►
From YouTube: Open Telemetry's Personal Meeting Room
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).
C
C
C
C
A
Can
you
see
my
screen
great,
so
upgrade
fix,
fix
the
issue.
We
have
three
open
at
PRS.
A
One
of
them
is
related
to
Samplers
I,
have
a
plan
to
include
this
logic
into
SDK
site
without
without
success.
Unfortunately,
I
think
we
should
not
wait
any
longer
for
it,
so
I've
decided
to
include
changes
here
and
I've
created,
separate
issues
to
remove,
remove
it
after
or
when
the
hotel
SDK
will
support
it.
I
do
not
think
it
will
be
one
for
all
versions.
So
please
take
a
look
into
it.
A
We
have
PR
for
Bureau,
it
is
pretty
simple
PR
extending
our
test
coverage
and
see
how
it
looks
more
or
less
good
Us.
In
other
things
we
in
in
other
places
we
have
issues
related
to
show
tests.
We
are
working
on
it,
but
no
clue.
What's
what's
wrong,
there.
A
And
there
is
the
ongoing
discussion
about
how
to
validate
open,
Telemetry
and
other
dependencies
version
by
our
loader
I'm,
not
sure
refresh
my
tasks.
No.
C
Yeah
rash
is
not
it's
not
here.
He
replied.
Some
of
my
questions.
I
didn't
have
time
to
read
yet
I
think
there
was
a
I
think
there
is
a
comment.
That's
about
the
the
pi
itself,
I
I
called
stuff
right.
I.
Think
the
relevant
thing
here
for
us
in
the
meeting
is
that
I
think
we
need
something
more
General.
I
think
this
is
a
good
start
to
have
right
now,
but
I
think
we
need
something
more
General
to
handle
this
kind
of
possible
conflicts
of
version.
C
I'm
I'm
wondering
if
I
can
do
something
related
to
these
with
the
issue
1646,
that's
the
one
about
that.
We
discussed
last
week
about
the
compatibility
on
the
framework
that,
right
now
we
are
requiring
by
binding
redirects
I.
Think
perhaps
we
can
piggyback
something
to
improve
the
situation
here,
while
in
handling
that
that
said,
one
is
for
the
setup
hook.
C
The
other
one
is
for
the.net
framework
that
depends
on
the
CLR
profiler,
but
I
think
that
the
part
of
Leverage
and
listing
the
dependencies
could
be
common
to
the
solution
of
both.
You
know
so.
C
Later,
with
what
graduate
replied
here
and
follow
up
on
that.
A
Great,
so
that's
all
right
now,
typically,
we
should
go
through
the
newly
created
issues,
but
I
see
one
or
two
new
faces
hello,
guys,
Daniel
and
Carlos.
Do
you
have
any
specific
topics
to
discuss
with
us.
E
I'll
start
first,
my
name
is
Daniel
ADH
I
work
at
the
company
called
observe,
which
is
an
observability
platform
on
top
of
powered
by
snowflake
stuff.
I
came
to
hear
this
company
recently
and
I
became
aware
of
the
hotel,
Auto
instrumentation
projects.
We
use
them
for
some
of
our
clients
and
I'm
here
to
listen
to
see
what,
if
anything,
our
company
can
do
to
help
contribute
to
this
project.
I
would
love
the
hotel,
Auto
instrumentation
quality
and
coverage
be
bigger
and
nicer
and
whatnot.
F
E
Side
of
things
fairly
successfully
to
instrument
some
of
the
commercial
Cuts
commercial-
oh
God,
I'm
spacing
out
anyway,
some
some
applications
that
out
there,
although
the
coverage
isn't
quite
there
yet
to
the
quality
of
what
I
was
experiencing
when
I
was
at
app
dynamics,
that
agent
was
fairly
fairly
strong
and
sturdy,
but
I
liked
the
Simplicity
and
the
ability
to
apply
it
to
multiple
Frameworks
that
this
agent
offers
so
I'm
here
trying
to
see
what
Cal
company
can
do
to
help.
If
anything.
C
Welcome
aboard
Daniel,
you
are
read,
did
some
contribution
by
I?
Remember
the
books
that
you
opened
very
detailed,
very
good.
Thank
you.
A
A
Dot
net.
We
have
a
lot
of
native
code
web
analysis,
which
are
absolute
right
now.
One
of
them
is
the
dotnet
runtime
stuff.
For
example,
you
are
not
able
to
use
I
CLR
profiler
version
13.,
it
is
nourage,
but
when
I
found
this
I
decided
to
open
issue,
we
should
update.
C
Of
the
Native
dependencies
that
we
need
to
to
update
exactly.
C
Yeah
I,
I,
I,
think
I,
don't
know
if
Zach
can
talk
right
now,
but
I
think
the
he,
because
we
come
from
the
Upstream
from
data
dot
that
they
are
the
ones
with
more
experience
of
doing
that,
any
tips
or
recommendations
for
us
when
doing
that
update
or.
B
B
So
it's
fairly
straightforward
but
I'll
see
if
we
have
the
docs
so
that
this
is
a
repeatable
process.
For
each
time
we
have
an
update.
A
One
more
issue
related
to
90
code.
We
are
using
as
data
doc,
regex
Library
called
Rito
I
suppose
it
could
be
replaced
by
standards,
C
plus
plus
Library.
It
is
standard
for.
A
If
it
is
not
possible
to
drop
this
Library,
we
could
we
should
upgrade
to
the
newest
version,
because
we
are
using
four
years
old
version:
yeah,
it's
not,
which
is
not
good.
A
H
C
A
Know,
oh
a
lot,
I
think
that
there
is
tons
of
issues
well
to
fix,
I
I
suppose
we
could
separate.
We
could
split
this
into
smaller
PRS
like
everything
except
the
open,
Telemetry
Library,
which
will
be
pretty
weak
and
then
cover
tests.
Also
by
by
this.
H
H
I
So
the
poor
Entity
framework
like
so
do
you
also
get
the
metrics?
Also
guys
I
mean
I
know
we
are
concerning
on
tracing,
so
don't
we
get
some
metrics
I
mean
I.
Can
I
can
update
what
metrics
I'm
talking
about.
I
I
There
is
a
way
you
can
get
metrics
out
of
it.
Basically
number
of
connections
and
all
those
things
active,
DB
context,
those
kind
of
stuff
so
I
I,
have
to
see
like
how
we
can
get
it.
A
H
A
And
I
saw
that
there
is
one
h
o
release,
so
it
will
be
great
to
check
if
they
included
native,
open
Telemetry
support.
H
C
I
think
this
is
in
in
general.
A
good
idea
on
your
observation
is
regarding
they
start
up,
because
that
is
embedded
as
a
resource
that
we
need
for
the
in
the
dotnet
framework
embedded.
What
I
mean
just
to
be
be
clear:
the
the
dll
it's
embedded
as
a
resource,
because
that's
the
way
that,
on
the
dotnet
framework
that
we
inject
the
bytes
of
the
startup
dll,
so
the
messages
of
that
one
should
be
kept
in
the
same
dll.
C
Otherwise,
it's
not
gonna
work
so
because
of
the
the
framework
that
module
needs
to
be
treated
especially
one
kind
of
unrelated
thing.
That
came
to
my
mind,
right
now
saying
that,
because,
in
the
past
the
we
embedded
as
as
a
resource-
and
we
use
it
on
Linux
2.
But
nowadays
since.
G
C
But
the
thing
is
I
think
I
think
you'll
not
be
possible
right
now.
Okay,
this
inquired.
I
I
I'm
in
the
I
I
was
just
looking
at
the
code
and
probably
that's
where
I
see
that
something
you
can
do
better
I
mean
it
doesn't
need
to
be
like
in
any
of
the
release
until
we
do
GA
but
like
when
we
are
doing
gmab
by
that
time,
I
think
we
should.
I
H
I
Think
this
would
actually
improvise
like
the
core,
like
whoever
is
looking
at
the
code,
they
would
say
that
okay,
this
is,
this
is
good.
A
G
G
A
H
Two
issues:
those
two
shoes
will
come
together,
basically
I'm,
not
sure
I
would
want
to
start
working
as
soon
as
possible.
But
right
now
we
have
a
failure
on
Main
Etc,
so
I
just
prioritize
these
things
over
this
one
right.
C
Robert
is
it
possible,
take
a
look
if
it's
possible
to
remove
the
plug-in
code
from
the
main
app
make
it
kind
of
a
library,
because
right.
H
To
be
honest,
to
be
honest,
I
think
of
a
one
plugin,
but
only
one
plugin.
That
would
basically
add
SQL
text
statements,
which
is
not
by
default,
enabled
because
there
was
even
a
ask
for
it,
and
at
least
we'll
have
some
tests
for
a
plugin.
What
do
you
think.
C
H
H
A
We
have
three
or
two
more
issues
related,
so
you
know
Five
One
release
it
is
upgrade
to
Auto,
but
still
we
are
waiting
for
the
official
release
and
fixing
the
filing
drops
related
to
the
Powershell
and
social
scripts.
A
I
Yeah
I'm
just
I'm
just
thinking
whether
I
should
go
for
it
or
like
you,
guys,
are
like
when
this
needs
to
be
fixed
by
the
way.
F
F
We
we
don't
know
why
that
configuration,
abstractions
was
library
was
not
present
or
either
we
had
a
lower
version
or
something
that
is
something
that
we
need
to
investigate
it.
It
is
like,
on
the
same
lines
of
the
open,
telemetry.dll
issue
that
we
had
it
earlier
now.
F
It
is
happening
with
some
other
Library,
so
I
can
take
a
look
at
it
because
I'm
working
on
this
theme
already
like
to
see
how
we
can
generalize
this
kind
of
exception
which,
because
even
it
is
getting
difficult
for
us
to
identify
even
I
on
the
other
error
message.
I
spent
a
lot
of
time,
debugging
and
figure
out
that
is
open,
Telemetry
dll,
which
is
of
logo
version
in.
D
F
That
I
think
we
are
trying
to
instrument
the.net.exe.
Also
here
I,
don't
know
like
that
could
be
the
issue.
It
tried
loading
and
crashing
there.
Let's,
let's
take
a
look
into
it.
These
are
the
things
we
need
to
try
and
fix
it
and
or
ensure
that
we
have
the
right
message
for
us
to
troubleshoot
or
even
the
customers
to
troubleshoot.
A
F
I'm
sorry
that
I
joined
little
late
because
some
like
contractor
has
come
and
I
had
to
just
be
away
for
that
time
like
there
are
two
things
I
wanted
to
discussion,
discuss,
one
I
had
put
it
one
I
did
not,
and
I
saw.
The
similar
discussion
happened.
F
The
other
one
is
about
exception,
crashing
the
process
or
continuing.
Last
week,
Robert
touched
upon
that
point.
We
did
not
dig
in
depth
on
it
now.
If
there
is
an
exception
during
the
instrumentation,
we
just
throw
and
crash
the
process
and
that's
the
design
decision.
We
have
it,
but
the
other
spaces
like
Java
and
the
open
Telemetry
SDK
completely
for
our
different
principle.
They
don't
crash
the
process.
So
it's
not
that,
like
we
need
to
discuss
something
now.
F
This
is
something
we
need
to
take
into
consideration
before
RC
release
or
something
what
What's
a
plan
on
that.
Should
we
align
with
the
other
like
the
spaces,
or
should
we
continue
doing?
What
we
feel
is
the
right
thing
for
customers.
From
the
exception
perspective,
the
second
topic
is
the
the
plan
for
nuget
package.
I
was
out
for
a
few
days,
but
I
saw
some
conversation
on
this
one
to
make
it
as
a
public
package
just
wanted
to
bring
up
the
conversation
and
get.
C
So,
let's
start
with
the
the
configuration
I
think
as
we
discussed
in
general
before
we
try
to
fail.
C
H
C
H
Honest
to
be
honest,
I'm
not
sure
if
that
makes
any
more
sense,
because
the
SDK
already
is
changing
their
mind
and
will
will
call
will
call
like
if
there
will
be
a
bad
configuration
for
some
environmental
variables
which
are
handled
by
the
SDK.
Then
they
will
handle.
They
will
set
the
default
instead
and
they
will
just
publish
warnings
to
the
less
events.
H
I
I,
actually
I,
I,
I,
didn't
I,
didn't
understand,
I
think
what
what
Java
is
doing
like
are
they
like?
Let's
say
one
library
or
one
one
jar
file
is
not
instrumented,
they
ignore
it
and
they
continue
with
the
next
jar.
Is
that
what
they
do,
and
that
is
something
that
we
don't
do
yeah.
F
Even
like
in
dotnet,
it
is
even
much
more
than
that.
Java
does
not
have
few
issues.
For
example,
we
have
like
dotnet
as
a
specific
issue.
If
there
is
a
lower
version
of
Library
available
with
the
application
and
if
Auto
instrumentation,
we
bring
a
higher
version,
we
tend
to
crash
the
process
in
the
startup
hook.
So
at
this
point
we
are
crashing
instead
of
checking
and
logging
it
we
just
log
and
crash
the
process.
F
We
can
just
continue
and
move
on
instead
of
scratch,
scratching
the
process
saying
that
hey
this
is
not
supported.
We
will
log
that
and
we
can
continue
the
process
to
continue
its
execution.
So
that
does
not
happen
at
this
point.
F
We
also
had
a
discussion.
This
can
be
controlled
with
an
environment
variable
like
we,
you,
we
provide
a
like
a
fail
fast
or
something
like
that
environment
variables.
Based
on
that,
we
can
take
a
decision
so
that
anyone
have
an
ability
to
control
the
exception
instead
of
following
one
model:
okay,.
I
F
H
F
C
Thing
so
so
I
think
that
in
in
a
sense
there
are
ideal
solution
is
to
have
this
configurable
in
the
end,
you
know
if
we
are
looking
for
a
solution
that
serves
everyone.
In
that
sense,
we
would
like
to
have
the
option
to
say:
hey
fail
fast
or
not,.
A
H
C
Yeah
but
but
Robert
you
said
that
then
the
SDK
is
not
going
to
free
you
anymore,
oh
misconfiguration.
They
are
changed
even
for
enumerations.
H
C
Yeah
I,
I,
I
I
still
think
that
in
general
Observer
here
it
should
be
considered
a
feature
if
it's
down
it's
down.
The
service
is
down,
but
I
understand
that
there
are
people
when
users
that
have
different
Vision
I
think
then
we
are
gonna
have
to
try
to
work
this
in
a
way
that
we
can
make
configurable.
And
then
we
have
to
start
that
on
the
SDK
level,
then.
A
C
Wanted
to
talk
about
the
nuget
package,
because
Raj
mentioned
that
before,
but
I
think
we
have
to
separate.
We
have
a
new
get
package
that
is
the
one
that
Raj
created.
That
has
the
dependence
that
you
can
use
just
so
if
they
start
up
hook,
and
we
also
want
to
have
a
nougat
package
that
should
include
those
dependencies
plus
the
crab
that
can
be
used,
then
for
that
that.net
framework.
C
So
we
have
to
have
a
plan
for
these
new
get
packages
in
a
way
that
kind
of
we
can
start
publishing
that
one.
But
we
also
have
to
think
about
names
and
platforms
to
add
those
orders.
You
know
kind
of
I,
don't
know
how
to
call
with
Native
profilers
or
bundle
I,
don't
know
distribution,
something
like
that,
but
it
seems
that
it's
a
good
idea
to
have
a
plan
specifically
for
the
new
get
package.
C
F
If
I
remember
correctly,
the
current
new
get
package
handles
both
dot,
NET,
Framework
and
Dot
net
core.
We,
we
bundled
the
CLR
profiler
libraries,
also
in
it
even
for
the
Linux
bills,
also
well.
C
I
I
look
at
that
the
package-
maybe
maybe
it
was
I-
look
at
some
time
ago,
but
I
look
at
the
package
didn't
have
the
content
files,
so
it
was
not
bringing
those
files
with
the
application.
Maybe
I
look
at
when
you
were
doing
the
the
initial
version,
but
I
didn't
see
the
the
content
files
there.
F
It's
the
the
entire
Tracer
home,
whatever
we
had
that's
the
same
content,
we
have
it
in
that
nuget
package.
Additionally,
it
includes
the
the
like
Linux
and
the
Alpine
version.
If
I
remember
on
it,
who
makes
everything.
C
F
Yeah,
sorry,
anyone
who
wants
to
look
at
the
nuget
package,
it's
a
part
of
any
PR
that
runs
so
you
will
find.
If
you
go
to
the
build
results,
you
will
find
the
nuget
package
to
take
a
look
at
and
if
you
want
to
play
with
it
and
I
completely
agree
with
Apollo.
We
need
to
decide
on
the
name
if
we
are
like
going
to
do
a
public
release
and
where
it
is
going
and
how
it
needs
to
be
published
and
everything
like
I
think
we
should
have
a
steps
for
these
going
forward.
F
I
think
most
of
the
this
is
one
of
the
publishing
model
which
will
resolve
many
of
the
issues.
The
recently
we
are
running
into
it
like,
even
if
maybe
the
other,
the
open,
Telemetry
lower
version
of
used
in
the
project
or
the
one
that
is
like
notified
in
our
issue.
So
similar
issues
can
be
easily
sorted
out
with
the
help
of
this
new
cat
foreign.
D
C
So
sorry,
I
I
went
on
that.
Did
it
go
over
what
you
just
asked
before.
C
Yeah
yeah:
okay,
are
you
getting
sync
with
Raj
and
we'll
match
the
descriptions
that
we
have
in
our
minds
that
right
now
are
different,
but
yeah.
C
All
right,
so
we
are
in
the
moment
of
silence.
If
somebody
doesn't
bring
anything
up,
it
means
that's
time
to
end
the
meeting.
Yeah.
I
I
have
one
quick
question:
if
anyone
I
I
see
lot
of
environment
variables
like
is
there,
is
there
a
plan
for
us
to
like
conceptualize
it
or
like
put
it
into
you,
know
some
sort
of
like
hierarchy
so
that
it's
easy
to
manage
I
mean
that's
what
I
have
been
thinking,
but
I
thought
I
would
just
bring
it
and
see
what
everyone
thinks
about
it.
C
Yeah
I
think
this
is
one
unfortunate
reality
of
the
way
to
configure
out
instrumentation
that
we
have
right
now
but
I
and
especially
that
a
lot
of
that
happens
kind
of
right
now
in
a
kind
of
ad
hoc
way.
Let's
say
somebody
wants
to
control
a
feature
of
some
instrumentation
and
then
we
have
to
add
on
the
instrumentation
and
then
come
back
and
kind
of
not
come
back
but
head
to
having
the
instrumentation
and
then
it's
supported
here,
I
think
more
medium
term.
C
We
should
be
looking
to
something
more
a
process
more
established
for
that
kind
of
that
Engineers.
This
configuration
path
whatever
it
is,
perhaps
the
files
that
Robert
was
talking
last
week
with
the
environment
variables,
but
right
now,
I
think.
The
best
thing
that
we
can
do
on
our
side
is
try
to
centralize
and
have
some
place
that
you
can
easily
see
every
environment
variable
at
least
that
we
created.
We
already
have
the
problem
of
the
ones
that
are
supported
via
the
instrumentations
and
SDK
that
in
principle
they
don't
have
visibility
in
our
repo.
C
They
are
maintaining
there.
So
it's
already
a
kind
of
abstraction
that
people
need
to
be
aware
about
all
the
possibilities
but
I
think
more
medium
term.
We
want
to
kind
of
somehow
to
find
a
centralized
way
to
handle
those
here
right
now.
I
I
just
think
that
having
a
central
places
where
we
can
start
I
do
I'd
love
to
hear
other
folks
thoughts
about
this
because
I
think
it's
a
it's
a
it's
a
really
challenge
for
us
right
now.
I
I
was
I
was
yeah
along
with
Central
documentation
like
so.
Do
we,
like
think
about
more
of
your,
like
hierarchy?
Kind
of
structure
like
you
set
up
a
configuration
file
or
Json
file
and
from
there
like
our
instrumentation,
would
like
set
up
those
environment
variables.
I
Something
like
you
read
something
from
a
configuration
or
a
Json
file,
and
then
you
set
up
an
environment
variable
so
that.
F
Did
you
visit
our
config.md
file
by
any
chance?
It's
all
the
environment
variables
are
documented
there
in
in
some
structure,
I,
don't
know
whether
it's
in
the
docs.
Maybe
you
need
to
visit
that
once
and
see
if,
like
how
you
want
to
get
that
changed
and
if
you
could
propose
a
change
to
that.
That
would
be
useful,
like
what
you're
asking
sure.
F
Yeah,
but
all
of
the
environment
variables
used
by
US
will
have
an
underscore
R2
in
it,
like
whichever
the
auto
instrumentation
uses
and
all
of
them
are
documented
in
that
page
and
I
feel
we
follow
some
hierarchical
like
saying
tracing
exporter,
however,
it
is
we
have
a
topic
and
listed
out
in
that
way.
F
So
if
you
are
thinking
of
a
better
way
to
improve
that,
that
would
be
helpful
for
everyone
else.
Yeah.
F
A
Yeah,
if
you
have
some
ideas,
it
should
be
great
to
discuss
it
with
Fabricio
is
his
dog
documentation
engineering
plant,
but
he
is
pretty
active
in
our
repository,
so
please
ping
him
and
he
will
be
the
the
best
person.
In
my
opinion,.
C
Yeah
one
one
thing:
it's
kind
of
obvious,
but
I
I
think
it's
worth
to
mention
in
that
regard.
We
need
a
few
environment
variables
for
the
setup,
but
ideally
our
users
should
our
default
should
be
such
that
they
should
need
to
touch
very
few
of
their
gazillion
of
environment
variables
that
we
offer.
You
know
we
offer
for
the
flexibility
or
somebody
asked
it,
but
ideally
we
should
be
striving
to
have
very
good
defaults.
C
That's
the
key
here
from
the
usage
point
of
view,
I
think
biraj
questions
have
have
a
very
good
point
about
also
how
we
can
ensure
in
our
process,
build
and
stuff
how
we
don't
use
any
environment
variable
from
the
documentation
from
both
sides.
For
now
we
rely
on
our
reviews,
you
know,
but
whatever
we
can
do
to
automate
and
improve
that
yeah
it'd
be
great.