Andrew Warren wrote: > > Steve Johnson suggested I commit MIFs instead. This in theory could > > work > > No it can't. Try this: > > 1. Save a Frame document in MIF format as test1.mif. > 2. Move a couple of paragraphs or a figure or something. > 3. Save the modified document in MIF format as test2.mif. > 4. Compare the two files using your favorite text diff program. > 5. Gaze in astonishment at the thousands of differences. > > Sorry, but I think you're stuck with SVN's Unlock-Modify-Lock model > for your Framemaker files.
Andrew beat me to it and is exactly right. But this isn't just an "I agree" post. There's a more basic point, IMHO, that makes what Joseph wants impractical and undesirable. A user manual, whether in binary FM, MIF, or whatever, isn't like code. You can't just merge together the new lines added by two authors who edited the same material. You'd end up with nonsense. Collaborative authoring has to use a lock model at some level of granularity. The question is whether you lock a book, a file, or a topic within the file. (I don't think paragraph-level locking can work -- too much chance of the edits of adjacent paragraphs conflicting or interacting confusingly.) Mechanisms/workflows for file-level locking already exist. And if you want to go to a more granular topic-level locking, I suspect the best solution is to go to DITA/XML. Richard G. Combs Senior Technical Writer Polycom, Inc. richardDOTcombs AT polycomDOTcom 303-223-5111 ------ rgcombs AT gmailDOTcom 303-903-6372 ------
