►
From YouTube: Octant Community Meeting - April 21st, 2021
Description
Octant community meeting is held weekly. We discuss and talk about the current state and future of Octant, demo upcoming features and releases, and preview new ideas we are considering for Octant.
Meeting agenda: https://hackmd.io/CzaPxtmXT_SW8nEpdwvGzw?view
B
All
right,
then,.
A
Let's
get
my
bearings
find
this
window:
okay,
cool!
All
right:
we
can
get
started
all
right,
hello.
Everyone,
sorry
for
the
late
start,
welcome
to
the
april
21st
octet
community
meeting
this
one's
a
pretty
exciting
one.
We've
got
a
few
things
on
the
agenda.
Does
anyone
have
anything
else
to
add
before
we
get
started
today?
A
All
right:
well,
if
anyone
does
anything,
you
can
always
edit
this
in
real
time.
You
just
go
on
to
the
middle
portion
here
and
you
can
add
them
to
q
a
or
bring
it
up
to
slack
or
any
of
the
other
mediums,
but
we'll
go
ahead
and
get
started
so
first
item
on
the
agenda.
This
one
has
been
an
issue
that
has
gotten
some
attention
from
a
few
people.
I
did
have
a
chat
about
it
yesterday,
so
back
when
we
bumped
our
client
go
dependencies
to
119.3.
A
I
believe
it
sets
off
a
lot
of
deprecation
warnings
for
a
lot
of
the
kubernetes
resources
which
are
now
deprecated
and
we
decided
to
one
just
say:
they're
not
officially
supported
or
we
implemented
a
version.
Skill
policy,
but
also
it
was
done
with
the
understanding
that
the
api
server
does
the
conversions
and
apparently
this
is
not
the
case
for
certain
versions
of
kubernetes.
A
So
in
this
issue
we
have
captured
some
of
some
of
the
discussion
here
and
what's
happening,
and
some
people
on
are
on
eks
clusters
and
really
it's
not.
We.
A
I
think
we
had
a
run
around
here
because
we're
just
working
off
different
kubernetes
versions,
but
people
on
1.8
and
below,
if
I
understand
correctly,
will
not
get
the
up
casted
ingress
version,
and
so
it
returns
nothing,
whereas
if
they
were
on
a
newer
version
of
kubernetes,
even
if
you
request
a
networking,
v1
beta
and
you've
applied
something
as
like
a
one
beta
one
ingress,
it
would
still
return
something.
So
vikram
did
the
detective
work
on
this.
A
So
thank
you
vikram
and
we
had
a
discussion
about
it
and
left
a
message,
and
essentially
the
thought
process.
Is
this?
It's
a
yes,
it's
a
problem.
I
think
it
does
break
a
few
users,
but
at
the
end
of
the
day
we
don't
know
how
important
this
this
is
right
like
this
might
not
even
be
like
a
valid
discussion
point
in
like
three
months
from
now.
Right
and
the
other
angle
to
take
this
with
is
like
we're
on
cloud
native
infrastructure
right.
So
it's
like
it's
kind
of
hard
to
imagine.
A
A
I
think
that's
the,
I
think,
that's
the
consensus
we've
made
and
if
there
is
overwhelming
community
pressure
saying
no,
we
really
need
to
support
ingresses
on
less
than
or
I
guess,
like
one
eight
and
below
then
we're
then
we're
happy
to
add
a
separate
key
in
here,
and
this
is
fixable.
But
if
not,
you
know,
I
think,
either
way
with
our
current
version
skew
policy.
The
next
time
we
update
client
go.
This
issue
becomes
invalid
anyways.
A
So
it's
you
know
so
we're
still
chasing
like
maybe
another,
two
three
months
of
at
most,
where
this
issue
is
valid.
So
does
anyone
have
any
questions
about
this.
A
A
Yeah,
I
think
we
could
totally
have
discussions
about
welf.
How
often
do
we
bump
client
go,
and
I
don't
think
we've
ever
explicitly
mentioned.
C
A
We
don't
have
timelines
right
like,
for
example
like
we
don't
we
like,
we
don't
really
say
like.
I
think
what
is
the
latest
version
like
taught
2-1
or
something
yeah
two
ones
are
c.
I
think
right
so
someone's
rc,
yeah
yeah,
so
okay,
so
we've
got
time
like
like
when
they
cut,
like,
I
don't
know
we'll
like
pin
right
away.
But,
yes,
we
do
have
a
version,
skew
policy.
So
keep
that
in
mind
as
opting
keeps
scaling
up
two
new
versions.
C
Yeah,
so
I
think
technically
right
now,
1.18
is
the
lower
end
of
our
of
our
policy,
so
anything
less
than
1.18
would
technically
not
be
supported
under
that
policy,
which
is
why
this
issue
has
come
up
and
yeah
anyway.
A
Yeah,
okay,
cool,
so
all
right!
Well,
if
anyone
has
a
strong
feelings
about
this,
please
comment
on
the
issue.
Let
somebody
know
I
think,
adding
a
separate
key.
Isn't
too
difficult?
It's
just
a
more!
You
know.
It's
turn
it's
it's
not
difficult
cool!
Moving
on
this
one
is
a
demo.
I
have
some
thoughts.
This
is
kind
of
more
theoretical
stuff.
It's
not
even
a
fully
working
demo.
Do
we
maybe
want
to
jump
to
resource
spear
improvements,
this
one's
5v1.
B
E
Can
you
see
that
yep
yeah?
So
it's
just
a
quick
video
recording
that
I
did
while
I
was
testing,
let
me
just
give
some
background.
We
one
of
the
oldest
and
the
hardest
issues
with
resource
fuel
that
we
had
a
long
time
is
when
you
sometimes
when
you
get
a
view
like
this,
it
would
just
start
updating
and
refreshing
and
changing
the
way
out
and
doing
things
like
that,
and
we've
worked
on
that
for
a
long
time
as
sam
kicked
this
off
a
few
months
ago.
E
I
think-
and
he
made
some
some
really
good
observations.
How
pods
and
services
are
basically
competing
for
cpu
time
as
he's
as
we
start
those
stress,
so
it
really
helped
and
then
also
louise,
took
took
a
stab
and
this
and
made
some
additional
observations
that
helped.
But
then
I
I
kind
of
started
with
that
and
spent
quite
a
bit
of
time,
trying
to
figure
out
what's
going
on
and
eventually
ended
up
making
a
solution
that
is
currently
under
review.
E
I
don't
have
it
right
now,
but
I'll
try
to
find
a
little
later.
But
basically,
let
me
start
this
video
here.
E
On
the
left
side,
we
have
the
the
old
resource
viewer
on
the
right
side
is
the
new
one
and
you
can
see
right
over
the
way
that
layouts
changed
quite
a
bit
and
basically
the
way
the
structure
the
layout
now
is
the
component
you're
looking
right
now
in
this
case,
this
is
deployment,
it's
always
going
to
be
at
the
top,
and
then
you'll
see
the
children
and
so
on,
which
is
a
additional
benefit
of
this
of
this
change,
and
let
me
start
this
video
on
the
left
side.
E
We
have
the
old
one
and
they
are
running
side
by
side
on
the
same
year
and
keep
an
eye
on
this
service
here.
So
there
is
the
first
update
second
update
and
along
the
way
you
can
see
that
that
services,
services
2
and
web
2
and
keep
swapping
places
a
few
times
and
at
the
same
time
you
see
on
the
right.
It's
that's
pretty
much
rock
solid.
E
So
I
recorded
this
for
48
seconds.
I
think
they
would
throw
updates
in
the
old
version
and
none
and
the
new
ones.
So
I.
B
E
C
E
C
Like
I
just
want
to
say
that
this
is
so
like
one
of
the
biggest
frustrations
I've
had
using
resource
viewer
in
the
past
is
like
you're
looking
at
something
and
then
you're
like
oh
okay,
yeah,
and
then
you
like
you,
have
it
selected
or
you
you're
zoomed,
in
to
a
section
because
you're
looking
at
a
large
set
of
of
connected
things
and
then
it
toggles
and
refreshes
and
like
zooms,
you
out
re-centers
and
like
the
thing
you
were
looking
at,
isn't
even
in
the
same
place.
It
used
to
be
it's
like.
E
C
Yeah,
this
is
so
great
yeah.
This
is.
This
is
like
a
definitely
a
team
effort
that
that
milan,
like
carried
carried
the
the
torch
the
last
bit
of
the
way
here-
and
it's
like
I'm
so
happy
to
just
see
this
solved
and
yeah
thanks
everyone
who
had
a
hand
in
working
on
this.
This
is
this
is
a
really
old
issue
too.
It's
like
probably
over
a
year
old,
right
right.
E
Yeah
yeah,
I
think
I
think
this
also
gives
a
lot
of
bass.
We
can
build
on
top
of
it.
You
know
the
the
resources
are
grouped
as
well
based
on
their
kinds
so
for
complex
layouts
that
that
opens
another
feature
that,
hopefully,
is
coming
soon,
which
is
expanding
and
collapsing
like.
If
you
have
16
services
here,
do
you
really
need
to
show
them
all?
Maybe
you
can
just
you
know,
show
one
box
or
things
like
that.
So
well,.
C
Yeah
like
in
the
old
one,
where
would
you
even
put
the
collapse
right
like
this?
It's
like
the
way
it
renders.
It
all
depends
it's
just
like
yeah.
This
is
great
and
another.
I
don't
know
if
you,
I
might
have
missed
it
if
you
didn't
call
it
out,
but
one
thing
to
notice
on
this
new
version
is
that
the
normally,
if
you
notice
the
pod
on
the
left
hand,
side
is
the
root
node.
C
The
deployment
now
is
always
the
root,
and
if
you
like,
go
to
the
pod
view,
if
the
pod
would
be
it
would
you
know
it
probably
reflects
what
resource
you're
looking
at
when
before
it
would
always
ultimately
always
boil
down
to
like
whatever
the
last
thing
get
associated
with
it
was
in
the
or
or
the
not
the
last
thing,
but
the
the
thing
that
was
common
would
end
up
being
the
root,
and-
and
in
this
case
it
was
pods.
E
Right,
so
it's
a
little
bit
it's
it's
still
somewhat
confusing
if
you
start
with
a
pod,
because
you
know
under
the
pod,
which
shows
secrets,
services
and
ingresses,
but
then
also
the
deployments
and
replica
sets
those
tend
to
kind
of
group
on
the
left
side.
But
you
know
still
that's
the
proper
way
to
to
show
it.
E
If
that
part
is
the
center
of
your
leave
yep,
so
yeah
yeah,
the
one
other
thing
I
wanted
to
mention
is,
unfortunately,
we
are
having
some
issues
with
the
customer
resources
right
now
and
and
this
this
is
really
helpful
with
custom
resources,
but
due
to
some
issues,
it's
it's
really
hard
to
test
for
complex
views,
but
one
once
that's
sold.
I
think
this
is
this:
is
gonna
really
help,
especially
for
custom
resources.
E
Yeah,
that's
that's
all
I
got
you
know
if
you,
if
you
had
a
chance
to
play
with
it,
please
let
me
know
what
you
think
and
if,
if
you
find
any
issues
or
something
that's
not
working
the
way
it
should
please.
Let
me
know.
A
Awesome
well
thanks
yamalan,
for
taking
on
that
issue.
That's
it's
a
tough
one
and
glad
we're
making
progress
on
it
and
yeah
take
a
look
and.
B
A
Cool
all
right.
Well,
I
will
jump
on
to
the
last
item
on
the
agenda.
This
is
something
bikram
and
I
worked
on
a
little
bit
and
this
one
is
a
demo
for
some
plug-in
work.
That's
been
happening
over
the
last
week,
or
so.
This
is
specifically
with
the
home
plugin.
The
url
for
this
is
under
blood
orange.
I
o
octane
helm
and
they
wrote
this
plug-in
more
than
a
year
ago,
a
year
and
a
half
ago,
and
since
that
initial
release
it
hasn't
really
had
a
lot
of
updates.
A
So
that's
a
little
nostalgic
if
you
want
to
really
jump
way
back
to
the
earlier
versions
of
octane
and
we're
just
adding
a
couple
new
features
back
when
this
plugin
was
written,
octane
was
mostly
read
only,
and
so
since
then,
we've
had
the
concept
of
actions,
and
now
we
can
actually
do
a
lot
of
write
operations
and
we
can
start
changing
how
to
you
know
how
we'd
essentially
interact
with
them,
and
so
one
of
the
things
I'd
like
to
demo
is
an
upgrade
from
the
initial
home
release
to
demo
some
of
the
new
functionality.
A
So
the
first
piece
that
I
want
to
show
is
actually
just
helmet
install
something
so
currently
right
now,
if
you
install
something,
you'd
have
to
do
it
from
the
command
line,
and
so
I've
added
a
little
note
about
this
saying.
Well,
what
work
needs
to
happen
before
helm
install
is
possible
right.
So
this
is
essentially
a
search
feature,
and
it's
also
in
a
way
a
ui
feature.
At
the
same
time,
right
we
have
to
figure
out
like
do
we
want
to
make
a
search
bar
component?
A
How
would
that
look,
and
even
yeah,
and
just
to
think
about
like
helm,
specifically
right?
If
you
do
a
home
search,
you
could
be
searching
multiple
repositories.
So
there's
a
lot
of
nuances
that
could
go
into
a
search
and
then
people
are
going
to
start
saying
like
well.
I
want
to
do
search
and
search
filtering,
so
it
can
get
complicated
really
quickly.
So
I
just
left
install
alone
for
now.
A
I
think
this
is
something
that
could
warp
discussion
in
the
future,
and
so
the
first
thing
to
show
is
now
we
have
an
uninstall
type
of
operation
fairly,
simple,
all
right.
So
now
you
can
remove
home
releases
since
that's
fairly
straightforward.
We
assume
that
the
cleanup
is,
we
assume
the
cleanup
works.
Essentially
there
could
be
cases
edge
cases
where
you
have
resources
left
behind
by
helm,
and
sometimes
I
might
warn
you
about
that.
A
So
this
still
assumes
the
happy
case
for
the
most
part
so
moving
on,
we
have
now
an
updated
chart.
We
have
the
history,
so
every
time
you
make
an
upgrade,
it'll
show
a
history,
so
we
could
even
try
something
like
hell
upgrade:
oh
sweet.
I
already
actually
have
an
older
version
of
this,
so
if
we,
you
know
set,
say
a
new
key
value.
A
To
help,
let's
see
what
we
get
home
upgrade
upgrade
failed
executing.
What
does
it
not
want?
You
must
provide
current
okay,
so
home
upgradeable.
Okay,
so
maybe
you
want
something
different
or
maybe
there's
something.
A
Key
value,
but
if
you
have
to
upgrade
a
key
value,
hypothetically
should
upgrade
and
you'd
see
the
history
up.
Here
you
get
notes
and
then
lastly,
it's
values,
there's
something
I'm
still
thinking
about
so,
and
this
is
kind
of
just
more
of
an
implement
implementation,
detail
versus
what
is
actually
shown.
So
here
you
have
the
concept
of
values.
A
What
values
in
home
typically
for
users
really
just
mean
this
section,
this
section
of
helm
here,
where
you
have
the
access,
keys,
custom,
annotations
deployment,
and
you
know
essentially,
values
that
you
can
override
compose
and
all
that
configuration
management
stuff.
But
internally
the
values
method,
or
at
least
the
values
object,
refers
to
a
whole
bunch
of
other
stuff.
Like
capabilities,
it
even
holds
the
chart
information
release,
and
so
it
holds
a
lot
of
other
metadata
pertaining
to
the
helm
release.
A
So
I'm
trying
to
figure
out
how
to
get
this
merged
up
so
that
it's
only
accessing
the
values
and
merging
it,
and
then
we
can
get
an
upgrade
command
and
we
can
essentially
do
the
helm
upgrade
through
a
ui.
So
those
are
those
are
what's
available
at
the
moment.
We
also
have
an
outstanding
issue
for
adding
a
resource
viewer
to
the
graphics
section,
so
we
could
have
a
tab
and
we
could
show
a
resource
viewer
of
helm,
related
objects.
That
seems
it's,
it
seems
straightforward,
but
it's
actually
not.
A
I
thought
about
this
a
little
bit
and
you
kind
of
have
to
because
of
the
way
that
octan
creates
the
resource
sphere
right
now
using
the
query
here,
it
looks
for
relationships
that
are
typically
like
parent-child
relationships
between
objects
and
for
helm.
That
may
or
may
not
be
the
case,
so
we
will
need
a
different
mechanism
for
drawing
lines
between
objects
in
a
resource
viewer
and
that
points
back
to
some
an
issue.
A
Brian
wrote
almost
what
last
november
yeah
that's
a
long
time
ago,
and
he
proposed
that
we
needed
to
have
a
mechanism
to
draw
lines
between
two
objects
that
would
be
available
to
users
and
it
might
look
something
like
an
annotation
or
you
know,
or
some
other
mechanism,
but
I
think
we're
going
to
need
something
like
this
implemented.
If
we
want
to
draw
a
resource
viewer
specifically
for
help
sam.
E
I
I
think
you
should
look
how
the
k
native
plugin
is
including
the
resource
view.
I
think
it's
using
it's
doing
a
slightly
different
level,
so,
basically
just
adding,
if
I
remember
right,
just
adding
nodes
and
edges
so
that
that
might
be
the
easiest
way
to
do.
This.
A
Sure
absolutely
yeah,
and
this
might
I
don't
know
on
go
plugins.
I've
never
tried.
I
don't
know
if
we
can
do
it
like
that
and
go
it
might
be
a
bit
more
restrictive
yeah.
I.
C
Think
right,
I
think
we
can,
but
not
with
like,
because
we
can
pack
into
the
content
response
a
resource
viewer
and
the
fields
of
that
resource.
C
Viewer
can
be
our
public
and
we
can
populate
that
component
with
the
edges
manually,
but
it
would
be
very
tedious
and
would
basically
require
you
generating
a
bunch
of
like
generating
everything
manually
and
then
like
it's
assigning
it,
whereas
on
the
javascript
side,
we
can
just
give
it
that
json
blob
and
it
does
the
right
thing
so
right
yeah,
I
do
think
I
do
think
having
some
formal
api
for
establishing
relationships
between
objects
is
a
good
idea.
C
Also,
I
think
that
sam
you
might
want
to
to
talk
to
luis
about
the
form
component.
I
think
that
is
a
good
case
for
like
you're
talking
about
a
search
feature
or
some
type
of
input,
just
generic
random
input
to
capture
doing
an
install,
and
once
we
have
the
ability
to
create
a
form
component
and
render
it
wherever
we
want
into
our
layouts.
A
Behavior
yeah
right
on
that.
That
is
something
that
yeah
for.
We
definitely
need.
We
have
four
refactors
and
it's
going
to
be
tied
to
a
lot
of
the
existing
work,
so
yeah,
but
I
think
ultimately,
the
what
we're
looking
at
here
right
now
is
probably
not
going
to
be
the
next
release
of
the
plug-in.
A
I
think
I'd
like
to
see
a
little
bit
more
maturity
before
we
kind
of
release
for
this
one,
but
as
always,
community
contributions
would
be
helpful
or
you
have
suggestions,
comments,
open
issues
and
all
that
good
stuff.
A
Let's
see
back
to
the
notes,
oh
and
file
system
operations
right,
so
helm
also
can
install
from
things
locally,
and
so,
if
you
want
to
do
a
home
install
from
a
file
that
also
kind
of
needs
to
be
a
thing
but
also
like
this
is
something
I
want
as
well
for
the
kind
plugin
where,
if
you
wanted
to
do
an
import
and
export
of
yaml
files
from
octan.
A
I
think
this
is
something
we'll
need
to
think
about
which
previously
wasn't
possible
on
the
browser
version
of
often,
but
now,
where
that
we're
an
electron,
we
have
a
way
of
interacting
with
the
file
system
on
the
front
end
and
on
the
end,
as
well,
so,
depending
on
how
we
designed
this,
we
are
going
to
slowly
get
file
upload
type
of
scenarios
as
well
cool.
A
Well,
that
is
it
for
helm,
plug-in.
You
can
move
on
to
open
q
a
if
no
one
else
has
any
comments.
C
Just
yeah
that
was
great
thanks
and
that
looks
awesome
looking
forward
to
seeing
that
as
it
matures
a
little
more.
I
like
the
like
yeah,
I
mean
just
bringing
it
up
to
date
with
the
latest
stockton
is
a
nice.
Is
nice
in
itself
that
to
me
that
might
be
worth
a
pr
upstream,
just
the
fact
that
we
brought
it
up
to
date
like
yeah
it
does.
It's
not
like
you
were
saying
it's
not
like.
Oh,
we
got
a
bunch
of
new
things,
but
also
like
hey.
C
A
Yeah
and
that
code
is
already
pushed
up.
It's
not
if
people
want
it.
They'll
have
to
cut
the
builder
themselves,
but
I'd
yeah
I'll
probably
have
to
paint
josh
to
see
like
what's
the
state
of
that,
because
he
added
me
as
a
maintainer
on
this
repo.
So
I
do
have
full
control
over
it,
but
just
kind
of
just
to
see
how
much
involvement
he
wants
to
have
over
time.
Sure
yep
sounds
great.
A
Cool
well,
that
brings
us
to
open
q
a
do
we
have
any
last-minute
questions.
Comments
concerns.
C
One
thing
that
I
just
want
to
like
get
people
thinking
about
is
it's
something
that
we've
had.
I
think
it
comes
up
every
now
and
then,
where
we
talked
about
getting
better
ci
cd
in
place,
this
is
going
to
become
really
critical
as
we
move
into
auto
updates
and
I
think
our
first
pass
of
auto
updating
is
going
to
be
fairly
naive
and
and
just
kind
of
be
like
a
hey,
we've
swapped
the
binary
and
that's
it.
C
But
I
would
like
to
eventually
get
us
to
a
place
where
you
know
you're
running
octant,
the
there's
there's
a
process
watching
in
the
background
for
new
releases
that
you
get,
you
get
a
global
alert
saying
that
there's
a
new
release
and
you
can
dismiss
it
or
you
can
click
right
there
to
update
and
then,
and
also
in
preferences,
there'll
be
an
update
section
which
would
then
let
you
see
like
what
version
you're
on
what
version
you
can
update
to,
but
before
we
get
to
that
point
of
like
allowing
users
just
to
kind
of
like
hit
that
button
and
get
those
updates
right
away.
C
I
would
really
like
for
us,
as
a
team
to
to
think
about
and
and
plan
a
like.
What
is
it?
What
does
testing
look
like
in
that
in
that
world,
where,
when
we
push
up
a
release,
tag
and
it
gets
built,
and
then
we
publish
it,
people
are
now
running
the
latest
right
and
and
we're
not
having
to
like
what
I
want
to
avoid.
Is
in
that
scenario
it's
great,
because
you
can
now
get
updates
out
to
people
very
quickly.
C
But
what
I
also
want
to
avoid
is
like
we
push
like
we're
doing:
oh
dot,
25
and
then
25.1
35.2
25.3,
like
just
because
we
can
doesn't
mean
we
should
and
and
so
having.
Having
really
good,
like
integration
tests
in
place
is
something
that
I'd
like
to
see
and-
and
this
is
more
of
just
like
a
call
out
to
the
team
and
anyone
in
the
community
who
has
done
this
thing
before
who
has
experience
with
like
doing
integration
tests
for
applications.
C
You
know
things
that
run
on
the
desktop
just
trying
to
capture
ideas
and
put
that
in
people's
brain.
Probably
I
would
say
this
effort
of
like
making
a
big
push
on.
It
is
a
couple
releases
out,
but
is
something
that
I
want
us
to
start
thinking
about
and
start
to
like
building
the
foundation.
When
we
see
opportunities
to
do
so.
E
Yeah,
I
I
totally
agree.
I
think
we
need
to
spend
some
time
thinking
about
this,
make
sure
we
get
it
right
feels
like
this,
that
you
should
probably
open
a
discussion
about
this,
and
then
you
know
try
to
put
all
the
comments
and
gods
in
theirs
kind
of
keep
track
of
it.
A
Yeah,
I
don't
quite
see
it
on
this
product
excellence
board.
It
might
be
a
good
place
for
it.
We
also
do
need
a
discussion.
I
think
this
is
something
that
needs
more
planning.
There
was
a
proposal
for
testing
that
I've
written
like
months
ago
and
never
executed
on
it.
Sorry
y'all
I'll
find
time
to
draft
up
a
proof
of
concept
at
some
point
one
of
these
weeks
and
kick
the
tires
on
it.
C
Yeah,
I
think
I
think
that
proposal
is
is
actually
still.
I
re-read
it
recently
and
some
of
the
outstanding
things
under
the
implementation
strategy.
In
that
proposal,
I
think,
are
a
little
easier
to
do
even
even
now
like
like.
That
was.
I
think
that
was
written
back
in,
like
I
don't
know
it
was
a
while
ago,
like
it's
a
year
ago,
was
it
a
year
ago.
C
Yeah
I
was
gonna,
say
I
knew
it
was
a
while
ago,
but
so
I
I
I
think
that,
like
getting
kind
set
up
in
via
github
actions,
what
it
is
now
versus
what
it
was
a
year
ago
is
like
a
completely
different
experience
and
like
it's
a
pretty
common
thing
to
be
able
to
do
now.
I
also
think
that
you
know
when
we
wrote
that
end-to-end
testing
we
there
was
still
a
lot
of.
C
I
think
there
was
some
concerns
in
thinking
about
both
the
like,
both
the
cli
version
of
auction,
that
we
were
shipping
and
the
and
the
electron
version,
and
I
think
now
we
can.
We
can
put
fresh
eyes
on
that
with
the
purely
the
electron
version,
as
our
as
our
target.
So
yeah,
I
think
I
think,
there's
there's
still
a
lot
of
really
good
things
that
were
captured
in
that
proposal
that
still
apply
and
will
still
be
very
useful.
C
A
All
right
yep
all
right.
Well
there
and
aren't
any
other
questions.
I
think
that's
we
call
it
a
day
and
thank
you,
everyone
for
coming
to
this
community
meeting
and
we'll
see
you
all
next
week.