►
From YouTube: 2021-06-17 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
Okay,
so
I
will
so
do
you
mean
in
the
in
the
notes
right,
yeah
yeah?
I
think
it's
because
yeah
we
had
the
someone
changed
it.
We
had
the
correct
one,
but
python
had
the
incorrect
and
you
had
changed
it
right.
Yeah,
yeah,.
D
C
C
B
A
Yeah,
okay,
so
I
think
we
are
all
already
and
we
can
start
with
with
it.
So
let
me.
A
A
The
notes-
okay,
so
yeah,
review
last
meeting
actions,
something
about
metric
fixes
that
I
I
have
been
doing
mainly
today,
and
some
changes
and
the
status
of
metrics
in
the
project,
because
I
think
I
have
been
reviewing
it
because
of
fixing
some
issues,
and
I
have
seen
it
it's
quite.
A
lot
of
anonymous
hamster
is
okay,
great.
A
Oh
yeah,
that's
right
great,
so
last
meeting
actions,
sweet
packages
to
registry-
I
created
up
here
and
still
not
merged,
but
I
think
I've
approved.
A
So
waiting
for
for
martin.
A
A
This
day,
maybe
tomorrow,
so
we
have
that
also
done
mobile
carrier
info.
You
wanted
to
track
it
here.
D
Yeah,
it's
it's
working
its
way
through
here.
Let
me
add
the
link
to
the
pr
here.
There's
been
a
lot
of
conversation
about
it,
but
I
think
it
we've
we've
iterated
on
it
and
I
think
we've
got
it
in
a
pretty
solid
spot
and
it
just
needs
to
settle.
D
Where
did
that
there?
No,
that's
not
it
here.
It
is
oh
that
sounded
either.
I've
lost
the
meeting
note
stock
there.
It
is.
B
D
There's
actually
a
bug
in
the
table
generating
tool
that
they
use
to
like
validate
the
validate
the
the
yaml
spec
and
it's
adding
an
extra
space
in
and
so
now
the
the
markdown
linter
is
mad
at
it.
That
there's
too
many
spaces.
So.
C
D
A
Okay
for
the
new
libraries
that
we
could
instrument,
bryce
created
the
issues
yeah.
D
They're
pretty
they're
pretty
skeleton
issues,
but
at
least
there's
something
to
track.
A
Yeah
yeah,
I
think
it's
it's
enough
until
someone
really
wants
to
take
on
it
right,
cool,
so
yeah.
I
think
it's
good
enough
and
we
were
talking
about
new
things
coming
to
swift,
the
asking
things
that
maybe
we
should
check
on
on
the
network.
Instrumentation.
A
A
We
can
also
maybe
create
a
an
issue.
So
do
you
think
you
will
have
time
this
week
to
do
that
or
we
prefer
to
have
an
open
issue,
so
anyone
can
take
it.
D
A
A
C
B
A
Okay,
the
maintenance
release
it
was
released
as
expected
and
nothing
got
broken,
nothing
new,
broken
and
also
this
one
point
to
change
some
classes
from
public
to
open
so
yeah.
I
I
have
I
have.
I
have
been
checking
other
projects
and
those
classes
that
are
that
are
in
the
sdk
are
not
usually
inevitable,
so
they
are
final
classes
and
things
like
that.
A
D
A
The
idea
is,
it
was
going
to
fix
the
problem
in
the
open,
telemetry
protocol
exporter,
but
it's
a
problem
that
could
happen
with
any
exporter,
because
if
you
use
the
batch
spam
processor,
then
you
never
know
when
it's
going
to
export
it.
So
I
created
a
a
version
of
the
constructor
just
added
another
parameter,
a
callback
parameter
to
the
constructor.
So
it's
called
it
when,
when
it's
going
to
export
with
all
the
data-
and
you
can
modify
it
if
needed,
so
I
I
think
it's
a
bit
more
flexible
and
should
fix
your
issue
so.
E
A
Okay,
so
those
are
the
last
meeting
actions
for
today,
yeah,
for
today,
I've
been,
I
started
yesterday,
mainly
working
on
metrics,
because
we
had
a
pair
of
issues
that
james
sherlock
has
open
final
one
of
them.
A
A
Okay,
sorry,
so
the
thing
is
that
there
were
a
pair
of
issues.
One
of
them
was
that
it
was
not
exporting
correctly
the
opportunity
protocol
to
promise
use
because
of
some
values
that
had
to
be
set
on
on
the
open
telemetry
protocol
and
and
that's
because
the
the
open
telemetry
protocol
spec
is
but
much
newer.
It's
based
on
the
new
aspect
of
the
metric
and
we
don't
have
the
spec
done.
A
We
have
something
previous
that
doesn't
match
perfectly
with
that.
So
james
said
that
with
a
small
change,
had
it
working
on
on
his
side.
A
We
should
have
the
new
aspect
for
solving
this,
because
it
can
have
different
different
usages
depending
on
on
the
type
so
yeah.
So
that's
something
that
we
will
have
to
fix,
but
until
then
I
think
it
at
least
solves
our
one
of
the
user's
problem.
For
the
rest,
I
think
it's
an
alpha
and
then
they
will
have
to
wait
until
we
have
something
better.
A
If
it
it's
not
hurting
anyone.
Currently,
at
least
we
don't
know
of
anyone
being
part
by
that.
So
for
the
other
thing,
the
other
fix
was
about
duplicated
metrics
that
were
happening
with
the
batch
span
with
the
sort
of
the
metric
exported
sdk
class.
A
It
was
because
when
it
was
exporting
the
matrix,
it
was
not
cleaning
the
previous
one,
so
I
have
fixed
that
and
created
up
here
and
that
that
should
be
solved.
So
the
point
is
that
it
was
working
in
another.
In
the
there
was
another
metric
exporter
in
the
break
that
now
have
reverted.
That
was
much
I
mean
it
was
like
the
sdk
exporter,
but
it
was
not
protected
against
concurrent
access.
A
So
the
thing
is
that
when
it
processes
all
the
metrics,
it
must
just
export
them
but
clean
the
the
ones
that
has
been
wet
and
that
was
not
being
done
on
the
sdk
one.
So
I
fixed
that
and
with
that
I
have
talked
to
james,
and
he
told
me
that
it
was
having
the
expected
behavior.
At
least
initially
with
that
change,
so
yeah,
so
that
should
be,
should
be
good
for
for
the
future.
A
D
A
Keep
an
eye
out
for
it,
though
yeah
I
mean
it
was
storing
all
the
metrics
that
it
read
it
so,
each
time
the
processor
ran
it
cutted
out
it
had
one
set
more
so
yeah.
The
change
is
very
it's
in
a
pr
also.
I
can
try
to.
A
A
Me
yeah
yes
here
when,
in
the
in
the
finish
collections
cycle,
it
had
to.
D
A
The
the
current
because
it's
time
is
processed
one
metric
is
appended
here
so
after
that
just
send
all
the
metrics
in
this
method,
so
can
be
read
the
before
this
before
this
exporter
was
exceeded
this
one
that
had
exactly
the
same
code
as
the
locks,
but
it
was
being
cleaned
after
returning-
and
this
was
this
is
the
one
that
was
running
on
the
in
the
sample
code.
So
it
took
me
it
was
working
for
me
until
I
I
didn't
know
why
it
was
firing
off
this
side.
A
So
until
I
he
told
me
that
he
was
using
this
one
and
then
this
just
this
line
was
missing.
So
I
I
think
you
should
be
also
receiving
many
metrics
it's
time
to
export
from
this.
D
Yeah
interesting
yeah,
I
wonder
I
I
wasn't
seeing
anything
weird
and
maybe
it's
because
I'm
using
a
gauge-
and
it
only
generates
one
metric
per
time-
I'm
not
sure.
E
C
E
D
D
A
Yeah
yeah
it
it
should
keep
give
you
one
more
it's
time
so
and
it
looks
like
promises
was
arguing
about
that,
and
it
was
this
so
yeah
that
from
my
side,
this
this
is
still
to
be,
I
think,
approved.
So
you
can
check
it
works
for
you
bryce
before
before
approving.
If
you
want
yeah.
A
Great
and
for
the
rest,
I
have
no
more
points
to
talk
today.
I
don't
know
if
you
have
any
other
topic,
we
not
see.
Sorry
sorry
yeah,
I
I
know
vinod
was
going
to
fix
the
or
was
planning
to
fix.
B
A
No,
I
think
it's
not
working
for
me.
It's
not
working.
A
But,
but
I
I
I
know
it
worked
in
the
past,
so
maybe
just
maybe
just
going
with
git
and
testing
it
in
previous
version
until
you
get
the
offending
commit
and
if
you
want
help,
I
mean
just
tell
me
finding
the
comments
should
not
be
very
difficult.
A
Okay,
and
just
let
me
know
if
yeah
I
I
without.
B
So
I
was
also
looking
at
other
projects.
You
know
when
I
was
looking
at
getting
started.
You
know,
especially
like
the
javascript
one,
because
you
know
it's
got
a
the
the
web
instrumentation
and
the
back
end
right.
I
mean
that
is
what
I
was
looking
at
for
for
swift
as
well,
and
they
all
seem
to
be
using
the
otlp
or
you
know
the
collector
right,
yeah.
A
Okay,
as
you
I
mean,
for
example,
you
know
I
I
have
been
as
I
was
trying
to
find
the
issue
with
metrics.
I
his
issue
was
with
with
the
collector.
I
I
checked
that
at
the
collector
project
they
have
a
docker
compose
that
has
the
collector
running
with
jager,
prometheus
and
shipkin.
I
think,
and
they
have
that
set.
A
So
yes
with
a
line
and
docker
you
get
that
running.
So
maybe
you
can
also
use
that
if
you
want.
B
A
B
In
our
examples
in
the
swift
repo,
there
is
no
example
for
the
tlp.
No,
you
know
with
enter.
A
A
So
yeah
it
should
work.
It
should
work.
Okay,
all
right,
so
I
think
that's
like
so
maybe
maybe
you
can
focus
your
your
tutorial
more
on
the
on
creating
a
sample
project
with
that.
D
I
think
there's
something:
okay
go
ahead,
no
show!
No!
No!
Yes!
Oh
I
was
gonna
say
I.
I
think
that
there's
something
to
be
said
about
air
states
in
the
in
the
sdk,
because
it
is
it's
hard
to
tell
when
it's
connected
properly
and
when
it's
not.
D
Yeah
with
the
collector
yeah,
because
it'll
it'll,
if
you
let's
say
like
you're,
trying
to
run
it
locally
and
you
don't
have
it
exposed
properly
or
something
along
those
lines
when
you're
running
the
sdk
and
it
can't
talk
to
the
collector,
it
just
will
be
silent,
it
won't
say
anything,
and
so
maybe
we
should.
We
should
add
some
better,
better
logging
to
that
or
something.
A
A
Yeah
the
trying
to
fix
that
I
have
the
first
thing
I
I
tried
to
set
something
similar
to
what
james
was
using.
I
set
the
docker
image
for
the
collector,
but
I
was
not
able
to
connect
the
tlp
exporter
to
the
collector
locally
yeah.
C
A
Know
how
to
create
the
channel
that
is
needed
in
the
collector
in
order
to
to
work
properly,
but
I
didn't
spend
much
time
on.
You
know
that
neo
grpc
things
I
didn't
spend
much
time
there
when
I
have
seen
that
it
was
taking
too
long.
I
try
to
other
focus
to
to
fix
the
issues
but
yeah.
For
me
it
was
not
easy
to
do
that
so
yeah.
C
A
Yeah
totally,
I
I
really
tried
to
create
the
year
pc
channel,
but
somehow
it
never
worked.
We
always
always
throw
errors
inside
swift.
So
with
the
bugger,
I
could
check
that
it
was
finding
nervous.
It
was
written
in
a
57
error,
so
it
couldn't
find
the
port.
I
don't
know
why
yeah
and
yeah-
and
I
have
think
that
maybe
we
could
in
the
open,
telemetry
exporter,
we
could
somehow
make
it
easier
for
the
user
instead
of
making
them
use.
A
The
channel
just
provide
the
channel
ourselves
default
channel
with
the
proper
configuration
done,
because
the
port
is
known
by
default
and
also
we
could
use
localhost
as
the
address
or
use
something
like
that
by
default,
so
it
can
be
used
yeah.
That
would
be
great
for
for
the
user,
because
I
I
was
not
able
to
create
a.
D
I'll
I'll
make
a
I'll
make
a
ticket
to
improve
vlogging
and
I'll
make
one
to
improve,
generating
a
default.
Grpc
channel.
A
A
We
should
also
expose
the
channel
to
for
for
adding
for
for
adding
security-
and
you
know
some
header
stuff
and
that
we
must
sp
still
keep
the
channel
public.
Somehow
right.
B
Okay,
another
question
I
had.
I
was
looking
at
the
directory
open
elementary
protocol
under
the
proto.
You
have
all
those
you
know
the
swift
equivalent
of
the
proto
buff
right.
A
B
A
A
Files
yeah
the
the
the
original
files
of
are
in
the
or
the
sdk
okay,
I
and
industry
in
the
spec
yeah
we
are
not.
We
are
the
only
copying
here
the
proto
generated,
because
I
get
it.
I
see
it
yeah.
I
think
I.
B
A
A
It's
not
an
easy
to
use
interface,
the
proto
staff
and
your
pc
on
switch
is.
B
B
So
is
it,
does
it
run
when
you
do
a
build,
does
it
use
a
generated
or
you
just
run
it
and
you
come
copied
over
the
no.
A
No,
no,
we
we
are
just
we
are
fixing
the
the
files
we
don't
want
them
to
be
updated
automatically.
Just
for
yeah,
we
could
have
a
yeah.
We
could
have
another
way,
but
currently
they
are
just
copied
the
final
versions,
just
in
case
we
need
to
fix
something
or
instead
of
depending
on
the
erpc
swift,
because
it's
it's
not
an
easy
configuration
to
generate
the
switch
files.
A
So
so
that's
why
we
are
keeping
on
only
the
street
files.
So
if
there
is
a
change
in
the
code
above,
how
does.
A
A
Ourselves:
okay,
so
we
can
be
sure
that
everything
works
with
the
spec.
We
have
that's
the
main
reason
because
automatically
having
that
we
could,
we
could
end
up
with
things
that
are
not
working,
yeah
and
breaking
change.
That
will
that
could
hurt
us
with
releases
and
things
like
that.
A
We
could
also
keep
a
copy
of
the
proto
files
and
generate
the
swift
files
each
time
yeah,
but
but
you
need
extra
libraries
in
your
system
to
generate
the
swift
files
from
prototype.
That's
the
reason
that
we
only
keep
the
sweep
files
because
it's
not
easy
to
configure
your
system
to
generate
them.
A
Yeah,
I
think
it's
easier
for
anyone
to
just
take
the
credit
and
build
everything
and
we
can
keep
the
version
in
just
in
our
git
folder.
That
was
the
main
reason.
B
B
Yeah,
let
me
let
me
also
you
know,
look
at
trying
to
use
this
open,
elevator
protocol
exported
as
well,
and
give
that
as
an
exercise
to
me
myself.
Okay,
great.
A
Nice
great
thanks
for
the
issues
created
yeah
any
other
topic.
D
Oh,
I
had
one
I
opened
that
tissue
tissue
that
issue
about
what
was
it
allowing
or
allowing
the
grpc
headers
to
be
set
through
the
metadata.
It
looks
like
there's
actually
a
spec
for
creating
for
for
passing
those
headers
in
via
a
environmental
variable.
I
I
made
a
note
on
the
on
the
ticket
that
I
opened
on
the
issue,
and
I
don't
I
just
wanted
to.
I
wanted
to
verify
that
that
something
like
that
isn't
actually
implemented
in
the
sdk.
D
D
I
I
tried
to
find
it
whichever
way
I
could,
but
here's
I
just
posted
a
spec
in
channel
where
you
can
yeah
pass
in
the
headers
that
you
want
to
set
in
environmental
variables.
I
think
that
might
be
a
better,
a
better
solution,
so
I
think
I'll
I'll
take
that
ticket
implement
it
through
through
this
mean,
rather
than
exposing
the
options.
Some
other
way.
A
E
D
It
is
so
what
I
want
to
use
it
for
is
setting
authentication,
so
our
our
host,
our
collector
host
at
the
moment,
needs
authentication
and
there's
no
easy
way
to
set
an
authentication
header
for
the
the
grpc
channel.
That's
getting
opened
so
exposing
this
would
be
would
be
useful
in
that
regard
and
I'm
sure
that
other
people
will
probably
want
a
you
know,
authenticated
endpoint
for
them
for
their
mobile
device.
B
D
Yeah
yeah,
they
have
there's
yeah
there's
certificate
config,
but
this
is
just
like
basic
basic
off
header.
A
Okay,
so
we
can
close
here,
yeah,
okay,
so
you
guys
have
a
nice
week.