►
From YouTube: 20200211 SIG Arch Conformance
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
B
Me
February
12,
2020
conformance,
subject,
meeting
of
cigar
protection,
I'm
your
host
hippie
hacker
today,
and
if
anybody
wants
to
take
notes,
please
feel
free
to
step
up
to
the
plate.
We've
tried
to
populate
the
agenda
accordingly.
Do
note
that
there
is
a
code
of
conduct
that
we
can
concisely
say
be
kind
and
first
on
our
agenda
today
is
some
updates
on
our
behavior
driven
plan
by
I,
assume,
John
and
Jeff
and
I'll
turn
the
meeting
back
over
to
them
shortly.
Unless
I've
missed
some
of
the
protocols
yeah.
A
Well,
that
sounds
good.
Well,
the
first
item
I,
don't
see
John
snarky
on
the
call,
but
I
was
hoping
to
get
that
this
PR
we
were
waiting
for
it
emerged.
We've
been
continuing
work
anyway,
right,
it's
based
on
that
PR,
but
it
was
approved
in
LGT,
m'd
and
like
there
were
so
many
flakes
it
was
just
continuously
flaky
for
like
a
week.
A
C
Hi
yeah
I
just
want
to
illustrate
some
of
the
stuff
we've
been
working
on
around
separating
the
behaviors
tests.
So
in
this
list
here
we
separated
the
behaviors
into
its
own
file,
such
that
it's
not
directly
linked
with
a
test
and
I
got.
Some
of
the
benefits
of
this
was
that
originally
I
didn't
want
to
one
mapping
between
the
behavior
and
the
tests.
Each
line
in
the
conformance
exe
file
represented
exactly
one
test
by
splitting
out
into
a
separate
behaviors,
but
we're
not
able
to
a
lot
of
times
the
test
tests.
Technically
they
tested.
C
Multiple
behaviors
was
enough
and
in
one
test,
because
we
wanted
to
shorten
the
run
time,
but
now
we're
able
to
sort
of
list
exactly
what
kind
of
behaviors
we
want
from
kubernetes
and
each
test
can
hit
multiple
behaviors.
So
hopefully
that
should
make
it
a
bit
more
clear
for
someone
just
looking
at
what
exactly
company
should
do.
This.
C
E
C
C
We're
able
to
have
people
only
reviewing
the
behaviors
without
reviewing
the
tests
because
they
live
in
separate
directories,
and
we
can
also
add
new
behaviors
that
we
don't
exactly
have
tests
for
right
now.
If
we
want
to
add
a
behavior,
we
need
to
first
write
the
test.
Let
us
sit
in
like
for
two
weeks
or
something
and
then
promote
it,
but
now
we're
able
to
directly
out
of
behavior
and
then
when
a
test
is
ready,
we'll
be
linked,
but
we'll
still
have
a
behavior
here
and
yeah.
That's
my
update.
B
B
B
Okay,
if
it
has
some
different
aspects,
I
mean
one
of
them
is
the
front
web
page
and
currently,
although
this
verification
code
seems
to
go,
is
a
is
a
static
page
that
served
up
that
retrieves
data
and
we
have
to
go
through
and
add
more
and
it's
an
older
version,
but
it's
we
were
wanting
to
do
that
where
what
how
we
use
it
in
cluster
and
it
returns
all
the
same.
So
this
is
our
coverage
over
time
that
we've
promised
this
I
guess
its
III
yeah,
so
we're
putting
passwords
on
stuff
for
now,
just
quick.
D
B
Zack
now
I
I,
not
co-op,
it'll,
show
you
some
points
in
time.
If
you
put
your
cursor
over
one
of
the
dots
it
says
at
this
point
there
was
twenty
nine
point,
four
or
five
percent
conformance
test
at
the
11th
of
November
as
we
grow
across
over
here
we
got
up
to
30,
and
this
is
just
depend
on
how
many
data
points
we
add,
and
so
this
is
still
in
progress.
Appreciate
some
there's.
Some
questions
that
I
have
in
our
I
know
it's
there
around.
B
What
we
want
from
this
this
is
Zack
is
on
the
call
as
well.
You
want
to
comment
on
any
of
these
things
that
this
is
currently
loading
last
three
months.
That's
the
questions
that
might
be
reasonable.
How
far
back
do
we
want
to
go
because
our
current
data,
only
everything,
drops
off
after
three
months?
So
if
we
want
to
go
back
to
earlier
releases,
we're
gonna
need
to
specify
different
jobs
that
are
still
around
three
months,
that
we
can
go
back
and
pull.
B
A
A
Are
we
reach
keeping
the
data,
because
we
should
at
least
you
know
like,
like
with
the
number
of
tests,
you
can
go
back
and
all
out
of
get
each
version
and
count
the
number
of
tests
since
we
started,
but
these
test
artifacts
like
we
should
keep
our
cooked
data.
If
we
don't
have
to
go
back
and
reanalyze
or
I,
don't
know
what
we
can
do,
that
stuff
is
cut
and.
B
A
B
We
were
just
to
go
back
and
get
like
stable
default.
What
is
this
is
this
takes
us
back
to
113,
and
that's
you
know,
and
depending
on
out
on
that
date
anyway,
but
I
think
if
we
look
at
these
jobs
and
just
do
all
the
way
back
to
the
third
version
of
stable
back,
it
gives
us
enough
of
a
background
to
see
where
we've
been
yeah.
A
B
B
But
I'm
I'm
open
for
feedback.
What
do
this
is
we're
trying
to
get
with
this
data
that
we
show
on
the
website
is
actually
coming
from
live
data
that's
available
for
other
people
to
start
from
source
to
and
differ
on
the
opinion
of
how
we
calculate
everything.
If
I
just
provided
cooked
stuff,
then
nobody
else
can
take
the
raw
ingredients
from
disagree.
Well,.
A
A
B
A
B
G
A
Think
that
would
be
fine,
Trini
I
think
that
the
it
sounds
like
hippies
concerned
that
he
wants
sort
of
more
transparent,
just
like
he
wants
people
to
be
able
to
it's
a
critical
than
anybody
would
be
able
to
go
back,
and
do
that,
though,
like
what
is
this?
This
isn't
financial
data
or
something
really
their
own
measurement
of
our
own
progress.
It.
B
A
A
E
B
This
is
still
the
report
it
wouldn't.
It
would
make
sense,
I
think
for
the
main
website
to
retrieve
the
the
results
when
that
every
four
hours,
when
that
repeat
so
the
front
website
should
be
current
and
then
I,
don't
know
how
many
data
points
we
want.
So
we
go
back
to
the
beginning
of
the
coverage
page.
This
coverage
overtime
could
show
the
data
points
and
maybe
just
show
it
when
it
changes
that
did
it
agrees
school
at
the
point.
If
it
didn't
it's
not
important,
so
we
can
see
well
when
it
did
increase.
B
B
G
B
So
the
coverage
of
that
we
won't
know
unless
we
look
at
it,
it
shouldn't
change.
Unless
there's
a
PR
against
tests
e
to
e
as
ours
is,
it
could
because
I
guess
the
other
parts
is
where
we're
adding
endpoints
or
anything
to
alpha,
for
example,
or
beta.
That
stuff
will
change
the
way
the
web
UI
or
without
us
focusing
on
the
covert
portion
and
those
things
can
happen
currently
kind
of
outside
of
anything
we
monitor.
We
just
want
to
be
accurate.
B
That
way
you
can
and
I
don't
know
if
we
want
to
go
to
the
point
of
if
there
was
an
increase
or
decrease
in
anything,
why
there
was
a
change.
Why
did
it
change
and
that
may
kind
of
help
give
it
more
of
a
temperature
of
things
that
change
within
KK?
As
far
as
the
API
surface
area,
overall
versus
the
things
that
we
decide
our
conforming,
because
we
also
have
some
things
that
are
just
because
it's
core,
or
just
because
it's
stable
doesn't
mean
it's
applicable
for
compartments
and
maybe
having
this
other
piece
of
metadata.
B
B
If
we
go
down
a
little
bit
further,
we
are
still
working
on
the
blocking
job.
It's
not
forgotten,
it's
just
the
lower
priority.
We
now
create
five
different
images
and
they
are
all
run
by
prowl
on
the
kubernetes
test
in
breath.
I
can
link
to
those
or
show
you,
but
it's
just
cool
to
see
them,
see
them
all
working.
If
I
go.
B
B
Here
is
a
graph
of
up
and
to
the
right
that
is
just
over
time
from
June.
July
is
when
we
started
writing
tests
and
before
that
we
were
just
measuring
it,
and
so
we've
really
been
able
to
see
that
Chris
like
this.
The
web
UI
should
show
this
suit,
like
that
over
time,
we've
just
got
to
figure
out
to
get
those
back
buckets
like
we're
talking
earlier
any
questions
over
the
blocking
jobs
or
the
manual
test
data
or
like
existing
a
case,
new
website,
I.
B
And
well
well,
we
were
going
to
do
a
demo
of
our
workflow,
but
I
think
there's
enough
on
the
agenda
this
week.
They
will
will.
Maybe
do
it
next
time
we're
trying
to
see
if
we
can't
do
a
workshop
for
using
our
workflow
at
them
who
come
you
either
way,
we'll
start
working
with
a
lot
of
folks
to
try
to
get
get
that
going.
But
for
now
one
thing:
that's
really
helped
us
is
our
triage
testing
and
promotion.
Workflow.
B
It's
somewhat
evident
in
these
first
two
I
was
trying
to
go
through
our
core
replication
controller
scale.
The
first
promotion
PR
was
these-
are
all
close,
but
trying
to
show
you
the
time
frame
and
the
reason
these
are
on
here
at
all
is
because
this
one
was
closed
just
recently
by
the
promotion
PR.
But
you
see
when
it
was
first
triage.
B
It
was
back
in
June
and
that's
a
really
long
that
took
us
a
long
time
to
get
to
the
point
where
we're
writing
a
test
and
I
think
we
may
or
may
or
may
not
have
the
promotion
here.
So
this
is
other
other
section.
I
was
trying
to
go
through.
What
is
the
promotion
on
this
particular
increase
in
coverage?.
B
Another
one
that
took
a
fairly
long
time
but
was
was
faster,
was
our
triage
issue
for
coordinating
space
that
took
30
days
to
get
through
and
that's
just
triaging
the
issue.
The
triage
is
always
open
until
the
the
PR
closes
and
so
the
PR
itself
actually
took
two
days
and
because
it
merged
on
Janet's.
That
meant
that
it
was.
We
could
go
ahead
and
get
smart,
get
clayton
to
approve
the
merge
for
the
promotion.
B
P
are
closed
because,
let's
not
break
the
test
apart
and
the
result
was:
let's
go
ahead
and
promote
the
existing
test,
so
it
took
us
from
June
12
2019
to
20th
of
January
to
agrees,
let's
quote
the
existing
test,
but
we
learned
a
lot
along
the
way,
but
it
was
definitely
the
process
has
been
hard.
I
don't
want
to
progress.
Past
I
kinda
want
to
find
a
way
if
anyone
has
any
questions
about
existing
things
before
I
go
too
fast.
Is
this
a
value
also
the
way
that
we're
doing
this?
A
B
Also
to
like
it's
part
of
a
teaching
tool,
I
guess
to
say
when
we
go
through
and
write
a
triage
issue
here
are
here's
the
process.
It
needs
to
be
created
and
then
in
one
of
our
meetings
we
need
to
take
it
from
triage
to
get
into
the
sort
of
backlog
and
then,
when
you
close
via
the
ticket
it
because
it's
I'm
trying
to
make
it
work
the
workflow
of
how
we
make
sure
the
work
is
ready
to
be
done
before
we
do.
B
It
is
something
that
other
teams
and
other
groups
can
do
as
well,
and
maybe
we
can
simplify
this
I
just
wanted
to
be
pretty
verbose
in
at
least
one
meeting
so
that
I
could
go
back
and
when
I
get
new
test
writers
on
I
can
say.
Look
here
is
how
we
run
meetings.
You
know
how
we
get
good
through,
because
the
target
here
is
increasing
coverage
and
it
just
takes
these
three
steps
and
I.
Don't
I,
don't
know
I'm
open
for
feedback
I'm
trying
to
be
really
verbose,
but
maybe
I'm
being
too
verbose.
I.
D
B
B
B
Let's
see
these
are
okay,
like
they
all
just
merged.
There's
no
problem
here
the
story
encore
be
when
limit
range
weather,
wasn't
a
settlement
on
the
tests
overall
and
I
think
this
is
where
we
started
to
see
the
increase
in
speed
and
velocity
where
we
started
in
the
19th
of
December
and
we
merged
the
promote
on
the
30th
of
January,
and
we
got
three
in
points
in
Greece.
B
If
we
look
at
the
actual,
you
know
mock
test
itself,
we
have
Emacs
or
file
that
generates
this
entire
thing
and
the
workflow
we're
using
at
that
point
was
deploy
a
P,
I
snoop
and
select
things
that
get
to
care
of
interest,
and
this
is
the
actual
three
employees
and
documentation
write
the
steps
and
run
this
particular
snippet
ago
code
that
results
in
this
particular
set
of
output
and
make
sure
that
your
test
actually
does.
This
is
it
hit
by
your
new
test
yeah.
B
B
Hey,
we
just
open
a
new
PR
promotions,
and
this
is
like
the
next
step
is
so
now.
We
know
like
the
next
time,
we're
only
going
to
have
things
that
are
in
the
triage
or
test
PR
and
we'll
get
to
see
more
of
it.
This
the
reason
this
causes
a
bit
long
with
screw
through
and
show
all
of
it
to
the
point
of
what
happened
since
the
last
I.
B
B
Did
any
of
this
stuff
that
just
get
done
get
done
in
the
last
two
weeks,
and
we
have
something
that's
in
need
through
view,
but
I
think
we
talked
about
last
week
because
we
didn't
do
anything
with
it
because
it
was
it
was
in
the
disk
area,
but
we
should
probably
get
back
to
to
our
friend
wish
I'm,
not
sure
if
it's
Michelle
or
my
Michael
I
don't
know
how
to
pronounce
things
really
well.
But
if
you
want
to
respond
here,
we,
it
might
be
good.
B
So
we've
now
assigned-
and
that's
in
the
needs,
review
and
conversation
I'm
working
on
PR
gate,
but
mainly
the
infrastructure
underneath
it
not
the
thing
just
yet.
This
was
fixed
tests.
We
opened
this
woman.
This
was
create
test.
The.
B
A
B
B
A
A
F
B
What
we
got
on
that
one,
and
then
that
was
I-
should
open
these
all
at
once
that
what
the
one
we
just
did
right
and
then
promote
I've
got
a
promotion
that
is
back
over
here.
So
this
bills,
I,
was.
A
B
A
B
Me
I
want
to
try
to
pick
notes
on
each
one
over
here.
So
we're
saying
that
this
is
in
progress.
Is
the
correct
state?
Can
we
mention
our
group
somehow,
like
employments
I,
can
just
effectively
touch
this
and
look
at
it
during
our
meeting,
because
I
feel
like
I
come
on.
It
looks
at.
Is
me
like
I'd
love,
to
comment
on
it,
as
the
group.
B
B
B
B
B
A
B
It
back
in
the
triage
and
keep
it
at
the
top
of
the
tree
off,
but
it's
not
prioritized
yet
like.
Where
does
it
say
currently
I'm
fine
with
that
yeah
I'm
gonna
open
this
thing,
so
this
is
create
priority
class
thing.
I'm
gonna
go
back
into
our
our
board
and
this
is
create
critic.
Last
I'm
gonna
put
it
back
into
the
triage
and
I
kind
of
want
to
have
like
maybe
it's
even
at
the
bottom
of
the
sort
of
backlog,
I'm,
not
sure
because
it's
it's
been
sorted.
B
F
B
B
B
B
B
C
A
B
B
That
was
what
we
just
did
rewrite
that
test
and
the
new
API
chunking
test
and
then
there's
a
promotion.
That's
just
after
this.
That's
next
one
with
it
was
so
a
few
minutes
ago.
He
was
born
so
for
this
one
we
test
retest
retest,
he
I
was
updated
because
we
wanted
to
do
pod
templates.
They
said
big
maps
would
be
easier
or
more.
Generic
I.
Remember
that.
B
B
D
B
B
Do
you
want
to
resubmit
these,
as
with
our
new
workflow,
just
to
make
it
quick
and
clear,
because
it's
it's
gotten
a
little
bit
hairy,
so
I'm
gonna
go
through
the
conversation
and
I'm
going
to
ask
the
right,
tutor
pen,
so
closing
and
I
wish
I
could
say
who
we
were
like
closing
via
this?
Is
our
cigars
and
performance
of
meeting
at
Bob's.
B
A
E
E
B
B
B
B
This
one
is
good
and
then
we
did
some
tests
that
failed
and
then
loading
all
the
hidden
things
renamed
it
retired
it
more
testing,
and
then
we
had
some
flakes
and
dims
top
started
hearing
about
it,
weighed
about
said:
I'm
done,
dude
rebase
it
here
II
based
it
and
it
got
to
be
small
when
he
based
it,
which
is
great
and
I,
think
we
just
need
to
retest
now
the
rebase
rebate.
So
that's
where
you
sing
that
so
quickly.
It's
this!
This
has
conflict.
It's
not
ready,
yeah,
really,
gray,.
A
B
F
B
B
B
Get
this
flow
down,
thank
you
and
you
dead
stack
and
I'm
like
the
scroll.
He
doesn't
work
as
well.
So
there's
a
pod,
spec
I
think
these
I'm
gonna
go
ahead
and
close
these.
These
were
ones
that
we
were
doing,
based
on
umbrella
issues
and
old,
workflows
and-
and
it's
just
good
so
I'm
just
going
to
go
ahead
and
say
close.
B
For
that
move
it
yeah
it
completely
moves
that
other
thing:
wonderful
and
I.
Don't
this
maybe
I
was
seeking
consensus
on
watch
and
I.
Think
we
we
said
no,
but
let's
go.
Consensus
is
needed
for
the
operation
end
points,
and
this
was
an
issue
right.
That's
an
issue
and
we
reviewed
it
and
I
think
this
may
be.
My
mic
cap
know
it's
an
issue.
I.
B
B
B
B
Cool
the
clothes
we're
rocking
it
out,
go
back
to
our
our
list
of
that
was
our
in
progress,
and
we
got
this
test
images.
Windows
support,
get
two
minutes,
I
think
you
didn't
done,
yeah!
No
thank
you,
let's
scroll
down
and
when
those
images
step,
one
I
suspect.
Next,
let
me
come
back.
We'll
have
a
whole
bunch
of
stuff
in
the
to
be
triage,
but
we
just
didn't
get
to
it.
This
week.
B
B
Go
back
to
make
sure
we
did
the
thing
so
in
general,
I
think
we
did
a
lot
of
really
good
stuff
today
and
learn
and
I
think
we
note
that
the
process
has
been
painful
in
the
past
and
it
will
be
less
painful
in
the
future.
I
look
forward
to
pairing
with
anybody,
who's
interested
on
our
workflow,
that
captures
increase
tickets,
all
with
it
in
cluster
UI
and
eventually,
hopefully,
something
that
everybody
can
use
so
have
a
wonderful
week
and
we'll
I'll
give
you
back
30
seconds
of
your
day.
All.