I'm currently trying to run TortoiseSVN as a "remote viewer" of an existing repository on a Linux RHEL server (running svnserver). This should in theory be possible, if the 'shell' interface works as described. Ideally I'd like to be able to run this under the svn+ssh://... configuration for basic security reasons.
Since our build space is rather huge, I'm reluctant to generate working copies of the repository on our wimpy windows hosts unless it is absolutely necessary. Even with extreme partitioning that could be nearly impossible to manage.
puTTY sessions into the svnserver host work perfectly. The existing SVN repository/working copies within the Linux host work perfectly. Pageant works perfectly. PLink works perfectly. All the preliminary steps described in the TortoiseSVN literature are in place (with the exception of tailoring /etc/-based configuration files for subversion. That's a no-no and I'd get my wrist slapped if I asked for it). All the SSH keys (private and public) are in place and work as advertised, on both ends.
TortoisePlink seems to work (when I insert the same entries in the Plink command line into it), but nothing happens visibly, and the i/o redirection of a simple command into a local file appears not to work.
TortoiseSVN crashes consistently - "Unable to Connect to a repository at URL ..." being the most typical message I receive in the working pane.
Either some of the instructions are in error, or there is some conceptual fault in my understanding of what exactly this package does or should do.
One comment I encountered earlier suggested that TortoiseSVN only works if you have a working copy on your Windows host. In this instance that could be problematic.
What say the gurus out there? Specifics can be provided (within reason), providing that proprietary corporate network/host details are excluded. I have an NDA that involves body-parts and first-born.