►
From YouTube: Argo Contributors Office Hours Feb 16th 2023
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
Hi
everyone
welcome
to
the
contributors
meeting
I'm
gonna,
be
your
host
today,
I'm
Leo
I'm,
already
sharing
my
screen
and,
as
usual,
starting
off
with
the
triage
in
discussions
we
had
for
the
previous
week,
Dan
and
Kate
I
see
you
added
a.
B
Yeah
some
answers
here
then
so
take
away
so
63
new
issues.
11
closed,
11
closed
that
were
open
this
week
and
then
a
total
of
85
open
issues
were
updated
this
week.
So
that's
like
kind
of
our
activity,
level
and
45
issues
were
either
closed
in
total
or
closed
were
updated
in
this
last
week
and
there
are
about
a
thousand
stale
issues,
so
these
are
ones
that
have
no
updates
in
over
180
days.
B
So
I
think
just
overall
one
thing
that
I'd
like
to
figure
out
at
some
point:
I
mean
looking
at
these.
Most
of
them
are
enhancement,
requests,
so
I'd
like
to
just
start
sort
of
categorizing
some
of
the
enhancement
requests
so
that
we
could
actually
just
say,
hey.
Everybody
seems
to
have
issues
with
this
section
of
Argo.
Maybe
we
should
take
a
look
at
that
so
with
that
I
actually
wanted
to
bring
up
just
Helm
and
values
issues
so
I
linked.
B
These
were
all
ones
that
were
open
this
week,
since
we
added
Helm
multi-source
Etc,
we've
just
had
a
lot
of
Helm
issues
and
if
I
could
summarize
them
and
then
I'll
go
to
this
random
stack
Trace.
One
that
I
stuck
in
there,
if
I,
could
summarize
them
all
they're,
basically
all
issues
that
are
related
to
the
fact
that
you're
not
using
Helm
natively
in
Argo
CD,
like
you're,
using
it
as
a
just
a
template
and
then
apply
so
these
range
from
multi-source
not
working.
B
Quite
the
way
people
expect
values
files
not
being
applied
the
way
they
expect
not
being
able
to
pass
in
like
values
blocks
like
you
might
expect
errors
not
necessarily
being
passed
in
the
most
straightforward
way
and
I
think
all
of
these
are
actually
relatively
small
issues
that
could
be
resolved,
but
I
wanted
to
just
get
like
a
feeling.
If.
B
I'm,
like
afraid,
to
bring
it
up
if
someone
were
to
make
in
it
would
it
would
probably
need
to
be
in
Argo
labs.
You
know
to
start
with,
but
if
someone
were
to
make
a
Helm
controller
that,
rather
than
you
know,
everything
in
Argo
CD
is
is
pre-processed
and
then
passed
over
for
application.
Basically,
just
as
a
coupe
CTL
applies
essentially
what's
happening.
B
What
if
we
had
a
Helm
controller-
and
you
could
pick
that
and
all
it
would
basically
do-
is
rather
than
do
the
config
build,
it
would
basically
just
pass
it
over
to
the
helm
controller.
Now
the
downside
of
this
is
you
essentially
have
two
controllers
that
apply
resources
now,
but
the
upside
is.
B
You
could
basically
support
a
straight
up
native
Helm
life
cycle,
because
this
controller
could
basically
just
it
can
do
server-side
lookups
all
that
stuff
and
it
would
be
based
on
you
know
it
would
have
to
be
you'd
have
to
you'd
have
to
re-implement
our
back
there
right
so
that
it
would
be
obeyed,
but
or
you
do
it
through
via
impersonation
or
something.
But
if
is
there,
is
there
any
interest
in
something
like
that?
Or
is
it
just
universally
a
bad
idea.
A
I
just
wanted
to
understand
one
thing:
first,
then
you
you
mentioned
the
controller
when
you
say
controller
is
because
you
want
this
to
take
care
of
the
whole
life
cycle
of
even
applying
the
Manifest
in
the
cluster,
or
it
could
be
just
delegate
that
to
to
a
more
specialized
CMP
specialized
Helm
seems
CMP.
Then
what
we
have
natively
supported
inside
our
rocd,
so
I'm
trying
to
understand
how.
B
Is
that
I
was
thinking
of
it
as
a
separate
controller,
but
since
you
bring
it
up
yeah
you
could
technically
have
it
then
send
what
it
finish
finishes,
rendering
with
server-side
lookups
to
Argo
to
apply,
but
it
you
have
to
implement
our
back
there,
because
otherwise,
Helm
lookups
could
go
beyond
scope
and
you
were
going
to
create
issues
that
way.
B
There's
security
issues
with
that,
so
I
actually
think
it'd
probably
need
to
be
a
separate
controller.
I
would
call
it
tiller
is
what
I
would
name
it
just
for
funsies,
but
but
yeah
that's
basically,
and-
and
that's
you
know,
that's
why
I
have
some
allergic
reaction
to
it
as
well,
but
thinking
about
it
in
terms
of
okay,
Helm
is
here,
people
want
to
use
it.
I
have
a
lot
of
problems
with
how
Helm
operates,
and
we
have
tried
to
opinion
our
way
out
of
Helm
issues
within
Argo
CD.
B
But
it's
like
look
if
they,
if
they
want
to
use
Helm
and
they're,
going
to
run
into
issues
from
Helm,
then
I
would
rather,
they
just
be
dealing
with
Helm
issues
and
not
Argo
CD
issues,
and
since
we're
trying
to
take
all
we're
like
trying
to
soften
the
edges
of
Helm,
it
feels
like
it's
a
losing
battle
and
one
that
we're
always
going
to
be.
You
know
adding
tweaks
and
things
to
make
it
operate
more
sanely
when
in
reality
you
should
be
like
yeah
Helm
does
some
stuff,
that's
weird!
B
C
Can
I
have
a
question
here
yeah
regarding
Helm
and
plugins
I.
Think
most
of
the
use
cases
are
well
a
majority
of
well.
Some
of
the
use
cases
are
is
plugins
are
being
used
to
handle
stickers.
It's
typically
AVP
the
World
plugin.
C
Would
it
be
possible
to
do
a
chain
plug-in
system
well
where
we
are
rendering
with
hammer
customizer,
whatever
grabbing
the
output
piping
it
to
the
next
stage
like
plugin,
that
does
you
know
some
kind
of
magic
like
replacing
the
secrets
and
then
goes
to
qctf
for
applying
it?
That
would
you
know
decoupled.
C
B
You
can
do
that
with
a
CM
with
like
a
config
management.
Plugin
like
you
could
have
it
do
Secrets,
look
up
and
shove
it
into
manifest.
C
I
know
it's
more
about
the
chaining
like
it
shouldn't
be
right:
now
we
can't
use
the
integrated
Helm
or
customize
or
whatever.
How
do
you
call
these
processors
Maybe?
We
have
to
bake
it
into
the
CMP.
The
CMP
does
the
hand
template
pipe
City
into
the
into
our
plugin.
Then
it
goes
back
to
Argo
CD
for
applying
the
point
will
be
letting
Oracle
CD
to
do
a
Helsing
or
customize
or
whatever.
It
is
then
separately
doing
the
plug-in
thingy
based
on
the
previous
stages
output.
D
I
think
your
PR
gurgly
is
actually
the
first
step
towards
that.
Basically
sharing
information
between
sources.
It
gets
a
little
bit
dicey,
because
then
you
have
to
resolve
dependencies
in
that
chain,
but
the
the
first
thing
is
just
sharing
some
information.
I
think
your
PR
starts
that.
C
Might
be
yeah
I
think
with
Britain
is
because
some
of
these
stages
are
not
linked
together
in
their
mechanics.
They
are
pretty
much
doing
independent
things,
but
we
can't
have
them,
as
you
know,
separate
mechanics,
change
together
and
working
together
right
now,
and
that
complicates
plugin
management.
We
have
to
do
a
separate
plug-in
definition
for
every
use
case
that
we
are
having
and
it's
kind
of
complicated
deployment
at
the
wall
internally,
exactly
are
ending
up
with
I.
B
Don't
I
don't
think
doing
a
server-side
controller
for
Helm
would
solve
your
issue
unless
you
could
resolve
it
using
native
Helm
stuff,
because
that's
really
what
I'm
talking
about
is
just
basically
just
having
a
full-on
native
Helm
support.
B
C
So
yeah
I
was
saying
is
the
relation
that
you
said
that
a
bunch
of
issues
got
opened
you
to
the
recent
time,
change
the
multiple
sources,
and
these
are
well
things
with
them
and
when
I
check
some
of
these,
these
were
related
to
plugins.
Usually,
and
people
wanted
to
do
things
that
the
new
thing
weren't
able
to-
and
there
are
a
couple
of
new
issues
opened.
A
Personally,
I
think
it's
an
interesting
idea,
but
I
think
at
this
point
we
would
need
a
proposal
to
really
understand
the
use
cases
itself
and
how
we
would
deal
with
security.
That
would
be
great
to
have
a
proposal
for
that.
But
yeah
sounds
like
an
interesting.
The
chaining
process
of
generating
manifests
sounds
interesting
at
first
to
me
at
least.
E
B
A
Are
you
familiar
with
the
proposal
process
that
we
have
I'm.
A
Cool
can
I
generally
sorry
yeah.
A
B
A
Also
yeah
related
to
that
I
guess
the
the
way
I
think
the
the
way
I
understand
now
is.
If
you
want
to
have
that
controller,
then
I
guess.
My
question
is
why
using
Argo
CD
at
all
right?
So
what
is
what
is
the
real?
What
what
is
the
benefit
of
using
Argo
CD,
if
you
have
a
controller
that
takes
care
of
everything,
so
it's
just
for
using
the
UI.
B
I
think
it's
both
the
UI.
You
can
still
have
automatic
reconciliation.
As
Michael's
pointed
out,
you
can
still
do
Post.
Sync,
you
know
stuff,
it
doesn't
have
to
all
be
on
Helm
hooks.
B
B
You
can't
do
like
a
diff
and
ignore
differences,
and
that's
something
that
I
think
we
could
Implement
so
anyway,
I'll
I'll
put
I'll
I,
have
a
little
write-up
I'll
share
it
with
a
few
people,
and
then
I'll
put
it
up
for
people
to
beat
up
I
think
it
would
have
to
be
an
Argo
Labs
project
to
start
with,
because
there's
just
so
many
issues
to
tackle,
but
that's
it
for
the
helm,
Helm
stuff.
One
other
thing
I
wanted
to
bring
up
from
the
issue
triage.
Was
this
random
stack
Trace?
B
This
is
something
I
haven't
seen
before.
Maybe
you
guys
have
seen
it,
but
it
basically,
they
got
an
error.
That
said,
go
make
an
issue
and
I
looked
at
the
issue
and
I
thought.
I
have
no
idea
what's
going
on
here
and
so
I
don't
know
if
somebody
else
makes
sense
of
this
or
or
you
know,
I
figure,
since
we
have
something
in
the
pla
project
that
actually
tells
people
to
make
an
issue
that
we
should
do
something
about
it,
but
which.
D
B
D
I
guess
so
they
click
on
a
resource.
To
failed.
It's
not
clear
to
me
where
they
did
that
click.
So
we
could
probably
reproduce
the
issue
by
doing
a
thing
causing
it
to
fail
clicking
on
whatever
they
clicked
on
to
get
this
error.
And
then
we
could
look
at
you
know
Network
Tools
or
whatever,
to
figure
out
why
the
stack
Trace
appeared,
but
I
think
we
need
more
information
to
actually
reproduce
this
because
I
don't
know
if
they're
clicking
on.
D
E
D
A
So
how
do
we
want
to
proceed
with
that
anything
to
add
in
the
ticket
suggest
with
my
earlier
okay,
you
added
a
comment.
E
E
A
Anything
else
for
triage.
F
It's
Keith
I
was
secondary,
yes,
I
want
to
bring
up
yeah.
The
one
issue
that
I
saw
appears
to
be
a
regression
of
this
put
it
here
in
the
chat,
12
404,
it
looks
like
it
was
discussed
already
on
slack
by
Michael,
and
so
yeah
I
just
want
to
bring
this
up
for
everyone's
awareness
and
the
other
one
I
wanted
to
bring
up.
Is
this
one
12
376
I'll
put
it
in
the
chat
as
well.
F
Yeah,
so
this
one
appears
to
me
to
be
a
bug
in
the
in
the
deletion
order
based
on
the
expected
behavior
and
so
I
want
to
bring
this
up
so
that
maybe
someone
who
is
more
familiar
with
the
sync
waves
Behavior
can
confirm
that
it
is
indeed
a
bug
and
maybe
have
a
look
at
it,
eyeball
it
and
make
a
comment
in
there
and
another
one
was
that
there
was
a
UI
crash
and
caused
by
the
ongoing
top
bar
redesign,
but
I
fixed
that
so.
E
F
Yeah,
it's
just
three
points
for
the
for
this
week
and
yeah.
That's
it
for
me.
A
Okay,
so
the
next
step
for
those
those
tickets
is
it
handled
Kate,
not
sure.
G
E
A
Something
we
discussed
briefly
in
the
the
containers
meeting
to
make
it
more
to
clarify
the
the
the
steps
that
is
required
for
the
triage
I.
Think
today
is
really
really
open
and
yeah,
so
we
need
to
better
Define
what
what
should
be
done
during
triage,
but
for
now
I
guess
what
we
need
is
to
define
a
direction
for
for
the
open,
Open
tickets
that
was
brought
to
attention.
So
what
is
the
direction
you
suggest
for
those.
F
Oh
I
just
want
this
one
for
this.
One
I
just
want
to
make
sure
that
this
is
indeed
an
issue,
because
yeah
I
I
didn't
have
any
comments.
I
didn't
add
any
comments
to
to
the
issue,
but
I
thought
that
it.
It
seemed
like
a
bug,
but
I
just
want
somebody
to
confirm.
That's
all.
E
A
Okay,
right
thanks
Kate:
okay,
is
it
done
for
triage?
A
Yes,
thanks,
Dan
thanks
Kate,
for
for
for
a
previous
week
work.
Now
we
need
to
collect
the
the
the
next
person
any
volunteers.
A
G
A
Thank
you
all
right
with
that.
Out
of
the
way,
let's
go
with
with
the
topics
that
we
have.
Let
me
just
see
my
time.
Okay,
we
have
a
few
minutes
yeah.
We
have
quite
a
few
topics,
so
if
you
can
yeah
be
careful
with
not
taking
a
lot
of
time,
we
have
a
lot
of
a
lot
of
topics
to
cover.
So
starting
with
you,
gergel
gargling
go
free.
A
D
Still,
no
good
I
think
it
was
a
hardware
issue
because
you
were,
you
were
off
mute.
We
just
couldn't
hear
you.
Oh.
C
Yeah
sorry,
yes,
now
I
can't
hear
you
I've
never
got
me
to
the
software
on
YouTube
and
vice
versa.
Yes,
go
ahead!
Thank
you.
So
the
DCI
link
is
also
connected
to
the
multi.
Source
doesn't
work
with
c
and
pzu
and
I
get
a
PR
on
the
way
fixing
it
with
the
help
of
Michael
he's
been
helping
a
lot
since
the
past
few
days.
Thank
you
very
much.
Michael
and
I
would
really
refer
to
have
it
a
candidate
for
2.7
and
I'm.
C
Looking
for
approvers
for
it
and
Michael
said
it
doesn't
fit
anymore
in
his
time,
he's
overburdened
and
I
would
be
really
happy
to
have
someone
to
review,
which
one
is
done.
It
doesn't
seems
to
be
a
big
change
by
the
way
all
right
check
the
source.
It's
not
a
biggie
I
just
have
to
extend
the
crds
that's
already
on
the
way
and
at
the
copying
logic.
D
I
can
summarize
the
change
in
sort
of
or
go
CDE
language
for
folks,
so
right
now
with
multi-sources,
the
only
connection
we
provide
between
two
sources
is
for
one
Helm
chart
to
pull
a
values
file
from
a
different
git
repo,
so
that
that's
done
literally
just
by
path.
Substitution
you're
only
allowed
to
access
that
one
file
on
the
other
repo.
This
change
would
be
to
basically
create
a
new
directory
for
both
of
the
sources
and
let
them
basically
access
each
other's
files.
D
As
wanted
I
think
it
would
open
up
a
lot
of
possibilities
for
combining
different
rendering
tools,
so
I
think
it'd
be
really
powerful,
but
it's
going
to
be
a
non-trivial
PR.
It
may
not
be
too
bad,
but
it
will
be
non-trivial
and
it'll
take
some
time
to
review
so
I
can't
review
it
for
2.7.
If
someone
else
can
then
great.
C
This,
yes,
it's
Linked,
In,
This,
Very
issue,
I,
think
a
bit
of
words
downwards
yeah.
My.
A
A
G
C
C
D
H
C
You
very
much
I
I
would
like
to
have
two
more
questions
slightly
connected.
One
is
the
contribution.
Locality
said:
I
need
to
use
my
final
female
as
my
primary
one
on
my
GitHub
account
and
the
corporate
guy
says
that
you
wouldn't
be
able
to
make
a
difference
between
any
of
the
email
addresses
unless
it's
also
authenticated,
and
why
is
that
particular
requirement,
because
I
into
a
corporate
biometric
regarding
our
account
setups
are
slightly
different.
C
We
know
exactly
contributing
MD
I
think
it
was
versus
filing.
C
A
It's
open
it's
to
just
you
just
follow
the
procedure
there
and
you're
gonna
be
all.
C
Or
no,
where
is
the?
Where
is
your
Flex
link
right,
select,
Channel
and
when
I'm
creating
an
account
I'm
being
told
that
you
can
use
any
accounts
with
the
domain
referee
but
CEO
and
QA?
Let's
come.
C
A
Okay,
let's
move
Justin
open,
ssf
scorecard,
there's
a
PR
open.
You
want
to
mention
anything
about
it,
go
for
it.
Yeah.
G
Can
you
can
you
do
me
a
favor
and
just
pull
it
up?
Unfortunately,
I'm
not
able
to
share
a
screen
at
this
time
sure
no
problem,
yeah
I,
just
wanted
to
discuss
adding
another
badge
on
our
main
readme.
This
is
a
newer
card
and
currently
they're
scanning
over
a
million
repositories,
and
it's
public
data.
G
Adding
this
badge
would
include
us
currently,
Argo
CD
has
a
score
of
8.7,
which
is
really
really
good.
That's
outstanding,
I
didn't
put
all
our
other
sister
projects
in
there
out
of
all
the
graduated
projects.
There's
only
two
projects
currently
used
in
this
I
believe
it's
SCD
and
tough.
G
G
So
I
just
wanted
to
discuss
it
with
the
maintainers
and
see
if
they're,
okay,
with
adding
this
I
know,
we
have
quite
a
few
badges,
but
basically
it
checks
for
security,
making
sure
we're
doing
a
proper
security,
okay
and
it
just
it
basically
just
shows
we're
doing
really
good.
If.
A
B
G
We
currently
have
best
practices.
This
is
built
upon
that
to
actually
improve
even
more
and
with
the
pr
where
it
runs
the
workflow.
It's
actually
for
the
approvers
they're,
going
to
see
in
the
security
tab
remediation
steps
on
how
to
improve
our
score.
G
One
of
the
few
things
we
need
to
do
is
maybe
pin
some
dependencies
in
our
Docker
file.
There
is
one
thing
currently
broken
in
their
scorecard
and
that's
the
fuzzing
check.
We
do
do
fuzzing,
but
it
doesn't
detect
us
doing
fuzzing
because
it's
listed
as
the
main
Ergo
project
repository.
Not
each
individual
repository
I'm
actually
going
to
open
an
issue
with
them
on
that.
E
G
Yeah
I
just
pushed
this
one.
Last
night
I've
been
working
hard
on
this.
During
the
release
process.
With
this
PR
we
will
now
be
generating
attestations
and
a
true
provenance,
so
users
can
verify
that
the
when
they
download
the
image
they
can
use
the
mission
controllers
they
can
integrate
it
into
their
CI
CD
process
and
actually
verify
these
images
were
built
by
us
on
GitHub,
using
our
workflows
and
not
by
anything
else.
G
I
not
gonna
do
a
demo
on
it
today.
I
can
do
that.
That
would
take
actually
quite
a
bit
of
time,
but
I
did
want
to
discuss
a
few
changes
that
I
made
on
this.
The
main
one
currently
right
now
when
we
push
a
release,
we're
pushing
to
Quay
and
Docker
hub
I
did
not
include
Docker
Hub
on
this
I
can
do
we
do
you
know
if
we
still
want
to
push
the
docker
Hub
there.
A
E
E
G
A
Well,
that
was
my
main
question
any
more
any
more
questions
related
to
that.
G
That's
how
I
designed
this?
Yes,
it
will
be:
okay,
I
created
a
reusable
workflow
which
you'll
be,
which
I'll
help
you
implement
it,
of
course,
I'm
actually
going
to
hope,
hopefully
help
all
the
sister
projects
implement
this
that'd.
A
H
A
Lot,
Justin
awesome
work,
no
problem
cool
right.
So
let's
move
ahead
with
Blake
Blake
is
bringing
up
a
topic.
Pr
backlog
is
starting
to
get
big.
Is
there
anything
that
we
can
do
on
that
front?
So
Blake?
You
want
to
start
that
discussion.
Yeah.
I
Sure,
well,
as
you
guys
can
see
it
as
342
open
PRS
on
the
rocd
repo
and
I
mean
one
I
mean
I
tried
to
triage,
but
that's
I
mean
I,
get
the
general
feeling
like
one
there's
a
lot
of
backlog
where
there's
there's
I,
guess:
there's
some
self-interest
there
with
some
of
my
own
peers
haven't
been
looked
at
I,
understand,
I,
understand
why,
but
then
I
also
get
the
general
feeling
that
there's
a
lot
of
I
mean
I,
see
lately
like
a
lot
of,
like
should
I,
say,
low
quality
PR's
coming
in.
I
It's
like
you
know
like
fixes
which
don't
really
fix
anything
or
and
or
they
don't
have
any
descriptions.
And
then
the
question
is
like
like
in
terms
of
triaging:
is
there
some
kind
of
process
to
do
or
is
there
somewhere
some
there's
something
where
we
can
just
be
ruthless
to
be
like
okay?
This
is
this
not
good
enough
clothes
or
is
there
something
other
means
we
can
move
forward,
I
mean
and
for
bigger
peers
as
well,
of
course,
where
there
would?
There
also
need
to
be
some
kind
of
triage
now
I.
I
Try
to
that
I
mean
that's
there.
There's
a
lot
I
mean
just
looking
at
it
just
gives
me
is
yeah
I
have
to
work.
A
Yeah
yeah:
this
is
something
that
is
being
discussed
so
in
the
last
maintenance
meeting.
This
is
a
topic
that
was
discussed
and
while
we
have
a
few
suggestions,
one
Implement
about
to
to
help
us
with
that.
Second,
to
clarify
the
the
steps
required
for
the
person
doing
triage
to
to
proceed
with
at
least
giving
some
directions
in
the
open
PRS
for
the
week.
So
those
are.
A
E
D
To
do
the
numbers
are
kind
of
interesting
at
the
bottom,
because
the
green
bars
show
the
gap
between
the
number
of
PR's
opened
versus
PR's
reviewed
and
even
though,
like
you
can
see
in
the
bottom
graph,
in
yellow
we're
doing
a
pretty
good
job
of
increasing
the
number
of
reviews
that
we
approve
and
get
merged.
The
Gap
is
still
growing
and
I
agree.
There
are
a
lot
of
low
quality
PR's,
but
even
the
low
quality
ones
like
they
claim
to
fix
something.
E
I
Yeah,
that's
what
I
was
getting
at
then,
is
because
there's
a
lot
of
things
where
I
mean
I,
guess,
like
my
spice
sense,
would
be
like.
Okay
close,
this
is
not
really
very
useful.
Of
course
you
don't
want
to
be
mean
or
anything
about
it,
but
then,
but
then
there
is
a
lot
of
time.
Consumption
so
I
mean
I'm,
not
a
fan
of
bots,
either
for
probably
for
the
same
reasons
you
Michael,
but
nonetheless,
I
mean
that's
happening.
A
lot
of
PR's
open
is
yeah,
it's
growing
and
it's
not
going
to
get
better.
D
Think
it's
going
to
get
significantly
worse.
Reviews
submitted
over
the
past
six
months,
I
average
per
quarter
four
times
as
many
reviews
as
the
next
highest
reviewer,
that's
out
of
people
who
are
designated
approval
enough.
The
people
who
are
necessary
to
actually
get
the
prsm
I've.
D
Into
it,
it's
going
to
take
more
of
my
time
over
the
next
quarter
or
two.
So
this
the
number
of
openpr's
is
only
going
to
increase
if
we
don't
get
more
approvers
looking
at
things
and
getting
them
merged
or
closed.
A
Yeah,
we
need
to
to
continue
that
discussion
and
have
some
action
items
for
sure
really
decide
about
the
bot
I'm,
not
necessarily
against
it.
As
long
as
it's
it's
helping
us
and
well
configured
and
not
discarding
good
work
from
from
our
contributors,
so
yeah
I
think
all
that
needs
to
be.
We
have
again
a
discussion
in
place
and
we
just
need
to
to
move
forward
with
that
and
come
and
come
up
with
some
action
items.
A
Okay,
thank
you
anything
else
regarding
triage
PRS,.
A
All
right
so
we're
Michael
replied
to
you.
Garagely
to
the
link
you
posted.
D
E
D
A
I'll,
take
a
look
at
the
contributing
markdown
I
I
know
that
this
is
not
a
real
requirement.
You
can
use
an
email
as
long
as
as
Michael
said,
is
signed
off,
and
that's
that's
the
only
requirement
that
we
that
we
have.
I
H
A
G
A
B
Is
there
if
I
so
sorry?
Are
you
guys
done
at
that
point
about
the
Imagine
I
think
so
yeah
go
ahead?
Some
of
these
PRS
are
like
pretty
simple,
like
there's
one
about
upgrading
the
coop
CTL
version
from
two
three
to
two
four
which
I
feel
like
is
uncontroversial
and
like
okay
I
can
just
improve
that
I'm,
not
an
approver,
but
if
is
their
way
to
tag
an
issue
is
like
like
it
like
tagging.
A
PR
is
like
this
is
kind
of
uncontroversial
or
like
I
expect.
B
B
Is
there
a
way
to
label
it
as
like?
This
is
like
a
dead,
simple
one
like
they're
fixing
some
documentation.
This
is
like
a
super
simple.
You
know
exactly.
A
That'll
be
great.
That
would
be
great
to
have
something
like
that.
I
know
kubernetes
folks.
They
use
some
some
bullets
to
help
with
that
yeah
right.
D
H
A
Yeah,
it
depends
on
how
to
label
sends
the
notification
right.
The
bot
might
send
a
notification
in
a
easier
way
for
us
to
track
the
label.
I
had
a
past
experience
with
github's
notifications,
so
not
always
easy
to
to
track
everything
from
GitHub,
because
it's
a
lot
so
yeah.
Maybe
we
can
improve
that
for
sure
and
make
make
it
more
evident
for
all
the
provers
and
maintainers
whenever
something
is
is,
is.
B
Is
there
is
there
a
label
like
that?
That's,
like
that's,
I,
feel
like
it's
a
grading
system.
You
know
it's
like
easy
change.
Medium
change,
hard
change,.
H
So
I
know
a
lot
of
projects.
I,
don't
think
it's
based
on
complexity
per
se,
but
to
do
small
medium
large
extra
large,
like
I,
know,
kubernetes
has
them
as
well
I!
Think
Bots
just
do
it
based
on
flying
count,
which
isn't
necessarily
great
for
complexity,
but
it's
just
something
I've
seen
in
the
past
I've
seen
the
auto
small
medium
large
XL
labels
get
added
to
Stars,
I,
don't
know
if
something
exists
out
there
already
that
we
can
just
do
that
with
some
thresholds,
but.
A
Yeah
yeah
yeah
I,
I
guess
my
complaint
is
about
the
notification
system
that
the
label
is
fine,
but
I
I
found
hard
to
configure
my
account
to
receive
notifications
when
specific
labels
are
added
to
so
there's
I'm,
not
aware
of
a
good
yeah.
A
D
The
difference
and
they're
they're
nine
active
approvers
on
Argo
CD
and
if
not
getting
notified
when
a
PR
is
marked
ready
to
review,
is
actually
keeping
people
from
going
through
and
reviewing
and
approving
PRS.
Then
we
should
probably
build
some
automation,
but
my
sense
is
how
do
the
nine
people
who
are,
who
are
reviewing
I'm,
not
sure
that
that's
going
to
make
a
dent
I
think
that
the
time
spent
building
the
tool
is
probably
going
to
be?
You
know
better
spent
actually
just
reviewing
the
PRS
well.
A
I'm
not
saying
building
any
tool,
there's
things
already
there
available
to
you
so
yeah,
but
again,
let's
discuss
and
let's
let's
see
how
can
we
improve
that
and
there
are
other
discussions
in
in
place.
So
let's
wrap
up
everything
and
I
think
yeah
for
sure
we
need
to
improve
that
and
we
can't
believe
with
that
super
huge
number
of
PRS
that
keep
increasing.
G
Go
ahead,
I
know,
we're
short
on
time.
Can
I
mention
something
real
quick?
Yes,
on
that
attestation
PR,
there
was
one
possible
breaking
change
as
far
as
the
release
assets
go.
They
are
now
coming
a
compressed
format.
Should
we
post
something
on
slack
and
let
people
know,
or
what
do
you
think.
A
G
G
The
CLI
binaries
before
they
come
in
a
compressed
format.
Now,
instead
of
just
the
raw
binary,
okay
I
know
some
some
people
may
have
CIS
CI
CD
systems
that
automatically
just
pull
it.
Yeah.
A
D
We
we
couldn't
I
mean
you're,
your
release.
Work
is
pretty
big.
Maybe
we
do
yeah
scary
pick
it
back
and
we
just
release
it
on
2.7.
H
G
A
Okay,
it's
preferable
if
it's
published
yeah.
E
A
All
right,
okay,
cool,
so
we
are
two
minutes
out
of
time.