►
From YouTube: 2020-09-01 .NET SIG
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
C
C
Hey
guys
c
joe
is
out
today,
so
I'm
gonna
kind
of
try
to
facilitate
for
him.
There's
not
a
whole
lot
to
cover.
If
you
guys
want
to
put
your
names
on
the
attendees
list,
the
big
news
is,
we
released
a
beta,
our
second
beta
last
friday,
cjo
updated
the
the
next
beta
is
targeted
for
the
end
of
september,
probably
with
net
five
release
candidate,
one
probably
a
little
bit
after
that
hits.
C
That's
pretty
much
everything
I
have.
I
know
we
have
some
new
members.
Does
anyone
want
to
introduce
themselves.
B
I'll
start,
maybe
I'll
turn
my
video
on
for
a
sec,
hey
everybody.
My
name
is
michael,
I'm
based
in
toronto,
and
I've
been
kind
of
lurking
around
the
chat
and
the
project
for
a
few
days
now
trying
to
start
using
open
telemetry
for
our
project
and
help
any
any
way
I
can,
if
you
guys
need
help.
I
just
got
it
up
and
running
earlier
today,
a
little
bit
so
I
don't
have
really
much
to
say,
but
it
would
be
interesting
to
get
involved
and
help
out.
D
C
E
Objected
with
having
it
it
just
a
question
of
who
will
be
doing
it,
so
I
think
the
most
work
is
initial:
okay,
publishing
and
publishing
first,
and
I'm
glad
that
eddie
just
sent
a
pull
request
visiting
names.
So
at
least
we
have
a
names
right
and
I
would
assume
that
names
will
stick
and
if
we
find
business
names
we
can
probably
push
it
and
get
the
first
version
out.
E
I'm
happy
to
do
this.
If
there's
no
objections.
E
Yeah,
the
only
small
question-
I
I
really
like
this
name
opencemetery.com,
the
only
small
thing
bugs
me
is
it's
open,
telemetry
namespace.
So
do
you
want
to
do
open,
selementary
country
without
dot,
so
we
can
clearly
distinguish
it
from
open
telemetry
or
we
find
with
this
dot
country,
I'm
just
bringing
it
up
in
case.
Anybody
has
strong
objections,
if
not
like
I'm
happy
with
dot
country.
C
C
C
D
Yeah-
and
I
think
that's
the
main
concern
about
rayleigh.
So
if
we
stick
with
the
open,
telemetry
namespace,
are
we
going
to
be
the
owner
if,
yes
or
if
no,
what
we?
What
are
we
going
to
do
with
that?
If
they
open
issues.
E
E
If
you
care
about
this
bug,
you
will
push
a
fix
for
that
and
like
it
can
it
probably
will
be
accepted
if
it's
good
enough
and
like
everything
there
is
no
like
it
doesn't
break
anything
so
same
in
country
repository.
We
will
just
have
even
less
people
care
about
it
and
we
will
just
articulate
it
clearly.
There
is
no
specific
people
behind
us
packages.
E
C
What
am
I
doing,
and
I
mean
it's-
not
a
ton
of
volume,
it's
a
little
more
challenging
like
something
like
mass
transit
that
came
up
like
I
don't
know
anything
about
it
so,
like
I
can
review
it.
You
know
for
conventions,
and
you
know
sanity,
but
some
of
them
I'm
gonna
have
to
rely
on
other
people
and
I'm
happy
to
like
pull
in
people
as
we
can
I'm
just
kind
of
hoping.
If
we
nurture
it
and
it
grows,
you
know
more
people
will
join
and
then
some
other
people
will
kind
of
step
into
those
roles.
C
G
C
G
What
we
have
is
like,
let's
say
for
exporters
and
things
like
this,
we
have
the
people
from
vendors
or
perhaps
the
the
key
contributor
that
bring
that
contribution
to
the
ripple,
be
how
is
involved
in
review.
So
we
maintain
the
repo,
but
a
lot
of
the
trusting
about
the
reviews,
and
this
stuff
are
for
this
kind
of
let's
say
between
quotes
owners,
office,
specific
parts
of
the
code.
E
Yeah,
I
remember
back
in
microsoft,
I've
been
reviewing
each
other,
basically,
on
country
repository
somebody
will
submit
the
apr
and
another
person
from
microsoft
would
review
it.
It
was
for
azure,
mind
or
exporter
specifically,
I've
been
that
reviewer
so
many
times,
and
I
think
my
worry
here
is
without
country
pv
and
without
this
place
for
this
music,
like
some
smaller
technologies.
E
Of
course,
like
not
microsoft,
provided
technology
support,
it
will
be
a
little
bit
hard
for
people
to
provide
this
piece
of
functionality
and
we
will
just
lose
it
and
the
community
will
not
get
it
away
at
all.
So
this
is
like
this
is
better
than
nothing.
From
my
perspective,.
E
So
I
I
read
these
notes
about
from
sieges
that
next
beta
is
end
of
september.
We
just
we
tried
to
implement
open
telemetry
for
one
of
the
internal
projects
in
google
and
there
is
a
nasty
bug
in
concurrency
for
metrics
and
we
really
want
to
have
a
beta
shift
like
at
some
point.
Maybe
earlier
as
you
do,
you
know,
I
think
it
is
closest
to
oh,
it's.
D
So
this
is
why
we
haven't
updated
a
lot
of
it.
E
Yeah,
but
I
mean
adoption
is
good
right,
so
I
just
curious
if
you
can
push
and
like
cut
another
beta,
maybe
sooner
than
end
of
september,
and
what
will
mean
so
maybe
we
can.
I
know
who
we
can
talk
about
it.
D
I'm
not
sure
about
how.
How
can
we
decide
it?
We
can
do
us
like
a
separate
issue
and
ask
siegel
and
riley
about
it,
because
since
I'm
I'm
helping
them,
but
I'm
not
actually
working
with
them
like
in
the
same
team.
So
I
really
don't
know
the
answer.
F
E
It
was
a
recent
pr
from
laura,
I
believe
it's
it's
fixing
the
concurrent
dictionary
in
matrix.
C
H
So
yeah
are
we
looking
for
justin
beta
before
end
of
september,
for
particularly
for
the
bug
fix
and
do
we
have
a
reason
for
that?
Like
is
someone
waiting
on
the
bug
fix.
E
Yeah
we
want
to
try
it.
We
have
a
project
in
google
that
we
onboarding
to
open
telemetry,
and
this
bug
is
so
nasty
that
it
crashes
a
process.
I
see.
H
Yeah,
if
there's
justification
for
that,
then
I
can
definitely
talk
to
brylee
and
really
nc
jonar
and
and
if
it
drives
any
adoption,
okay,
you
can.
E
Take
that
I
can
file
an
issue
and
we
can
discuss
it
there.
I
will
mention
you
in
the
issue.
D
H
D
Sergey
what
you
can
do
in
this,
meanwhile,
is
use
the
myget
version.
I
know
that's
not
the
official
release,
but
at
least
you
can
test
with
lara
bugfix.
E
Yeah
we
already
tested
in
the
private
version,
so
yeah
and
my
git
is
good
enough
for
for
a
while.
I
really
hope
that
we
can
get
something
that
we
can
push
into
codebase
and
and
it
will
have
some
long,
lasting
package
reference.
E
Yeah,
okay,
I
will
create
an
issue
to
discuss
it.
End
of
september.
Doesn't
sound
too
scary
if
you
can
get
it
earlier,
it's
it
would
be
great.
F
This
is
tangential
but
related
to
my
get.
I
was
actually
experimenting
with
something
last
week
and
was
looking
to
pull
the
package
from
my
git,
and
I
didn't
realize
until
recently
that
all
the
the
links
to
the
feeds
and
so
on
have
been
removed
from,
I
think
all
of
the
documentation.
So
I
actually
had
to
kind
of
figure
out
what
the
url
for
the
feed
was.
F
On
my
own,
I
mean
I
got
it
working,
but
does
anybody
know
the
history
between
behind
removing
that
or
has
it
just
moved,
and
I
just
had
didn't
find
it.
D
Hi
alan,
I
think
we
we
remove
it
it.
I
think
we
don't
add
it
in
another
place,
because
if
I'm
not
mistaken,
the
idea
was
for
after
the
first
release,
the
idea
would
be
for
people
you
to
use
the
the
beta
release
instead
of
the
nightly
release.
F
Page
yeah,
I
think,
that's
fair
kind
of
pondering
whether
it
makes
sense
to
have
maybe
a
readme
for
like
a
developer
setup
kind
of
read
me
where
maybe
that
would
be
desirable
information,
yeah.
F
E
Talking
of
adoption
since
everybody
here
and
we
don't
have
any
other
topics,
anybody
knows
people.
A
E
So
anurag,
do
you
have
more
information
if
you
receive
a
trace
with
a
random,
fully
random
id,
how
your
sdk
will
behave
like
you
would
try
to
think
that
this
id
is
not
random.
We
will
somehow
fix
it.
A
E
A
I
mean
x-ray
traditionally
uses
a
custom
header,
but
as
long
as
the
id
starts
with
the
epoch
prefix,
so
the
four
first
four
bytes
have
to
be
the
time
stamp,
which
is
sort
of
a
subset
of
w3c
ids.
So
as
long
as
the
generation
happened,
okay,
all
the
other
propagation
formats
still
work
well
with.
A
A
E
Yeah
yeah,
I
think
one
thing
we
discussed
long
ago
is
some
customers
need
to
restart
trace
id
because
they
don't
trust
whatever
they
see
in
headers
so
and
we
discuss
like
whether
we
can
allow
to
to
do
something
about
it,
and
the
answer
was
always
to
get
into
one
of
the
one
of
the
callbacks
of
activity.
E
I
think
it's
should
sample
callback
from
asp.net
and
in
this
in
this
callback
you
can
start
your
own
activity
and
then
new
activities
will
be
parent
of
a
child
of
your
activity
so
and
this
activity
can
be
set
with
trace
idea
of
whatever
you
want.
So
this
kind
of
magic
can
be
used
in
your
case
as
well.
A
Investigate
and
that
new
trace
would
not
be
set
to
like
is
a
remote
or
whatever.
That
would
be
a
new
trade
like,
even
if
you
provide
a
trace
id,
it's
still
not
a
remote
spam
context,
I'm
not
sure
about
that.
Okay,
but
that
could
I
mean
right
now.
We
don't
use
remote
versus
local
spam
context
for
anything,
so
that
could
still
be
a
workaround.
C
We
have
the
same
challenge
in
the
net
sdk
because
it's
really.net
runtime
handling
that
incoming
message.
So
if
the
user
says
they
want
to
do
b3,
propagation
or
any
other
type
of
propagation,
we
do
what
sergey
is
saying.
We
just
create
a
child
span
with
the
correct
id
and
it's
just
running
under
the
runtime
one.
So
it's
it's
a
challenge
for
even
the
sdk,
because
that's
that's
provided
by
the
runtime,
so
we'll
have
to
work
with
that
team.
If
we
want
to
totally
change
that
id
generation,
if
that
makes
sense,
yeah.
C
Deeper
could
also
in
the
net
sdk
I'm
just
trying
to
think
through
it
mentally.
If
you
just
register
like
an
activity
processor
that
will
get
called,
it
might
be
a
little
bit
easier
because
that
sampling
call
you're
limited
in
what's
passed
in
and
out.
So
if
you
created
an
activity
in
there,
I
think
the
return.net
will
will
overwrite
your
activity.current
with
the
one
that
it's
creating,
so
you
might
want
to
look
at
an
activity
processor.
It
might
be
better.
It's
going
to
be
tricky,
though,.
E
C
E
E
Designed
specifically,
so
you
can
replace
activity,
I
don't
have
example
of
code
and
this
code
pass
may
have
been
changed
because
nobody
asking
about
this
scenario
after
that,
but
yeah.
If
we
need
to
change
something,
probably
it's
time
to
bring
it
to
the
net
team.
E
C
A
E
Another
activity:
this
is
something
that
customers
asking
about,
but
I'm
not
sure
about
customer
id
generation
because
it
opens
up
for
lots
of
potential
security
issues
since
speed.net
propagates
this
activity
automatically
through
all
the
layers.
If
somebody
will
put
some
ip
address
into
it
or
user
identifier,
then
it
opens
up
for
some
liabilities
on
all
service
owners.
A
A
E
I
think
this
will.
This
is
why
we
do
like
we
create
a
trace
state,
so
anybody
can
propagate
any
extra
piece
of
information.
You
need,
and
I
really
hope
that
you
will
be.
You
will
get
into
to
the
point
when
you
can
use
it.
Otherwise
I
mean,
I
think
this
is
exactly
why
we
do
outreach
date,
so
it
will
be
the
best
case.
A
E
So
yeah,
I
want
to
repeat
like
do
we
know
about
any
adoption?
Is
anki
dropped?
Yeah?
I'm
kid
dropped.
E
G
C
C
E
You
have
a
mix,
do
you
have
any
examples
when
one
service
is
monitored
by
both
backhands
like
two
or
more.
C
We
have
that
in,
like
our
development
stuff,
like
people
can
in
configuration,
can
say
like
you
can
export
to
yeager,
zipkin
and
google.
If
you
want,
I
don't
think
anyone's
really
played
with
like
sampling.
Yet
it's
pretty
much
just
like
we're
just
kind
of
playing
with
it,
so
everybody's
just
doing
100.
So
then
we
get
a
noise
situation,
so
I'm
doing
like
education
on
like
okay.
Now
we
need
to
sample
and
hear
how
that's
going
to
work
and
stuff.
C
G
I
I
had
customers
doing
the
forking
in
devs
and
rollout
environments,
kind
of
forking
to
our
production
and
also
on
their
dev
environment.
They
had
a
zip
king
that
they
maintain
you
know
so
they
could
they
use
that
to
some
kind
of
initial
shack,
but
now,
actually,
I
think
they
disabled
that
last
month,
so.