Please keep replies on the mailing list.
On Mon, Sep 10, 2012 at 11:21 AM, John Maher <JohnM_at_rotair.com> wrote:
> Thanks Mark.****
> ** **
> Hereís what I am trying to do: learn subversion. To do this the best I
> could come up with is making some test projects and going through the
> commands to learn them. The problem I am having is that each step is
> making MORE questions than it answers.****
> So in a nutshell hereís three things I am currently working on:****
> ** **
> **1) **Finding a client side gui.
If you are on Windows, I would start with TortoiseSVN. If you use an IDE
there are integrations like Subclipse for Eclipse and AnkhSVN for Visual
> **2) **Determining the use of local and server
There is basically no such thing as a "local repository". Repositories
live on the server only. Clients checkout from the repository which
creates a "working copy".
If you haven't, the place you should be starting is the book. It is an
easy read. You should at least read the first two chapters as many times
as needed to understand them. The rest can serve as a reference:
> **3) **Prohibit programmers from stepping on each
> otherís ide settings.
Not sure what the issue is, but usually you can avoid committing IDE
setting files. It depends on the IDE. I like to commit my Eclipse
settings and they can usually be setup in such a way that they work for
> **(1) **We are using VisualSVN Server so I donít think we
> can use subversion edge. A client gui would greatly enhance productivity.
VisualSVN Server and SVN Edge are both serves. So correct, you only need
one of them.
> **(2) **Can you use hooks on server side repositories?
Yes. Hooks live in the hooks folder of the repository on the server.
I see no hooks directory on the server.
Where are you looking? When you create a repository on the server, the
repository folder should contain a folder called hooks. SVN Edge provides
a web UI to manage the hooks. I think VisualSVN has something similar in
its GUI. But you should also be able to access it via the server file
A local repository has a hooks directory but I read somewhere that a
> directory based repository is not recommended for multi users and they
> should use a server. Is this true?
Again, no idea what you mean by local repository. There is only one kind
of repository. The protocol by which you access it can vary, but that is
> **(3) **There are some files that are part of each module
> that we donít want in the repository. According to the book, I have to
> issue the same command on each directory which is not only tedious but
> error-prone. I will have to manually track each directory and hope I can
> regain my place after each and every distraction.
Again, do not know what tool you are using. But usually the files you want
to ignore are not contained in every folder. You can set some
repository-independent ignores in the client configuration too. For
example, you can ignore all files named *.dll or *.obj etc. This has to be
configured on each client though. svn:ignore is superior because it lives
in the repository and clients get it automatically when they checkout.
Received on 2012-09-10 17:32:46 CEST