Hi again, reply to myself.. On Mon, Sep 27, 2010 at 1:31 AM, Christian Lohmaier <cl...@openoffice.org> wrote: > On Mon, Sep 27, 2010 at 1:22 AM, Harold Feit <dwkni...@depthstrike.com> wrote: >> [...] >> The base64 code had a one-line bug in it that prevented it from >> compiling on GCC 4.4.x that has been corrected in the latest subversion >> revision. No it doesn't break 4.3.x gcc. > > Another one than the line from my patch (the const_cast)?
No, not a seperate one, I just was too impatient ... >> I'll be committing the changes to the main code once I've made sure that >> the patch does the following: >> 1> works on 32-bit as well > > It should - I checked the tracker-link with a 32bit version at least :-) But what I see now after looking at it again: For big-endian systems, make sure to define SHA1_BIG_ENDIAN Either in the makefile or in bnbt.h (where BNBT_BIG_ENDIAN is set with the guess for APPLE (probably assumption wrong nowadays, as most Macs are intel ones) or SOLARIS). I'd not modify sha1.h, but rather keep them "vanilla"... ciao Christian --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@distribution.openoffice.org For additional commands, e-mail: dev-h...@distribution.openoffice.org
