IF YOU WOULD LIKE TO GET AN ACCOUNT, please write an
email to Administrator. User accounts are meant only to access repo
and report issues and/or generate pull requests.
This is a purpose-specific Git hosting for
BaseALT
projects. Thank you for your understanding!
Только зарегистрированные пользователи имеют доступ к сервису!
Для получения аккаунта, обратитесь к администратору.
As discussed while tweaking power feature, e2k's powerbutton
handling scheme differs slightly at the moment so it does
need some special handling.
No need to toss firmware-linux around if it's not there though.
Kernel flavours differ on secondary arches (and there's a bunch
of platform-specific kernels for ARM systems), so let's provide
some more flexible scheme than hardwiring un-def into images.
This might be further enhanced to e.g. set KFLAVOURS to
$$(KFLAVOURS_LATEST)) with it being preset to un-def
where available so rpi images could set it to rpi-un,
and still get reasonable starterkits for free.
The corresponding commit in RELENG-e2k-p9-starterkits
private branch just tore un-def apart => unmergeable.
This is a controversial feature right from the start,
but keeping jeos alive on non-x86 (that is, catching
every case when a removed removed package would bring
something critical along with it) is somewhat harder
so just skip this part there for now.
There's some peculiarity in writing ISO images for e2k
onto USB Flash media: we need to arrange a partition
and ext2 filesystem with its label passed to propagator;
the former part is what's expected, and the latter one
is what this script does.
Thanks to aerodisk.ru guys for catching a brown paper bag
class typo making format() try to affect the image and not
the drive, by the way.
There are two different versions of the script with the same
name for different subprofiles, sharing some of the same bugs
over the years, *sigh*.
Shoving those into stage2 would yield online repositories enabled
there too which *might* be wanted but will likely need installer
support (at least some better error handling, the current culprit
with unavailable repositories being apt-pkg's size estimation).
Fixes: d2f892abda
Fixes: 0f4ee98fa4
- x11: Added missing xorg-dri-armsoc for armh
- oem: Added rootfs-installer-features
- server-v: Added more commits by andy@, shaba@
- Set default timeout 60 seconds for syslinux and grub
- Set default item to install2 for syslinux and grub
- main.mk: Added vm/, ve/ targets into everything target
- Extended e2k support in distributions (thanks mike@)
- education: fix build and install
There was something wrong about that order of additives
and the base configuration; let's make it straight,
putting the foundation first and architecture-dependent
additions on top of it.
Note that double-colon syntax is specifically allowing
for multiple rules adding up their recipes (and prereqs),
but all of those rules must be specified as double-colon
(this was violated in the previous state of alt-server.mk).
See-also: http://www.gnu.org/software/make/manual/make.html#Double_002dColon
Fixes: 33436b6b1b
Fixes: c06f87539d
This package contains both the fan control service
(disabled by default so sort of safe) and the manual
fanspeed.sh (that we rather need to provide out-of-box).
The problem at hand is that sem@ explicitly disables acpid
within installer-distro-simply-linux; early shots at
ALT Education 9.0 for Elbrus suffered from this as e2k's
power button handling differs slightly regarding events
generated so we *do* need acpid{,-events-e2k} to poweroff
on power button press.
Hope this doesn't introduce regressions or woes for anyone.
Finalize the move to shared multiseat support on e2k
following education.mk; should have been done properly
in the first place but I couldn't manage that for 9.0,
let's at least fix this now.
This includes merge of RELENG-e2k-edu private branch current state
along with the move to shared multiseat support on e2k and assorted
updates following pkg.in/.
This should have been done when reworking Workstation,
but better that late than making a copy for Education.
NB: directory name has been chosen to both sort late
(so that functionality related checkboxes come first
and graphics setup related ones sit down there), and
to relate to X11 as well.
There are literally no office packages on riscv64 so far,
let's ease the need to count that in within pkglists like
it was done in slinux/misc-base since this approach makes
for unreadable conditional chunks if we look at LO-still
being absent on e2k; a common place should be in place.
