On 07.10.11 17:18, "Joshua J. Kugler" <jos...@azariah.com> wrote:
>On Friday, October 07, 2011, Lubos Kosco elucidated thus: >> yes, you need a checkout (for every SCM type, no exception) > >That is extremely disappointing. Granted, SVN Searcher (which I tried >first) took up way to much disk space, it would access the repo >directly. Bummer. Why is it extremely disappointing that you need to check out a single copy of your source? OpenGrok isn't tied to svn, but supports a wide range of scm systems (git, mercurial, bazaar, cvs, subversion, monotone, clear case, sccs etc). It makes the rest of the code a lot easier (and less dependent on the actual scm system) to be able to operate on a checked out copy of the code instead of having to go through a "virtual filesystem" every time we want to look at a source file. In addition to that checked out clone we create a pre formatted html file (but you can disable this if you like to) of the source files. Then we're generating the index databases used to search in, and a "database" containing the history comments for the various revisions. > >> also symlinks are not a good thing to index and currently some of >> them are ignored and I discourage usage of symlinks > >A symlink in src pointing to the root of the repo seemed to work. > >> source root is where the checkouts live > >OK. > >> OPENGROK_REMOTE_REPOS is a trigger variable to disable the remote >> repo support, which is on by default > >OK...then can you explain "remote repo"? If a checkout is required, >then what is a "remote repo?" A remote repo is when you access it from a remote host and not from file:/// Cheers, Trond _______________________________________________ opengrok-discuss mailing list opengrok-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/opengrok-discuss
