►
From YouTube: 2022-03-02 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
A
B
In
case
you
missed
the
recording
from
last
week.
I
believe
I
blamed
you
for
something.
I
don't
remember
what
I
I.
C
C
E
G
D
D
First,
one
is
for
executing
integration
tests
for
net
force
to
point
in
ca
pipeline.
I
also
added
comment
that
maybe
we
could
create
some
kind
of
job
matrix
that
basically
will
execute
the
four
different
frameworks
in
parallel,
but
it's
just
nice
to
have
by
the
way
was
it
only
about.net
framework
which
were
not
executed
for
integration
tests
or
there
were
more.
D
Okay,
I
will
start
working
on
on
misintegration
tests,
etc.
So
I
also
try
to
keep
an
eye
on
it
and
I
will
add
it
to
the
beta
okay.
D
So
yeah
so
basically
one
of
the
tests
uses
web
requests,
which
is
deprecated
and
there's
a
proposal
to
use
http
client.
I
took
a
quick
look
and
it
was
not
straightforward
because
the
api
is
totally
different
and
basically
they
will
detail
at
least
maybe
even
few
hour
factoring
or
something
like
that.
D
F
Yeah,
just
one
thing:
we
in
a
sense
we
perhaps
wanted
to
keep
instrument
those
for
longer
than
that.
So
it's
not.
F
B
G
D
D
Okay,
so
so,
post
post
one
post,
rc
right.
D
B
Yeah,
so
I
took
a
look
at
the
unhandled
exception
stuff
and
yes,
we
can
flush
things
that
are
already
in
the
in
a
processor,
but
even
then
there
can
be
a
timing
issue
where
the
app
might
still
shut
down
before
we
can
flush
everything
in
there.
B
It's
just
edge
cases,
and
then
oh
there's,
there's
some
other
edge
cases
around
that
I
I
threw
some
sample
code
out
there
that
we
won't
be
able
to
handle.
So
those
are
all
things
that
would
make
sense
to
to
go
into
some
sort
of
documentation
explaining
how
telemetry
could
be
lost,
but
it's
not
going
to
be
a
comprehensive
list.
B
B
I
mean
I
normally
like
to
to
have
tests
for
for
any
code
changes
that
I
make,
but
I
feel
like
with
the
state
of
the
integration
tests.
B
I
almost
want
to
break
that
rule
of
thumb
and
handle
the
test
differently,
because
I
still
got
to
think
more
about
whether
or
not
we
want
this
to
be
a
separate
test
app
or
if
it
would
make
sense
to
start
building
on
to
kind
of
our
and
a
simple
console
test
app
where
we're
able
to
run
multiple
tests
using
that
single
console
app
something
that
we've
talked
about
before.
F
So
so
this
this
is
also
kind
of
you
already
mentioned
that,
of
course,
it
is
not
possible
to
call
over
our
cases
so
kind
of
it's
a
best
effort
done
when
the
application
is
shutting
down
right,
so
correct,
yeah,
so
improving
and
documentation,
I
think,
is
the
best
that
we
can
hope
in
the
short
run,
and
I
I
agree,
this
kind
of
integration
test
specific,
for
this
is
kind
of
a
bit
on
the
edge
case.
It's
not
as
simple
as
a
unit
test
right.
F
You
want
to
break
the
app
and
be
sure
that
it's
doing
the
the
right
stuff.
So
I
think
it's
very
reasonable
what
you're
set.
B
Make
sense,
okay,
so
I'll
go
and
split
up
that
work
into
multiple
pr's.
B
So
that's
it
for
this
issue
and
oh
and
yeah
robert.
I
had
responded
to
to
some
of
your
questions
about
the
finalizers
and
yeah
from
what
I've
read.
There's
just
there's
a
lot
of
implementation,
specific
details,
depending
on
the
runtime
and
possibly
even
the
the
host.
So
yeah
there's
no
guarantees
there.
F
I
I
always
find
the
life
of
the
robert's
kids
interruptions.
I
I
like
their
voices.
G
D
Okay,
next
one
rasmus
have
you
created
the
issue
here
on
the
run.
D
B
Yeah
so
there's
one
follow-up
thing,
so
I
know
that
we
just
merged
the
pr
for
the
net
six
and
running
some
of
the
integration
tests.
I
think
the
startup
hook
is
still
a
special
case
as
far
as
which
frameworks
it's
running
against,
and
so
I
don't
know
if
the
startup
hook
test
was
updated
for
net
six.
D
I
do
not
remember,
if
do,
will
you
check
double
checklist
or
do
you
want
me
to
do.
G
Also,
there
is
another
small
issue
in
the
status
quo,
so
I
recently
noticed
that
dot
net
core
2.2
also
has
support
for
the
startup
work,
but
in
the
runtime
everywhere
it
is
called
as
net
3.1
onwards.
They
have
this
startup
hook,
but
if,
unfortunately,
if
a
customer
comes
and
tries
this
out,
it
might
crash
them.
Then,
in
that
itself,
probably
we
should
have
a
check
in
our
like
startup
hook
to
check
whether
it's
dotnet,
2
or
not.
B
D
B
Now
yeah,
I
think
the
native
tests
are
only
gonna
run
against
dotnet
framework.
B
G
It
is
like
unsupported
yeah,
we
are,
people
does
not
even
know
about
it.
We
just
come
and
hook
in
something
to
dotnet
to
and
we
we
just
crash
it.
So
so
I
have
seen
it
in
one
of
the
scenarios,
so
we
have
it
for
our
internal
implementation,
so
I've
seen
customers.
It
is
a
very
difficult
issue
to
troubleshoot
because
they
will
not
even
have
a
proper
stack
to
identify.
This
is
where
it's
coming
from.
It
will
be
some
stack
in
the
dot
net
area.
It
will
throw
up.
G
So
so
it's
better
to
handle
that,
even
if
some
guys
like
run
into
it,
they
won't
be
able
to
figure
it
out.
It
is
the
auto
instrumentation
which
is
causing
it.
So
it's
better
to
leave
it's
a
one
or
two
line
of
code,
but
I
think
we
are
good
to
have
that,
but.
B
Yeah
yeah.
No,
it
definitely
sounds
like
it's
easy
enough
to
do
that.
It's
worth
it
yeah
yeah.
I
was
just
more
curious
about
customer
uptake
on
different
versions
because
at
least
on
the
new
relics
side,
dot-net
core
2.2-
is
not
really
used
anymore.
G
It's
not
even
used
it
in
microsoft,
but
still
there
are
people
external
customers,
they
have
it
in
their
like
vms
or
in
the
app
service
space,
even
though
it
is
supported,
they
still
run
on
it
and
a
single
button.
Click
will
come
and
integrate
everything
for
that.
That's
when
they
don't
even
know
that
this
is
what
is
causing
the
issue.
The
app
just
goes
down.
G
And
I
have
another
update,
also
like
I
said
I
will
try
out
the
the
other
custom
exporter.
How
does
it
work
without
plug-in
and
everything?
So
I
could
take
the
azure,
monitor
exporter
and
tried
it
with
the
like
all
right
documentation.
Everything
worked
and
no
issues
at
all
anywhere,
so
I'll
be
taking
a
learning
from
there
and
like
do
the
document
update,
so
it
will
make
it
easier
for
people
who
I
want
to
try
it
out
in
those
kind
of
scenarios.
G
It's
not
only
the
exporter.
We
have
the
control
for
to
hook
up
anything
over
that
with
that
plugin.
B
So
I
I
ran
into
something
interesting
when
I
was
working
on
a
test
app,
and
so
I
just
did
a
simple.net
6
console
application
and
run
it
in
a
linux
docker
container
and
tried
to
instrument
it
with
with
our
project
and
so
with
dotnet
6.
I
figured
all
of
the
dependencies
should
be
up
to
date,
but
when
I
try
to
use
the
otlp
http
exporter,
I
was
getting
errors
with
trying
to
load
the
logging
abstractions
assembly.
F
Was
that
related
to
that
issue
that
you
identified
quite
some
time
ago
about
the
dependencies
right,
because
the
package
switch
the
way
that
they
build
instead
of
referencing
the
package
directly?
So
they
don't
bring
that
stuff
anymore.
Yeah
that
I
didn't
expect
to
happen
on
dot.
B
B
So
I
haven't
fixed
the
issue
but
to
work
around
it.
I
just
updated
my
application
to
reference
that
specific.
G
If
that
is
going
to
like
cause
an
issue,
these
are
the
issues
like
we
should
identify
and
update
our
like
additional
devs
and
the
chat
store.
So
even
though,
if
it
misses
from
the
machine,
it
should
come
and
pick
it
up
from
there.
So
this
could
be
the
next
candidate,
which
is
eligible
to
be
a
part
of
the
additional
depths
for
us.
B
F
Yeah
that
may
be
worth
opening
a
issue
for
for
us
to
not.
B
Yep
I
can,
I
can
open
that
and
I
can
even
reference
the
error
message
that
I
was
seeing.
F
They
changed
it
to
have
a
reference
to
say
that
it
was
the
web
sdk,
basically
asp.net
or
when
they
changed
that
we
start
to
have
this,
because
before
we
are
shipping,
the
sdk
was
shipping
the
the
assembly
itself.
Then
it
starts
to
build
just
with
the
reference
to
say:
hey,
I'm
a
asp.net
core
application,
so.
F
Score
yeah.
I
think
I
think
that
sounds
like
a
a
reasonable
thing
for
us
to
tackle
before
any
any
attempted
release
or
anything.
You
know.