►
From YouTube: 2021-01-20 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
B
C
A
Other
side
right,
never
mind,
I'm
in
I'm
in
oregon,
I'm
in
portland
oregon.
C
A
C
A
Well,
you'll,
if
you
can
see
the
poster
on
my
wall,
which
is
vancouver
bc
yeah,
it
depends
on
what
part
of
oregon
you're
talking
about.
I
think
oregon
is
very
bipolar.
Anything.
A
D
A
D
B
A
D
B
A
A
C
A
Yeah,
my
we,
my
wife
and
I
and
my
family
usually
go
to
mexico
every
spring,
very
disappointing
that
we
couldn't
go
last
year
and
we
can't
go
this
year.
A
All
right
now,
it's
9
35.,
not
much
on
the
agenda.
I've
switched
over
right
now
from
the
ga
tracking
project
update
to
the
1.0
milestone
update.
So
I
created
a
one
point
milestone
and
it
has
a
few
things
in
it.
I
removed
all
of
the
open,
tracing
and
open
sensor
stuff
from
it,
because
we've
tagged
those
as
alpha
and
I
don't
think
the
requirements
are
really
done
in
the
spec
and
we
would
like
to
get
to
1.0
as
quickly
as
we
can
at
this
point.
A
A
Yes,
there's
not
much
left,
I
am
currently
even
as
we
speak,
I've
been
working
on
the
fallback
service
name
for
the
exporters,
so
that's
going
to
be
done.
Well,
hopefully
done
very
soon
and
the
same
thing
with
this
fallback
there's
a
pr
for
this
one.
This
is
just
waiting
for
the
spec
to
get
merged
to
merge
that
nikita
we're
looking
for
feedback
on
the
configure
on
your
issue
here
on
the
configuration
issue,
see
whether
you
think
things
are
in
a
good
state
for
1.0.
B
A
A
Yeah
I
mean
the
current
state
of
things
is
the
sdk
by
itself
is
purely
and
fully
programmatically.
Configurable
has
nothing
else,
just
programmatic
configuration
and
then
the
auto
configuration
module,
provides
all
of
the
environment
variables
and
sbis
and
etc.
In
order
to
do
auto
configuration.
A
So
the
only
so
there's
this
putting
synth
tags
on
everything.
This
is
actually
should
be
api
and
sdk,
since
we
don't
have
any
synth
tags
anywhere
at
the
moment,
I'll
go.
A
And
yeah,
so
someone
will
need
to
pick
that
up
it's
a
bit
of
grunt
work.
I
don't
know
if
it's
automatable
or
not,
and
then
let
this
last
item
which
is
currently
assigned
to
josh,
I
did
see
we
have
a
benchmark.
A
A
C
My
last
week
was
on
the
open
senses
stuff
which
got
bunted
so
yeah.
That
would
be
this
this
week.
I
don't
know
if
I'll
be
able
to
finish,
but
if
you
punt,
that's
fine,
I
I
really
do
would
well.
Okay,
two
questions
related
to
that.
I
was
reading
through
the
spec
to
make
sure
that
we
actually
built
the
right
thing.
C
Do
you
care
where
this
runs
at
all
I
mean
for
v1.
I
think
it's
just
going
to
be.
You
run
it
on
your
machine.
C
Okay,
the
next
question
is:
do
you
care
about
requirements
for
how
you
run
it,
because
we
need
to
run
a
service
that
it
exports
to,
and
so
I
could
try
to
keep
that
in
in
jvm
if,
if
that's
desired,.
A
C
All
which
I
think
the
specification
calls
for
having
a
service
that
takes
a
request
and
just
pulls
in
the
bytes
and
asks
them
that's.
It
does
nothing
else,
and
that's
actually
what
I
was
spending
my
time
on
last
friday
on
this.
But
if
that's
not
a
requirement,
I
don't
need
to
do
it.
It's
just.
That
is
what
the
spec
says.
So
I
was
going
to
try
to
make
that
happen.
A
A
E
Yeah,
maybe
I
just
didn't
find
it.
So
that's
sorry!
No!
No!
No!
Don't
just
no
worries.
I
think
it's
called.
E
B
A
Yeah,
what
we're
doing
in
the
benchmark
right
now
is
we're
just
using
the
docker
image,
the
actual
docker
image
from
the
collector
and
using
it
for
real.
A
Shouldn't
it
shouldn't
matter
for
benchmarking,
except
it's
all
running
in
the
same
hardware,
so
it
all
it
could
matter
of
course
yeah.
So
I
think
what
I
was
looking
at
for
this
story
here
was
less
than
that,
because
I
think
that's
in
place-
and
this
is
more
like
micro
benchmarks
on
the
batch
strain
processor
itself.
A
A
C
I'm
planning
to
dive
into
it
thursday,
so
I
can
just
ping
you
in
chat.
If
I
don't
find
anything
useful
in
it
perfect
that
sounds
great.
Okay,.
A
Sounds
perfect,
so
I
think
we
are
on
target
then
to
so
we
released
last.
What
was
wednesday
really
we
released
on
monday
whenever
we
released
last
week.
I
lost
track
of
time,
and
I
think
we
would
like
to
release
next
week
as
our
kind
of
1.0
release
candidate,
so
we'll
probably
still
call
it
0.15,
but
we
will
make
an
announcement
that
it
will
be
our
1.0
release.
A
Candidate,
we'll
still
have
a
few
things
that
we
need
to
delete,
deprecated
items
that
we'll
need
to
delete
before
we
hit
1.0
for
real,
but
I
think
releasing
next
week
as
a
release.
Candidate
is
the
current
plan.
A
D
D
A
When
we
cut
oh
15,
the
snapshots
will
be
out
of
we'll
be
oh,
but
then
we
will,
when
we
do
the
release,
assuming
that
0.0016
whatever
that
would
be,
is
our
1.0
we'll
just
bump
it
to
1.0
as
a
part
of
the
release
process.
I
think
that's
probably
fine.
A
A
C
Okay,
that
that's
another
topic
of
discussion.
I
I
just
released
open
sentences
and,
oh
my
god,
yeah
they're,
they're
they're
things
that
anyway,
releasing
can
be
completely
automatic
if
we
want
to
do
it
based
on
the
tag
so.
C
A
You
can
do
rc,
you
can
like
we're
using
the
we're
using
the
the
name
of
the
release
plugin,
but
you
could
do
our
c's
absolutely.
The
issue
is
the
alpha
versions
are
going
to
end
up
looking
very
strange
because
they're
going
to
look
like
1.0
rc
alpha
and
then.
C
A
C
Has
anyone
evaluated
how
maven
does
resolution
with
that
crap
it
works
because
it
works.
Does
it
oh
yeah?
Okay,
what
works
fine,
what
happens
with
eviction?
Sorry,
I
I
am
that
used
to
be
my
life.
I
was
a
build
tool
guy
I
worked
with
sona
type
worked
with
jfrog
all
that
fun.
It
does
weird
things
right
based
on
based
on
the
dash
with
eviction.
So
if
you're
using
the
bomb,
everything
should
be
gravy.
C
What
do
you
mean?
What
do
you
mean
by
eviction,
so
it
it's
possible
that
the
dash
alpha
could
override
an
actual
released
version
in
some
weird
scenarios
like
if
you're
doing
a
plot
like
if
you're
playing.
C
If
that
makes
sense,
so
there's
all
sorts
of
weird
things
where
it'll
say:
oh,
these
versions
are
compatible.
I
can
replace
this
and
then
just
do
something
crappy
we
can
it's
not
worth
changing
anything
you've
done.
Yet
it's
more
worth
I
like.
If
we
need
to
write
some
documentation
on
this.
The
the
too
long
didn't
read
is
make
sure
anyone
using
open
telemetry
defines
a
dependency
in
their
application
for
every
artifact
and
that
will
resolve
all
issues
so.
A
C
C
You
want
to
pull
in
the
major
version
right
or
another
library
pulls
in
the
major
version.
You
could
accidentally
get
the
alpha.
Instead,
when
the
alpha
goes
1.0
now
the
question,
I
guess,
if
alpha
and
version
will
never
ever
conflict,
if
you'll
never
have
like
a
1.1-alpha
and
a
1.1
artifact
everything's,
fine
yeah,
we
should
never.
We.
C
A
Is
the
repo
will
always
ev
all
of
the
base
versions
will
go
in
sync
and
so
that
when
we
release
1.2,
if
there
was
any
1.1
alpha
that
got
promoted
to
the
stable
signal
they'll,
there
will
no
longer
be
a
one,
but
that
two
alpha
would
just
be
1.2,
and
so
you
shouldn't
ever
have
that
issue.
1.2
should
always
be
the
latest.
C
Okay,
yeah
that
that
works,
then,
as
long
as
you
never
try
to
make
a
1.2
and
a
1.2
alpha,
I
think
you're,
fine
yeah.
A
C
Yeah
yeah
and
that
the
way
you're
doing
it's
totally
fine.
I
just
having
run
into
all
sorts
of
stupid
issues.
I
am,
I
no
longer
trust
dependency
resolution
myself,
but
that's
a
fair
enough,
fair
enough.
A
I
think
anurag,
and
I
have
also
been
taking
the
tact
that,
if
you're
using
anything
alpha
you're
asking
for
trouble
like
you're,
you
you're
you're
preparing
to
shoot
yourself
in
the
foot
as
often
as
possible,
because
we're
gonna
break
stuff
if
it's
alpha
we're
gonna
break
stuff
and
it's
gonna
break
things.
But
this
dependency
thing
is
an
interesting
one
that
I
don't
think
we
necessarily
completely
thought
through,
but
I
think
we
should
be
good
because
we're
never
going
to
try
to
we're
never
going
to
take
1.1
alpha
and
make
it
into
1.1.
C
Yeah,
that
should
be
great.
That's
that's
good!
Okay.
Can
I
just
ask
why
this
is.
This
is
just
me
out
of
curiosity,
why
put
alpha
in
the
version
string,
as
opposed
to
in
the
artifact
name,.
A
I
don't
have
a
particularly
good
answer
trying
to
think
through
if
there
would
be
anything
negative
in
there.
So
oh.
This
is
why
this
is
why
this
is
exactly
why,
because
then
we
would
be
releasing
something
called
open,
telemetry
api
metrics
alpha
at
version
1.0
and
which
means
that
from
senver's
perspective,
we're
stable
with
that
alpha
and
we're
not
stable
with
that
alpha
we're
going
to
break
graph
and
we're
not
going
to
bump
a
major
version
on
open,
telemetry
api
metrics
alpha.
If
we
make
breaking
changes,
does
that
make
sense.
C
Yeah
yeah,
interesting,
okay,
and
I
mean
it's
it's
it's
kind
of
annoying.
Well,
the
problem
is
then
they're
the
same
artifact
as
well.
So
again,
you
know
maven
ignores
semver
unless
you
specifically
request
it
to
pay
attention
to
it.
Anyway,
I
can
I
I
can
write
up
a
here's,
a
weird
scenario.
We
should
never
let
people
get
into
and
it
should
be
fixable
with
docks,
but.
A
Okay,
I
don't
know
whether
that
would
be
helpful
or
harmful,
but
it
certainly
seems
like
it
would
be
at
least
make
users
a
little
bit
more.
They
have
a
few
a
few
fewer
things
to
have
to
to
fuss
about
and
it'll
be
easy
for
them,
at
least
to
keep
their
alpha
dependencies
in
line
with
what
sdk
they
depend
on.
A
Java
modules
make
a
lot
of
things
a
lot
more
annoying
like
we,
we
just
we
had
so
we
have
this
issue
right
now,
where
we
have
a
so
we
have
a
testing
on
back
right
that
provides
some
j-unit
extensions
for
doing
testing
using
open,
telemetry
apis
and
in
order
to
do
that,
the
sdk
needs
to
have
this
method
to
reset
itself
or
sorry.
The
api
needs
to
have
a
method
to
reset
the
global.
A
So
that
means
that
the
api
has
to
have
a
public
method
to
reset,
and
I
was
like
well,
let's
not
make
it
a
public
method.
Let's
make
it
a
package
access
method
and
then
have
the
testing
artifact
just
have
something
in
that
same
package,
but
you
can't
do
that
with
the
module
system.
You
can't
have
them
both
both
exporting
things
in
the
same
package.
A
Sure
yeah
I
mean
I
think
at
this
point,
we're
just
like
modules-
are
they're,
they're,
necessary
evil,
probably
an
unfortunate
thing,
because
we're
never
going
to
get
off
java
8
right,
we're
never
going
to
be
able
to
move
beyond
java
8..
So
we're
going
to
be
stuck
with
this
weird
inter-module
world,
where
there's
always
2.0
yeah,
but
I
mean
open
telemetry
is
going
to
have
to
be
on
java.
Everyone
you're
going
to
be
people
using
java,
8
forever,
like
we're,
never
gonna
get
rid
of
job.
Eight!
C
C
Two
more
years
september,
23
yeah.
Well,
here's
a
question:
do
you
have
more
faith
in
enterprises
trying
to
not
invest
in
renewing
old
software,
which
I
also
I
I
like
that?
Or
do
you
have
more
faith
in
oracle's
monetization
strategies?
I
think.
A
E
A
A
D
So
tyler
to
your
point
about
five
and
six,
I
mean
that
thing
that
that
ben
evans
put
together
last
year
about
the
java
ecosystem
stuff
that
was
pre-seven
accounted
for
less
than
one
percent.
So
I
mean
they're
definitely
still
out
there
and.
B
B
A
I
was
last
year
I
was
hacking
on
the
nand
to
tetris
project,
which
is,
you
know,
learn
hardware
using
some
java
tooling
and
worked
on
taking
the
source
code
for
that.
Their
swing
gui,
I
think
it's
swing.
It
might
even
just
be
old
school
pre-swing,
but
that
code
base
doesn't
compile
under
modern
java,
like
you
can't
compile
it
in
java,
11.
it'll
run
just
fine,
but
it's
using,
I
think
the
default
they
have
variables
named
default,
which
or
is
no
longer
allowed.
A
A
C
A
A
Because
some
people
want
all
the
alpha
things
to
never
hit
1.0,
but
at
this
point,
unless
somebody
who
has
incredible
build
expertise
wants
to
jump
in
at
the
last
minute
and
figure
out
how
to
make
all
this
stuff
work
in
the
next
week
and
a
half
we're
sticking
with
the
plan.
C
C
Yeah,
would
it
break
semfer
completely
if
the
alphas
were
just
added
a
zero
dot
in
front
of
the
rest
of
the
version
string?
I
think
it
does.
I
think
it
does
if
I
recall
the
spec
correctly,
but
it's
been
a
while
since
I
touched
it.
So
what
would.
C
Library
0.1.0.0
so
like.
If
you
have
a
1.0.0
artifact,
then
all
your
alphas
would
just
put
a
zero
in
front
of
it
to
be
semantic
version
compatible.
The
downside
is
that
last
dot
zero,
I
think,
breaks
it.
You
have
to
do
like
a
dash
and
and
strings
that
are
lexicographically
increasing
and
that
I
think
also.
A
C
A
C
A
Benefit
is
just
not
there,
like.
The
benefit
is
to
people
who
are
using
alpha
software,
that's
going
to
break
in
other
ways
and
hopefully,
for
a
relatively
short
amount
of
time,
yeah.
So.
C
I'm
gonna
noodle
with,
though,
is
moving
splitting
your
multi-project
build.
So
all
your
alpha
dependencies
are
in
one
directory
and
everything
else
is
in
another
one
effectively
in
the
multi-project.
So
you
have
two
nebula
plug-ins
one
for
everything
and
one
for
everything
alpha,
but
I
still
think
that's
going
to
be
I'll.
Probably
do
one
project
and
then
be
like
this
took
too
long
and
give
up
but
I'll,
try
it
and
see
what
it
looks
like.
A
I
did
ponder
something
like
that
with
basically
splitting
you'd
have
to
split
the
project
into
I
was
unable.
Well,
I
didn't.
I
worked
on
it
for
like
three
or
four
hours
and
was
I'm
unable
to
coerce
nebula
into
even
making
that
work,
but
I
am
also
not
a
gradle
expert,
so
I
I
suspect
there
are
definitely
if
you
really
split
the
project
into
two
and
there
was
no
top-level
build
at
all,
then
you
could
probably
do
it.
A
A
C
Yeah,
I
think
that
just
ruins
the
dev
experience
so
anyway,
I'll
take
I'll,
take
a
10
minute
crack
and
if
I
can
get,
one
project
migrated,
we're
good,
otherwise
I'll
again
I'll
ping,
you
for
both
the
the
benchmarking
stuff
and
the
build
stuff
on
sometime
tomorrow,
cool,
that's
right.
Tomorrow's
thursday.
A
Already
holy
cow
cool
anything
else,
anyone
wanted
to
talk
about.
I'm
really
hopeful
that
we'll
be
able
to
move
into
customer
support
mode
here
for
1.0
very
soon,
which
will
be
a
completely
different,
a
different
kind
of
work.
A
A
No
worries
so,
if
we
finally
stabilized
the
api
at
this
point,
the
api
has
been
stable
for
quite
like
two
releases.
Now
nice
at
least
not
three,
I
mean
four
four
tracing
baggage
context.
A
Right,
nothing
has
changed
in
a
couple
releases.
All
of
the
work
we've
been
doing
is
in
the
sdk
side,.
A
C
So
I
do
have
one
quick
question
now
that
it's
stable
I'm
going
to
be
updating
the
stackdriver
exporter
for
google
up
to
the
stable
version
question
on.
That,
though,
is
I
don't
want
to
contribute
this
into
open
telemetry,
because
I
don't
want
to
give
you
the
burden
of
keeping
our
stuff
working,
but
I
do
want
to
advertise
it
somewhere
yeah.
So
do
you
have
a
again
because
I'm
new
to
the
open,
telemetry
community-
and
I
should
know
this
already-
but
where
do
you
advertise
exporters
that
come
from
third
parties.
A
So
the
official
way
to
do
it
is
if
you
go
to
open,
telemetry,
dot,
io
and
click
on
the
registry,
everything
all
of
the
third-party
stuff
everything
can
be
registered
in
here.
Okay,
so
I
think
we
might
be
in
there
already
okay.
So
if
you
like
search
for
new
relic,
you
don't
get
anything
probably
because,
actually
I
don't
know
why
you
don't
beneath
it.
A
I
added
it:
oh
you
had
it:
oh
it's
new
space,
really,
okay
anyway,
so
that
stuff
is
all
here
and
you
can,
I
think,
yeah
it
has
links
to
the
original
repo.
A
I
think
it
would
be
fine
if
we
also
wanted
to
have
that
sort
of
documentation
in
the
java
repo
itself,
but
we
haven't
up
to
this
point.
This
has
kind
of
been
the
central
place.
Is
the
registry?
A
C
A
A
A
So
this
is
a
fantastic
question
right
now
we,
the
the
stability
of
telemetry
of
the
actual
generated
telemetry,
is,
of
course
in
question,
and
we
talked
last
night
on
our
late,
the
late
night
java
special
topics
group
meeting
about
what
that
means
for
the
agent
and
the
instrumentation
project,
and
I
think
the
it's
a
little
bit
up
in
the
air,
because
if
they
say
1.0,
it's
not
clear
right
now.
A
What
that
means
from
a
stability
perspective
like
if
they
change
the
name
of
a
span,
do
they
have
to
rev
everything
to
2.0
or
if
or
do
they
need
to
version
every
single
instrumentation
module
separately
or
like
all
of
that
is
a
big
open
question
right
now.
So
I
think,
if
you're
interested
in
discussing
the
instrumentation
group
meeting
tomorrow
morning,
9
00
am
pacific.
Time
will
be
a
great
place
to
ask
those
questions.
A
C
That's
way
worse,
it's
9
p.m.
Night
time,
no,
no
yeah
noon!
I'm
I
have,
I
have
to
go,
run
an
errand,
so
I
have
to
sign
a
bunch
of
documents
yay
and
then
so
I
can't
I
can't
make
it
this
week,
but
I
may
try
to
attend
one
in
the
future.
A
C
A
Than
gitter,
that's
not
hard
anymore,
exactly
because
gitter
is
awful
and
I
think
it's
really
nice.
So
if
we
can
get
people
using
it,
I
would
it's
much
better
for
me
as
a
maintainer
than
stack
overflow,
because
it's
one
place
for
me
to
monitor
and
I
get
notifications
in
my
github
inbox
about
discussions.
A
So
as
a
from
a
maintainer
perspective
is
great.
It's
one
place
all
the
discussions
happening
in
one
place.
All
of
the
questions
you
can
vote
up.
Questions
like
you
can
on
stack
overflow
or
vote-up
answers,
and
it
seems
like
a
really
it's
a
big
step
forward
for
having
one
place
to
handle
all
that
support,
stuff,
cool.
C
A
Yeah
and
I
think
we've
the
maintainers
at
least
in
on
java
onorag
and
I
have
started
using
them
for
all
the
announcements
for
release
announcements
or
all
that
kind
of
stuff.
So
yeah,
you
see
there's
a
bunch
of
things
here
and
so
I've
asked
questions
rather
than
it
see.
It
seems
a
little
bit
better
for
me
to
put
it
here
rather
than
creating
an
issue.
That
is
a
question
like.