> Sorry this review took a while (well from your POV it might have been over-
> night no matter how long it took :-) ).
> 
> Generally looks good already, a few requests for info or changes at the bottom
> of this.
> 
> Changelog:
> - [√] old content matches Debian
> - [√] changelog entry correct version and targeted codename
> - [√] changelog entries correct
> - [√] update-maintainer has been run
> 
> Actual changes:
> - [√] upstream changes look ok
>       I have double checked the orig-tarball without an import yet
>       I compared your PPA vs upstream vs git and they match.
> - [√] no further upstream version to consider (1.7 is dev, ok to take 1.6)
> - [√] no "new" debian changes to consider as we don't merge but go ahead
> 
> Old Delta:
> - [√] dropped changes are ok to be dropped
> - [√] nothing else to drop
> - [√] changes forwarded to debian (I know you are in contact, let that be
> enough for now)
> 
> New Delta:
> - [√] I went through the new changes one by one and they seem ok, some
> feedback detail below
> 
> Build/Test:
> - [√] build is ok, but x86 only
> - [√] verified PPA package installs/uninstalls
> - [√] autopkgtest against the PPA package passes
> - [√] sanity checks test fine
> 
> TODOs:
> - the bugs 1858347 and 1809262 need to be mentioned in the changelog

They were kind of fixed upstream, so I put them next to the "new upstream release" entry.

> - This is "Architecture: any" but I see only amd64 test builds, before a final
> upload probably better also test cross arch build?

Ups, they were in the ppa I was using for testing, just not this one. Fixed.

> - debian/patches/mysql8-mybool-ubuntu.patch should get LP: #1795314 in its
> header

done

> - for "enable generate messages" the reasoning seems off, when reading
>   https://kea.readthedocs.io/en/latest/arm/install.html
>   https://kb.isc.org/docs/kea-build-on-ubuntu
>   It seems a better reason would be "to re-generate docs on build" or
> something like it.
>   If we have another reason we should mention that one

When looking for users of the message compiler, I found this bug: https://github.com/zorun/kea-hook-runscript/issues/15

They fixed it another way, but it shows a third party could need it. I thought about putting it in the kea-dev package, but that would introduce a package transition scenario which is a potential delta I wasn't ready to take on, specially if we have to revert it later.

> - --enable-perfdhcp: why is it default off is it known to be broken or bad in
> anyway?
>   I found upstream 8ee2c24be1383819fe1d79a5c75e993cbbc5e243
> cdc89731ccb2b4aed2c01573bf02ffdd71622be5 and https://gitlab.isc.org/isc-
> projects/kea/issues/340 yet they just state the fact, but no reason :-/
>   If you know a reason please add it.

I figure it's not something you would run frequently, that's why it's optional. Maybe we could put it in another package, but that would also introduce a package transition delta which we might not need, depending on what debian will do eventually. I haven't run the tool myself.

> - your PPAs seem to have proposed disabled I see in buildlog libboost-
> system1.67.0 linked but focal proposed has 1.71.0.0ubuntu1. Please make sure
> it builds and tests against the new version as well

Ups, I had that in a previous ppa where I was experimenting with this, but uploaded a bad version to it and didn't want that to accidentally land in the archive, and forgot to enable those things in this new MP ppa. I just did it and will upload a new package anyway, so we will get those results.