The reason is that this is more appropriate if for example we modify
patches on commit not yet build but released older version. Such state
would not compile so instead we have to choose packages version. The
changes in updater's lists should not have such effect.

This installs blkdiscard and fstrim no matter on board.

This also removes resiserfs as upstream removed reiserfsprogs and only
kernel module is now present. In the end it makes no sense to have it in
lists as it is unusable.

It compiles for now, let's enable it.

This reverts commit 9a43c4ca.

I am able to compile it locally and buildbots in OpenWrt compiles it as
well.

This reverts commit 0857a679.

Turris 1.1 was missing USB3 driver for its front USB port.

turris-survey is not available in Turris OS 4.0 so it can't be
installed.

The patch is included in the kernel (master branch) and it was cherry-picked and included in
kernel 4.14.158.

More details:
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/drivers/net/dsa/mv88e6xxx?h=v4.14.158&id=ef36d52351c96506328043423655cf674173d8e1

Some packages could fail to run postinst because libc is already
replaced but busybox is not yet replaced. This is just potential
problem.

The reason for this is to have atleast somewhat working shell on Turris
1.x when migration fails.

This removes provide of wget of itself.

Karel Koci authored 5 years ago and Karel Koci committed 5 years ago

This is not clean patching and we do not pretend it is so just ignore
those warnings.