►
From YouTube: 2020-10-21 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
B
D
A
A
And
riley
has
a
meeting
conflict
this
morning,
so
I'll
help
him
host
today's
community
meeting.
Maybe
we
can
wait
one
two
more
minutes
and
then
we
can.
A
A
Today,
everyone
help
and
add
your
name
on
to
the
attendee
list
and
if
there's
any
topics
feel
free
to
add
it
to
the
agenda
list
too,.
A
A
E
A
Meeting
because
he
has
a
conflict
this
morning
and
yeah,
I
will
help
him
host
today's
meeting.
A
Yeah
feel
free
to
add
your
topics
to
the
agenda
list
and
let's
get
started
yeah.
I
think
the
first
one.
The
first
item
in
agenda
is
the
login
api
update,
yeah
mark
and
karen
in
also
both
are
in
a
meeting
and
they
are
working
on
their
draft
and
prototype
in
the
last
few
weeks,
and
I
think
in
in
the
last
week,
there's
progress
and
the
would
you
meant
give
us
an
update
or
about
your
climbing
or
for
the
logging
api
for.
D
For
sure
yeah,
so
last
week
karen
and
I
we
made
an
initial
commit
with
the
api
and
we
made
that
to
a
private
repository
just
locally
for
our
aws
team,
and
so
we
got
some
reviews
from.
I
think
you
had
access
to
it
tom
and
max,
and
we
had
some
comments
on
it.
So
we
made
a
couple
changes
and
right
now
it's
in
a
good
position
to
be
like
committed
upstream,
so
everyone
can
see
it,
but
we
still
need
a
bit
of
confirmation
from
our
manager
about
that.
D
So
by
the
end
of
this
week
it
should
be
committed
upstream.
So
you
guys
can
see
and
in
terms
of
other
progress,
karen
she's
been
working
on
the
exporter
and
processor
component
of
it,
so
she
has
a
working
o
stream
exporter
and
I'm
working
on
the
logger
sdk
analog
provider,
sdk
component,
where
I'm
still
writing
up
a
design
draft
on
that.
But
I'll
have
some
example
code,
probably
by
the
next
two
plus
plus
sig
meeting
on
monday.
A
Okay,
so
you're
getting
getting
permission
to
publish
your
current
draft
and
the
prototype
api
and
pr
right,
yep,
yep,
okay,
great
okay,
any
questions
on.
A
A
C
Item
yeah,
so
this
is
this
is
basically
for
me.
I
mean
just
not
this
pr
was
just
hey.
I
can
I'm
audible
right,
hey
tom,
can
you
hear
me
yeah?
I
can
hear
your
hair
yeah
yeah.
So
basically,
this
just
wanted
to
understand
on
what's
the
fate
or
the
status
of
this
pr,
I
mean
it
was
probably
raised
before
I
really
started
being
active
in
this
group
and
just
going
through
the
review
comments.
C
I
really
don't
see
anybody
objecting
to
that,
but
I
just
wanted
to
understand,
because
right
now
I've
been
working
on
the
http
api
and
the
things
could
have
been
different
had
we
definitely
had
a
lib
event,
async
design
api
there
in
the
so
we
just
wanted
to
understand.
I
mean
the
people
who
were
there
earlier
as
part
of
the
review.
I
mean,
I
definitely
see
max
and
your
hands
there.
So
is
it
it's
something
that
we
are
not
going
to
proceed
further
or
something
still
under
discussion.
C
B
Words
to
that-
and
that
was
done
quite
a
time
ago
quite
a
while
ago-
and
the
intention
was
to
use
the
lip
event
for
basically
timing
in
our
batch
spam
processor
and
also
for
yeah
http
stuff.
B
B
So
I
think,
checking
out
and
compiling
lip
event
via
the
bezel
make
system
doesn't
work.
A
B
C
E
E
So
pretty
much,
we
have
a
full
feature
to
be
working
with
youtube
client
that
we
can
contribute
today
that
works,
and,
yes,
we
need
to
refactor
the
api.
There
were
comments
about
how
then
this
web
event,
which
practical
and
maybe
a
better
option
which
requires
a
bit
more
work.
Then
you
would
need
to
figure
out
how
to
wire
this
for
all
the
corresponding
platforms
that
we
already
support
the
networking
http
client.
E
So
what
I'm
trying
to
say
is
that
this
should
probably
be
two
parallel
tracks
and
I'd
say:
let's
go
with
something
that
gives
us
the
real
world
first,
which
is
the
working
initiative,
client
and
then
see
how
to
merge
the
web
event
and
if
it
works,
use
that
instead
and
my
comment
was
that,
for
some
reason,
chromium
browser
doesn't
use
it
on
windows.
I
don't
know
what
practical
reason
for
that.
It
could
be
anything
related
to
builds
or
other
issues
like
it
might
be
a
best
option
for
posix
or
us.
E
It
may
not
be
the
best
one
for
windows,
so
it's
like
as
a
fallback
plan.
Maybe
you
will
end
up
having
to
use
clipboard
rental
projects,
but
retain
the
existing
like
http
client
implementation,
that
we
can
contribute
for
windows,
and
then
we
have
the
full
set.
So
it's
like
two
tracks
and
then
some
some
sort
of
conversion
see
what
I'm
saying
right.
B
C
C
B
A
C
So,
which
exporter
so
basically,
any
network
communication
file
based
even
based
communication?
We
can
leverage
the
lib
event
library
for
any
of
the
exporters
I
mean
in
case
they
need
a
networking
functionality
or
file
reading
functionality.
Instead
of
waiting
for
file
to
read
the
file
from
the
disk
io,
we
can
use
the
label
event
eventually
even
event,
troop
approach,
which
basically
is
brought
by
lab
event.
C
E
C
A
Okay,
let's
move
to
the
next
one.
The
next
one
is
tracer:
sdk
doesn't
use
library,
name.
D
But
yeah
I
put
that
there
it's,
it
might
be
a
stupid
question,
but
inside
this
file,
if
you
look
at
line
which
number
14
and
15
the
two
parameters
for
this
function
are
the
library
name
and
library
version.
But
when
you
create
the
tracer
instance,
it
doesn't
use
that
anywhere.
So
I'm
wondering
if,
like
it's
a
future
intent
to
include
that,
but
it's
just
not
implemented
right
now.
B
Yeah
I
mean
I,
I
already
consider
a
few
words
to
this-
that
that
is
a
future
indent
and
basically
that
will
go
hand
in
hand
when
we
support
kind
of
enhanced
configuration
possibilities
for
the
sdk,
because
the
idea
here
is-
and
that's
I
think,
that's
the
same
for
the
media
provider.
That
probably
will
be
the
same
for
the
logging
or
logger
provider
that
that
you
say
yeah.
You
pass
the
library
name
and
library
version
there
and
that
via
some
configuration
option,
you
can
disable
like,
for
example,
tracing
for
certain
libraries.
For
example.
B
B
B
E
A
E
E
A
E
So
namespace
really
remains
the
same,
that
we,
how
we
expose
interfaces
really
remains
the
same,
but
when
the
binary
is
produced,
it
just
aliases
the
classes
to
the
standard
implementation
which
may
be
of
benefit
for
statically,
linked
or
header
own
implementations
of
the
sdk.
E
It's
just
there's
a
bit
of
planning
needed
for
testing
as
well
to
make
sure
that
in
both
parts,
all
the
tests
passed-
and
I
want
to
clean
this
up
and
refresh
before
the
next
signature
I'll
do
that.
A
Yeah
there
is
an
issue
that
the
pr
uses
some
new
cmake
functionality,
which
I
think
that
johannes
mentioned
only
available
in
3.14
and
above
and
yeah,
if
we
want
to
which
is
http
fetch,
which
fetches
the
grpc
reaper
from
github.
I
think,
but
if
we
want
to
take
this
pr,
it
means
we
need
to
update
the
global
minimum
cmake
requirement
onto
to
at
least
3.14,
but
I
think
this
this
version
is
quite
new
right.
A
If
we
update
it,
maybe
could
it
cause
compatibility
issue
yeah
like
where
the
default
remake
are
installed
in
my
ubuntu
18.04,
which
is
3.10,
I
think,
which
means
if
I
clone
it,
I'm
account
build
open,
flammability
plus
plus,
if
we
upgrade
it
or
in
this
one
or
concern,
I
mean
to
to
neural
simic
version.
We.
E
E
A
E
C
E
By
the
way,
I
have
seen
fairly
good
snappy
results
for
downloading
and
building
simulink
from
source
using
google
ninja.
You
can
fetch
summate
then
use
google
ninja,
and
it's
like
so
if
needed,
I
can
contribute
the
script
for
that.
A
Do
you
mean
download
the
repo
as
a
file
clone.
E
E
The
only
the
only
reason
why
I
suggested
ninja
is
regular
scenic
build
is
slow
if
you
just
build
configure
magnetic
installations,
if
you
do
the
configuration
with
ninja
and
then
you
build
with
new
jack
juice
faster,
that's
just
an
observation:
ninja
is
actually
generally
much
faster.
E
You
see
cash
class
ninja
is
a
perfect
combo
to
speed
up
our
ci
and
build
loop
types
since,
like
many
build
systems,
android,
for
example,
ndk
by
default,
now
uses
ninja
and
chromium
business
ninja,
and
I
tried
it
out,
and
it's
gonna
add
a
little
bit
of
build
time
to
ci
loop
if
we
fetch
the
c
mic
from
source
and
build
it
up.
Maybe
two
minutes.
A
Okay,
I
think
max
you,
you
are
going
to
take
a
look
at
this
vr
and
more
right
and
yeah.
A
A
Any
outstanding
blocker
here
or
almost
ready.
C
B
I
just
posted
some
review
comments
on
that
like
20
minutes
ago.
Oh.
C
B
A
C
A
A
Okay
thanks
so
the
last
one.
I
think
this
is
a
implementation
of
the
http
api.
C
Yeah
so
basically,
I've
just
used
one
header,
only
http
cross
platform,
history,
lib
library,
for
implementing
the
abstract
api.
C
C
Now
so
this
the
http
client,
so
it
has
two
header
files.
One
is
the
http
lib.h,
which
is
basically
coming
from
the
other
repo.
It's
a
single
header
file,
yeah
this
one,
the
last
one
I
mean.
E
E
The
what
can
I
check
the
license
over
there?
What's
the
license,
the.
E
E
Also,
does
it
support
https?
C
B
E
E
A
B
C
C
An
issue
good
point:
yeah.
We
we
have
to
see
if
we
can,
we
have
to
ignore
to
remove
this
file
from
the
coverage.
I'm
not
sure
how
I
mean.
B
The
coverage
will
be
more
like
the
third
party
folder,
so
once
this
is
moved
into
the
third
party
older,
the
coverage
should
not
consider
this,
and
if
so,
then
we
should
change
our
the
kickoff
configuration.
E
I
am
a
bit
skeptical
about
it
because,
judging
by
the
amount
of
subtle
bugs
across
different
urban
electric
versions,
for
example
in
the
last
three
years,
I
can
comfortably
tell
that
this
code
is
probably
not
addressing
all
of
the
possible
situations,
especially
on
the
negative
path.
When
connection
is
reset,
when
the
libraries
are
able
to
create
a
handle,
all
this
fun
and
tricky
stuff,
not
a
perfect
case
when
everything
works
and
you
get
200,
okay,
but
the
other
corner
cases
and
that's
the
minefield.
A
A
C
C
B
It's
good
to
bring
this
up
because,
usually
when
reviewing
I
other
reviews,
we
don't
like
review
draft
prs
because
we
think
that's
still
kind
of
work
in
progress
and
we're
waiting
to
bring
that
up.
Because
then
I
know
okay,
so
I
will.
I
will
give
that
a
review.
A
C
I'll
check
that
I
think
that
okay,
I
think
that's
the
I'll
check
that
probably
why
it
is
failing,
but
I
think
some
other
files
got
because
of
I
think
merge
conflict.
I
probably
put
some
other
files
the
merge
conflict
files
which
I
removed
it
I'm
not
sure
why
it
is
still
showing.
C
B
A
Yeah
sure
I
saw
yeah,
you
you're,
you
put
some
keyword
like
within
progress
in
the
title
and
for
for
such
pr
I
really
are
ignored
to
take
a
look.
A
A
Opened
in
the
last
seven
days,
rename
okay,
this
rename
spam
count
to
spy
reference.
I
think
I
opened
it
in
the
last
week.
I
think
there's
someone
who
claimed
to
do
this,
but
there's
no
update
for
for
this
vr
a
few
days
ago.
So
probably
I
will
ping
this
pr
and
the
yeah.
This
should
map
your
big
change,
just
preview,
mostly
stream,
replacing
and
take
care
of
some
some.
Maybe
there
are
some
edge
cases
like
not
strictly.
It
could
be.
Oh
two
words
right,
so
there
are
some
cases.
A
I
saw
some
cases,
not
a
direct
string
replacement.
So
I
hoping
this
issue.
A
Okay,
okay,
this
this
one.
That
sounds
like
a
a
big
change
right,
maybe
just
change
some
some
numbers
in
the
constructor
or
is
it
maybe
we
can
mark
it
as
a
first
good
issue
or
something
like
that
so
to
help
help
people
to
pick
up
this
kind
of
issue.
I
think.
B
B
E
A
E
Yeah
imagine
a
case
where
a
hacker
corrects
a
packet
which
contains
more
than
a
thousand
records
right,
like
anybody
can
do
that
right.
Somebody
can
download
the
open
source
system
and
that's
just
the
limits
valid
and
adjust
the
packet
right.
So
from
that
perspective,
collector
must
enforce
the
limited
base.
That's
fair!
So
then,
what
if
we
go?
E
If
we
don't
find
we're
trapped
in
another
pocket,
we
allow
the
users
to
fall
into
trouble.
We
don't
want
the
event
to
fail
and
it
depends
on
how
exposure
works,
whether
it
gets
the
response
code
or
not,
but
we
bundle
up
the
response
or
something
from
educating
along
with
me.
That
request
has
been
rejected,
for
example.
E
E
A
A
C
A
C
A
Okay,
let's
move
to,
I
think,
the
last
one
this
one
this
one
comes
from
karen.
I
think
that's
the
issue.
F
Yeah
so
when
I
was
doing
the
pr
to
move
key
value
interval
and
the
related
key
value
interval
view
class
from
trace
to
common
layla,
and
you
as
well
made
a
comment
that
all
of
the
scope
resolutions
should
be
consistent
and
and
or
maybe
some
generic
short
form.
F
What
do
you
call?
The
name?
Spaces
should
be
added
as
well
for
simplicity.
So
I
filed
a
separate
pr
or
a
separate
issue
for
that
and.
F
F
I
think
it
might
require
going
into
I'm,
not
I'm
not
sure,
but
I
like,
because
some
of
them
are
prefaced
with
open
telemetry
and
some
of
them
aren't.
I'm
not
sure
if
there's
a
bigger
change
there,
other
than
just
adding
like
using
like
lines
or
like
adding
aliases
for
name
spaces,
I
think
like
it
could
have
like
with
the
build
files
need
to
be
changed
as
well.
I'm
not
super
sure
on
that.
A
C
A
E
So
that's
that's
only
for
this
cpp
file
right
for
headers.
We
would
typically
use
the
full
scope,
because
what
if
there
is
another,
our
library
where
it's
also
using
a
trace
or
common-
and
we
would
get
a
name
space
clash
so
for
depends
the
depends
where
headers
we
should
use
complete.
C
A
We
only
don't
use
type
or
like
name
space,
alias
in
header
files
or
sounds
like
a
good
suggestion
to
me.
E
Yeah
it's
like
we
can
also
check.
I
think
this
is
because,
when
somebody
codes
the
cpp
file,
they
know
that
this
is
self
contained
unit
and
they
do
using
whatever
and
then
it's
alice.
They
will
use
the
short
short
name
space
whatever
and
there's
no
clash,
there's
no
profit,
because
it's
a
isolated
unit
you're,
not
adding
any
random
vibrations
to
it
anymore.
E
In
other
cases,
people
are
not
so
sure,
because
this
may
be
a
big
other
external
library
included
and
then
you'd
use
more
complete
scoping
declaration
here,
like
a
full
path.
Yeah,
we
need
to
clean
it.
Yes,
I
agree
that
maybe
some
consistency
would
be
this.
A
Okay,
that
makes
sense
thanks,
so
I
think
we
just
went
through
all
the
prs
and
issues.
So
is
there
any
other
other
topics
we
want
to
discuss
here.