►
From YouTube: 20200130 SIG Arch Community Meeting
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
A
All
right,
so
there
were
no
action
items
from
the
last
meeting,
looking
just
our
standard
agenda,
so
we
don't
need
to
worry
about
that.
I
would
like
to
request
that
anybody.
It's
been
a
while,
since
we
had
an
update
from
the
tech
debt
research
team.
So
if
anybody
is
on
that
team,
I'd
like
to
just
put
it
out
there
that
we'd
love
to
hear
from
you
next
next
meeting,
which
is
in
two
weeks
and
next
item
on
the
agenda-
that's
pretty
light
agenda
today,
but
we'll
move
through
it
as
we
can.
A
There
is
the
contributor
summit
coming
up
in
in
the
March
and,
of
course,
Q
Khan.
So
anybody
here
that
is
interested
in
submitting
a
talk.
The
CFP
is
still
open,
I
think
until
February
7th,
so
please
go
ahead.
I
think
that
we'll
probably
submit
one
for
production
readiness.
But
beyond
that
I'm,
not
aware
of
anything
at
the
contributor
summit
or
cigar
ch,
we
do
have
a
intro
and
status
update
in
the
larger
Q
con
conference.
A
B
Alright,
so
I
wanted
to
just
highlight
the
things
in
the
118
milestone
that
involve
API
changes.
There
are
the
two
right
at
the
top:
are
the
oldest
ones
in
the
list.
These
are
the
certificate
signing
request,
API
and
ingress
v1.
That
is
the
priority
for
API
review.
This
ties
in
to
the
conformance
efforts
to
make
sure
that
all
the
API
is
that
we
depend
on
for
bringing
up
clusters
are
at
v1
or
higher,
and
so
getting
these
two
specific
API
was
graduated
will
enable
that
ingress
is
targeting
graduation
to
b1.
This
release
certificate.
B
There
are
various
and
sundry
other
proposals
that
are
still
making
their
way
through
the
review
process.
One
that
has
been
particularly
thorny
is
the
IP
family
handling,
which
effects
pod
re
effects.
Services,
Clayton
and
Tim
and
I
have
been
working
through
some
of
the
issues
with
that,
and
so
that
is
influencing
how
we're
looking
at
other
proposals
to
add
fields
to
existing
types
so
expect
to
see
some
of
the
guidelines
updated.
B
We
we
made
some
mistakes.
Fortunately,
it
was
still
in
the
alpha
phase
for
the
IP
family
field,
but
in
thinking
through
how
to
fix
those
mistakes,
it's
just
getting
corny
or
Dornier.
So
this
is
why
we
do
alphas
and
why
we
figured
things
out
before
we
enable
them
for
everybody,
but
the
likely
outcome
will
be
that.
B
Making
changes
to
existing
objects
will
be
more
restricted
to
the
future,
which
is
unfortunate,
but
that
is
those
are
the
main
things.
I
wanted
to
call
out
feature
freeze
or
the
design
freeze
was
this
past
week,
and
so
a
lot
of
these
they're,
their
final
designs
are
in,
and
some
of
them
passed
a
peer
review
already.
Some
of
them
are
deferring
a
peer
review
to
the
implementation,
which
is
fun,
but
we
working
through
those
as
we
proceed
in
180.
A
So
Jordan
do
we
have
one
thing:
I
haven't
never
went
clear
on
from
the
API
review
program.
It
was
do
we
have
an
idea
of
that?
We
have
the
reviewer
bandwidth
like
I,
don't
know
if
we
have
I
guess.
You've
got
a
kind
of
bench
of
folks
that
you've
you've
run
up
over
the
last
year
or
two
I
guess.
My
point
is:
if
we
need
like
extra
review,
does
this
look
normal
to
you
the
amount
and
it
will
have
enough
capacity
or
we
need
to
start
thinking
about?
B
Always
thinking
about
recruiting
that's
why,
whenever,
whenever
we
are
doing
reviews
or
trying
to
pair
with
people
from
that
sig
and
that
component
to
make
sure
that
they're
getting
context
and
getting
drained
so
that
they
can
start
doing
those
reviews
as
well?
It's
always
hard
to
tell
at
this
point
in
the
release.
There's
a
lot
of
ambitious
things
and
a
fair
number
of
them
don't
actually
even
make
it
to
the
limitation
point
to
where
they
are
ready
to
be
RAM.
Okay,
okay,
I
would
say
this.
A
C
Think
we're
good
with
that.
We've
got
some
links
if
you
want
to
follow
them,
but
they're.
The
main
thing
is
on
the
front
page
of
API
snoop
that
Co
our
core
conformance
or
our
stable
coverage,
is
up
2%
from
30
0.37
over
the
holidays,
to
32
point
four
two
and
a
lot
of
that
coming
from
finally
merging
PRS
that
that
the
team
has
been
working
on
and
I'm
getting
them
promoted.
So
we
had
a
lot
of
promotions
and
a
lot
of
the
velocity
is
increased
and
that's
been
in
part
due
to
the
test.
C
Writing
workflow
improvements
that
we've
been
in
grading
and
the
theme
to
the
start,
with
issues
that
are
well
researched
for
a
particular
endpoint
and
look
at
the
documentation,
and
we
use
our
bi-weekly
informants
meetings
to
to
move
those
from
the
unsorted
into
the
sorted
backlog
so
that
those
we
kind
of
get
approval
that
this
is
a
good
test
and
they
that
we
have
sample
code
blocks.
That
say
we
will
actually
run
those
code
blocks
to
see
that
they
actually
hit
in
cluster
and
increase.
C
So
the
ticket
itself
can
say
precisely
how
many
endpoints
this
approach
will
cover
as
far
as
increasing
our
endpoint
coverage,
as
well
as
getting
agreement
on
how
the
test
will
operate
before
we
write
the
test.
So
those
issues
then
can
very
straightforwardly
turn
into
pull
requests
that
are
well
written
and
easy
to
get
merged
because
we've
already
agreed
it's
a
good
approach
and
then
two
weeks
after
those
merge
we're
able
to
easily
get
them
promoted
because
it's
already
been
agreed
that
it's
a
good
idea.
C
C
We
finally
got
this
CN
CF
API
snoop
and
the
CN
CF
conformance
working
group
repos
connected
to
prowl,
and
we
have
the
cow
and
the
various
small
things.
But
the
first
jobs
that
have
been
set
up
are
to
create
all
of
the
API
snoop
images,
and
then
that's
now
done.
The
next
steps
are
establishing
our
baseline
for
comparison
and
we've
got
shared
agreements
on
c
I
kubernetes
GCE
GCE
gcci
gcci
DC
as
the
that's
the
job
that
we
will
compare,
and
that
runs
every
four
hours
so
in
in
the
near
future.
C
Api
snoop
that
csefel
will
be
dynamically,
updated
every
time
that
that
job
runs
to
give
us
a
new
baseline.
That
will
compare
the
PRS
against
and
list
the
specific
endpoints
that
are
either
added
and
untested
or
added
and
conformance
so
each
PR
you
will
know
which
PR
increased
and
then
later
on,
we
will
start
disallowing
ones
that
removed
coverage
and
that's
that's
the
conformance
sub-project
update
back
to
you,
John.
A
D
Don't
see
him
on
the
call,
but
I
read
the
issue
and
I
commented.
The
issue
is
about
the
naming
of
taints
due
to
the
urban
dictionary
definition
of
it
versus
the
well-established
computer
science,
literature,
so
I
I'm
I
think
at
this
stage
of
the
game,
just
because
there's
so
much
history
there
and
it
would
require
making
such
a
drastic
changes
across
API
that
it
would
break
so
many
people
I'm
opposed
to
this.