►
From YouTube: NixOS 20.03 Markhor Go / No-Go #2
Description
Blocker review meeting.
A
That's
plenty,
that's
probably
plenty.
Okay,
let's
get
it
going
hello,
everyone.
This
is
the
Knicks
of
us
no-go
meeting.
This
is
the
meeting
where
we
go
over.
Blockers
identify
if
Knicks
of
us
is
ready
for
release
and
I
am
wrote.
The
piece
2003
release
manager
and
we
have
Samuel
here,
who
is
2009
release
manager
as
micro
release
manager
and
if
we
can
quickly
identify
who
is
present
for
the
meeting
for
certain
subsystems.
A
So
if
I
have
already
identified
you
as
like
a
participant
or
if
you're
up
just
observing
I'm
just
going
to
go
over
the
participants
right
now
so
Linus,
is
there
any
subsystem
that
you're,
representing
in
this
meeting
today.
C
A
D
A
A
D
Okay,
wonderful,
so
this
is
a
problem
with:
let's
encrypt
certificates,
we
changed
the
let's
encrypt
certificate
provider
from
sip,
led
to
legal
and
there's
currently
a
problem
when
you,
when
you
have
multiple
certificates,
which
have
the
same
name
but
I
actually
like
have
different
different
additional
names.
Now,
there's
basically
some
internal
state
overriding
and
in
this
discourse
regressions
when
people
switch
from
1909
to
2003
and
I.
Wasn't
aware
of
this,
but
I
think
this
is
definitely
something
critical
that
needs
to
be
fixed
before
we.
A
Exactly
I
think
we
merged
on
a
PR
that
changed
this
module
greatly
rightly
released
was
back
partner,
I
think
I.
D
D
D
A
A
C
A
Was
there
like
me,
sir?
It
was
like
just
briefly
summarize
what
this
is
actually
for.
I
guess
the
the
opie
of
this
pull
request
would
be
sufficient,
or
maybe
it's
repeated
here,
I'm,
not
sure
if
both
of
these
are
needed.
So
it
says
this
is
about
a
change
from
upstream
DHCP.
Cd
is
backwards
incompatible
as
this
locked
me
out
of
my
box.
This
is
my
setup
and
this
today
nothing
else
is
what
they
upgraded
from.
A
This
is
a
configuration
and
when
I
rebooted
the
system,
I
notice
that
vr0
didn't
doesn't
get
an
ipv4
address
anymore.
The
reason
was
that
DHCP
CD,
didn't
hoc,
didn't
consider
a
purchase
interface
that
would
get
an
address
by
default,
and
then
they
certainly
made
some
configuration
change.
So
I
believe
this
PR
does
documents
such
configuration
change.
C
A
A
So
I
think
that
both
of
these
are
probably
fine
when
you
get
someone
who
review
the
assertion
that
is
the
proper
raps
to
do
this
is
the
document
and
then
assert
at
the
same
time
in
case
someone
doesn't
read
the
release,
notes
so
I'm
guessing
that
probably
both
these
are
fine
to
be
proud
today.
What
do
you
think
Samuel.
A
A
Okay
and
I
think
the
last
thing:
besides
the
more
failing
to
build
packages
and
turning
or
three
years
broken,
that's
just
something
we're
gonna
do
when
we
actually
get
ready
to
release
it,
because
that's
just
something
that
has
to
be
done.
Sort
of
all
at
once
is
just
finishing
the
release,
notes
and
I.
Guess
that
is
something
I
haven't.
You
have
obviously
been
putting
off
because
of
you
know
docbook.
A
So
I
will
try
to
look
through
this
thread
and
hopefully
try
to
get
as
many
sort
of
like
attended
upgrades
that
are
in
there
are
things
that
would
be
important.
It
would
be
helpful
if
someone
could
just
do
that.
For
me,
I
know
that's
a
bit
embarrassing,
but
just
writing.
Dog
walkers
sometimes
is
so
difficult
inconvenient.
D
A
Great,
that's
actually
my
preferred
way.
Yeah
I'm,
like
us,
my
preferred
way
to
actually
write
doc,
look
is
to
translate
it
out
of
markdown,
so
yeah
I
can
definitely
do
that.
I
was
considering
that
before
this
call,
if
someone
could
just
translate
that
for
me,
so
I
will
do
that.
Probably
after
this
call,
since
I
have
the
time
for
that
and
I
guess,
maybe
it
like
sorts
of
Locker
PRS
differently
know.
A
So
we
have
went
over
all
the
blockers
and
it
doesn't
appear
that
were
exactly
finished,
but
it's
definitely
we
could
mark
a
goal
date
and
how
this
be
done,
but
we're
not
going
to
release
without
it
so
say,
that's
technically,
not
a
goal
like
because
we're
not
exactly
ready,
but
if
we
just
set
a
goal
date.
We
can
just
probably
have
this
done
in
like
a
few
days.
C
C
C
A
A
We
already
have
a
similar
sort
of
errata
that
would
be
in
the
release,
notes
or
like
Intel
GP
is
the
newest
with
the
default
kernel,
which
I'm
not
sure,
I
followed
up
on
that
with
having
sort
of
patches
to
fix
those
issues.
A
So
you
might
have
missed
that
I
think
we
already
decided
that
we
would
release
with
those
issues,
issues
with
Intel
GPUs.
B
A
Think
I
talk
gram
that
we
would
like
sort
of
out
a
tracking
issue,
because
we
agree
that
it
did
like
effect
a
lot
of
people
but
like
you,
can
just
use
4.19
and
that
just
works
fine
and
doesn't
have
the
regressions.
But
the
default
colonel
will
have
those,
but
so
I
just
think.
We
thought
that
was
just
something
that
we
added
to
the
release
notes.
But
I
didn't
get
to
like
adding
the
sort
of
issue,
because
adding
with
many
references
is
kind
of
difficult.
Once
it's
since
it's
a
wide
spread.
A
B
Seeing
why
do
we
think
we
have
an
issue
with
the
five
five
colonel
I'm,
not
quite
understanding
that,
with
until
GPUs
as
I
call
them
tell
GPUs.
A
B
A
D
D
Yeah,
so
this
is
an
issue
that
popped
up
while
we
we
changed
the
nginx
process
to
by
default
not
run
as
root,
but
instead
like
run
it
as
the
user
and
it's
supposed
to
run
der
worker
slaves,
and
this
caused
some
problems.
If
people
still
want
to
have
nginx
master
process
run
a
fruit
for
some
use
cases,
but
I
think
that
the
fix
should
be
somewhat
trivial.
We
just
need
to
add
a
little
bit
more
logic
in
handling
the
the
files
that
nginx
tries
to
write
to
in
vast
pool
nginx
and
it
yeah.
D
It
basically
needs
some
some
our
documentation
and
maybe
like
a
small
fix
understa
said
he
is
gonna,
take
care
of
this
hopefully,
and
it
it
should
be
quite
trivial.
I
just
wanted
to
mention
it.
So
if
there
is
a
back
port
for
this,
so
that
people
don't
wander
but
I,
think
it's
it's
gonna
be
quite
easy
to
fix.
D
C
A
D
A
I
often
definitely
approve
on
I'm
sure,
okay,
cool
and
I
think
the
person
that
who
proposed
this
last
time
that
you
weren't
so
sure
about
hasn't
I
might
have
to
ping
him
about
it.
So
I'm
not
exactly
sure
what
it
wasn't
exactly,
but
it
was
at
Raspberry,
Pi,
3,
B
pluses
in
boot,
and
that
is
a
device
that
we
support
officially
and
I.
Think
last
meeting.
It
wasn't
reproduced
by
Samuel,
I,
think
and
I'm,
not
sure
where
we've
gotten
since
then.
A
You
know
it
seems
be
yeah
I'm,
not
sure
about
that
I.
Don't
really
specialize
in
these
sort
of
things.
A
A
So,
okay,
back
to
the
Walker
decision,
you
do
agree
that
we're
ready
to
make
some
sort
of
decision
Samuel
right,
yep.
B
A
Let
me
think
I
could
probably
be
fine
with
like
next
week,
Wednesday.
A
It
is
kind
of
a
bit
late,
so
I
think
we
should
try
to
push
it
out
as
best
as
we
can
would
be.
Okay,
if
I
could
proceed
without
you.
B
A
B
B
D
C
D
Is
kind
of
already
documented,
like
in
inside
this
issue,
it
probably
is
just
an
hour
to
of
taking
a
look
at
it
and
testing
ensuring
this.
This
works
in
the
test.
Okay,
like
if
you
have
someone
who
can
take
a
look
at
it,
that
would
be
great,
but
otherwise
I
would
try
to
sum
well,
somehow
put
it
inside
Easter
weekend.