►
From YouTube: 2022-04-06 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
B
B
B
B
B
E
Nice
but
but
but
to
be
fair,
I
don't
know,
I
don't
know
it
is
not
as
bad
as
the
bout
scary,
but
I
can
tell
you
that
yesterday,
look
at
the
forecast
was
saying
that
there
was
a
chance
of
snow
on
sunday.
E
Okay,
I
think
yeah
chris
chris
can't
join
us
today.
He
sent
me
a
message,
so
I
think
you're
ready
to
start.
If
you
want
to
drive
robert.
B
Yes,
I'll
share
my
screen
all
right,
so
let's
go
first,
please
yeah
feel
free
to
put
your
names
here,
but
you
can
do
it
always
afterwards,
if
you
prefer.
B
So,
let's
go
quickly
through
the
items
here,
pull
request,
18!
So
first
question:
is
there
any
open
pull
request
that
anyone
wants
to
discuss.
B
B
F
F
F
So,
basically,
what
I
try
there
currently
I
have
a
load
from
file.
Then
I
use
the
assembly
load
context.
F
Default
and
load
from
stream
so
basically
giving
in
the
file
stream-
and
this
should
actually
load
it
to
the
default
assembly
load
context
instead
of
the
file
load.
Context
that
is
completely
separate
from,
but
still
the
type
get
type
is
not
working.
F
Yeah,
basically,.
E
But
assembly
load
context:
oh
no!
Okay!
So
now
I
see
the
the
if
f,
on
top
what
you
are
doing
framework
just
for
me
to
have.
E
You
mean
yeah,
I
mean
because
this
is
net
core,
okay
yeah.
This
is
it.
Is
it
working
for
framework
just
it's
a
side
kind
of,
but.
F
Yeah,
it's
definitely
working,
but
just
the
question
is
in
this
second
time:
if
it's
a
loading
thing,
it
should
be
like
a
faster
option,
because
the
type
should
be
already
in
default,
load
context
or
default.
G
Domain,
so
that
is
the
first
time
load
here,
because
I
see
one
load
file
right.
So,
if
you
did
you
change
that
to
load
from
or
something
like
that.
Instead
of
note.
F
C
B
A
D
Also
apollo,
to
answer
your
question:
on.net
framework
we're
not
running
this
instrumentation
right
now,
the
above
you'll
see
some
grade
out.
If
def
we
just
return
yeah.
F
The
problem
there
is
share
the
assembly
store,
so
basically
just
looking
for
the
library
there,
because
the
assembly
store
is
kind
of
working
already,
and
it's
very
easy
to
include
this
library.
There.
G
G
If
you
use
a
load
from
context
here,
we
may
run
into
an
issue
later.
If
customer
also
brings
a
similar
file,
we
will
be
in
the
same
dependency
hell
problem
and
we
might
crash
out
when,
when
you
use
a
load
from
context
because
we
are
trying
to
load
before
the
customer
load,
whatever
the
line
number
you
have
in
64
for
the
first
time
whenever
it
executing
the
code,
you
will
always
get
null
because
application
was
not
initialized
at
all.
At
that
point
in
time,.
F
G
B
D
F
F
E
Yeah,
but
but
before
the
the
optimization,
the
the
the
question
that
I
have
in
back
in
my
mind,
sorry,
I've
been
kind
of
late
on
this
pr
is
able
to
work
right
because
I
think
after
you
get
to
the
point
that
the
assembly
is
loaded.
E
E
G
E
Yes,
we
want
to
avoid
calling
for
nothing,
but
my
my
first
question
is
kind
of:
is
it
working
in
the
in
the
let
us,
let's
assume,
that
make
useless
calls
after
load
the
first
time,
but
it's
working
that
case
yeah.
F
Also
work
for
a
document
framework
right.
No,
because
the
how
the
dependency
packing
is
working
is
through
the
additional.
I
think
it
was
this
one
here.
E
But
should
it
shouldn't
be
the
case
that
when
we
we
have
assembly
load
failures,
the
handle
should
also
look
at
the
location,
not
the
environment,
variable
but
the
location.
E
We
we
do
inject
the
handle
right
when
the
assembly
fails
to
load
and
then
we
could
add
the
same
folder,
so
kind
of
gets
independent
of
the
net
shared
store.
E
Yes,
that
that's
what
I'm
saying
so
we
also
have
the
the
that
handle
that
handle
red
checks
in
the
additional
depth
that
we
ship
right.
E
So
the
the
dotnet
shared
store
is
kind
of
more
compliant
helps
with
the
stuff.
But
my
compliance
for
the
net
core
doesn't
work
for
the
net
framework,
but
we
should
still
be.
E
So
let
me
step
back
one
thing
here:
why
we
need
a
specific
code
to
load
the
assembly
here
shouldn't
be
how
this
on
the.
E
Yeah,
I
I
mean
to
get
the
type
and
do
the
stuff
needs
to
be
specific
to
each
implementation
right,
but
handling
the
load
itself
of
the
assemblies
should
be
a
common
code.
G
I
agree
with
apollo
here
and
again,
as
you
are
using
the
additional
devs,
you
may
not
even
need
the
code,
only
the
line,
number
64..
Whenever
you
call
the
line
number
64,
you
should
always
get
a
type,
because
it
knows
the
probing
part
that
it
has
the
library
in
that
location.
G
We
don't
need
to
have
look
whatever
you
are
doing
from
67
to
70.
That
is
not
needed.
That
is
done
by
the
dotnet
itself
for
us,
so
64
should
always
have
a
value
whenever
we
are
using
additional
depth,
because
it's
in
the
probing
path.
F
E
F
E
Sorry,
I'm
I'm!
I
I
really
not
being
really
close
to
to
this
pr,
but
but
then
pressure
it.
It
can
fail,
but
our
the
the
handle
that
we
we
add.
So
when
we
fail
to
load,
then
we
do
the
code
to
search
and
load
the
specific
version
right.
The
specific
assembly
that
code
should
be
common.
F
I
I
think
I
already
got
it.
I
raj
was
mentioning
the
additional
maps,
but
I
was
including
share
store,
but
I'm
not
sure
if
it
was
included
in
additional
dev
station.
So
if.
G
The
library
correct
it
should
be
part
of
two
things.
One
is
in
the
additional
tips.
Json
it's
manually
built
it's
in
the
build
steps.
I
believe
you
need
to
include
it
that
and
then,
if
you
have
it
in
the
store,
then
only
it's
gonna
work
and
look
into
that
probing
path.
G
F
F
D
F
E
So
we
already
packed
the
dll,
so
our
we
need
to
look
at
that
code
why
it
will
fail.
It
can
fail
for
some
unknown
reason
you
have
to
handle
that
on
the
instrumentation,
but
any.
C
E
Yeah
but
but
we
can
still
in
our
probing
path
there
on
the
handle
we
can
kind
of
hey.
We
are
in
the
framework.
We
want
to
look
at
this
location
here
and
try
to
load
from
there.
B
B
B
F
When
time
permits
I'll
be
addressing
paolo
comments,
probably
tomorrow
yeah,
that's
that's
all.
B
Okay
and
now
I
have
a
question
about
this
bumping
prs:
does
anyone
have
time
to
look
at
them
and
do
we
have
any
strategy
which
of
them
should
be
bumped
and
should
not
be
bumped
because
from
we
have
four
examples?
I
think
that
in
examples
we
in
theory
could
be
able
to
bump
these
ones
unless
they
are
explicitly.
B
E
B
A
Robert,
I
think
we
should
split
the
issues,
everything
what
is
related
you,
which
is
used
in
the
test
folder.
I
think
it
should
be
able
to
be
bumped
yeah.
A
E
E
So
so,
for
the
examples
yeah
seems
real
simple:
we
can.
We
can
bump.
B
B
E
Yeah,
so
I
was
looking
at
at
mateo's
pr
and
also
remembering
what
I
talked
last
last
time
that
I
need
to
debug.
I
think
adding
the
code
to
other
event.
Listening
is
something
that
we
should
look
before
the
beta,
because
it's
going
to
help
our
life
to
debug
itself.
We
can
just
hey,
send
us
the
the
logs
you
know,
without
that
we
have
to
go
to
that
steps
to
enable
to
the
sdk
kind
of
to
get
the
event
source
and,
I
think,
makes
harder.
So
I
okay.
B
E
B
Further
you
we
can
read
it
later.
We
can
put
it
to
our
scene,
but
what
would
you
rc,
but
we
can
discuss
it
later?
I
think
it
might
be
needed
you
know
even
for
sake
of
knowing,
then
you
know
the
versions
of
the
diagnostic
source
etc.
So
I
put
it
here
right
now
right
now.
E
Really,
I
I'll
not
say
that's
a
blocker,
but
I
remember
no
reviewing
that
stuff
when
chris
added
and
it's
there
by
default,
I
think
we
really
should
disable
by
default,
it's
better
at
our
stage
to
kind
of
lose
some
traces
that
shut
down
then
have
this
the
application
hanging
you
know,
so
I
think
either
we
you
don't
know,
my
preference
will
be
to
add
a
configuration
to
enable
handling
the
shutdown
but
by
default
is
off
because
I
I
didn't
have
time
to
investigate.
E
E
E
E
It's
I
it's
just
that.
I
think
that
is
a
rather
exceptional
handler,
if
I
remember
correctly,
but
instead
we
just,
I
think
we
log
a
message.
That's
very
generic
and
let.
B
E
I
don't
think
any
of
them.
I
I
don't
think
defects
on
the
product.
I
I
opened
a
few
of
them,
but
they
didn't
look
like
the
facts
of
the
product.
E
There
was
one
that
was
about
a
port
already
been
using,
so
perhaps
something
that
test
infrastructure
that
we
would
improve,
but
in
general
I
I
didn't
find
any
flake
test.
That
was,
that
was
a
product
defect.
E
B
E
So
I
I
I
think
we
have
to
point.
I
don't
think
we
have
time
to
do
this,
putting
especially
the
timeline
that
you
are
trying
to
to
reach.
I
think
we
have
to
point
this.
B
E
Even
the
one
from
the
application
is
not
working.
I
didn't
look
why
I
didn't
investigate
that.
Perhaps
we
can
limit
to
investigate
for
the
example
application
to
get
the
example.
Application
working
you'll
know
that
on
the
example,
application
are
kind
of
cheating,
because
the
plugin.
B
E
B
B
B
F
B
G
B
I
propose
to
put
it
into
rc,
because
then
I
double
checked
the
specification
and
I
think
it
was
something
like
there
should
be.
Headers
not
must
be
headers.
As
far
as
I
remember.
Maybe
I
can
find
it
quickly,
but
I
don't
yeah.
B
F
F
B
To
test
containers,
this
is
definitely
a
pro
post
format.
B
B
E
Yes,
so
so
the
issue
that
russia
is
going
to
investigate
is
related
to
that
all
right,
so
so
from
what
is
in
progress
is
because
you
are
still
missing
the
configuration
right,
because
chris
added
the
the
handler-
and
I
think
the
only
thing
missing
was
the
configuration
to
enable
disabled,
but
we
wanted
this
to
be.
Actually,
we
wanna
very
likely,
let's
wait
for
a
large
investigation,
but
very
likely
we
wanted
this
to
be
opting
by
the
time
that
we
ship,
I
suspect,.
B
We
will
see
what
will
come
from
the
investigation,
so
david
is
in
progress
of
creating
integration
tests,
so
he'll
try
to
make
us
basically
a
big
smoke
integration
test
that
will
cover
a
lot,
a
lot
of
cases
mata,
which
is
also
working
on.
So
it's
kind
of
the
the
changes
in
the
logs
basically
is
about
about.
If
it's
about
to
fix
this
issue-
and
here
I-
and
here
is
something
that
I
want
to
gather
with
mature
to
discuss,
I
think
so.
When
I
created
this
issue,
we
thought
that
basic.
B
So
if
they
are
in
the
parent
folder
the
props
file,
it
is
isn't
it
isn't
bumped,
and
even
if
we
put
the
path
to
the
location,
where
was
the
when
the
props
directory
built
props
fire
was
located,
then
the
panda
boat
was
complaining
that
there
is
no
project
fight
there
so
baby.
Basically,
there
are
two
possibilities
that
we
could
use.
We
could
create
some
dummy
cs
pro
for
pro
cs
profiles
or
contribute
to
the
panda
bot.
B
I
have
assigned
before
the
meeting
myself
to
end
user
documentation.
Basically,
when
I
was
working
on
some
stuff
before
I
have
already
covered
a
lot
of
things
here,
I
think
that
installation,
usage
instructions
are
updated.
Troubleshooting,
I
think,
is
also
up
to
date.
B
I
will
ask
our
technical
writer
to
review
this
stuff,
because
I
think
that
he'll
have
a
lot
of
comments
here,
supported
environments,
os
frameworks,
etc
is
covered
list
of
provided
instrumentation
libraries.
Here
I
have
a
proposal
to
remove
the
mongodb
because
of
the
issues
that
we
have
and
just
move
the
mongodb
stuff
out
of
the
scope
of
the
beta.
We
can
work
on
it
in
the
background.
I
do
not
fixing
any
problems,
but
I
do
not
think
we
need
the
commitment
to
fix
it.
B
B
B
B
G
Not
yet
I
have
a
meeting
with
like
noah
today,
so
I'm
going
to
take
a
discussion
with
him
before
coming
with
with
any
proposal
over
here
last
week
we
decided
to
put
a
design
document
or
a
one
pager
in
a
repo
like
listing
the
design
challenges
and
how
we
can
tackle
it.
So
I'm
internally
working
with
noah
to
come
up
with
that.
F
So,
basically,
for
dotnet
core
at
least
one
of
my
days
is
to
use
custom,
load
context
and
reused
ability.
If
the
library's
version
is
the
same
in
default
load
context,
it
will
be
the
same,
the
same
dll,
basically,
which
means
it's
the
same
instance
of
a
system
diagnostic
diagnostic
source.
E
You
how
how
we
are
gonna
communicate.
Things
like
the
current
was
updated
in
one
context,
if
it's
a
different
version,
you
know
that
I'm
not
saying
that
we
need
to
answer
now,
but
this
kind
of
question
needs
to
be
kind
of
on
top
of
our
mind
when
we
think
about
this
kind
of.
F
Problem
yeah,
so
basically
we
I'm
trying
to
address
at
the
moment
the
same
way
like
we're
doing
it
for
now
the
using
the
appeal
time
reference
just
for
this
package.
G
I'm
sorry
I
said
like
yeah.
It
would
be
difficult
with
this
approach.
Maybe
it's
a
thought
I'm
giving
like,
because
we
are
bringing
in
in
auto
instrumentation.
We
are
bringing
open,
telemetry
sdk,
which
has
a
like
a
direct
reference
to
diagnostic
source.
So
I
think
what
we
are
trying
here
is.
You
are
saying
you
are
going
to
modify
the
diagnostic
source.
So
if
you
even.
F
Not
exactly
diagnostic
source
will
be
exactly
the
same,
so
basically,
let's
say
user
needs
to
bring
in
the
version.
Six
and
sdk
brings
the
version
six.
That
means
it
will
be
the
same
instance
in
both
load
contexts.
That
means
it
will
work
in
custom
load
context
that
it
will
work
in
default,
node
context.
G
Yeah,
build
time
is
definitely
very
easier,
actually
yeah.
If
we
had
a
new
get,
everything
will
work.
So
what
I
am
trying
to
tackle
is
the
runtime
issues.
Whatever
you
have
the
diagram,
I
have
a
simple
repo
in
my
personal
repo.
I
have
that
I
have
this
like
small
demo
or
a
prototype
created
on
that
one
also.
So
what
we
do
is
if
there
is
an
activity
created
by
the
open,
telemetry
sdk
like
we
and
that
will
be
like
proxied
to
the
auto
instrumentation
brought
sdk,
and
we
try
to
track
that.
G
So
that's
how
I
have
the
prototype
done,
but
like
it's
better
to
get
the
guidance
from
the
dotnet
team
itself
on
what
we
do
and
how
we
should
take
forward.
So
probably
in
like
another
couple
of
weeks,
we
should
see
some
design
from
my
end
and
currently
how
like
we
tackle
in
application
insights
it
we
spoke
about
the
il
repack
and
to
tackle
the
diagnostic
source.
G
So
we
have
the
least
supported
version
of
diagnostic
source
dependency
with
application,
insights,
auto
instrumentation,
that's
how
we
are
supporting
all
the
across
all
the
dot
net
supported
environments.
F
Right
yeah,
so
basic!
That's
why
I'm
saying
the
customer
send
reload
context,
because
it
will
be
not
the
only
library
and
if
you
need
to
start
aisle
back
in
10,
it's
getting
very
tricky
because
there
are
will
be
a
lot
of
different
sub
dependencies
even
dependencies
of
the
dependencies,
and
I'm
not
sure
that
we
want
to
proxy.
All
of
that.
B
G
A
few
things
in
my
mind,
so
I
need
to
put
everything
together
and
probably
we
will
take
one
of
the
sig
for
a
discussion
on
it.
F
Basically,
what
I
did
I
shared
it
to
the
three
steps.
One
is
the
customer's
application.
So
basically,
this
is
simple
case
that
we
get
to.
The
google,
then
is
the
startup
for
the
custom
load
context.
So
all
the
this
is
quick
hack
to
load
the
libraries
that
the
sdk
is
actually
looking
for.
So
you
can
see,
there
are
already
a
lot
of
different
dependencies
and
in
the
startup
work,
basically,
I'm
just
creating
the
custom
load
context
and
loading
in
the
let's
say
the
client,
which
is
actually
bootstrapping
all
of
the
sdk
codes.
F
I'm
running
this
is
five
and
six
currently,
but
probably
should
also
where
3.1
also
work.
G
F
C
B
B
I
think
that
I'll
propose
to
do
it
manually
in
the
beginning,
just
copying
the
build
outputs
and
just
manually
creating
documentation,
thursday,
entire
attiring
for
linux,
the
binary,
outputs
and
probably
mac
outputs
and
zip
the
windows,
binaries
etc,
but
we
will
see,
I
will
just
propose
it
at
this
for
now
all
right.
So
that's
for
the
most
important
here
on
the
board.
B
B
All
right,
so
I
I
want
to
discuss
one
thing
I
probably
already
mentioned
it.
I
want
to
cut
some
scope
from
the
from
this
board
as
we
intended
to
have
a
release
next
week.
At
least
this
was
our
target
and
I
just
want
to
go
through
the
stuff
and
puts
change
the
minus
some
percy
for
some
of
them,
even
if
they're
in
progress.
B
I
just
do
not
want
to
have
like
full
attention
and
just
to
have
have
one
board
when
I
will
know
when
we
can
release
a
beta
so
going
from
from
this
side
a
question
to
you
guys:
do
you
think
that
integration
tests
are
a
must
have
for
better
or
not.
B
E
I
I
I
think,
because
of
that
symptom
that
I
said
I
kind
of
oh,
it's
okay,
to
have
the
unhandled
exception,
but
it
should
be
disabled
by
the
phone,
probably
or
at
least
we
should
have.
E
B
F
B
F
F
D
This
one
that
can
go
to
rc,
even
if
I
did
finish
it
soon,
it's
kind
of
a
big
change
so.
D
Yeah
yeah,
I
am
I've
started
it
just
quite
a
bit
quite
a
bit
changes.
B
I
haven't
done
anything
for
two
weeks
myself
as
well,
so
this
one
should
be
here.
This
one
should
be
here,
update
license
third
party.
To
be
honest,
it
should
not
be
big.
I
would
not
like
to
lie
regarding
licensing,
so
I
prefer
to
put
keep
it
here.
Dotnet
is
supposed
to
be,
is
supposed
to
be
removed
in
may.
I
suppose
we
can
remove
it.
E
B
B
E
A
E
Yeah
we
could
ask
morgan,
but
at
most
we
are
gonna
get
some
guideline,
you
know
I
don't
you
have
to
do
any
of
the.
If
anything
is
required
to
be
done,
we
have
to
do
the
work.
You
know.
B
E
C
G
President
apollo,
should
we
have
a
like
environment
variables
here
to
write
to
a
specific
file
or
we
are
selecting
a
location
and
we
will
write
it.
E
We
already
have
files
that
we
create
under
program
data,
opentelemetry.net
and
another
specific
location
on
linux,
so
it's
just
adding
those
messages
there.
The
only
thing
that
crossed
my
mind
is
that
perhaps
we
we
wanted
to
have
we
end
up
with
two
time
stamps,
because
there
is
the
timestamp
that
we
write
from
our
log
and
since
we
are
receiving.
E
I
think
we
should
also
log
the
message
just
in
case.
If
there
is
a
delay,
you
know
the
time
stamp,
not
the
math.
The
message,
of
course
we
need
to
log,
but
stamps.