Daniel Gröber (dxld)

Yes `--with-stack` should fix this if the problem is `stack` not being in `PATH`. The `stack exec -- ghc-mod ...` thing only works properly with >5.6 though where I added...

What are you trying to do? OpenWrt already has a fully integrated (kernel module based) CLAT called `nat46`(https://openwrt.org/packages/pkgdata_owrt18_6/kmod-nat46). IMO a misnormer but here we are :) https://github.com/openwrt/openwrt/blob/ea609fe486a99af6540dcc7a18f0827481d0c4b0/package/kernel/nat46/Makefile It may be...

Then `opkg install kmod-nat46` and it should Just_Work^TM but I've never tested it myself.

FYI: You can just boot up one of the x86 images for testing: https://downloads.openwrt.org/releases/23.05.3/targets/x86/generic/openwrt-23.05.3-x86-generic-generic-ext4-combined.img Something like this is what worked last I did this: ``` $ qemu-system-x86_64 -m 128M -M...

FYI: I'll be following this up with a patch enabling a whole slew of warning/hardening flags.

Hi Vít, Doesn't really matter. As long as they exists there's all sorts of regex knobs to get the format adjusted to what the Debian workflow likes. Me personally, I...

Excellent, as you can see from PR activity, I've been busy reviewing and packaging tundra :)

Hi Vít, removing print-config is fine with me, I don't think tundra's config is complex enough for it to be necessary. I'm probably going to shoot more patches when I...

After initial packaging and subsequent more thorough code review I'm not sure anymore tundra is suitable for deployment yet. I'm looking at developing some kernel patches to support SIIT/NAT64 instead....

Maybe instead of disabling the connection limit for such cases mox could support killing the oldest established (and probably stuck) connection to allow the new one to go through?