Re: Unable to increment revision number

[Larry Jones]

Dave Korn writes:
> 
>   Still, as long as you're doing that, you may as well open the file in a
> text editor and look through it.  The format is fairly self explanatory:
> there are some headers and branch/revision info at the top, then there's a
> full copy of the current HEAD revision of the file, and that's followed by
> backwards anti-deltas from that to the earlier revisions.  It's quite
> possible that you may be able to spot, for example, a chunk of lines that
> have been somehow duplicated, and just cutting them out might repair the
> file.

It may be quite possible, but it's extremely unlikely.  I can't think of
any possible failure mode that would just result in duplicate lines. 
All the likely failure scenarios involve missing data.

>   You should, however, still be very very worried indeed about how it got
> that way in the first place.  It seems likely you have a disk fault.

Or you're using a network file system.  You really need to figure out
what happened and fix it -- your entire repository is at risk until you
do.

-Larry Jones

Everything's gotta have rules, rules, rules! -- Calvin