Nothing serious of course - AND the advantage is that I do not have to do a new build to switch to pre-fork (which was the old way iirc).
So - was I the first to find a bug in the new release :P btw - I am much more interested in the ssl tests and whether it is a failed test (going back to MC4 128-bit) when the initial connection was much better. I assume this is not logjam (or some other horrible recent OpenSSL TLS renegotiate CVE) - but it is something we want to prevent (as far as I know LibreSSL has no support for RC4 as it is too weak - hence these will fail by definition if the test (client) is forcing a renegotiate to that level of cryptography (key exchange?). In other words - think more about my other post please! On Thu, Jul 16, 2015 at 5:26 PM, Yann Ylavic <[email protected]> wrote: > Yes, and with --enable-load-all-modules (not so common, I think, when > not testing with the framework...). > > On Thu, Jul 16, 2015 at 5:19 PM, Jim Jagielski <[email protected]> wrote: > > Yeah... grrrrr. > > > > In any case, this would affect only those w/ virgin builds, right? > > > >> On Jul 16, 2015, at 11:01 AM, Yann Ylavic <[email protected]> wrote: > >> > >> On Thu, Jul 16, 2015 at 4:41 PM, Michael Felt <[email protected]> > wrote: > >>> My comment is that with 2.4.12 the same configure did not do this. > This is > >>> new behavior. > >> > >> Probably a consequence of [1] which may not play very well with > >> --enable-load-all-modules. > >> > >> [1] http://svn.apache.org/r1661848 > > >
