Mail Archives: cygwin/2009/12/14/11:55:15
Marc Girod <marc.girod <at> gmail.com> writes:
> Renaming a label type is an O(0) operation.
There's no such thing as O(0).
> It doesn't depend on the number of labels of that type applied.
Just because renaming a label is O(1) doesn't make applying the label in the
first place O(1). I want a VCS that tracks changesets. In other words, if I
can point to a given file that changed on a certain date, then I want my VCS to
find all other files that were changed at the same time. Atomic renames mean
nothing if you don't have atomic commits in the first place. And yes, you can
get changeset information in clearcase, just like you can get changeset
information from CVS, but in both of those systems, it costs some painful
overhead to reverse-engineer, and the cost is most certainly not O(1). Or you
can use modern VCS, git being my favorite, but several others will do, where
obtaining the information about a changeset is O(1).
And to keep this cygwin-related: yes, it is true that one of the reasons I
maintain the cygwin port of git because it is open source, while there is zero
chance of clearcase ever being ported to cygwin because it is proprietary. But
it is more than just the free source aspect - one of the reasons I maintain
git, even though I'm forced to interface with clearcase at my work, is because
the concept of atomic changesets is such a fundamental improvement over any
features that clearcase may claim to offer amidst all its bloat. I find that I
am much more productive if I manage my own work in git, then convert it to
clearcase at the end of the day, rather than suffering through the slowness of
clearcase throughout the day. And the fact that a VCS distributed with cygwin
will understand cygwin paths and interact well with cygwin ptys, while closed-
source behemoths like clearcase do not, is another argument in favor of a
cygwin VCS solution.
Finally, I feel justified continuing to argue the merits of programs offered as
part of the cygwin distribution on the cygwin mailing list, but I don't think
it is appropriate to argue the merits of proprietary non-cygwin programs on the
cygwin list. This is for the same reason that we tell people to take their
questions about 'gcc -mno-cygwin' to an appropriate list - this mailing list is
for promoting the use of cygwin, not for promoting the use of non-cygwin. If
you insist on touting clearcase, at least move the conversation to the talk
list.
--
Eric Blake
--
Problem reports: http://cygwin.com/problems.html
FAQ: http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
- Raw text -