It got missed somehow that it's a bullet list indeed...

- regular tweaks

Carrying over metadata and packages downloaded by apt while
being run off a LiveCD is implemented as suggested in #29192.

It's the only installer in this collection
and it did lack any branding indeed.

"prompt" and subsequent first "label" were not separated
in any way while second "label" and forth were; let's make
the resulting isolinux.cfg a tiny bit more pretty.

- regular fixes

NM/TDE is fixed (#25713), ARTS is needed.

This is to avoid NM messing with network interface
involved in NFS root filesystem being operational
(see alterator-netinst); thanks sem@ for the hint.

Make it automatically start in desktop ones,
and let it be available in rescue too (there's
a risk of gpm picking up a wrong protocol and
selecting/pasting at random which is not exactly
the right thing for rescue environment with root
shells all around).

alterator-netinst currently relies on "default"
being specified explicitly; it's wrong and it should
cope with the first "label" clause as well but we're
better off being strict to this script, not that one.

This commit should be no-op regarding syslinux itself.

sshd is rather requisite, wonder how it got missed.

It's based on the package list provided by ldv@ and intended
for barebone OpenVZ template cache serving as a base for any
customizations needed.

It's a great tool giving the ability to at least debug
the novel problems that weren't there before systemd.

Good that it doesn't want half of GNOME or python yet...

It was a desperate kludge to warrant fallback localization
for cases when livecd-setlocale failed to work out properly;
"thanks" to systemd paralizing startup the order was chaotic,
see #28991 for some details.

Now that livecd-setlocale >= 0.3.1 looks like working this
should be dropped for good.

And the proper preset solution will be l10n feature.
That one requires part of alterator-sysconfig backend
factored out into a standalone package along with its data.

- armh/p7/ve fixes

Setting up apt configuration in every container
is pretty annoying, let's enable it by default.

Some provision to add/enable local mirror would
be beneficial but not there yet...

This one adds another image class as well.
Hope this kludge will not get self-conscious...

etckeeper was suggested by george@ along with vim-console;
the latter should come by server,base tags but so far does not
(splitting server-base list into common, bare metal and virtualized
parts in haste looked suboptimal).

It should actually work better than manual specification
now that it's fixed.

That is, no need to pull in systemd as syslogd-daemon provider
when an unspecified one has been requested by interactivesystem
or anything else.

The tricky issue is that THE_LISTS will get expanded separately
and too late to specify a particular provider which will have been
auto-chosen while expanding e.g. BASE_PACKAGES.

rootfs got handy either though managing modules to preload
might be enhanced further.

This proxy makefile is still needed unfortunately...

So the autobuilds were failing for these two weeks since
commit e43386c1 was *totally*
braindead: there's no "arm-e17" target neither "arm-kde4" one.

Fixed while awake and adjusted (in sync with build node)
to reference particular nexus7 and not just "arm".

The sad thing is that it should now read "nexus7old" TBH,
2013 model is not Linux friendly at all at the moment...

- minor fixes

The behaviour that sort of settled didn't actually follow
the principle of the least surprise when one really wanted
to have BUILDDIR available for inspection; DEBUG=2 would be
effective to achieve that but CLEAN=0 would not.

Thanks led@ for spotting and reporting this.

aen@ has suggested that gvfs support be present in regular-lxde
as pcmanfm2 can make use of it; let's factor it back in here.

gvfs pulls gnome-online-accounts and dconf in;
these add considerable bloat that well may be
undesirable in a lightweight distro, just pull
this into a separate pkglist.

It was a temporary hack actually, and is better dropped long-term:
things like predefined root accounts with remote access are *evil*
and this hook was a half of that "solution".

Use of oem feature to integrate first-boot setup is recommended
to deal with this issue, at least when graphics are available.

There's no such thing as xorg-dr[vi]-intel on non-x86 AFAIK,
and the rest of the videodrivers is only generic on x86 too.

- vm-net retired

It's merged into firmware-linux, thanks lakostis@
for the explanation; P:/O: should be added
but less line noise is better anyways.

basesystem doesn't warrant an init these days apparently,
let's make sure one does land here.

This one was asked for and is pretty reasonable common base
to play with cubox from scratch (being ALT); the intermediate
targets had to be refactored with:

- vm/.cubox-bare becoming *that* bare (it doesn't even
  warrant an init anymore, let alone xorg);
- vm/.cubox-desktop accomodating most of desktop bits;
- vm/.cubox-base becoming vm/.cubox-oem as it should be.

The just-introduced tty feature is employed either.

The initial suggestion that any cubox image is a desktop one
didn't hold out for long; and xorg related bits are not that
related to boot script setup in terms of neccessity.

This one only cares for serial console(s) in sysvinit case now
but might end up controlling the number of gettys for ve/vm, etc.