RE: Ghost of missing fonts mystery--SOLVED...GIGO

["Deborah Snavely" <dsnavely@xxxxxxxxxxx>]

Dear Framers,

Recap of issue:
>*None of the files in the book reports any missing fonts upon opening.
.*The book updates successfully with only the book file open.
>*The book file itself is new (re-created from a chapter file).
>
>Yet when I update the book, for 7 of the 40-odd files in the book,
Frame
>(a) opens console windows that report the former fonts missing from
>these files...and (b) CLEANS OUT the missing fonts reported for those 7
>files from the console windows by the time it completes the book
update.

#$%^#&*%$ MS Outlook!  I had this lovely message all written and it hung
on me...in the course of saving my draft!

Short form: turned out there were ABSOLUTE PATH x-refs in some files in
the book to other files in an earlier version (archive-cloned to a
date-named directory). So, when I copied the contents of individual
files from the old book's files into the new-template book's files,
these absolute x-refs came along. And they didn't affect anything in the
appearance of the individual files, so they didn't trigger "missing
fonts" message in the individual files. And they didn't affect anything
in the FINAL pass that Frame makes when updating the book file, so no
"missing fonts" message in the book file. 
(And when it's not in the file, you won't find it by searching the MIF!)

I got rid of it only after I searched to ANY x-ref when opening the
suspect files; double-clicking these x-refs and examining their sources
revealed the fact that the path led to outside the current directory. 

My reconstruction of the Frame sequence of actions, which matched my
symptoms:
1 scan files sequentially to update numbering and pagination
2 scan files sequentially again to resolve x-refs
3 sequentially resolve relative path x-refs en passant
4 encounter an x-ref to an external file not already open
5 scan external file 
6 ENCOUNTER MISSING FONTS
7 sequentially report missing fonts in Frame console windows (1 per
file)
	with window title <filename> 
8 sequentially scan now-final content of all files, generating TOC & IX
files
9 sequentially encounter current book's file of matching <filename>
and...
...correctly display no missing fonts, thereby removing list of missing
fonts
from the open console windows for identically named <filename>
10 save all files 
11 close all files

So...my lesson is, ALWAYS teach your authors/collaborators: NEVER open a
Frame file from outside the book file.  Repeat early and often, until
they get it. 'Cause the only way I can see this thing occurring is 
...<Joe/Jill Programmer> opened individual files in his/her newly cloned
current draft with newly dated directory name
...receives reports of unresolved x-refs (another reason to work only
from book file!)
..."fixes" those unresolved x-refs to files that <Joe/Jill Programmer>
opens when cued by Frame so helpfully offering: "Do you want to open the
source file for this reference?" 
...and those files are in the OLD directory name of the file

Whooof.

Many thanks are due to Thomas M., Tom N., and Ananda for their
suggestions. None of 'em quite hit the gold, but they encouraged me to
keep looking, and to help me "sweep the corners" in making sure I
settled it. 

Deborah Snavely
Document Architect, QA & Documentation, 
Aurigin Systems, Inc. http://www.aurigin.com/ 

** To unsubscribe, send a message to majordomo@omsys.com **
** with "unsubscribe framers" (no quotes) in the body.   **