[darcs-users] Re: whatsnew and record not catching changes

[Stephen J. Turnbull]

>>>>> "Timothy" == Timothy Webster <timw at outblaze.com> writes:

    Timothy> Yes, using timestamps checks gives a quick dirty check.
    Timothy> but if you are like me I tend to touch a lot of files I
    Timothy> and leave them unchanged.

It's not quick and dirty if your editing discipline is adapted to it.
For example, I almost never see empty diffs (except under CVS, which
generates a lot of them when merging intertwined branches), but I see
a fair amount of inadvertant changes.  Sounds like switching editors
would be a good idea.  Or learning to use the revert function of the
one you've already got.

Thing is, you are going to save at best half of the time you're
losing.  A darcs diff of identical files uses almost no CPU, because
it doesn't actually do any diffing IIRC (it truncates the common
header, which is the whole file.  So it's all IO, and since the files
are identical by hypothesis, none of that is screen output.  That
means that unless you touch without changing more than half of the
files (by sector count) with time differences, you're actually going
to lose by doing the MD5.

OTOH, you can save all of it by learning to use revert.  It's up to
you, of course, but I doubt you'll find much interest in adding this
feature unless there's a more broadly applicable reason for it.

-- 
School of Systems and Information Engineering http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
               Ask not how you can "do" free software business;
              ask what your business can "do for" free software.