On 30/10/2012 10:46, Duncan wrote: > ... I've been running gnutls-3.x for some time (at one point it was > needed for the live-git pan I run), tho I had to remask gnutls-3.1.3 as I > experienced some problem (IDR what) with it. But I'm running 3.1.2 > without issue.
I've been using gnutls-3 on one of my devsystems as well, it's just as you can see from a tracker that some software is still working properly. But today's screwup with libtasn1 3.0 shows that we really can't keep it masked much more. > What gnutls-3.1.x are you planning to unmask? Do I need to try 3.1.3 > again and file a bug (if there's not one filed already) if the problem > still exists, or is 3.1.2 good enough? Given that 3.1.2 is not in tree anymore there's no choice uh? Beside, I don't go masking micro versions around. If you think there's a problem with 3.1.3, please test and let us know as I haven't hit any (that's what I've been using myself, and testing the tinderbox against). > FWIW, I also recently did a full emerge --empty-tree @world too, so there > shouldn't be any hidden problems lurking around to bite on either > package, at least with my @world and USE flag combo, either. The only big problem we're going to hit as I said is that Boost 1.50-r1 and 1.51 don't use eselect boost any longer, which means that the reverse dependencies need to be updated. Scarabeus was looking into it earlier today, I was waiting to hear from him as I don't want to go near Boost in the near future if I can avoid it. -- Diego Elio Pettenò — Flameeyes [email protected] — http://blog.flameeyes.eu/
