what I'm doing
copy/pasting into XML Mind, a deeply nifty application that's intelligent enough to convert a FM table into a neat CALS table when pasting
the style for the content of the table specifies Forest Green as colour; this is in the Normal view, not least because I've never had occasion to explore colour views anyway
the paste works, but the (elusive) FM console pops up and tells me
AppData\Local\Tmp\Frm20FA.tmp (785) Value of separation out of range (10)
curiosity, not desparation :-} What does Value of separation out of range (10) mean? I know from other miffing that the error doesn't seem to have any impact on the content or appearance.
Not convinced, I'm afraid :-} taking my investigation that bit farther, I've turned up an allusion to Separations in the online documentation and it seems to be about plates for colour printing – logical enough, I suppose. But … the printer I use every day doesn't support separations; and to confuse me farther, the colour that most often shows up in an offending
<FSeparation n> element seems to be flagged as 'do not print' in the separations set-up anyway.
I wondered if it might be an array-ish problem: original file decides to assign Forest Green to
<FSeparation 10>, I edit the .mif to remove some rogue colours, FM concludes it not longer needs 10 separations…
Regrettably, I finished working on the book in question yesterday and apparently took the trouble to remove the problem in all the affected files; so I can't reproduce it. I'll make a note next time.
Aha – found a file that threw the error, and it does look array-ish: IF Forest Green is automagically assigned
<FSeparation 10> AND tampering with the .mif file leaves less than 10
<DPrintProcessColor> entries, THEN FrameMaker complains.
Still some considerable way from a full explanation, but I can live with it :-}
Same thing this morning … distilling a set of files fails, editing .mif copies to remove
<CSeparation 10> gives new .fm files that distil without any problem. All begins to smell rather like a bug.
And again this morning … but before I'd done anything =8-}
This time, the only definitions with Separation 10 were the FM8 change tracking conditions. I deleted these from the .mif file, opened it, saved it as .fm and was able to distil without problems. Close the new .fm file, re-open it … bang goes Distiller. Again.
Same problem again … a file which (when examined in .mif) included the unused FM8 conditions crashed Distiller. Editing the .mif to delete the conditions gave me an .fm I could distill without any problems.
Can anyone give me any idea why this definition is so damaging?
<Condition <CTag `FM8_TRACK_CHANGES_ADDED'> <CState CShown> <CStyle CUnderline> <CSeparation 10> <CColor `Forest Green'> > # end of Condition
The colour 'Forest Green' is alive and well, fwiw
Well, I entered a bug report Bug#4093614 - files with referenced graphics are rejected / flushed by Distiller and now have no idea whether anything is a) happening; b) ever going to happen
In the meantime, as part of limping towards an imminent deadline, I've used the 'publish to webhelp' trick – thanks, Arnis! – to make mifs of all the files in the latest troublesome book. Each individual .fm file distils to .pdf without problems, but distilling the book (even with all the components open) crashes … and the log file for the webhelp conversion is a long, long list of "Value of Separation out of range " messages.
Not an encouraging start to the new working year!
IIRC, the book file itself also contains colour and separation settings. It might be easier to create a new book file and import the colours from one [good] file into all of the files of the book to ensure consistency.
The suspect lines in the .mif are either in style definitions or in conditions, and the first occurrence is always in the generally unaccessible FM8 conditions. With (regrettably!) practice, I'm becoming pretty efficient at zapping through a set of .mif files with a text editor. However, I will now check the .book as well as the component .fm files.