Work continues to progress well but I've hit a few snags: * My current recovery model is just not working for cross-cluster transactions. * I can't fix the write performance issue due to the way the cluster localization and spike code works. * Handling nearly full filesystems is proving to be a problem due to the inability to estimate space requirements which in turn are related to the cluster localization the current model uses. * Efficient real-t...

Yah, that's fine. FreeBSD did something similar. -Matt

* The floating point state now being saved on the signal stack and restored on return. * Probably a few other things that I can't put my finger on at the moment. I think the FP changes may be the biggest issue with regards to threaded pkgsrc binaries. -Matt Matthew Dillon <dillon@backplane.com>

I'm getting it together for building the 2007Q4 pkgsrc release on pkgbox, so that we have a new set of binaries for the next release. Are there folks who want each quarterly release to last half a year, or should I stick to only the most recent release? I'm leaning towards not doing this, as the previous quarter's release is not updated, and anyone trying to stick with it will not be updating frequently anyway. If the answer to that is yes, I was thinking I'd take the existing 2007Q3 binaries an...

Aha! I remembered you had said something, but I couldn't find the message. This is a good idea. However, does pkg_add and make bin-install expect a certain hierarchy? I'll have to experiment.