On Fri, 25 Jun 2004, Ben Collins-Sussman wrote:
> Well, the published book doesn't *exactly* correspond to any tagged
> revision. The publishing process was yucky and messy:
>
> 1. we gave O'Reilly a tagged snapshot
>
> 2. months later, they gave us hardcopy to proofread. This hardcopy had
> their own changes (which weren't enumerated, we had to hunt for them!),
> and meanwhile, the book's source had continued to evolve on its own.
>
> 3. we spent a very painful evening trying to "merge" changes back and
> forth between the two lines of development. Everytime we spotted an
> editorial change, we updated the docbook source. And we ended up
> sending verbal descriptions of source-changes back.
>
> So the result is that the book and /trunk are very, very close. But a
> bunch of lossy, fuzzy editing and merging happened. There are very
> likely a few small changes that weren't ported in one direction or
> another.
Sounds suboptimal. I don't see why they couldn't have directly edited the
docbook sources and sent them back to you. Or better still, created a
separate branch of the repository and done commits there. :-) Surely
whoever was editing the book would need to know how to use subversion
anyway?
Faheem.
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Received on Fri Jun 25 19:46:16 2004