Things that had to be done by hand are now generalized
and made convenient, hopefully.

Minor fix: /boot directory is not cleaned for livecd if there are
homeros-install or luwrain-install files (in addition to live-install
and livecd-install being already checked).

- armh related fixes

The automated build relies on particular names
to be buildable and this broke during some sync;
http://nightly.altlinux.org/sisyphus-arm/snapshots/
were empty as autobuild was failing since 20130710.

p7 aliases go in too.

This inobvious buglet has been spotted shortly after
1.1.0 refactoring was roughly over; the problem is
that armh-specific targets would get ignored without
explicit ARCH=armh.

It's commit 768df7e9
which streamlined introspection within main.mk
and a bug in Makefile which left ARCH empty
even upon having autodetected it for ARCHES.

The current version is more straightforward as a bonus.

Thanks glebfm@ for nudging me to do this; initial draft
was the very firsh armh distro target successfully built
and tested but the effort has refocused onto nexus7 ones
with cubox images joining the party a bit later.

Now it's the time for all good servers...

It was sitting next to the wrong line after refactoring
of cubox related target and was referring to "king"
instead of "kind", very kind of me but not so kingly.

george@ spotted gqview in regular-lxde.iso and wondered why;
it's not being developed since 2006 or so while there's a fork
named geeqie which has continued to improve upon it.

Intro: NetworkManager-wait-online.service would, well, wait
for some network interface to become online or for timeout
to kick in.

Problem: if a LiveCD is tested in offline environment
that timeout will only impede the boot.

Proposed solution: use/net/nm/nodelay target has been implemented
to disable that service as proposed by sem@ and done in Simply;
"+nm" target changed to be an alias to this one.

- assorted fixups

kvm fork is no more, and qemu-kvm is provided by qemu-system.

Thanks glebfm@ for spotting and reporting bitrot;
these pkglists were copied from m-p-d during
server-light reproduction.

It's old, it uses consolekit (even if not neccessarily),
it borders obsolescence *but* removal of udev-alsa has caused
massive regressions (e.g. regular-gnome3 had soundcard mixer
levels dropped to zero from the start, regular-razorqt added
inability to poweroff to that...).

Just get it back.

Thanks boyarsh@ and cas@ for pointing out that the branding
that takes MATE peculiarities (e.g. background settings)
into account is in Sisyphus/armh already.

Following cubox-tde it's just plain reasonable to add
OpenSCADA to the regular one ;-)

- armh fixes and tweaks

This one is IMHO best suited DE for cubox' meagre performance
(especially on I/O side due to microSD) as it incurs quite modest
I/O, CPU and RAM footprint quite suitable for PII-300 times
(cubox feels quite like that but has lots of RAM compared to
even maxed-out Pentium II workstation).

The image includes OpenSCADA as a nice and unique feature
which has been developed with TDE environment in mind
and was used on ARM hardware like N900, incidentally;
thanks aen@ for this suggestion.

Thanks rom_as@ for the entire OpenSCADA project and
packages as well as for keeping TDE packages alive
and kickin' in Sisyphus.

Most of these are slated to employ oem feature by now,
no sense to call it in almost every dependency chain;
the only image left with predefined locale and credentials
is cubox-xfce-ru.

The nuance being that:
- alterator-setup package would change default.target
  for systemd providing a symlink of its own and making
  a backup of what was there (rc3 basically);
- 40-x11-autostart would ignore that backup;
- 99-oem-setup would do nothing about it all either;
- alterator-setup removal would restore rc3 symlink.

It's not pretty either, something more robust should be
invented some day.

Basically forgot to add after initial testing.

It came unnoticed that USE_IFPLUGD clashes with
CONFIG_WIRELESS which is the default for eth ifaces.

It's a bit heavy but still works.

That one conflicted with alterator-setup
by removing what it needs... should be done
differently.

Really no need to test and sleep when there's sync mode.

- control and sound features

It used to be a hasty static string; something like
"ALT Linux regular-e17/x86_64" seems more relevant.

The extent of this feature's action should be clear
as well as its prerequisite.

rootfs presented a special case when there is no resulting
directory at all as it gets merged with the target subprofile
by design.

Still those features adding only rootfs scripts need to depend
on it but this resulted in an attempt to process a missing subdir.

This is brought back to sanity now.

`make check' must be included in my release process :-/

vm/.cubox-gtk already pulls +nm in, no need to do it again.

Few things:

- extend feature specification
  + SysVinit can be chosen explicitly via init feature,
    no need to keep sysklogd in yet another pkglist;
  + power management should be included too
    (both cpufreq setup and power button handling);
  + LILO seems to be heavily preferred among the
    target audience :)

- use desktop installer for regular-server
  + the seeming controversy is explained easily:
    installer-distro-altlinux-generic has very few
    modules to the point of being inconvenient for
    anything but quick rounds of basic testing,
    and distributions rather do need network setup
    along with a non-privileged user.

The funny thing while debugging this was "how the heck
could a sound related change induce privilege related shift?"
-- turns out that udev-alsa (which pulls in ConsoleKit)
was the culprit... looks like LXDE hasn't dumped it yet :)

Factor of 2 is not enough for kernel upgrade due to filesystem overhead.
Add BOOT_SIZE_FACTOR on top of that.

Maybe the conditionals shouldn't be too terse, by the way...

...instead of setting up pam_mktemp ourselves
even if that was no big deal too.