►
A
Okay,
hi
welcome
to
a
breakout
discussion
to
talk
a
little
bit
about
the
status
of
Patrick.
All
these
efforts
on
refactoring,
the
e2b
framework
sort
of
where
we're
at
and
what
direction
we
were
headed.
I
think
this.
You
talked
a
little
bit
about
those
traits
and
testing
on
Tuesday,
but
since
the
main
objections
were
raised
by
Tim
st.
Clair-
and
he
wasn't-
president,
we
felt
like
all
we
were
doing-
was
playing
guessing
games
so
yeah,
no
just
take
it
away,
Tim.
So.
B
The
biggest
objection
I
have
was
directly
so
inside
of
the
intent
testing
framework.
We'd
use
things
that
are
kind
of
a
little
bit
weird
and
that
we
do
binary
smashing
so
binary
smashing
is
basically
you
luck,
all
the
symbols
of
X
into
your
binary
and
then
the
way
we
build
the
testee
to
e-file.
It
is,
is
kind
of
different
right
like
it's
this,
it's
an
executable
that
that
people
can
run
independently,
but
they're
actually
they're,
actually
tests
right.
So
like
the
way
goaling
does
a
split.
B
Pieces
and
I
can
understand
that
in
the
long
run,
if
we
have
a
plan
to
piecemeal,
take
out
those
integration
pieces
as
a
separate
entity,
but
this
actually
conflates
the
dependency
graph
right
now
so
by
been
smashing
those
extra
pieces
of
all
the
different
providers,
that
included
as
your
and
OpenStack
and
other
stuff,
and
that's
the
opposite
objective
of
what
we're
trying
to
get
to.
So
what
I
was
missing
here
was
if
this
is
an
incremental
step
for
a
larger
plan
to
move
this
out
of
tree
or
if
this
was
like.
B
C
B
C
B
C
That's
right,
that's
right!
So
my
and
my
my
solution
is
what
you,
what
you
observe
that
the
decision
to
still
have
those
provider
support
in
the
end-to-end
test,
binary
that
is
now
done
at
V
test
suite
level.
So
it's
not
a
framework
anymore.
That
depends
on
a
certain
provider.
It's
V
test
suite
and
therefore
it
is
another
control
of
a
test
suite
offer,
whether
he
wants
say
Google
Cloud
support
for
Cuba
net
is
the
current
situation
is
as
you
as
well,
because
it's
a
hard
dependency,
the
the
test
binary.
C
The
offer
doesn't
really
have
a
choice.
He
has
to
have
those
different
providers
and
my
pull
request
doesn't
change
that
status,
because
I
don't
want
to
break
for
testing
as
it's
currently
done
in
Cuban
it
is.
It
was
supposed
to
be
just
refactoring
at
all
I'm,
not
trying
to
change
the
scope
of
what
Cuba
Nettie's
is
testing.
I
know
that
there
are
lots
of
discussions
now
happening
or
starting
in
this
new
work
group
where
people
are
trying
to
figure
out
whether
Cuban
it
is,
should
be
blocked
by
a
test.
C
B
You
kind
of
stepped,
you
stepped
into
a
very
thorny
mess
and
tried
to
do
the
best
thing
to
clean
it
up
and
I
totally
get
that
it
makes
a
lot
more
sense.
What
you're
doing,
though,
I
understand
you're
pushing
the
the
dependencies
over
to
here?
I.
Think
one
thing:
that's
missing,
though,
that
we
we
can
probably
address
in
a
separate
issue
which
we
should
log,
is
that
what
are
we
going
to
do
with
this
cloud
provider?
Integration
code,
because
it
is,
it
is
a
real
problem
on
the
intent
test
framework,
so.
A
I
feel,
like
that's
part
of
the
road
map
that
or
or
you
know
you
used,
Tim
I
feel
like
I
heard.
You
say
you
weren't
really
sure
what
the
plan
was
and
maybe
what
I'm
hearing
from
Patrick
is.
His
plan
didn't
go
quite
as
far
as
that
and
so
you're
talking
about
putting
together
a
separate
plan,
but
it
did
seem
to
me
like
Patrick's
efforts
at
refactoring
or
such
that
conceivably.
A
C
And
that's
that's
exactly
it!
That's
exactly
what
I
wanted
to
add
myself
with
that
first
pull
request
in
place.
What
you
can
do
is
now
go
to
a
cloud
provider
and
tell
them.
Ok,
we've
split
out
your
code
into
this
one
particular
package,
and
if
you
want
to
run
your
this
framework
yourself
on
your
own
see,
is
it
set
up,
then
you
become
responsible
for
your
cloud,
provide
a
specific
code,
we're
going
to
take
it
out
of
Cuban.
B
I'm,
totally
cool
with
the
game
plan
of
opening
a
separate
issue
for
a
long-term
plan
of
finding
a
way
to
refactor
the
cloud
provider
integration
for
long
term
support
you
are
going
to
have
because
the
framework
has
this
plumbing
for
provider
through
people
are
going
to
loop
in
these
dependencies
in
some
way
shape
or
form,
which
will
be
non-trivial
so
long
as
we
log
the
issues
to
deal
with
this
I
think
that
seems
reasonable
for
right
now,
it's
a
step
in
the
right
direction.
I
think
what
was
unclear
to
me
is
like
it.
B
A
I
had
that
maybe
is
somewhere
in
the
direction
either
of
you
are
ahead
of,
is
what
would
it
take
to
turn
this
into
a
repo,
a
staging
repo?
Basically,
I
am
led
to
believe
that
if
you
get
your
code
and
staging
you
write
it
in
such
a
way
that
it
doesn't
really
depend
upon
any
of
the
tangley
bits
that
are
elsewhere
in
KK
and
that
might
make
it
even
easier
to
reuse
and/or,
eventually,
one
day
extracting
its
own
repo
yeah.
B
I
think
that's
a
tiny
city
on
a
hill
which
doesn't
exist,
but
the
it
does
give
you
some
benefits,
but
they
come
at
a
different
cost
right.
So
if
we
do
this,
you
need
to
create
a
separate
repo
because
the
way
people
import
it
changes
now.
Yes,
so
so
you
can
import
it
right
now
by
just
having
the
full
long
ridiculous
path.
He's
a
framer.
C
Now
they
have
to
and
I
agree
that
this
is
not
ideal.
Besides
for
conceptual
or
dvd,
besides
dependency
issues
we
can,
which
can
be
avoided
by
just
being
careful
in
in
what
you
allow
in
that
particular
package.
It
is
a
it
is
a
practical
problem
and
trying
to
do
vendor
this
with
depth,
for
example,
because
you
are
forced
to
well
actually.
If
this
is
the
only
package
that
you
pull,
you
can
still
do
it.
Whatever
you
can
do
book,
you
can
pull
whatever
revision
you
want.
C
A
D
A
A
Maybe
that's
why
I
Hans
is
here
about
like
trying
to
develop
reusable
frameworks
that
encourage
or
facilitate
testing
of
kubernetes,
both
at
an
integration
or
et
level
and
half
those
live
completely
out
of
tree,
so
that
we
can
start
to
make
it
easier
for
anybody
to
write
and
tests
against
kubernetes
or
integrationist.
Whatever.
C
Yeah
sorry
I
can
certainly
sorry.
I
can
certainly
help
with
moving
about
forwards
and
I'm
interested
also
personally
in
getting
it
into
a
staging
and
under
a
different
import
path.
So
I
can
certainly
help
with
that,
but
only
event.
I
would
still
prefer
to
get
bad
pull
requests
as
it
is
now
get
merged.
First,
yeah.
B
Of
course,
we
we
can
I
can
unblock
that
this
afternoon
or
right
after
this
meeting.
The
the
second
step
of
staging
is
not
as
clean
and
clear.
You
will
need
to
do
a
couple
of
things
and
I'd
highly
recommend
opening
an
issue
first
before
you
do
that,
because
you
need
a
separate
repo
and
that
repo
needs
to
get
approved
and
then
there's
all
kinds
of
other
mechanisms
and
jiggery
as
part
of
staging
that
does
this
photo
replication
into
this
fake
of
the
repo.
So.
A
If
I
can
make
a
proposal,
I
feel
like
Timmy,
you
have
a
much
more
articulate
vision
for
what
the
next
steps
are
to
carry
Patrick's
work
to
forward
to
sort
of
the
universe
you
described
where
we
stopped
in
smashing
everything
together
and
I.
Think
like
it
might
be
appropriate
if
you
made
those
those
follow-on
issues,
I'm
just
trying
to
make
sure
we
we
lead
to
actual
outcomes
here.
Yeah.
A
E
Quickly,
safe,
you
like
this
I
like
joint,
so
we
in
certain
manager
basically
had
a
similar
thing.
Well,
not
quite
a
similar
thing,
but
we
needed
an
end-to-end
testing
framework
so
that
require
students
lined
up
keys
and
be
able
to
do
the
majority
of
the
things
that
happened
in
that
framework.
But
in
our
own
project
and
we
went
down
the
route
of
actually
like
I,
we
don't
import
kbk
anywhere.
E
So
that's
taken
a
while
to
get
my
head
around
everything,
but
I
would
definitely
like
to
so
I've
spent
a
few
days
over
the
last
few
days
for
us
refactoring
and
doing
some
work
there
and
I've
gotten
to
the
point
where
I've
been
into
thinking
as
well,
that
this
needn't
be
a
part
of
our
own
repository.
So
whatever
work
is
going
on
here
to
pull
this
out
of
KCK
I'd
love
to
get
involved
with
that
too,
because
you
make
much
more
sense
for
us
to
standardize
on
something
there.
A
Right
yeah,
like
I,
don't
know
if
maybe
the
meeting
topic
was
super
vague
and
that's
why
a
bunch
of
people
showed
up
I
did
try
to
put
in
the
specific
PR
that
we
wanted
to
discuss
in
the
description.
So
it
was
clear
like
we
were
blocked
on
this
here's
the
comment
that
said
we
were
blocks
and
here's
what
we're
going
to
talk
about,
but
I
am
super
interested
in
the
fact
that
a
bunch
of
people
like
the
word
ETA
framework
and
refactor
and
want
to
come,
show
up
and
talk
about
it.
A
So
we
do
have
like
this
testing
common
sub
project
that
I
think
I
guess
Tim
runs
but
I,
don't
know
how
often
yell
at
the
meeting
I've
seen
the
thing
drop
off.
But
if
there's
like
a
7:30
a.m.
Pacific,
which
is
kind
of
European
friendly
since
I
see,
this
is
also
relatively
European
friendly
timeslot.
Maybe
it
makes
sense.
A
Don't
know
if
I'm
counter
to
the
opinions
of
Tim
here,
since
the
project
kind
of
was
born
out
of
a
proposal
he
put
forward
but
I
think
like
having
more
people
talk
about
framework
II
things
in
general
and
if
they
don't
want
to
post
them
in
their
own
bespoke
repos,
but
put
them
up
in
the
kubernetes
SIG's
org
under
the
sponsorship
of
sig
testing,
like
we're
super
I,
think
we're
super.
Ok
with
that
and
seeing
like
what
wins
out,
what
most
people
adopt,
etc,
etc
and.
B
A
The
analogy
I
would
think
of
would
be
client
go
and
all
of
its
intricacies
that
have
been
born
out
of
years
of
experience,
with
working
with
the
kubernetes
API
versus
something
that's
I
just
want
to
go
client
to
talk
to
kubernetes
I,
don't
need
a
ton
of
heavyweight
stuff,
so
there
might
be
room
for
both
of
these.
There,
probably
even
more
approaches,
I'm
encouraged
to
see
people
interested
in
both
of
these.
E
A
I,
don't
know
that
it
matters
a
ton
like
honestly,
we
were
one
of
the
first
six.
They
create
a
sub
project
with
a
separate
meeting
time.
It
was
really
just
a
time.
You
know
Hans
and
Maria
had
an
integration
test
framework
that
they
wanted.
I
think
this
was
within
the
context
of
coop
CTL
testing,
and
so
it
made
sense
to
just
like
have
a
separate
rating
to
talk
about
pushing
that
forward.
Yeah.
B
There
was,
there
are
a
couple
different
parties
who
are
all
interested
in
basically
the
same
thing.
We
needed
a
separate
vendor,
Abal
library
that
was
agnostic
to
the
KK
Depp
graph.
That
folks
could
import,
as
we
start
to
break
apart
some
of
these
tools
right
and
we
wanted
it
to
be
relatively
minimalistic.
B
A
This
could
be
that
time.
I
it
conflicts
for
me,
maybe
Tim
I,
don't
know
on
the
is
the
effort
to
migrate
project
infrastructure
over
the
CN
CF.
That
is
also
a
Friday
at
9:00
a.m.
call
on
alternating
weeks.
So
we've
kept
this
slot
as
bi-weekly
it's
kind
of
pacific
friendly
kind
of
european
friendly.
That's
one
option.
We
could
also
throw
up
a
doodle
to
talk
other
to
see
if
there
are
other
times
that
work
thursday
at
7
a.m.
doesn't
work
for
me
because
of
india,
but
basically
like
I.
A
A
If
we
have
a
sub
project,
that's
let's
have
the
battle
royale
of
frameworks
or,
let's
just
have
everybody
talk
about
frameworks
and
stuff
or
if
each
framework
it
has
people
who
are
super
motivated
to
sort
of
chart
out
roadmap
before
we
can,
you
can
do
that.
I'm
not
super
picky
about
it.
I
just
I
do
like
to
see
the
interest
and
I
think
they're
recurring
meeting
to
talk
about
testing
frameworks
would
be
a
good
idea.