On Sun, 17 Dec 2017 01:22:06 +0100, Branko ?ibej <brane_at_apache.org>
>We'll have to dispel some misconceptions. Subversion's data model is
>significantly different from CVS's. Tags (and branches) are not
>properties of files, they're just (sub)trees within the repository.
>Creating a tag or a branch is exactly the same as copying a directory:
>the only thing that happens at the repository level is that a new
>version of the directory is created.
>The structure of the repository is entirely free-form. The
>trunk/branches/tags convention is exactly that: a convention, nothing
>more. Of course it helps to pick a convention and stick with it.
I have a hard time getting to understand this...
Do you mean that using "trunk", "branches" and "tags" as directories
is entirely voluntary? Does this also mean that files inside a
tags/something directory can be modified and committed, thus changing
the content of the tag?
In my CVS life there is a clear distinction between a tag and a branch
in that it is not possible to commit changes into a tag but it *is*
into a branch. So tags are immutable snapshots of the situaton at a
specific moment in time. Is this not the case in svn?
>I'd very strongly recommend to use one of the existing Subversion mailer
>scripts. There's one in the Subversion repository:
>but there are others. You can certainly use this script to see how it
>gathers commit info from the repository.
I looked at it but could not make much of it since it is in Python, a
language I am very much not familiar with. I am doing ObjectPascal
(Delphi and FreePascal) for windows and linux platforms and also C
(not C++ though) for embedded controllers...
The Python code was very hard reading for me so I will try my way
along the Pascal rouad to get a commithook mailer that replicates the
emails we got out of CVS.
More experientation has shown up how I can use "svnlook tree" properly
to extract the files in a new tag, so that part is at least covered
Developer in Sweden
Received on 2017-12-18 00:35:59 CET