Andrew Lentvorski wrote:
I believe that Mercurial has a Trac plugin, but I have no experience.
I have heard this also. I should take a look at how mature it is.
I have no idea what "auto-versioning" is.
Auto-versioning is where you can just dump a file onto a webdav share
(using your GUI file manager all of which speak webdav and present users
with what looks just like a local file share these days) and it
automatically overwrites the existing file and a delta is kept. It's
version control for people who don't know how or don't want to know that
they are using version control. A useful feature in some environments.
Webdav really isn't required for Mercurial since you can pull directly
from a basic ordinary URL.
Yes, you can use hg and pull from a normal url but that url references a
mercurial specific cgi which spits out the data mercurial needs if I
recall correctly. You can't browse or submit files with a browser. I
also like using apache's mod_ldap to authenticate requests off of our
ldap server so developers automatically get their accounts when they get
their email set up.
Well, I believe you *can* do this, most people just don't like to do it
anymore because it changes the source files and then makes merging
particularly obnoxious. My bias: I don't like it because it conflates
metadata and source into the same object. Although, your use of it to
punch the value into the motd is probably reasonable.
Yeah, this makes sense. Keywords aren't a big deal and perhaps I
shouldn't use them at all.
Well, you have to be careful here. This is a philosophical difference
in defining an "atomic unit".
The DVCS systems tend to define an atomic unit as a changeset. SVN
defines an atomic unit as effectively a repository snapshot.
I'm not clear on all this atomic unit stuff. How does it affect the VCS
and why do I care how it is defined?
You can fake a central repository with the DVCS systems, but why? If
you really want a central system, then using SVN is the right way to go.
If you're really the only developer, I believe that there are
"offline" tools that allow subversion to run in an unconnected state and
then later "sync up".
I guess I am looking at this from two points of view: I would like to
learn and specialize in using one VCS. I just have too much stuff to
keep on top of as it is. But my use cases for the VCS vary between
only-developer style where I am the only sysadmin modifying the only
copy of the source, that being system config files to one developer
among potentially several working in remote locations where they may not
have reliable net access or may not want to have to commit to the main
repository all the time but will definitely want to merge their work
with the main repo eventually.
--
Tracy R Reed Read my blog at http://ultraviolet.org
Key fingerprint = D4A8 4860 535C ABF8 BA97 25A6 F4F2 1829 9615 02AD
Non-GPG signed mail gets read only if I can find it among the spam.
--
KPLUG-List@kernel-panic.org
http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
