In a ghostwriting project I completed recently I acted as the liaison between the author, copy editor and book designer. That is, I was responsible for entering all corrections and keeping the master manuscript copy.
As so often happens in a project like this, the author found half a dozen simple changes she wanted after what we all thought was the final manuscript had been sent to the book designer.
I typed those up more or less like this:
Sam, we’ve got two changes:
on page 21, change the name in the second paragraph from Smith to: Jones
on page 30, the last sentence of the last paragraph should read: She was surprised that she hadn’t finished last.
Things went well until I spotted another error. I got back the .pdf with that change, but the first two had been lost. I realized that the book designer had probably scrambled versions and we soon sorted it out. I also realized that since I hadn’t used this book designer I should have had a conversation with him in advance about how he wanted corrections, and maybe even version control.
There are two versions of version or revision control. The formal and the informal.
Formal Version Control
The formal involves keeping a separate copy of each revision. The goal is to be able to get back to an earlier version if something goes wrong. This approach can also be used to track who makes what changes – more appropriate, probably, when designing bridges or software.
This mean each revision gets its own, unique, filename. Numbering is probably the most common. Something like Annesms – 1, Annesms – 2, etc. I’ve also used dating and appended my initials. The problem for me with this system is when I get tired; it’s awfully easy for me to forget to rename the file. I’ve found if I do that as my very first action I’m much more likely to have a complete record.
If you’ve got multiple authors and a big project how you approach version control is critical and you may want to adopt at least some of the conventions developed in engineering.
Informal Version Control
Left to my own devices and assuming the author doesn’t need to see every single change, I don’t try to control the versions at all. What I want, after all, is the most recent iteration.
It’s heresy to some, but I simply overwrite the manuscript with each change. I keep the manuscript itself in it’s own folder; its title includes the abbreviation, ms after the client’s name, like this: annems.doc
I evolved this simple system because I find it so hard, particularly toward the end of writing a book, to remember to save each change as a new file. This way the one file is always the best.
Since I use an automatic offsite backup system I do have a way to get to the almost most recent copy if my computer crashes.
However you do it, finding a way to be sure you’re working on the correct version is a must.
How do you handle revisions?