[AccessD] Source Code Control

Josh McFarlane darsant at gmail.com
Wed Jan 18 11:31:22 CST 2006


On 1/17/06, Charlotte Foust <cfoust at infostatsystems.com> wrote:
> Yes, the behavior is different is you use the add-in.  The shop I work
> in uses SourceSafe, so I'm not likely to switch to something else.  I'm
> not sure I would like creating multiple stems from a base file.  Sounds
> to me like an even better way to get versions screwed up.  At least the
> way we have SourceSafe set up, only one person can have an object
> checked out at a time, and all project that use that object will get the
> same version when it's checked back in.  If we've branched it for a
> particular project, then the changes don't get applied to that version.
> Works for us, and it works nicely with VS.Net, so we're happy.

Well, branching is made to be unintusive. Most other versions of
source control use merging rather than single-user checkouts. This
way, multiple people can work on the same source, and then when they
go to check back in, if changes collide, the second person reviews the
new code and modifies the version they are checking in accordingly.
Otherwise, if it's the same file but non-related functions, the merge
is seamless.

Branching is used if you're working on a major change, say adding a
new module to the program, but don't want to disrupt other people's
work. You work on your change, and then once it's finished, you can
rejoin your branch to the main version and merge changes accordingly.

Oh, another thing I like is in Subversion you can create tags. They
basically make a little marker with a title that you have that pulls
all files down from a certain revision, so when we send software to
our customer, we tag it with the date and name, and then if they ever
have a version, we can pull the exact code they used down with a
simple click.

--
Josh McFarlane
"Peace cannot be kept by force. It can only be achieved by understanding."
-Albert Einstein



More information about the AccessD mailing list