►
From YouTube: 2021-11-30 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
A
C
Good
morning
sitting
here,
just
ignoring
the
reminders,
as
they
keep
popping
up
and
then
somehow
didn't
get
a
reminder
for
the
actual
10
o'clock.
Though.
C
I
was
messing
with
my
trying
to
figure
out
how
to
sync
google
calendar
to
outlook
that
was
fun.
B
C
Yeah
for
the
most
part,
since
you
know
everybody's
using
that,
and
we
have
good
support
and
good
setup
things,
work
pretty
good,
except
when
trying
to
interrupt
with
other
things
and
the
fact
that
everybody
in
the
world
hates
teams
so
they're
like.
C
C
I
actually
like
teams,
actually
I
mean
they've,
come
a
long
way
since
the
they've
put
a
ton
into
it
since
the
pandemic
with
everybody,
although
my
everybody
at
microsoft,
doing
everything
on
teams
now.
A
C
D
C
Yeah
they
have
a
cool
program
in
the
u.s
called
teals,
which
is
a
lot
of
like
computer
sciencey.
Education
working
with
schools
and
stuff
also.
E
B
C
D
D
C
C
So
patches
so
specifically
191
and
then
in
general,
our
release
strategy,
automation.
C
Yeah
I
have
time
tomorrow
afternoon,
if
nobody
wants
to
wants
to
mess
with
that,
which
I
I
I
would.
C
I
think
I
have
probably
the
most
context
from
from
and
and
have
probably
the
person
who
broke
whatever
is
not
working
so.
D
C
D
F
C
C
C
I
think
we
would
merge
that
to
maine
after
we
release.
C
The
discussion
I
remember
before
was
just
a
a
preference
for
not
having
not
promoting
sort
of
the
late
at
this
latest.
Am
I
sharing
what
happened?
You
stopped
sharing,
so
I
was
sharing
okay,
yeah.
C
Well,
actually
I'll
wait
until
what
I
want
to
do
is
I
want
to
list
out
all
the
steps
in
the
releasing
doc
the
manual
steps.
So
I
can
basically
just
point
the
governance
committee
at
that
of
these
are
the
steps
that
we
want
to
automate.
D
C
Yeah
or
keep
the
protection
but
remove
the
cla
you're,
not
quite
sure
on
that.
C
C
Yeah
we
had,
we
didn't
have
this
at
all.
We
had
like,
or
this
the
easy
cla
bot
keeps
like
reapplying
these
rules
from
time
to
time,
or
they
did
at
one
point
and
we
went
in
and
kind
of
did
some
work
around,
but
now
that
it's
manual
and
now
that
they
had
that,
I
had
kind
of
synced
it
back
to
what
it
was
before.
C
C
I
remember
seeing
on
an
issue
that
the
sort
of
yeah-
I
think
that
was
like
a
one-time
thing,
but
with
the
idea
that
it
should
probably
be
ongoing
like
if
they
give
us
maintainers
like
that,
was
sort
of
the
a
balancing
act
of
yes,
okay
maintainers,
you
can
have
more
rights
but
from
time
to
time,
we'll
verify
that
you,
you
haven't
done
anything
horrible
with
the
cla
check,
for
example,.
D
D
D
D
C
C
D
Yeah,
especially
with
aws
stuff,
when
it's
one
of
those
things
where
I'm
sure
like
aws,
cares
how
cool
you
can
get
those
fixes
up.
So
that's
the
only
reason
I
care,
otherwise
I
wouldn't
care,
but-
and
so
it's
like,
if
it
takes
a
week
of
course,
I
mean
like
the
longer
it
takes
for
upstream
to
be
able
to
release
patches
the
less
confidence
they
have
been
open-thumbed
in
general.
I
guess
less
sort
of
conversations.
D
C
With,
if
there's
a
if
there's
a
break,
and
we
need
a.
D
C
C
B
B
When
using
those
test
containers
cloud,
I
almost
was
able
to
run
all
our
tests
with
just
like
unlimited
gradle
workers,
like
how
many
cores
do.
I
have
use
all
of
them
and
it
works
so
much
faster
than
than
without
probably
not
just
because
of
like
speed,
but
because
your
local
machine
is
not
overloaded
anymore,
but
by
all
those
test
containers.
B
D
B
Yes,
so
build
just
oh,
oh
there
is
that
integration.
Then
I
will
redirect
proxy,
like
docker
commands
to
remote
docker.
If
there
is,
if
there
is
not,
then
it
all
works
as
usual
for
your
local
doctor
yeah,
so
no
no
changes,
no
actual
changes
in
build,
except
for
test
containers
version
upgrades
that
I
did
that's
all.
C
C
D
C
Yeah,
this
in
particular
has
been
why
I
can't
I
want
to
enable,
like
our
test,
suite
on
windows,
just
so
that
we
can
make
sure.
B
C
C
Sorry,
oh
yeah,
so
that
we
have
tests
that
don't
pass
on
windows,
so
local
developers
working
on
windows,
some
of
our
tests,
don't
pass
or
haven't
passed
in
the
past.
D
B
C
B
So
this
containers
cloud
are
currently
like
in
beta
and
invite
only
I
do
have
an
invitation
link,
but
it's
like
they
currently
like
target
companies
and
they
want
to
like
sneak
into
companies.
B
C
I
don't
think
yeah,
I
don't
think
I
would
run
it
locally,
but
if
you
know
at
the
point
where
we
could
run
it,
like
you
said
from
the
the
github
actions
and
you
know
so
they
have
it.
C
C
Both
of
these
would
be
great
either
either
one
or
both.
D
B
B
B
B
D
C
C
Or
do
you
want
to
go
over
to
get
materia
and
anarch's
thoughts
on.
A
C
C
Fair,
we
can
chat
on.
I
I
I
just
would
like
to
you
know,
unblock
the
contributor,
maybe
by
end
of
the
week.
A
Thank
you.
Okay,
so
just
like
the
final
tip,
yeah
camel
endpoint
yuri
is
a
different
thing
from
http
url.
It's
an
that
camera
endpoint
is
a
completely
internal
thing.
It's
internal
to
come
up.
It
can
contain
like
various
different
things
as
this
scheme,
which
is
not
a
real
scheme,
but
it's
it's
used
internally
by
by
camel
to
direct
the
calls
to
different
components.
A
So
in
this
case,
if
you
put
http
4,
it
uses
apache
http
4
client,
but
if
you
use
http
it
uses,
I
think
apache,
two
http
clients
something
like
that
and
if
you
use
nerdy
hyphen
http,
then
it
uses
neti
for
http
calls
which
is
totally
confusing,
but
that's
how
it
works.
B
A
C
C
And
this
is
almost:
could
this
be
an
http
client
test
or
maybe
not
now?
That
would
be
probably
too
much
for
this
pr,
but.
C
D
A
D
B
D
B
B
D
D
D
Like
our
maria
does
away
to
use
custom
schemes
to
register
different
things
like
grpc
is
http
plus
jrpc,
and
I
think
in
the
brave
instrumentation
people
liked,
seeing
that
that's
what
I
vaguely
remember
just
as
one
random
data
point.
I
don't
know
what
to
do
here,
though.
Let's
see.
D
Yeah
so
at
least
in
the
army
instrumentation,
I
can
say
that
we
do
record
weird
schemes
also.
E
C
Sure
I
don't
see
like
ftp
like
anything:
that's
not
http,
really
working,
but
things
that
are
yeah
things
that
are
still
http.
F
D
F
A
I
can
try
to
do
that.
I
I'll
try
to
reserve
some
time
this
week
or
next
week
like
a
day
or
two
and
that,
but
I
I
think
it
will
have
pretty
interesting
structure
since
there
are
so
many
different
kinds
of
conventions
used
there
messaging
http
or
everything.
D
A
The
idea
I
had
is,
like
you
see,
saw
this
decorator
registry
and
in
the
previous
vr,
and
we
could
just
have
a
separate
instrument
for
each
component,
maybe
yeah
it
could
work
instrumental.
C
I
also
think
it's
like:
we
don't
need
to
block
instrument
or
api
being
stable
on
those
last
like
two
or
three
instrumentations
from
you
know,
from
a
prioritization
perspective,.
A
C
D
C
I
don't
feel
super
strong.
I
I
I'm
still
weakly.
B
In
favor
of
it,
no
so
that
that
pull
request
is
open
long
enough,
it
has
four
four
approvals
or
three
approvals,
only
two
balls
again,
so
I
thought
either
nurture
clothes.
Okay,
there
is
no.
D
C
C
B
C
I
was
gonna
ask
if
anybody
had
a
chance
to
look
at
the.
I
tried
to
take
notes,
but
the
main
them
you
know
most
of
the
discussion
was
just
what
what
what
are
we
trying
to
accomplish.
C
C
C
Yeah,
it
hasn't
made
it
to
the
calendar.
Yet,
okay,
I
guess
I
probably
have
I
don't
know
how
to
do
it
with
the
new
google.
I
think
m2.
What's
m2's
name,
morgan,
morgan,.
D
C
Supposed
to
add
it
to
the
calendar,
but
it's
it's
alternating
every
other
week
monday.
C
10
o'clock,
pacific
time
and
wednesday
10
o'clock,
central
european
central
european
time.
C
A
C
So
I
think
the
action
items
were
async
to
get
different,
observability
vendors
in
particular,
I
think
datadog,
the
splunk
ones
already
published
their.
C
C
C
Yeah
this
was,
it
was
asked
jonathan
asked
about
requirement
for
metric
instrumentation
to
be
declared
stable,
so.
C
He
went
back
to
yeah.
The
the
current
wording
is
still
the
same,
where
we're,
basically
not
supposed
to
instrumentation,
should
must
not
be
marked
as
stable
until
telemetry
stability
is
defined.
C
C
E
Well,
I
was
actually
talking
about
this
before
with
my
team,
and
I
find
I
find
that
the
kind
of
instrumentation
gets
defined
before,
like
the
actual
convention
and
like
the
actual
data
itself,
we've
found
that
kind
of
backwards
like
I
would
think
that
the
convention
would
be
defined
stable
first
and
then
the
instrumentation
would
be
defined
still
because
I
can't
see
how
you
can
make
an
instrumentation
stable
if,
like
the
data
emits,
is
not
stable.
E
If
that
makes
sense,
I
assume
that's
what
you
mean
by
telemetry
stability
yeah,
because
that
is
what
like
makes
the
data
right
like
the
and
a
system
is
only
good
as
like
the
data
it
emits
so
yeah.
I
would
that
I
think
the
push
to
stabilize
the
convention
needs
to
go
a
bit
faster
or
do
what
and
do
whatever
we
can
to
get
a
stable,
especially,
I
feel
like
there's
not
much
happening
in
the
metrics
convention
space
as
well,
so
that
yeah,
that's
my
view.
E
Well,
yeah
because
you
have
say
I'm,
I
have
a
dashboard
that
I've
got
and
I'm
filtering
by
some
attribute
and
then
the
attribute
changes
then
my
dashboard
breaks.
So
it
is
a
breaking
change
for
the
user.
If
you
change
that
and
yeah.
So
that's
why
yeah?
I
just
think
the
semantic
convention
should
be
stable,
like
as
soon
as
possible
and
the
same
with
metrics
as
well.
E
D
The
past
I've
generally
tried
like,
even
if
I
made
a
breaking
change
or
metric
and
some
framework,
like
even
army,
edited
it
sometimes
before
1.0.
I
don't
know
if
they've
done
it
since
one
point,
it
was
always
important
to
give
a
couple
of
releases
to
be
able
to
migrate
yeah.
I
would
say
that
stability
is
important,
at
least
to
some
degree,
if
not
a
100
guarantee,
though.
C
Yeah,
what
do
you
think
james
from
so
there's
getting
getting
it
stable
and
then
so,
once
it's
stable,
can
you
remove.
C
E
Yeah,
I
guess
you
need
to
keep
once
it's
there
and
it's
stable.
You
need
to
keep
it
there
unless
there
is
some
kind
of
automatic
translation
mechanism
which
I
know
there
is
something
like
the
schema
in
the
spec.
So
if,
if
the
collector
say
can
handle
a
translation
that
will
you
know,
make
the
same
kind
of
and
make
the
experience
the
same
for
the
end
user,
it
doesn't
like
break
their
dashboards
and
all
kinds
of
things.
E
C
Yeah
there
was
in
the
meeting
today
there
was
some
concern
about
things
that,
like
there's
some
things
that
are
better
known
and
some
that
here
this
there
was
some
consideration
of
well
what
about
things
that
we're
not
sure
about,
or
I
think
in
this
case,
though
it
was
more
about
temporary
things
like
like,
say
on
temporary
jvm
metrics,
that
only
apply
to
an
older
version
of
the
jvm
that
we
know
is
going
to
be
phased
out
like
say
a
cms
collection
type
of
metric.
C
Not
and
maybe
maybe
we
were
in
this
meeting,
getting
confused
between
stability
of
what
the
jvm
itself
emits
like,
there's
no
yeah,
so
the
jvm
say
that
jfr
emits
certain
metrics
and
java
8
emits
those
but
starting
at
java.
8?
U
300!
It's
not
getting
emitted.
E
C
E
Changing
your
applications,
so
you
kind
of
expect
something
like
that
to
happen
right,
but
if
I
say
upgrade
from
hotel
one,
you
know
0.9
to
1.10
and
all
of
a
sudden
my
dashboards
don't
work
anymore.
I
think
that
is
that's
probably
not
ideal
is
what
I
mean,
but
if
you're
upgrading
say
you're
upgrading,
you
know
some
library
or
you're
upgrading
some
you
know
jvm
or
or
whatever
it
is.
You
can't
like
the
the
convention
says
you
don't
well.
If
it
can't
meet
the
convention,
you
upgrade
something
and
the
instrumentation
can't
meet
the
convention.
D
Are
you
differentiating
between
a
jvm
versus
an
hotel
version
upgrade?
Those
still
seems
similar
to
me,
it's
more
is
where
that
discussion
comes
from.
D
E
Yeah
yeah,
I
think
that
makes
sense
like
that.
That
makes
sense.
In
that
case,
I
think
the
convention
will
require
some
attributes,
but
you'd
probably
only
require
ones
that
you
know
will
always
be
there
and
the
rest
will
be
obviously
optional
and
yeah
it's
more
about.
I
guess,
convention
stability.
E
I
suppose
you
wouldn't
like
remove
things
from
the
convention,
but
maybe
yeah
you
upgrade
versions.
You
upgrade
instrumentation.
E
I
think
the
user
has
to
be
kind
of
willing
to
accept
some
breakage
there
yeah,
I'm
really
sorry.
I
have
to
go
to
a
meeting
right
in
the
middle
of
this,
but
I'll.
C
E
D
D
I
don't
know
because
normally
http
is
like
to
notify
normalize,
very
different
http
frameworks,
and
if
you
only
had
one
http
library
ever,
you
wouldn't
have
cement
to
convince
you
just
instrument
that
library
will
be
done
so
jvm.
I
guess,
falls
sort
of
in
the
middle
of
those
two
extremes
but
still
seems
to
lean
towards
just
one
thing:
not
an
abstraction.
D
F
C
Discussion
of
that
and
how
they
have
different,
especially
growl,
having
very
different
and
missing
metrics.
In
many
cases.
C
I
think
the
goal
is
to
define
the
define
the
metrics
define
mappings
from
these.
I
don't
think
we're
expecting
the
jvm
makers
at
this
point
to
emit
what
we
standardize
and
then
what
ben
calls
t-shaped
api.
B
C
He
means
like
the
common
stuff
that
works
across
all
of
them
and
but
still
allowing
to
go
deep
and
differentiating
across
things.
Okay,.
D
C
It
could
be
for
sure
I
personally
really
like
the
idea
of
defining
the
defining
them
in
the
spec,
just
as
a
as
a
place
where
we
can
drive
that
discussion
and
agreement
from
since
there's
going
to
be
a
lot
of
opinions
about
jvm
metrics.
D
Yeah,
that's
probably
where
I
have
not
like
it's
just
too
many
opinions.
Can
you
actually
reach
something?
That's
the
question.
I
guess,
but
hopefully
that's
not
the
situation
that
we
would
run
into
so
rather
than
that
like
leaving
each
jvm
to
have
a
bit
more
freedom
is
another
way.
I
guess
to
approach
that
if
that's
required,
I'm
not
too
sure.
C
I
I
think
the
harder
thing
to
to
get
agreement
on
is
what
is
sort
of
the
the
set
of
metrics
that
are
worth
collecting.
C
I
mean
there's,
there's
a
billion
metrics
yeah
and
so
that's
kind
of
the
a
bunch
of
the
meeting
kind
of
rabbit
holed
into
that
area
of
well.
Why
are
we
we
need
to
define
why
we
want
to
collect
metrics
first?
What
are
we?
What
problem
are
we
trying
to
solve
before?
We
know
which
metrics
we
want
to
collect,
but
I
think
ben
ben
did
a
good
job
of
you
know,
kind
of
pulling
it
up
being,
like
you
know,
hey,
there's
a
lot
of
a
lot
of
people.
D
C
Already
collect
a
lot
of
vendors
already
collect
certain
sets
of
metrics.
They
already
know
these
are
useful
to
their
customers.
That
seems
like
a
really
reasonable
place
to
start.
C
So
I'm
fairly
optimistic,
more
optimistic
than
I
was
before
the
meeting,
even
though
there
was
a
lot
of
kind
of
sideways
discussions
at
times.
C
But
it's
great,
I
mean
ben
ben
is
doing
open,
telemetry
and
hopefully
the.
C
Jason
asked.
C
Micrometer
yeah,
so
micrometer
has
2.0
coming
up
and
they
are
interested
in
aligning
with
the
these
conventions,
these
namings,
potentially
breaking
stuff.
C
C
Yeah
so
jason
asked
you
know
if
there
were
interested
if
they
were
thinking
of
aligning,
because
I
think
jonathan
brought
up
the
2.0
release
and
oh,
I
didn't
the
aaron
didn't
sign
in
either.
Aaron
said
that
the
they
wouldn't
be
here
if
they
weren't
interested
aligning
so.
A
D
D
C
You
can
you
can
update
us
on
on
thursday.