►
From YouTube: 2021-05-19 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
A
I'm
trying
to
see
if
sharing
half
of
the
screen,
how
that
works.
How
is
looking
for
you.
D
A
A
So
so
I'm
I'm
trying
to
share
here
the
screen,
but
I'm
sharing
half
the
of
the
screen.
How
is
looking
for
you
guys.
F
A
Okay,
so-
and
I
think
david
is
also
not
gonna
join
today,
so
then
we
can.
I
was
hoping
that
maybe
dave
was
gonna
join,
but
then
I
will
send
a
message
to
him.
I
have
some
questions
about
the
vent
pipe.
A
E
Yeah
speaking
of
the
event
pipe
stuff,
paulo,
you
remember,
alan
west
he's
been
playing
around
with
that
plug-in
in
the
collector.
Oh.
A
E
With
dot
net
five
apps
and
let's
run
into
some
rough
edges,
and
so
he's
probably
gonna-
be
submitting
some
issues.
If
we
can
identify
it.
E
A
And-
and
we
are
also
missing
the
discovery
part
you
know
that
needs
to
be
implemented.
You
know
we
are
using
the
thing
that
makes
it's
good
for
us
to
experiment
right
now,
but
not
good
enough
for
people
to
use.
That's
a.
E
Yeah,
I've
recently
run
into
someone
that
was
interested
in
using
the
performance
counter
receiver.
E
A
Okay,
I'm
I'm
gonna.
Yes,
I'm
planning
to
get
the
episode,
send
a
message
to
you
and
joe
allen
so
because
I
also
have
some
some
user
interested
in
trying
that
so
I
want
to
flush
it
out,
but
I
just
wanted
them
to
try.
After
we
have
the
automatically
discovery.
You
know.
A
All
right,
I
I
put
the
the
poc
on
the
agenda,
but
if
you
guys
have
other
things
that
you
guys
want
to
talk
before
that,
because
I
robert
created
the
poc
issue
and
I
think
erasmus
did
some
trying,
I
think
the
motivation
is
kind
of
clear
it's
kind
of
preparing
for
a
dot
net,
five
and
afterwards,
and
also
to
really
because
we
had
this
idea
of
that
you'll
be
a
bunch
of
potential
problems,
but
we
don't
have
actually
a
measurement
of
the
problems
happening
in
practice.
A
You
know
we
anticipated
them,
but
we
don't
know
how
and
when
they
are
gonna
occur.
One
thing
that
our
ad
came
to
surface
on
this
is
that
the
way
that,
right
now
that
the
sdk
is
doing
versioning,
I
I
find
for
for
people
looking
from
outside
a
bit
complex
because
the
instrumentations
don't
follow
the
same
version
as
the
main
package,
and
sometimes
they
require
different
stuff.
You
know.
A
A
So
I
think
that
was
a
interesting
start
and
one
thing
that
rasmus
showed
me-
and
I
I
would
like,
if
he
could
talk
more
about
that,
is
that
he
did
some
tries
injected
the
the
sdk
and
we
saw
the
same
version
of
december,
the
same
assembly
being
loaded
with
different
versions
and
but
to
our
surprise,
everything
worked.
So
the
parents
saw
the
child
and
actually
I
I
personally
don't
understand
how
that
happened.
I
saw
the
two
modules
I
kind
of
surprised
that
it
worked
but
happy
that
it
worked.
A
So
I
kind
of
if
erasmus
would
like
kind
of
to
describe
the
tests
that
he
did
and
the
effects
that
we
saw.
I
think
it
would
be.
G
Oh,
you
cannot
share
okay,
because.
G
G
Then
you
can
see
two
open
telemetry.
G
G
G
F
So
yeah,
okay
x,
I
think-
and
this
is
on
you
said
a
donut
five
app
right.
G
This
is
dotnet
3.1.
F
Oh
okay,
yes,
I
think
the
instrumentation,
our
asp.net
core,
one
that
instrumentation
tries
to
load
system,
dynastics
4.0,
so.
G
No,
it
should
not
load
anything.
If
I
remove
the
package
reference
hotel,
the
sdk
package
reference,
then
it
should
go
away.
C
Even
though
we
are
using
two
different
packages
from
the
diagnostic
source
yep,
because
this
is
something
really
interesting
right.
A
Yeah
so,
as
I
said,
the
first
concern,
when
I
saw
this
kind
of
thing,
was:
oh
I'm
afraid
that
the
parents,
not
the
child,
is
not
going
to
be
able
to
see
the
parent,
but
the
type
is
the
same.
So
everything
working
that's
kind
of,
I
think
better
than
what
I
expected.
I
was
anticipating
problems
when
we
had
this
multiple
versions.
You
know
so.
E
C
G
G
Just
here
I
have
the
previous
session,
and
so
that's
the.
G
Here
you
can
see
that's
the
activity
source
directly
starting
the
activity
and
then.
G
F
F
E
Maybe
we
could
also
try
well
the
the
interesting
thing,
though
it's
using
the
sdk
behind
the
scenes,
and
so
the
sdk
itself
is
going
to
be
loading
a
in
this
case.
It's
probably
the
activity
listener
class
and
that's
going
to
have
to
listen
for
the
activities,
and
so
that
has
a
reference
to
activity.
E
F
Yeah,
I'm
interested
to
figure
out
who
is
yeah
referring
to
each
one,
but
it
looks
like
so
what
we're
getting
right
now
in
this
test
is
it's
all
just
you
know,
statically
from
the
app
at
compile
time
where,
like
the
app
is
creating
an
activity,
and
then
the
sdk
which
is
linked,
statically
is
also
creating
another
one,
so
be
really
interesting
to
try
and
generate
another
event
like
through
the
automatic
instrumentation
and
see
where
that
goes,
because
that
one
tries
currently
tries
to
load
an
existing
diagnostic
source
and
then
tries
to
yeah.
F
It's
only
asp.net
core
on
net
core
right
now:
okay,
so
specifically
the
asp.net
core
integration,
where
it
listens
to
the
like
the
asp.net
core
diagnostic
source,
like
events,
but
it's
certain
it's
currently
under
an
if
def.
So
if
we're
building
for.net
framework,
we
don't
put
that
in
because
we
just
didn't
want
to
cause
any
assembly
loading
issues
with
donna
framework
being
pretty,
but
the
way
they
load
assembly
identities
pretty
strictly.
F
F
If
you
wanted
to
just
see
how
the
current
asp.net
core
integration
would
work
with
activity,
you
could,
theoretically
you
get
it
in
the
instrumentation.cs
class,
it
gets
the
currents,
diagnostic
source
type
and
that's
sort
of
what
we
use
for
the
listening
to
it,
and
so
from
there.
You
can
use
that
type
to
get
the
activity
class
and
start
stop.
A
There
so
so
I
I
think
the
next
steps
is
to
try
the
instrumentation
with
that
and
also
clarify
some
of
this
question.
I
tried
the
best
that
I
could
to
kind
of
write
the
using
in
the
in
the
notes,
but
please
take
a
a
look
after
and
and
complement
if
you
guys
wanna
some
of
those
questions.
You
know
yeah,
but
actually
I
I
was
expecting
more
problems
on
this
first
trial
than
what
I'm
seeing
the
fact
that
I'm
seeing
it
working
it's
kind
of
a
good
surprise.
For
me
at
least.
E
Yeah,
I
still
feel
that
I've
run
into
the
type
mismatch
stuff
in
net
core,
but
I
think
that
was
when
I
was
dealing
with
the
microsoft
extensions
logging
and
trying
to
do
something
with
ilogger
and
dealing
with
different
versions
being
loaded
at
runtime.
There
interesting.
F
What
I've
experienced
so
far
is
one
on
dot
net
coil
when
it's
loading
assemblies,
if
you
try
to
load,
say
a
newer
version
like
2.0
and
then
someone
else
has
a
reference
on
1.0
and
or
or
tries
to
load
like
a
1.0
version,
you
immediately
get
an
assembly
load
exception,
or
that's
not
the
name,
but
you
get
an
exception
since
it
already
loaded
one
and
it's
newer,
so
it
can't
satisfy
the
requests
and
then
I
have
had
interesting
things.
F
A
Yeah,
and
actually,
what
comes
to
my
mind,
is
that
I
would
like
to
do
a
debug
of
this
kind
of
seeing
what's
triggering
the
load
of
each
one.
No,
but
I
I
think
yeah,
I
pick
up
a
native
debugger
and
put
breakpoints
on
mobile.
C
We
have
been
even
debugging
this
one
here
we
put
a
debugger
before
each
usage
and
we
saw
that
in
this
modules
window
this
you
know
these
dealers
are
are
being
popped
up.
I
think
that
you're
gonna
erasmus,
you
can
show
it
right
now.
Just
put
a
debugger
between
you
know
the
first
use,
oh,
if
you
know,
and
there's
they're
being
loaded
magically
without
any
issues.
G
F
F
Yeah,
I'm
trying
to
figure
out
the
right
one.
I
think
I
thought
assuming
I
was
actually
only
loaded
on
failure
to
resolve
one.
Let
me
see
yeah.
C
And
we
have
and
we
have
in
the
profiler,
we
have
an
event
which
occurs
when
it
happens.
Oh.
F
A
Yeah,
I
I
I
was
thinking
through
all
that
I
was
thinking
load
exception
on
the
wind
dpg
and
because,
when
you
read
the
image
then
I
I
can
put
a
breakpoint
on
the
clr
or
anything
that
I
want.
You
know.
G
A
Yeah,
so
so
yeah,
that's
that
is
interesting.
I
I'm
seeing
a
bunch
of
things
there
so,
but
what
I
was
I
was
thinking.
I
I
basically
want
to
see
what
the
clr
is
is
thinking
in
terms
of
kind
of
why
I
need
to
bring
this
other
dependency
here.
You
know
basically
that
same
in
my
mind,
it
seems
that
the
activity
is
being
the
one
that
is
first
loaded.
A
You
know
we
are
not
having
silos
for
the
types
so
the
first
first
one
loaded
is
five
zero,
but
then
for
some
reason
we
triggered
the
load
of
four.
You
know
and
I
actually
not
understand
how
the
assemble
resolution
assembly
resolution
kind
of
comes
to
the
conclusion
that
hey,
I
still
need
to
load
this
guy.
You.
A
A
A
A
All
right
this,
this
is
great
stuff.
You
know
kind
of
make
it
concrete.
A
lot
of
the
questions
that
we
have
about
this
stuff,
you
know
so
experimental
stills,
is,
is
really
great
and
also
perhaps
also
get
us
to
the
point
that
we
may
get
give
some
feedback
for
the
sdk
guys
regarding
the
versioning
on
this
kind
of
stuff.
If
we
needed
to.
A
Oh,
so
maybe
we
are
very
tempted
to
repeat
something
that
we've
been
talking
about
here
and
raising
just
presented,
but
since
you
joined
it,
I
think
I
think
perhaps
you
can
give
some
clarity
for
us.
So
just
to
give
a
little
bit
of
context,
we
are
trying
to
do
a
experiment
with
loading,
the
open,
telemetry,
sdk
and
the
open
element.
Sdk
uses
the
diagnostic
source
version.
5.
A
and
rasmus
did
the
experiment
mixing
different
versions,
and
we
end
up
with
the
other
versions
loaded.
However,
some
versions
loaded
and
but
to
to
great
relief,
at
least
for
me,
everything
worked
as
is
one
and
everything
their
parents
saw
the
the
child,
the
child
saw
the
parent
and
their
type
is
using
the
same
async
or
local
static,
so
everything
seems
to
be
working,
and
actually
I
don't
understand
how
and
why
that
happens.
A
B
Yeah,
so
it
depends
yeah
it.
It
probably
shouldn't
work.
So
what
did
load
multiple
versions,
so
that
means
that
something
did
a
a
load
from
and
then
it
gets
loaded
into
a
separate,
separate
context
and
and
you
shouldn't
it
shouldn't
work,
and
so
I
don't
I'm
surprised
too.
F
I
can
confirm
just
looking
at
that
path
that
that
middle
one
with
the
5.0
that
one's
coming
from
the
instrumentation
and
so
that
one
was
a
load
from
okay.
Although
it's
still
unclear
how
the
it's
loading,
one
that
seems
to
be
from
the
output
of
the
build
and
one
from
a
new
location,
so
it
didn't
resolve
it.
Didn't
the
second
one.
F
F
G
A
Yeah
so
so,
basically,
I
think
I
interesting
casey
will
be
if
there
is
some
library
involved
that
uses
the
activity
and
then,
if
you
are
going
to
be
able
to
listen
to
it,.
F
F
B
F
That's
actually
probably
correct:
we
jankly
get
the
first
jit
compilation
and
invoke
our
own
manage
code,
which
then,
at
that
point
in
time,
instantiates
tracer
and
tries
to
find
a
diagnostic
source
type.
So
yeah,
I
probably
wouldn't
be
loaded
yet.
F
F
I
think
that's
probably
why
greg's
work
was
to
try
and
do
an
assembly
load,
but
do
it
in
the
try
to
do
it
from
the
default
load,
contacts
by
just
doing
assembly
load
and
just
the
name
like
system
dynastics
and
try
to
get
whatever
the
application
will
load
and
then
implementing
fallbacks
and
stuff.
Like
that.
A
Yeah,
but
really
interesting,
we
we
have
to
dig
into
this,
but
so
dave.
I
do
remember
that
you
mentioned
some
time
ago
that
a
few
folks
in
microsoft
that
really
understand
the
assembly
load.
Of
course,
then
we
we
have
to
prepare
kind
of
very
precise
questions,
but
then
could
you
kind
of
forward
to
then
or
get
somebody
to
look
at
this
after
we
organize
and
kind
of
have.
I
Very
straightforward
questions
for
them,
so
that
is
an
option.
The.
B
Those
people
are
it's
when
we
have
very
specific
questions
that
I
would
want
to
bug
them,
and
so,
if
you,
if
you
have
questions,
I
you
know
basically
I
can.
I
can
answer
them
to
mess
my
ability
and
I'll
forward
to
them.
If,
if
I
think
that
it's
the
type
of
thing
that
they
should
answer,
but
a
lot
a
lot
of
times
on
desktop,
is
you
just
have
to
try
it
and
see
what
happens
because
it's
complicated
and
it's
not
going
to
change,
and
so
whatever
you
can
get
to
work
is.
B
A
E
G
E
G
G
Yeah,
so
here
you
can
see,
the
profiler
here
is
25.001
okay,
but
in
the
eastern
it's
printed
500
after
I
downgraded.
E
F
It's
probably
provided
by
it's,
probably
because
it's
provided
by
the
framework
at
this
point,
it's
probably
in
like
the
runtime
cache
or
whatever.
So,
even
though
there's
a
nuget
package
reference,
it's
just
kind
of
disregarding
it.
It's
probably
just
a
ref
assembly
and
the
donut
5
sdk
is
providing
it.
Oh
wait,
you're
on
dynamic,
oh
crap,.
F
C
G
Yep,
so
that's
the
same
old
test
here.
C
F
B
Yes,
I
would
expect
it
to
because
you
don't
know
where,
where
the
assembly
that
full
name
is
going
to
be
at
runtime,
because
there
might
be
all
sorts
of
things
redirecting,
you
know,
binding
redirects,
all
that
sort
of
thing.
A
So
so,
to
temper
a
little
bit
my
excitement,
it
seems
that
this
is
working
because
by
specific
accidents
we
are
actually
referenced.
The
same
activity
version
that's
coming
out
from
the
same
assembly.
A
All
right,
but
but
it
it's
very
good,
we
are
doing
the
experiments.
We
are
getting
some
questions,
so
I
think
I
think
it's
a
a
great
great
kind
of
start.
I
I
was,
I
was
very
surprised,
positively
surprised
and
still
am,
but
now
we
have
more
questions
to
follow.
A
Up,
okay,
do
you
guys
have
a
further
questions
regarding
this
plt.
F
A
He
he
hadn't,
he
did
that
demo
for
us
with
a
smaller
type
than
diagnostic
source
right.
What
was
the
name
of
the
type
but
okay
and
he
he
didn't
update
the
branch
recently
you
know
so,
but
I
think
also
is
because
there
is
other
part
of
the
question
that
we
want
to.
Answer
is
kind
of
the
the
way
that
he
proposed
works
for
for
the
auto
instrumentation,
but
it
still
doesn't
allow
us
to
you,
reuse,
the
sdk,
and
I
think
we
have
been
getting
a
lot
of
questions
about.
A
Why
is
not
possible
to
use
the
the
the
sdk
and
why
kind
of
we
are
going
to
end
up
with
duplicating,
therefore
you
know
so
we
wanted
to
have
kind
of
concrete
answers
to
give
kind
of.
If
it's
not
possible,
it's
not
possible,
but
if
we
can't
do
it,
we
we
should
explore
and
make
the
problems
that
we
think
our
problems
we
we
need
to
make
them
concrete,
make
sense.
F
Yeah
yeah,
so
in
that
case
I
would
also
suggest
maybe
trying
to
maybe
load
the
sdks
also.
I
guess
they
have
a
couple
ways
in
that
that
that
they
create
activities
ones
just
from
diagnostic
source
kind
of
sources
and
then
also
activity
sources.
F
A
So
using
diagonal
source
directly
and
the
other
types
from
diagnostic
source.
F
Yeah,
so
I
there's
there's
still
the
the
order
in
which
the
assemblies
are
loaded
which
might
cause
problems,
but
I
think
that
if
you
are
able
to
successfully
load
some
types
from
the
sdk,
or
at
least
you
know,
try
to
load
them
and
then
duct
type
them
then
with
the
the
typing
stuff.
That
tony
did
that
might
bring
about
some
good
results,
which
is
like
starting
starting
and
stopping
different,
like
activity.
A
Typing
via
the
profiler.
E
Yeah,
you
know
similarly
another
thing
that
you
can
try
doing
and
this
might
make
it
easier
to
reproduce
the
different
diagnostic
source
versioning
would
be
using
instrumentation
to
basically
bootstrap
the
sdk.
A
So
so
let
me
try
to
rephrase
here
instead
of
using
the
new
get
reference
try
to
build.
You
use
that
build
time
to
get
all
the
dependencies
that
you
may
need,
but
you
you
are
talking
about
loading
these
dependencies
directly.
I
I'm
not
sure
if
I'm
following.
E
Yes,
basically,
you'd
have
to
bundle
the
sdk
with
the
with
the
tracer
and
then
instrument
some
method
around
startup.
That
would
basically
bootstrap
the
sdk,
which
is
now
embedded
within
the
tracer.
E
A
Okay,
but
I
once
more,
I
think
it's
a
it's
a
great
start
and
we
are,
we
are
making
these
questions
and
answers
concrete.
This
will
be
very
good,
at
least
to
have
this
kind
of
information
at
the
end.
Also,
one
thing
that's
related
to
this.
A
That
is
the
final
we
are.
We
are
also
looking
to
the
the
most
complicated
cases.
The
those
are
framework
net
core
app
3.1,
but
I
also
would
like
to
look
at
dot
net
5
and
moving
ahead
because
movie
ahead,
I
mean
after
dot
net
5,
because
I
think
there
we
shouldn't
have
all
these
problems.
Okay,
we
have
applications
that
may
bring
dependencies,
but
we
kind
of
have
a
bunch
of
this
stuff
from
the
runtime
cell.
A
So
I
also
don't
want
to
lose
the
focus
of
what
we
can
do
with
net
five
make
sense.
E
A
E
Yeah
yeah,
I
I've
just
been
seeing
more
and
more
people
use
the.
What
is
it?
Yes,
the
trimming
feature
that
was
introduced
with
dot
net
5.,
and
so
people
are
asking
hey,
which
assemblies
do
we
need
to
add
to
the
keep
list
to.
H
A
A
Yeah
so
so
I
think,
I
think,
on
on.net
5
we
you'll
be
interesting
kind
of
to
see
the
case
with
streaming
and
perhaps
then
we
have
a
set
of
assemblies
that
we
say:
hey
keep
those
if
you
wanna
help
implementation.
B
With
the
trimming
stuff,
the
there's
an
option
in
the
single
pile
to
skip
trimming
and
that's
probably
going
to
be
necessary
for
so
the
the
exclusions
for
the
trimming
list,
I
think
happen
at
the
method
level
and
not
the
assembly
level.
So
the
if
you
wanted
to
do
go
that
approach.
I
think
you
have
to
identify
all
the
methods.
Sorry
I
quit
it
so
anyways.
I
think
you
would
have
to
identify
all
the
methods
you
need,
not
just
the
assemblies
if
you
wanted
to
prevent
them
from
being
trimmed
out.
F
Might
actually
be,
that
would
actually
probably
be
a
good
exercise
for
us
to
do
just
to
have
like
as
a
automated
kind
of
build
step
to
just
emit
the
list
of
type
rest.
We
use,
because
I
mean
the
ielts
trimming
is
probably
going
to
be
even
more
prevalent
in
the
future.
F
Yeah
so
like
once
once
we
build
our
assembly
as
like
a
post,
build
step,
just
iterate
over
it
with
one
of
the
reflection,
one
of
the
metadata
libraries
and
just
kind
of
just
write
a
list
of
here's
all
the
type
reps
that
we
need.
F
A
E
E
But
do
we
intend
on
actually
merging
that
pr
or
what
is
our
intention
for
it?.
B
A
I
think
I
I
I
I
was
actually
kind
of
leaning
in
in
a
different
way,
robert,
because
I
think
they're
falling
instead
of
but
instead
of
doing
all
the
branch
I
would
like,
because
the
branch
is
pretty
large.
I
would
like
to
have
a
version
that
has
the
native
profiler,
the
the
loader
one
or
two
instrumentations,
and
just
call
target,
because
we
know
that
yeah.
C
D
C
C
A
I
think
we
had
a
bazillions
questions
if
we
had
a
common
base
of
code
that
somebody
can
do
and
do
one
test
and
show
to
us
against
that
branch,
what
was
changed
and
then
I
think,
then
we
can
have
a
kind
of
ask
and
discuss
brainstorm
between
ourselves
to
to
to
see
exactly
how
the
things
are
going.
You
know,
so
I
think,
keeping
that
branch
and
kind
of
trimming
down
to
the
exact
what
we
need
to
to
do
the
test.
I
think
I
think,
should
be
what
we
should
be
doing.
D
C
C
A
Yeah,
I
I
I
agree
for
robert.
I
think
we
should
do
prs
that
perhaps
when
there
are
mixing
when
we
are
merging,
we
can
say
hey
these.
I
know
that
work
in
this
scenario
that
I
tested,
but
since
it's
for
us
to
test
stuff,
we
can
perhaps
not
add
the
the
test
or
anything
like
that,
but
we
just
described
the
scenarios
and
we
we
give
visibility
and
history
for
for
the
whole
thing.
Yes,.
C
So
in
my
previous
work,
we
were
doing
sometimes
even
like
for
experimental
stuff.
We
were
often
even
doing
branches,
sorry
prs
and
merging
them,
but
always
someone
can
add
comments
later
and
discuss
it
later.
That's
also
one
reason
why
having
fears
is
good
because
you
can
always
edit
comments
at
you
know
a
description
afterwards,
which
you
cannot
do
it
on
committing
it.
A
Yeah,
so
I'm
gonna
put
if
everyone
is
agreement,
I'm
gonna
put
that
let's
keep
the
plc
branch
and
do
prs
with
the
with
the
changes
that
allow
the.
A
I
think
I
I
have
one
question
for
days
regarding
event
pipe
encounters,
and
I
also
that
we
are
going
to
follow
up
on
the
work
that
pablo
did
on
that
and
we
are
going
to
start
to
listen
to
the
processes
and
start
at
least
with
metrics.
A
My
question
in
that
regard,
dave
is
about
net
five
dot
net
five
on
windows.
I
I
I
can
support
the
counter
support
both
via
vent,
pipe
and
via
windows,
counters.
B
I
think
just
a
vent
pipe,
I
don't
think
windows.
Can
I
guess
I
don't
know.
I
don't
think
windows
counters
work
on
core,
but
I
I
don't
know
for
sure.
A
A
Yeah,
I
I
that
crosses
my
mind
just
because,
since
on
windows,
the
support
for
counters
is
is
pretty
good
for
the
framework
I
was
kind
of,
if
doesn't
have
that,
I
think
people
will
feel
like
a
downgrade.
You
know,
so
I
I
I
I
that
popped
in
my
mind,
so
because
then,
if
it's
just
event
pipe,
then
one
thing
that
becomes
kind
of
priority
to
this
stuff
on
the
collector
is
to
implement
the
event
pipe.
B
Yeah,
so
my
other
people
on
my
team
are
the
ones
that
are
working
on
the
perf
counter
stuff,
and
so
I
can
you
know
I
know
who
to
talk
to.
If
you
need
people
to
talk
to
it's,
I
just
haven't
been
working
on
it.
Cool.
E
Cool
yeah,
so
paulo
I've
played
around
with
the
perf
counters
and
the
event
pipe
stuff
before
3.0
was
released
and
at
least
in
the
2.1
time,
time
frame
and
even
2.2,
so
you
weren't
able
to
get
the
performance
counter
data
even
when
you
were
running
on
windows.
As
far
as
that,
I
was
aware
of
from
the.net
core
apps,
but
then
using
the
event
pipes
prior
to
3.0.
E
You
would
run
into
memory
issues
and
you
it
also
didn't
have
the
full
set
of
data
that
I
would
label
interesting
as
far
as
debugging
applications.
B
B
A
All
right,
very
good
meeting,
guys
any
I
I
was
going
to
say
final
comments,
but
then
somebody
may
make
a
question
something
that
keeps
us
going.
So
any
other
things
that
you
guys
want
to
bring
up.
A
A
All
right,
then,
nice
to
see
everyone
let's
keep
in
touch,
let's
keep
moving
the
the
plc
with
some
quick
interactions.
So
we
can
have
this
discussion
going
fast.
F
I
just
remembered
something:
sorry
andrew
locke
has
just
created
a
pr
to
which
is
like
a
first
step
in
consolidating
a
lot
of
our
build
steps
into
nuke.
It's
definitely
not
the
end-all
be-all,
but
it
is.
It
brings
a
lot
of
it
into
nuke.
So
if
you
want
to
check
that
out
on
the
data
dog
repo
feel
free,
but
that
is
progress.
A
I
just
a
quick
comment.
I
was
sometime
without
building
the
ripple
and
last
week
I
did
the
upstream
the
pool
from
upstream,
and
I
don't
know
why
it
took
a
long
time
for
the
test
that
I
typically
run.
So
I
don't
know
if,
because
I
I've
been
without
building
for
like
two
or
three
months,
I
don't
know
if
it's
a
bunch
of
simple
applications
or
perhaps
it's
because
we
are
testing
out
target
and
not
called
target.
A
I
had
the
impression
that
we
increased
a
lot
of
this
stuff
there,
but
this
this
change
is
really
welcome.
I'll,
take
a
peek
there
too.