diff options
Diffstat (limited to 'zarb-ml/mageia-dev/attachments/20120611/2ee669f4/attachment.html')
| -rw-r--r-- | zarb-ml/mageia-dev/attachments/20120611/2ee669f4/attachment.html | 23 |
1 files changed, 23 insertions, 0 deletions
diff --git a/zarb-ml/mageia-dev/attachments/20120611/2ee669f4/attachment.html b/zarb-ml/mageia-dev/attachments/20120611/2ee669f4/attachment.html new file mode 100644 index 000000000..8fcfdd061 --- /dev/null +++ b/zarb-ml/mageia-dev/attachments/20120611/2ee669f4/attachment.html @@ -0,0 +1,23 @@ +<p>On Jun 11, 2012 1:14 PM, "Colin Guthrie" <<a href="mailto:mageia@colin.guthr.ie">mageia@colin.guthr.ie</a>> wrote:<br> +><br> +> 'Twas brillig, and Guillaume Rousse at 11/06/12 10:19 did gyre and gimble:<br> +> > Le 11/06/2012 09:56, Colin Guthrie a écrit :<br> +> >> Well it the .so name does not change then it *should* be ABI compatible.<br> +> >> It seems in this case it's only API compatible but that doesn't rule out<br> +> >> certain values, macros or flags changing such that everything needs a<br> +> >> recompile.<br> +> > Just rebuilding openssl with different build options already triggered<br> +> > such kind of issues. I remember adding SNI support broke ntp in the<br> +> > past, until this last package was rebuilt. The safe way should be to<br> +> > rebuild everything right now, but given than almost everything in the<br> +> > distribution uses this library, I think just letting normal update flow<br> +> > should be enough, unless actual problems are found.<br> +><br> +> Probably true and I think we should ensure that a full rebuild is done<br> +> before the beta cycle anwyay. We did moot this for mga2 but it was<br> +> deemed (correctly IMO) too late to start messing with that.<br> +><br> +> But until then a full rebuild is likely overkill.<br> +What about gcc 4.7? I think we should switch to that one as well. And then rebuild is needed anyway. So full rebuild at the moment is probably not the best idea.<br> +--<br> +Sander</p> |