►
From YouTube: 2022-01-10 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
B
A
Boo
covid,
hey
feeling
better
yeah
I
ate.
Luckily
my
partner
had
mild
cases
of
it
but
very
typical
deal.
We
had
a
small
family
gathering
for
christmas,
like
you
know,
maybe
seven,
seven
or
eight
of
us,
and
then
we
all
had
it
afterwards,
in
spite
of
the
fact
that
we're
all
vaccinated
and
a
lot
of
people
were
boosted.
C
A
Yep
yeah,
you
just
got
unlucky
one
fun
fact.
I
don't
know
if
people
saw.
I
did
see
like
one
scientific
article
about
this,
but
we
all
clearly
had
it,
but
when
we
tested
using
pcr
tests
going
down
and
getting
the
proper
pcr
test,
only
about
half
of
us
came
up
positive,
even
though
we
all
have
the
same
symptoms.
So
obviously
we
all
had
it,
and
I
was
reading
a
thing
saying
that
the
pcr
tests
don't
appear
to
be
as
accurate
with
the
omicron
variant,
at
least
as
far
as
giving
false
negatives.
A
And
on
that
note,
maybe
we
should
just
charge
on
in.
I
see
people
have
been
filling
out
the
sig
check-ins
and
morgan's
out
today.
So
I'm
happy
to
to
just
run
through
this
with
the
maintainers.
A
D
D
D
A
A
A
All
right
looks
like
typing
has
stopped,
let's
probably
go
through
this
starting
with
specification.
A
Here
we
are
all
right
specification,
so
december
release
was
skipped,
which
is
fine,
but
a
release
soon
is
most
likely
and
there's
a
lot
of
important
metrics
stuff.
Any
details
about
like
where,
what's
in
this
new
metrics
release,.
F
Yeah,
I
think
that
one
of
the
ideas
is
that
we
wanted
to
go
stable
with
the
sdk
at
least
doing
some
parts
as
experimental.
However,
we
we
wanted
to
clarify
more
the
documents.
Josh
stuart
has
been
driving
this,
but
he's
he
won't
be
around
for,
for
he
has
been
around
for
a
couple
of
weeks
with
personal
stuff,
so
we
will
try
to
think
about
how
to
handle
that.
I
will
probably
talk
to
riley
on
private,
but
yeah
matrix
is
the
important
thing
now,
so
we
will
we'll
see
when
we.
D
D
Sorry,
I
mean
carlos
again
definitely
because
I've
been
working
with
josh
and
riley
as
well
as
bogdan
on,
you
know,
figuring
out
what
we
can
release
in
a
phased
way
for
metrics
and
obviously
the
you
know
clear
gap
right
now
is
some
of
the
spec
work
that
needs
to
be
completed
in
specific
areas,
both
for
the
sdk
implementations
to
be
complete,
as
well
as
some
of
the
prometheus.
F
Yeah
that
totally
makes
sense
yeah.
I
think
that
that's
a
fair
complaint,
yeah,
we'll
think
about
what
things
to
do,
as
I
mentioned
before,
the
things
that
wanted
to
be
done
were
mostly
clarifications.
Some
of
the
things
riley
already
proposed
as
experimental
like
exemplars,
for
example.
So
we
will
try
our
best,
I'm
sorry
for
the
delay
unexpected,
but
yeah.
I
totally
agree
with
you
with
what
you
are
saying.
F
D
Yeah,
and,
and
also
carlos,
just
to
your
point-
I
mean
I
chatted
with
josh
suretz.
D
Also,
and
you
know,
one
of
the
areas
as
you
know
which
I've
been
driving
is
the
prometheus
interoperability
and
making
sure
that
the
collector,
as
well
as
the
other
sdks
you
know,
have
either
pull
exporters
or
push
exporters,
but
you
know
making
sure
all
the
components
are
prw
compliant
and
I
have
I
do
have
a
gap
analysis
which
I've
posted
as
well
as
a
solution
to
fix
anything,
that's
missing,
but
that
also,
you
know,
for
the
full
exporters,
especially
is
dependent
on
the
spec
being
complete.
So
again
long
story
short.
D
This
affects
the
you
know
all
the
implementations
that
exist
for
prometheus
exporters,
as
well
as
in
the
sdks
as
well
as
the
collector.
The
collector
is
actually
in
good
shape.
We
do
have
we've
verified
everything,
so
it
is
ready
actually
for
prometheus
interrupt,
but
the
full
exporters
and
the
spec
needs
to
be
complete.
D
So
we,
you
know
again:
I've
been
looking
at
how
we
can
help
in
you
know,
working
with
the
prometheus
work
group
to
complete
that
out
and
submit
you
know
or
update
the
pr
that
is
already
there
for
draft.
A
It
seems
like
this
might
be
a
good
time
when
we
release
this
version
of
the
spec
to
also
put
out
a
a
short
blog
post.
Yeah.
Definitely.
D
A
All
right,
awesome,
yeah,
okay,
next
up
is
metrics.
We
were
just
just
discussing
this.
Anyone
have
any
additional
metrics
notes.
They
want
to
voice.
G
I've
updated
the
the
matrix
project
board
on
the
main,
open
telemetry
project
last
week
and
we'll
be
doing
the
daily
issue
change.
One
thing
I
noticed
is:
there's
a
dependency
on
the
language
implementation.
So
so
far
my
understanding
is
like
we
have
dependency
on
python
on
javascript
and
non-go.
G
G
H
Sorry
go
ahead;
this
is
this
is
tyler
from
the
gosig.
This
is
something
that
we've
been
discussing
last
week
at
our
sig
meeting
as
to
like
how
we
can
communicate
our
current
status
of
the
metric
implementation.
Our
plan
is
this
week
to
have
some
sort
of
project
created
to
try
to
communicate
this
to
you.
H
H
That
being
said,
the
current
status
is
that
we
still
have
to
work
on
implementation
of
the
views
that
some
of
our
bigger
things
and
our
api
stability
is
still
a
question.
We're
still
making
sure
that
we're
providing
users
with
what
they
they
really
need,
and
so
you
know,
I
would
also
put
us
on
maybe
like
the
the
slower
path
as
compared
to
some
of
the
people
that
have
already
accomplished
this
fee.
D
K
J
Project
in
our
in
a
repo
with
the
metrics
issues,
bundled
in
then
also
that
will
be
a
a
good
way
for
community
and
products
to
do.
A
A
Better
but
yeah,
hopefully
we
can
keep
that
compliance
matrix
up
to
date,
because
this
is
kind
of
like
the
prime
period,
where
that's
useful
to
people.
H
So
ted,
I
can
only
speak
from
the
the
traces
release
that
we
did.
I
definitely
remember
we
in
the
gosig
audited
the
specification
itself
and
then
tried
to
update
the
the
compliance
matrix,
but
they
were
just
things
that
were
very
out
of
sync
and,
and
it
turned
out
that
there
were
essentially
like
two
sources
of
truth.
There,
like
one,
was
the
compliance
matrix.
One,
that's
the
specification,
so
I
it's
it's
value
I
think,
is
helpful
and
a
quick
overview,
but
to
say
that
you're
complying
with
the
specification.
A
I
see
I
mean
is
that
again,
is
that,
just
because
of
the
way
things
are
broken
down
in
the
matrix
are
being
kept
up
to
date
with
changes
in
the
spec
yeah.
I.
H
Think
all
of
the
above
and
there
were
actually
instances
where
the
specific
or
sorry
the
the
compliance
matrix
included,
things
that
were
not
included
or
not
really
well
specified
in
the
specification,
and
so
it's
gone
yeah.
It's
kind
of
all
three.
A
Okay,
maybe,
as
part
of
the
the
metrics
spec
sig
meeting
just
cleaning
up
the
metrics
part
of
that
compliance
matrix
might
be
a
helpful
helpful
item.
A
If
we
just
can't
keep
it
up
to
date
or
something
I
feel
like,
we
should
get
rid
of
it
frankly
like
it
should
be
either
something
that's
accurate
enough
to
be
helpful
or
you
know
something
we
don't
do.
It
seems
like
the
middle
ground
of
it.
Being
out
of
sync,
is
you
know
not
helpful
to
anybody.
L
Great,
it's
a
big
piece
of
work
and
diego
from
lightstep
has
essentially
prototyped
it
and
shown
us
a
few
pr's.
I
started
following
his
practice
just
kind
of
informally
when
I've
written
two
of
my
recent
spec
changes,
and
I
find
it's
actually
really
helpful.
So
it
might
improve
the
spec.
If
we
would
do
that,
yeah.
A
I
think
that's
the
other.
The
other
side
of
the
fence,
right,
like
the
spec,
is
not
necessarily
organized
to
to
make
it
easy.
So
if
that
sounds
like
a
great
project,
diego.
L
L
I
don't
know
I
feel
like
this
is
a
place
for
the
maintainers
to
discuss,
or
you
know,
I'm
not
sure
if
we
can
discuss
it
here
if
you'd
like
ted,
do
you
think
that's
worthwhile.
A
Well,
let's,
let's
maybe
move
through
all
of
the
everybody's
updates
and
we
have
time
at
the
end,
it'd
be
great
to
look
at
it,
but
yeah,
I
would
say
the
the
hotel
specification
channel
is
probably
the
best
place
to
put
it.
A
A
D
A
H
D
Yeah,
diego,
I
mean
again
I'd
love
to
see
a
proposal.
You
know
in
terms
of
what
you're
proposing
to
do,
because
I
mean
again
having
done
poc
prs
is
good,
but
also
having
an
larger
proposal
is
useful
for
folks.
To
just
understand,
I
mean
just
to
ted's
comments.
We've
attempted
this
before
also
so
just
just
to
make
sure
that
we're
successful.
A
Okay,
moving
on
logs,
nothing
back
from
break
great
to
have
you
back
logs
with
metrics
moving
through
the
pipe
I
do
feel
like
the
logs
group
is
been,
I
don't
say,
working
off
in
a
corner,
but
it's
been
like
because
the
a
lot
of
the
people
interested
in
logs
are
like
a
different
group
of
people
than
who've
been
interested
in
tracing
in
metrics.
A
D
I
just
wanted
to
give
a
quick
logs
update,
because
I
actually
have
been
working
with
some
of
the
stakeholders
in
the
logs
community
and
there
is
an
otep
which
is
being
worked
on
right
now,
which
I
I
am
the
owner
for,
because
I
you
know,
tigran
wanted
an
owner
for
that
effort,
where
we
are
actually
proposing
the
need
to
incorporate
the
elastic
common
schema
into
the
specification
or
at
least
consider.
You
know
how
it
can
be
leveraged
for
complex
use.
Cases
and
elastic
actually
has
been
working
with
us.
D
So
again,
just
to
give
you
an
update,
you
know
so
that
everyone's
aware
of
that
that's
an
overtap
that
is
being
worked
on.
A
D
C
D
Yeah,
it's
possible
john.
I
mean
absolutely
that's
been
discussed
in
the
log
sig,
but
it's
also
good
to
look
at
existing
standards
right.
So
no.
C
A
Yeah,
so
elite
I'll
try
to
help
figure
out
how
to
wrangle
that,
because
I.
D
A
M
A
M
Definitely
like
the
yeah
we
can.
We
should
definitely
talk
we've.
The
group
has
been
working
on
the
data
model.
It
put
no
updates
because
there
was
really
nothing
since
december,
but
yeah
just
to
reiterate
status
for
anyone
who
hasn't
followed
along
basically
we're
trying
to
get
to
feature
freeze
on
the
data
model.
We've
got
two
open
issues
and
we've
been
tracking
those,
basically
whether
to
have
logger
name
and
whether
to
have
a
name
field.
M
There's
been
a
lot
of
discussion
about
these.
I
think
we
have
some
consensus
and
we're
pretty
close,
but
we're
sort
of
waiting
for
people
to
to
confirm
that
they're
they're
happy
with
the
current
proposal
so
but
yeah
definitely
would
would
welcome
feedback
on
how
to
how
to
mainline
our
conversations
as
well.
Yeah
yeah.
A
Because
I
think
what
john's
saying
is
it's
like
there's
there's
like
agreement
and
it's
like
almost
feature
freeze
and
like
everyone's
happy,
but
then
there's
this
like
wider
audience,
and
then
you
know
we.
We
have
this
pattern
that
we're
surely
going
to
go
through
again,
which
is
when
that
whiter
audience
shows
up
and
looks
at
it.
A
They're
going
to
have
possibly
prob
most
definitely
a
rehash
of
some
of
the
same
conversations,
but
then
possibly
also
like
additional
input,
probably
mostly
having
to
do
with
like
how
does
this
stuff
relate
to
the
existing
architecture,
for
example,
things
like
that
so
and
to
make
it
complete
often
when,
as
we
did
with
metrics,
when
we
started
being
like
all
right,
we
got
a
model.
We're
basically
done
we're
gonna
start
implementing
it,
and
we
start
like
shouting
about
that.
A
The
the
rest
of
the
logging
universe
may
thundering
herd
into
the
sig
at
that
point,
which
is
sort
of
like
what
happened
with
matrix.
So,
just
just
a
heads
up
mental
preparation
that
that
that's
probably
gonna
happen
and
there's
as
a
result
of
that
there
will
be
rehashing
yet
again
about
some
of
these
decisions,
and
it's
not
that
you
all
are
making
bad
decisions.
But
you
know
how
it
is
when
a
whole
bunch
of
new
stakeholders
show
up.
N
I
wonder
if
it
might
be
a
good
idea
to
have
like
a
logs,
kickoff
meeting
type
of
thing.
You
know
as
a
maintainer,
that's
not
at
all
involved
in
logs,
I'm
not
sure
where
I
would
even
begin
to
try
to
implement
logs
in
js,
and
I
don't
know
if
others
feel
the
same
way.
I
might
be
able
yeah.
D
K
A
Just
whipping
up
a
like
a
slideshow
or
just
a
quick
presentation:
that's
just
like
open,
telemetry
logs!
Here's
where
we're
at
here's,
what
we're
thinking
about
it,
something
that's
like
a
nice
overview
for
the
maintainers
and
other
people
to
to
ingest.
So
we
don't
have
to
pester
you
y'all.
With
the
the
same
set
of
questions.
I.
D
A
Yeah
cool
cool
great
looking
forward
to
it:
okay,
moving
on
php
working
towards
beta
great.
How
are
things
going
over
there?
Bob.
A
Great
to
hear
also
congratulations
on
moving
over
to
intuit.
I
noticed
that
in
your
bio
I
don't
know.
A
B
A
E
A
Yeah
congrats
on
that
as
well
by
the
way
yeah
good
job
mailchimp
I've
always
been
reading
for
you
guys.
Okay,
java
java
java
metrics,
stable
api
congrats.
C
Yep
and
as
I
said
here,
the
sdk
is
not
stable
because
we
don't
have
a
stable
spec,
so
we
will
probably
trail
at
least
a
month
behind
the
spec
stability
with
the
stable
sdk.
D
C
D
A
That
that's
great
that's
definitely
when
we
put
this
blog
post
out
telling
people
like
go
start
using
metrics
in
java,
like
we
need
beta
testers,
it
sounds
like
we
are
like
in
like
very
stable
beta,
a
point
this.
So
that's.
That's
really
awesome.
I'm
really
happy
to
see
that
yeah.
C
I
mean,
as
I
said,
the
api
is
now
locked
in
stable
and
the
sdk
works.
Fine,
I'm
sure
there's
still
bugs
in
there.
We
don't
have
enough
right.
D
A
Awesome
and
otlp
log
exporter,
so
we
don't
have
a
log,
don't
have
a
log
api
in
air
quotes
so
there.
C
To
log
for
j
or
anything
like
that
in
the
instrument
I'll,
let
the
instrumentation
folks
talk
about
that.
Since
that's
an
instrumentation
project,
not
okay,.
A
Great
well
moving
on
then
to
java
instrumentation
release
plan
for
this
week,
including
first
logging
instrumentation
any
details
there.
H
Yeah,
so
we
do
have
instrumentation
for
log
back
log4j
and
java
util
logging
well
log
back
and
log
for
j.
We
have
both
library
instrumentation,
like
appenders,
that
people
can
manually
hook
up
as
well
as
auto
instrumentation
in
the
java
agent
and
then
java,
util
logging
there's
no
like
manual
appender
instrumentation,
but
there
is
auto
instrumentation
awesome.
Awesome.
A
Great
and
instrumentation
api
stability,
so
this
is
the
the
instrument
or
objects
that
you
all
have
been
developing.
I'm
curious
is
this
something
you
are
making
public
yet
or
you're
planning
on
making
public
aka
use
for
for
the
general
population
to
start
using,
not
just
internally
yeah,
it
is
public.
H
Already
we
do
have
a
couple
of
people
that
we
know
are
using
it
already,
including
spring,
which
is
part
of
the
push
towards
making
it
stable.
A
Yeah,
it
would
be
great
to
get
some
of
this
into
the
the
actual
spec.
I
know
we
haven't
had
much
bandwidth
from
other
cigs
to
to
prototype
this
stuff
there,
but
I
did
just
want
to
throw
that
out,
because
this
is
api.
Surface
area
we're
gonna,
stand
behind,
I'm
not
proposing
that
we
should
go
muck
up
the
stuff.
B
Yeah,
do
you
mean
something
different
than
honorable
tap.
A
If
there's
more
going
on
there
than
what
was
in
that
otp,
I
think
maybe
a
reboot
of
that
otep
would
be
nice.
I
do
think
that
the
hold
up
here
is
actually
bandwidth
from
other
cigs
to
to
try
to
implement
something
similar,
because
the
the
biggest
piece
of
feedback
about
that
otep
is
like
this
looks
cool,
but
it
it's
very
java
ish
in
what's
being
proposed,
and
we
don't
know
like
quite
how
we
would
implement
this
in
in
go
or
php
or
javascript,
or
something
like
that.
A
A
We
do
have
an
instrumentation
sig
that
is
trying
to
sort
out
semantic
convention
stuff,
but
we
would
like
to
sort
some
of
this
part
out,
in
particular
as
part
of
getting
the
maintenance
of
all
the
contrib
instrumentation
contrib
packages
off
the
plates
of
the
core
maintainers,
having
some
amount
of
harness
and
and
utilities
around
how
to
write
that
stuff.
A
We
think
will
make
it
easier
to
to
keep
a
big
pile
of
instrumentation
packages
like
up
to
spec.
So
if
people
do
have
time
would
like
to
join
the
instrumentation
sig
or
at
least
join
the
the
slack
channel
there
and
say
hi.
That
would
be
great
along
with
logging.
I
see
this
is
like,
like
the
last
bit,
a
big
api
surface
area
that
we
need
to
think
about
before
we
can
say
we're
totally
stable.
A
I
do
think
it's
as
important
as
the
rest
of
the
api
work
we've
been
doing,
just
because
maintaining
instrumentation
that
does
all
of
this
stuff
is
just
a
huge
pile
of
work,
as
we
all
know
cool.
Unless
people
have
any
any
comments
or
questions
about
that,
gonna
move
on.
N
Things
are
good
we're
moving
along
with
metrics
we
had
some
may
have
heard
of
the
colors
js
bug
that
occurred
earlier
over
the
weekend.
It's
kind
of
killing
all
rci
at
the
moment.
I
believe
we
actually
just
fixed
it,
though
we're
working
with
the
airline
group
to
create
a
1.0
blog
post.
Since
we
went
to
1.0
just
before
the
new
year
and
our
you
know,
metrics
is
approaching
a
workable
state.
I
know
I
initially
said
by
the
new
year,
which
we
obviously
didn't
hit.
K
N
Yeah
definitely
we
actually
have
an
instrumentation
api
that
we've
been
using
also-
and
I
have
not
personally
joined
the
instrumentation
group.
So
I
don't
know
how
far
off
ours
is
from
that
yeah,
but
we
have
like
40
or
50
instrumentations,
so
it
would
be
probably
pretty
difficult
to
migrate
them
all
over
to
a
new
api.
A
From
looking
at
it,
I,
the
details
are
very
fuzzy
in
my
head
because
it's
a
while
ago,
but
I
think
actually
what
you
guys
call
an
instrumentation
api
is
a
bit
different
than
what
they're
doing
in
java
like
like
attacking
a
different
part
of
the
problem.
If
that
makes
sense,
oh.
N
I
understand
okay,
what
time
is
the
instrumentation
meeting.
A
N
O
A
Great
okay,
moving
on
python,
welcome
back
metrix
sd8
sdk
is
a
work
in
progress.
Is
the
api
metrics
api
stable
in
python.
J
Yes,
it
has
been
merged
into
main
already.
It's
still
something
that
we
don't
expose
publicly,
but
it
is
pretty
much
there
and
pretty
pretty
stableness.
J
No
not
yet
we
pretty
much
the
next
big
step
that
we
have
now
is
aggregations.
We
have
kind
of
a
series
of
peers
that
pretty
much
include
a
full
implementation,
but
we
are
still
working
out
details
after
aggregations.
It
should
come
views
and
well
it
should
be
more
more
or
less
usable.
A
Sweet
yeah,
okay,
dot
net
welcome
back
dot
bet
are
there.net
people
in
the
call.
I'm
curious
about
metrics
in.net,
I'm
here
yeah.
What's
the
status
of
metrics
and
dot
net.
G
A
G
We
we
have
everything,
except
for
the
full
implementation
of
exemplar.
A
G
A
H
A
Okay,
great
c,
plus
plus
sounds
like
metrics
is,
is
well
underway
in
c
plus,
plus.
Likewise,
is
this
something
people
can
start
beta
testing?
Is
it?
Is
it
that
far
along.
O
No,
no
so
api
is
there
and
it's
stable.
The
sdk
implementation
is
going
on.
As
mentioned
the
view
implementation
inverse
aggregation
work
is
going
on.
So
it's
a
steady
progress.
We
plan
to
have
an
export
pipeline
for
sync
matrix
by
this
month.
So
that's
that's.
That's
the
plan.
As
of
now.
Okay,.
A
O
A
Okay,
yeah,
so
so
shout
out
to
to
everyone
on
the
call
c,
plus
plus,
is
really
important.
I
feel
like
the
orgs
people
work
for
should
have
c
plus
plus
programmers
sitting
around
somewhere,
and
it
would
be
great
to
get
more
people
getting
their
hands
on
this,
since
it's
going
to
be
such
a
critical,
critical
component,
specifically
when
it
comes
to
integrating
open
telemetry
with
a
lot
of
big
services,
this
is
going
to
be
a
critical
sig.
So
it'd
be
great.
If
we
could
get
more
help
here,.
D
New
release
bargain
did
you
want
to
give
a
quick
update,
or
I
can
at
least
on
what
we
have
completed
for
targeting
some
metric
stability
announcement.
D
D
The
other
aspect
that
has
been
completed
is
the
prometheus
compliance
for
all
the
remote
right
pipeline,
as
well
as
the
and
and
the
pull
exporter
is
also
functioning,
but
we
need
to
make
sure
that
it
is
meeting
all
the
implementations.
Now.
The
question,
though,
that
we
went
around
and
asked
folks
is:
do
we
really
need
a
pull
exporter
in
the
collector
and
that,
I
think,
is
only
for
what
I
found
the
use
cases
to
be
is
only
for
debugging
or
just
testing
collections.
D
D
So
that
said,
those
are
the
two
areas
that
are,
you
know
pretty
stable
and
and
can
be
called
out
as
stable.
The
other
area
that
is
in
progress
is
configuration
and
configuration
components,
but
bogdan.
I
think
that
the
idea
was
also
to
identify
specific
components
that
can
be
designated
stable
in
a
rolled
phase,
so
I
think
that
work
is
in
progress.
K
D
Think
that,
right
now,
the
use
cases,
even
you
know,
are
pretty
much
debug
debug
and
testing.
It's
the
prw
exporter
that
is
actually
being.
A
A
Great,
I'm
glad
to
hear
that.
That's
that
so
basically
the
remote
right,
you
would
say,
is
in
production,
ready
beta
yeah.
D
K
Yes,
I
think
that
the
full
prometheus
pull
to
remote
right
exporter
pipeline
is
working
relatively
stably.
There
are
there's
a
I
think
today
or
this
weekend
there
was
another
new
bug
that
was
reported
regarding
how
it
deals
with
some
metrics
families,
but
I
think
we're
at
the
point
where
we're
picking
around
the
edges
of
of
defects
and
then
not
significant
structural
problems,
so
we
should
be
good
there.
K
A
F
A
Thanks
erlang
1.0
congrats,
great
congrats
to
erlang;
that's
awesome,
good
job,
ruby
and
js
yeah.
That's
great
to
have
another
batch
of
1.0
excited
to
announce
that
okay,
moving
on
okay,
we're
moving
on
past
paso,
reportbacks
15
minutes
left
in
the
meeting
otlp
mvar
renames,
who
added
this
item.
F
I
have
that
one,
it's
just
for
your
information
if
you
yeah.
Thank
you
so
much
for
that.
Basically,
tristan
opened
this
pr.
There
are
a
pair
of
basically
incorrect,
all
environment
names
and
we
want
to
update
them.
We
will
keep
the
previous
ones
as
deprecated,
so
we
don't
break
people
but
yeah
that's
planned.
This
is
just
since
they
will
be
packing
on
the
sigs.
I
just
wanted
to
mention
that
out.
We
will
merge
it.
It
has
no
approvals.
A
And
okay,
so
with
the
last
bit
of
the
meeting
diego,
would
you
like
to
to
go
over
your
proposal.
I
J
Right,
okay,
so
what
I
have
here
is
a
pr
that
is
closed
already,
because
it
went
still
it
pretty
much
tries
to
make
open
to
level
three
specification
compliant
with
this
document.
So
the
w3c
has
this
document.
It
is
pretty
much
a
specification
for
specifications.
J
J
J
J
A
non-normative
section
is
pretty
much
everything
else.
So
every
statement
like
that
includes
an
rfc
keyword,
it's
a
normative
section
and
what
this
pr
tries
to
do
is
it
tries
to
add
normative
sections
into
our
specification
and
tries
to
make
these
sections
individually
identifiable,
and
I
think
it's
easier
to
see
this
in
an
example.
So
here
is
oh
sorry,
the
pr
that.
J
J
The
sections
are
identifiable
because
they
start
with
the
word
requirement.
They
have
a
unique
name
here
and
they
include
one
and
only
one
rfc
keywords.
The
advantage
of
this
is
that
it
makes
it
possible
to
refer
to
every
section
with
an
url
so
that
it's
easy
to
communicate
between
specification
developers.
Okay,
I'm
referring
to
this
part
of
the
section,
so
this
part
of
the
specification-
and
they
no
longer
have
to
try
to
find
out
the
the
rfc
keyword
in
the
middle
of
a
big
paragraph,
or
something
like
that.
J
Okay,
this
is
related
to
the
compliance
matrix,
which
is
what
we
were
discussing
so
right
now.
The
compliance
matrix
is
pretty
much
this
it.
J
The
problem
that
I
find
with
these
compliance
matrix
is
that
it's
very
unspecific.
It's
hard
to
know
what
does
this?
What
does
this
mean?
What
does
this
mean?
For
example,
we
have
a
feature
here,
get
a
tracer
and
according
to
this
table,
that
feature
is
supported
in
in
airline,
but
a
feature
can
be
composed
of
several
smaller
components.
J
So
it's
hard
to
know
if
exactly
every
one
of
those
components
is
implemented
and
some
of
those
components
may
be
optional,
so
this
column
also
doesn't
quite
accurately
describe
what
how
optional
this
this
feature
really
is.
So
when
I
was
when
I
started
implementing
metrics,
I
realized
that
I
pretty
much
ended
up
reading
the
the
metrics
specification
looking
for
every
keyword
and
then
trying
to
implement
by
following
every
requirement
that
there
was
so
in
in
that
way,
it's
very
straightforward.
J
That
is
hard
to
do
right
now,
because
the
rfc
keywords
are
inside
of
just
one
big
section,
so
you
need
to
read
all
the
specifications
to
find
all
of
them
and
we
also
have
not
been
completely
regards
with
this.
So
they
are
statements
that
are
normally
by
themselves,
but
they
are
not
written
down
in
using
a
normative
language.
J
So,
for
example,
you
may
find
a
statement
like
this
thing
is
mandatory
and
that
is
equivalent
to
using
must,
but
when,
if
you
try
to
extract
all
the
requirements,
you
you
can't
just
search
for
all
the
occurrences
of
the
word
must
because
there
is
some
other
mandatory
statement
there,
that
is
not
using
this
language.
Okay.
J
So
the
idea
that
I
have
is
that,
instead
of
having
this
matrix,
what
we
should
have
is
a
list
of
all
the
requirements
as
defined
they
are
here
and
in
this
way
we
can
be
much
more
specific.
J
We
can
pretty
much
say:
okay,
python
implements
this
particular
requirement,
and
we
can
also
say:
is
this
requirement
mandatory
or
not?
And
based
on
that,
we
can
pretty
much
say
how
compliant
with
the
specification
and
implementation
is.
You
can
say,
for
example,
it
implements
82
percent
of
the
the
must
requirements
and.
L
Wow,
that's
that's
great,
and
can
I
just
add
add
on
to
the
end
there
that
that
you
did
show
us
at
least
one
change
to
the
spec
that,
where
you
translated
from
current
state,
as
you
described,
to
new
state,
as
well
as
a
generator
program
to
convert
that
standardized
text
from
the
specification
into
the
compliance
matrix,
so
that
we
can
just
have
an
automated
test
framework
for
the
specification
and
the
compliance
matrix
could
be
generated
from
ci
pipeline.
For
example.
J
J
That
shows
how
we
can
extract
the
requirements
into
a
json
file
that
is
easy
to
parse.
You
can't
find
it.
Oh,
maybe
it's
this
one.
No
all
right!
Just
trust
me,
it
should
be,
should
be
here,
but
but
it's
easy
to
imagine
just
json
file
that
pretty
much
has
all
the
all
the
requirements
listed
there,
so
they
can
be
parson.
From
from
that,
you
can
generate
the
the
compliance
matrix
automatically.
A
Awesome,
I
would
love
to
hear
feedback
from
from
maintainers
thoughts
on
switching
this
model,
but
the
one
thing
I'd
like
to
just
say
diego
is
besides
thank
you
for
putting
in
all
this
work.
I
do
really
appreciate
that
the
the
direction
this
work
is
coming
from
is
coming
from
spec,
practically
making
the
spec
easier
and
more
straightforward
for
maintainers
to
be
able
to
implement
and
to
have
agreement
with
each
other
about
what
the
precise
requirements
are
correct.
J
Yes,
I
think-
and
I'm
going
to
make
a
very
old
statement
here-
that
the
people
who
are
mainly
writing
respect
that
are
right
now
very
expert
in
their
in
their
area.
So
they
write
a
specification
and
they
have
such
a
great
knowledge
that
it's,
I
think,
easy
for
them
to
grab
the
the
whole
meaning
and
understand
everything
that
that
is
written
down
there.
J
But
this
specification
will
be
implemented
in
other
languages,
also
by
people
who
may
not
have
that
expert
knowledge.
So
it
is
more
important
for
for
them
to
have
things
as
clear
as
possible,
so
that
implementations
can
can
go
on,
because
if,
if
the
people
who
are
implementing
this
specification,
don't
have
that
universal
knowledge,
it's
they're
gonna
rely
much
more
on
on
what
is
exactly
written
in
that
document.
So
it's
it's
very.
H
Great
so
as
a
as
a
maintainer,
I'm
a
I've
been
a
supporter
of
this
pr.
I
think
for
a
while,
and
I
like
the
changes
included
here.
It
helps
as
a
maintainer
understand
it
and
exactly
as
diego
just
said,
like
you
have
people
who
are
experts
in
a
particular
topic
or
their
field
or
their
language,
and
it
makes
it
a
little
bit
more
digestible
and
it
makes
you
write
the
specification
with
your
audience
in
mind.
So
I
really
prefer
it.
H
You
can
read
through
the
pr,
though,
there's
definitely
resistance
from
active
specification
authors,
because
it's
going
to
impose
a
different
stylistic
approach
to
writing
things
which
I
think
is
appropriate
because
of
all
the
positive
things
we
just
said.
But
I
think
that
is
going
to
be.
H
The
uphill
here
is
when
you're
going
to
have
to
revise
a
specification
which
is
going
to
include
changes
which
is
always
going
to
include
bugs,
which
I
think
is
going
to
be
a
concern,
and
it's
going
to
put
a
restriction
on
new
additions
to
the
specifications
to
conform
to
this,
which
I
I
think
is
all
worth
it
personally.
But
it's
it's
just
a
subjective
standpoint
on
this,
and
so
I
just
want
to
say
I
support
this.
I
do
have
to
drop
off,
but
thanks
again,
diego
for
working
through
this.
A
Yeah,
if
everyone
could
lead
further
feedback
as
comments
on
this
on
this
pr,
that
that
would
be
great
and
continue
the
discussion
in
slack.
It
would
be
good
to
understand
if
people
do
have
concerns
about
switching
to
this
style,
like
what
the
specific
concerns
are
so
thanks,
but
we're
out
of
time
on
this
call
happy
monday.
Everyone.