Noel L Yap writes [quoting me quoting him]:
> 
> >> It seems you've been more thorough than I have.  Looking at it another way,
> "cvs
> >> up -C file"  should be equivalent to "cvs up -p -r <base-rev> file >file".
> If
> >> other flags (eg "-j") were specified, the command should be equivalent to
> "cvs
> >> up -p -r <base-rev> file >file; cvs up <other-flags> file".
> >
> >That's the parallelism I was looking for.  I agree completely.
> 
> I forgot to mention (but I think it was implied) that the two-part statement
> should conceptually be atomic.

Actually, I missed the fact that there *was* a two-part statement -- I
read what I thought you wrote instead of what you did write.  What I
*really* think is that "-C" should be equivalent to "-p -r <base-rev>
>file".  So, I think the last case should be equivalent to the one-part
statement "cvs up -p -r <base-rev> <other-flags> file >file", with the
understanding that a "-r" or "-D" in <other-flags> overrides the "-r
<base-rev>". 

Perhaps it's simpler to say that "-C" moves the existing file out of the
way and then does a normal update with the remaining options (while
suppressing the "file was lost" message).

-Larry Jones

Who, ME?  Who?! Me??  WHO... Me?!  Who, me??? -- Calvin

Reply via email to