"Ben Collins-Sussman" <sussman_at_red-bean.com> writes:
> Last week, a googlecode user manually modified some svn:date revprops
> on his repository. He got the format incorrect, but wasn't aware.
> However, things started breaking as a result. Our viewvc-like source
> browsing choked on the invalid svn:date format (which we're fixing),
> but even plain old 'svn log' produced 'invalid date' errors, rendering
> the command useless.
>
> I'm wondering if it doesn't make sense to have libsvn_repos (or fs?)
> do some simple server-side validity checks on incoming svn:date
> values, and reject the revpropset if they fail. We already do
> validity checks on filenames and log messages ("no control chars
> allowed"), and validity checks on things like 'svn:eol-style' . After
> all, subversion controls the values of svn:* props, we have the right
> to refuse bad values.
Completely agree. (Also agree with Mike's observation that it should
happen in libsvn_repos, not libsvn_fs.)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe_at_subversion.tigris.org
For additional commands, e-mail: dev-help_at_subversion.tigris.org
Received on 2008-05-06 02:15:06 CEST