►
From YouTube: 2021-01-29 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).
B
A
This
is
a
blast
from
my
my
my
childhood.
There
was
this
so
terry
nation
who
is
a
writer
for
doctor
who
and
other
bbc
serials
came
up
with
this
idea
of
well
what,
if
the
it
was
referring
to
star
trek
and
he's
like
what,
if
the
federation
is
actually
fascist,.
A
Right,
what
if
all
of
this
stuff
about
the
primary
directive
and
being
all
nice
and
helpful,
is
just
a
facade
and
actually
it's
a
totalitarian
state
interesting
anyway,
and
so
he
took,
he
took
the
starfleet
logo
and
he
rotated
it
90
degrees
and
turned
it
into
the
logo,
for
a
tv
show
that
he
pitched
called
blake's
7,
which
is
a
scrappy
band
of
resistance
fighters.
Trying
to
you
know
they
steal
a
starship
and
run
away
from
prison.
A
A
Yeah,
if
it
was
a
tardis
console
you,
you
know,
everyone
knows
the
tardis,
but
yeah
it's
a
little
more
excuse
me.
D
Yeah,
it
looks
more
like
the
the
old
tardis
inside
too
than
the
new
tardis.
A
Original
yeah,
now
the
the
one
I
and
I
think,
let
me
pull
up
this
one
from.
A
D
D
So
I
started
working
on
the
like
the
first
item,
a
major
thing
for
1.0.
I
started
work
on
the
open,
telemetry.io
documentation,
oh
yeah,
and
it's
been
bringing
up
some
loose
ends
in
the
api,
so
I'm
gonna
get
a
pull
request
out
for
that,
but
I
think
we're
pretty
good
for
1.0,
even
with
the
change
that
I
have
to
make
because
it's
more,
it
became
so
simple
versus,
like
open
senses
that
breaking
we
won't
have
breaking
changes.
D
A
D
So
I
want
some
reviews,
hopefully
on
that
soon,
maybe
early
next
week
and
simple
stuff,
it's
like
getting
started,
how
to
add
the
dependency
and
then
how
to
started
spam.
A
Sweet
yeah,
I
I
I've
been
trying
to
get
back
into
the
my
dog
food
repo,
and
so
I
revisited
it
the
other
day
and
immediately
ran
into
there
was
no
published
version
of
the
ecto
integration
that
works
with
the
latest
open
telemetry.
A
So
you
know
I
had
to
pull
a
dependency
on
and
I
think
there'd
be
a
it'd,
be
nice
to
at
least
get
everybody
to
publish,
as
some
point
version,
some
patch
version
of
0.5
and
then
and
then
we
can
all
rerun
our
dog
fooding
and
make
sure
that
everything's
actually
tidy.
And
then
we
just
take
that
and
bless
it
as
one
point
naught
or
something
I
don't
know.
D
A
Have
a
a
rela
it
turns
out.
We
have
a
relaxed
attitude
towards
simva,
I
mean
I'd
gotten
a
bit.
There
are
some
parts
of
the
node
community
who
are
very
strict
on
it,
my
favorite
one
being
the
happy
framework
which
got
up
to,
I
think
version
17.
While
I
was
using
it.
That
was
their
major
version
because
they
were
like
no
literally,
everything
is
like.
A
They
are
really
quite
good
at
it,
whereas
on
the
other
side
of
things,
the
phoenix
framework
for
elixir
they
will,
they
will
take
a
a
a
a
monitoring
or
like
something,
that's
not
in
the
main
api
path,
but
something
that's
that
you're
using
to
monitor
your
work,
saying
it's
just
how
I
ran
into
this
this
phenomenon
and
they
will
call
it
deprecated
and
what
they
mean
by
deprecated
is
deleting.
A
A
A
A
Yeah
and
it's
like:
oh
okay,
fine
it
it
just
just
sort
of
teaches
you
that
you
know
it's
like
if
it's,
if
anything
change,
you've
got
to
treat
it
as
possibly
breaking
just
in
case.
You
know
it
teaches
you
to
be
careful
anyway,
so
yeah,
zero,
0.6,
maybe
but
yeah.
If
we
can
get
everyone
to
line
up
on
some
minor
revision
and-
and
and
you
know,
then
that
would
be
when
I
would
try
to
migrate
our
main
app
cool
all
right.
A
An
old
boy
will
actually
nothing
like
putting
in
prod
to
find
out
how
it's
gone.
D
A
A
Oh,
it's
like
a
so
there
was
an
event
for
ending
a
span,
but
it
was
being
in
the
log
process
in
the
processing
pipeline.
So
it's
getting
cold
from
the
same
as
part
of
the
same
call
stack
as
the
userland
code
ending
the
span,
but
you
know
so
it's
it's
in
the
same
process,
for
example,
but
the
span
is
already
immutable
like
no
more
no
more
changes
to
the
span
will
be
entertained,
so
we
had
one
event
for
that,
which
I
think
they
were
calling
span
end.
A
It
was
like
too
late
read
only
we're
like
well.
How
are
we
going
to
put
in
other
stuff?
How
are
we
going
to
do
math
and
insert
you
know
dynamic
attributes
for
counting
cpu
usage
or
integrations
or
sorry?
What's
the
erlang
term?
For
that
it's
not
integrations.
No,
like
it's
a
a
unit
of
cpu
work.
Oh
you
mean
the
reductions
reductions.
Thank
you
yeah.
I
want
to
count
reductions
and
anyway
they
they
ended
up
in
the
in
the
spec
discussion
going
well.
Okay!
A
Well,
we're
too
late
to
actually
change
the
spec
before
the
spec
is
one
naught,
but
we'll
we'll
entertain
it
as
kind
of
like
an
rfc
two
one,
one
nine
may
you
might
yeah
like
like
implementers,
may
define
a
before
span
end,
which
is
called
at
the
same
time,
but
just
before
the
span
is
immutable
and
we
will
yeah
which
would
suit
our
purposes.
Did
we
ever
do
that.
A
Yeah
just
that'd
be
handy,
but
like
actually
finding
that
okay,
so
I
found
your
chat
in
getter.
A
D
A
Yeah
I
mean
I
was
doing
my
best
to
work
with
the
the
spec
community
and.
A
It's
kind
of
funny:
they
they
it's
it's
the
nature
of
that
kind
of
a
body.
But
there's
there's
a
lot
of
stop
energy
to
be
had.
A
Very
cool
yeah
anyways.
If
we
can
sneak
that
f106
that'll,
make
it
don't
make
me
more
enthusiastic
for
dragging
everything
away
from
open
senses.
A
And
then,
and
then
on
the
baggage
side
of
things.
That
would
that
you
know
the
because
the
actual,
I
think,
with
the
baggage
the
time
to.
A
So
the
idea
there
is,
you
can
hook
up
your
processing
pipeline
to
in
your
app
to
or
a
service.
I
guess
is
the
other
way
to
use
a
better
word
to
that's
where
you
hook
up
code
establishing
a
convention
within
that
service.
That
goes
like.
If
someone
shoves
a
user
id
into
the
baggage,
then
you
use
your
your
spam
start
stage
to
copy
stuff
out
of
the
baggage
and
go
oh.
If
that's
a
user
id.
A
I
always
want
to
include
that
the
spam
attributes
copy
it
out
there
that,
but
that's
up
to
that's
up
to
whoever's,
writing
the
code
that
gets
shoved
into
that
pipeline
right
right
and
that's
it.
So
there
was
one
other
obstacle
in
the
elixir
side
of
things
where
we
the
only
way
to
hook
up
those
stages
required,
manipulating
a
record
which
you
hadn't
exposed
as
its
own
struct.
Yet
that
was
the
tracer.
A
D
A
D
Oh,
for,
like
the
processor
yeah
start
for
when
you
start
the
span,
you
want.
A
A
Right
well,
and-
and
that's
that's
what
I'm
intending
with
the
you
know,
so
my
dog
fooding
repo,
I
was,
I
was
treating
it
sort
of
commit
by
commit,
was
basically
a
how-to
guide
and
the
the
tail
end
of
that
was
going
to
end
up
very
phoenix
live
view.
A
But
to
get
there
it
would
first
have
to
go
through
getting
it
set
up
in
the
first
bloody
place,
right
and
and
carrying
through
phoenix
and
ecto
integration.
So
yeah
and
that's
going
to
cover
a
fair
amount
of
the
getting
people
comfortable,
stuff
yeah.
At.
B
A
A
D
A
A
A
D
Need
for
some
dog
fooding
is
been
talking
with
the
otp
team
about
how
we
can
integrate
contexts.
Oh
yeah.
D
D
If
you
just
do
a
bang
like
a
elixir,
doesn't
even
have
a
bang,
so
it
might
not
matter
at
all
in
elixir,
but
so
the
context
won't
be
carried
across
processes.
If
you
just
do
a
bang,
but
when
you're
using
the
core
you
know
of
gen
server
or
something
you
can
have.
This
context
carried
along
with
it
and
that'll.
A
A
Yeah,
so
I
mean
if
we
started
yeah,
so
I
mean
elixir-wise
the
things
that
get
us
so
there's
a
there's,
a
whole
task
mechanism
which
it'll
be
nice.
If
that
carried
across
where
you
like
start
a
task,
it'd
be
really
nice.
If
any
span
started
in
that
task
process
was,
you
know,
took
the
its
parentage
from
whoever
started
it.
A
A
Even
that's
a
bit
kind
of
rude
like
you
can
imagine
that
you
might
want
to
be
able
to
tie
that
execution
to
both
of
those
contexts.
And
you
know
which
brings
us
back
to
links
right,
yeah
and.
A
D
Right
now
in
the
the
the
documentation,
I'm
writing
I'd
show
how
to
do
a
link
when
spawning.
I
guess
in
make
a
note
that
in
the
elixir
version,
I'll.
D
Instead
of
just
a
start
link
like
I'm
doing
in
the
erlang
example
and
then
linking
to
the
context
from
the
outside.
A
Yeah,
I
mean
some
of
that
I'll
get
through
with,
with
my
live
view,
dog
fooding,
just
because
there's
a
there's
a
whole
bunch
of
options,
all
of
them
really
bad
under
certain
circumstances,
and
so
yeah
you've
really
got
to
be
staying.
Aware
of
you
know.
Why
am
I
doing
this
and
what
kind
of
side
effects
am
I
willing
to
tolerate
like
if
it
if
it
crashes
out?
Do
I
want
my
view
dying?
If
my
view
dies?
Do
I
want
it
dying
and
you
gotta,
you
know
it
turns
out.
A
You
cannot
ignore
the
otp
underlying
it.
Otherwise,
when
something
goes
wrong,
you're
gonna
get
stuffed,
and
it's
just
one
of
the
you
know
things
with
you
know
all
these
web
apps
is
we're.
So
often
talking
to
external
web
services
to
get
stuff
done.
Stuff
is
going
wrong
all
the
time
right.
A
A
Oh
yeah,
you
cannot
ignore
it
anyway,
yeah
so
the
but
yeah
so
wow
I
mean
getting
that
all
the
way
into
otp,
so
that
you
know
the
any
wrapper
we've
got
then
around
it
is,
is
just
passing
it
through.
That
sounds
super.
Is
there
any
performance
advantage
that
comes
with
that
or
just?
Is
it
just
great
that
having
the
meaning
that
if
you
call
into
some
library
and
that
calls
into
some
other
library,
then
all
the
spam
yeah
and
then
that
third
library
calls
you
back.
F
A
D
D
A
That's
super
news
very
handy.
I'm
just
surprised.
I
mean
like
we're
a
team
of
we're
a
small
team
at
amplified
and
yeah
even
we're
running
into
this,
this
kind
of
a
problem.
So
it's
not
like
it's
a
you
know,
niche
issue.
It's
you
know
that
only
you
know,
uk
based
sports
betting
organizations
are
going
to
run
into
you
know
it's
yeah,
oh
yeah.
Anyone
trying
to
build
anything
on
this
platform,
yeah,
I'm
glad
they're
they're
onto
it.
D
Yeah,
it's
hopefully
I'd
love
to
see
it
in
24,
but
maybe
24.1.
B
Can't
wait
for
that.
One.
E
E
Well,
I
can
see
you
in
two
weeks:
yeah
run
and
hide
for
now,
good
to.