantankerous color definitions using Toolbox - Format -
> Delete All Unused Color Definitions. They all disappeared without a
> further squeal of protest.
> 
> I haven't attempted to replicate this on any other files yet, though I
> most certainly will be doing so in the not-too-distant future. I pass
> this along, though, in the hopes it may prove to be a solution for
> others in a similar situation. 


Yes, washing via MIF is the standard solution for removing spurious RGB color definitions. But you'll sometimes find that it won't remove *all* of them from a given file, and that always seems to relate to graphics. I've got a handful of files where I've just learned to live with two or three RGB color definitions.

 

But recently I discovered a new type of autogenerated color definitions in some of my files (FM8) that cannot be purged a MIF wash. These show up with names like fm_gen_106529 and fm_gen_98334. Some files had two or three of these, and one had over 40. When you selected any of these colors from the list, you'd see that the Add and Delete buttons were grayed out. Washing via MIF wouldn't remove any of them. I whent into the MIF with a text editor, and found that each color defintion had the line <ColorAttribute ColorIsReserved>, which is what made them undeletable. But I still had no clue where they were coming from.

 

Finally I realized that the files that had the largest numbers of these spurious defintions were all "owned" by a writer who likes to manually highlight text in color for her own purposes in addition to applying the dozen or so defined conditions that we use to tag three different types of comments, plus changes, additions, deletions, and content to be reviewed. What I realized at that point was that these reserved colors are ones that FrameMaker is defining on the fly to paint multiply-conditionalized text with the average of the color definitions for all of the overlapping conditions. 

 

So now I know why they exist, but I still haven't found any way of making them go away other than (a) manually editing the MIF to delete the definitions, or (b) cleaning up the overlapping conditions and importing the content into a clean template. Maybe the "all overlaping conditions indicated in magenta" scheme used in older versions of FrameMaker wasn't so bad after all...

 

-Fred Ridder