On Thu, Oct 4, 2018 at 7:45 AM Juliusz Chroboczek <[email protected]> wrote: > > > Also it would be really helpful if that remaining PR (prefsrc) could be > > considered. I am not sure if anything is missing there. A merge would be > > the driver for me to take steps to get the version bumped in openwrt. > > The plan is for the patch to get into 1.9.0 -- I'm only merging bug fixes > in the 1.8 branch. Don't hold your breath, though, the code that's > destined to become isn't complete and hasn't even been merged into master > yet (it's in branch "unicast").
I'm cool with that. But are the source specific and crypto branches ready? In particular, I can get new source specific running on a spare comcast gw... fiddle more with bird.... My "plan" is to finish updating the whole network to 18.06 in the next two weeks. And then try to stick with it for 18 months. I am *not* planning to build openwrt again from scratch, just (sigh), build absolutely the minimimum new stuff I need from their sdk. There's a ton new stuff landing for fq_codel for wifi that I mostly test on x86_64 apu boxes, and I do need to look harder at what's going wrong in the mcast front in particular. In the lab I'd *really* like to do "all up testing" of all the new features at the same time, because that's the fastest (if the noisiest) way to get from point a to point z. I can sit down and try to merge up this stuff myself if you like. I'm pretty good at that (and at blowing up the results. :) ). I would so love to re-accellerate babel functionality and features now that the ietf process is winding down some! In terms of personal focus, however, I'm going to try and focus on operational issues with my deployment, and fix those. Which is far from just babel. oy, dns is being a pita at the moment, openwrt 18.06 "feels" noticibly slower post spectre, the whole comcast ipv6 issue, I have a long list). Still, most of my babel biggies are now in: https://github.com/dtaht/babeld/issues And (bad from your data collection standpoint) I probably am going to fully deploy the 3 patches you refused on top of 1.8.3 because they result in a much stabler network for me. I will try to rigorously prove the ecn point in particular in the coming weeks, but it's easy to see if you stress out the network with flent. At this point honestly - I'd encourage the openwrt folk to deploy those patches on top of 1.8.3 also and get more field data on what happens with the fq_codel version they're shipping. > > Sorry for that. > > -- Juliusz > > _______________________________________________ > Babel-users mailing list > [email protected] > https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/babel-users -- Dave Täht CEO, TekLibre, LLC http://www.teklibre.com Tel: 1-669-226-2619 _______________________________________________ Babel-users mailing list [email protected] https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/babel-users
