I didn't catch that the capability is reported by the client.
At the moment this will be enough for me, as I'd like the
users to use at least version 1.5.
At 02:56 31.03.2009, you wrote:
>Ryan Schmidt wrote:
>>On Mar 30, 2009, at 05:09, Alain Palamara wrote:
>>>Is there a way to know the client version from the server part (i.e in
>>>a hook script). I know this question was already posted, but I don't
>>>saw a clear answer (the thread diverged on other subject).
>>I think the answer is no.
>>>My point is to ask the user to upgrade their client to a minimal
>>>version the next time he commit. I know it should not be a problem
>>>to use older client with newer server, but I don't want to support
>>>questions about missing features of older version (we mainly use
>>>Tortoise as client). Of course our users are all "in House", this is
>>>not question of open project repositories.
>>I believe there is a way for the server to know if the client has the
>>new merging support in Subversion 1.5. So there is a way you can
>>block clients < 1.5. But I don't think there's any way to get more
>>specific than that, or to know exactly what version the client is.
>The start-commit script is passed a colon-separated list of capabilities reported by the client.
>Here's the text from a newly created start-commit script:
># The start-commit hook is invoked before a Subversion txn is created
># in the process of doing a commit. Subversion runs this hook
># by invoking a program (script, executable, binary, etc.) named
># 'start-commit' (for which this file is a template)
># with the following ordered arguments:
>#  REPOS-PATH (the path to this repository)
>#  USER (the authenticated user attempting to commit)
>#  CAPABILITIES (a colon-separated list of capabilities reported
># by the client; see note below)
># Note: The CAPABILITIES parameter is new in Subversion 1.5, and 1.5
># clients will typically report at least the "mergeinfo" capability.
># If there are other capabilities, then the list is colon-separated,
># e.g.: "mergeinfo:some-other-capability" (the order is undefined).
># The list is self-reported by the client. Therefore, you should not
># make security assumptions based on the capabilities list, nor should
># you assume that clients reliably report every capability they have.
To unsubscribe from this discussion, e-mail: [users-unsubscribe_at_subversion.tigris.org].
Received on 2009-03-31 07:57:52 CEST