►
From YouTube: 2023-02-27 meeting
Description
Instrumentation: Messaging
A
B
A
A
B
He
had
some
good
points
on
the
on
the
spam
links.
A
A
B
B
A
Also,
where
I
think
the
go
people
are
there,
so
I
think
there
is
some.
It
seems
to
me
some
consensus
there
that
the
event
solution,
like
adding
things
as
an
event,
I
think
people
say
it's.
It's
like
from
the
design
point
of
view,
a
solution
that
would
be
okay
on
its
own,
but
the
coexisting
with
the
existing
spanning
solution
creates
a
real
mess.
A
Is
some
some
consensus
there
and
I
think
without
going
to
actually
a
version
V2
of
the
whole
hotel
API?
We
will
not
get
rid
of
the
Legacy
spendings
so.
B
Yeah
and
then
there
will
be
also
another
feature.
Probably
that
will
come
and
then
what
what
will
happen
right.
A
Yes,
so
so
I
think
I
mean
Carlos.
Thankfully
leads
many
discussions
there
and
I
think
it's
the
main
yeah,
the
main
stumbling
block
he's
just
yeah:
getting
to
go
people
to
kind
of
accept
this
I
mean
Carlos
even
like
offered
this
and
I
thought
about
this
before
he
offered
this
kind
of
compromise
where
he
said
yeah,
because
the
code
people
said
yeah,
we
can,
we
can
add
links
just
this
add
as
a
kind
of
arguments
on
the
on
and
spam
method,
because
that
would
not
break
their
API.
A
B
A
Also
from
a
user
point
of
view,
basically,
if
then,
you
kind
of
want
to
create
the
link,
you
have
kind
of
to
wait
until
the
span
ends.
You
have
to
save
your
link
context
somewhere
and
maybe
I
don't
even
have
access
to
this
on
end
method,
maybe
like
the
messaging
like.
Maybe
the
control
that
you
have
doesn't
even
extend
that
far
to
the
end
of
this
and
so
I
think
that's
not
a
feasible
method,
but
what
the
Carlos
suggested
is
that
you
are.
A
Maybe
we
convert
the
spec
kind
of
vague
enough
that
kind
of
go
can
Implement
their
like
on-end
solution
and
other
languages
can
can
add,
like
an
ad
link
method.
A
B
B
I
mean
if,
if
this
softening
that
Carlos
proposed,
if
that
goes
through,
then,
if
there's
another
way
to
do
this
without
having
to
break
the
API,
then
I,
guess
that
would
be
more
acceptable
than
doing
the
on
the
end
on
the
end,
because
I'm
not
sure
that
will
work.
Yes,.
A
Yes
and
the
meaning
other
and
other
thing
I
thought
is
that
one
could
also
bring
in
a
discussion
with
the
code
people
that
one
says:
okay
yeah.
We
what
makes
this
change
to
the
spec
but
I
mean
I
mean
that's
what
other
people
don't
need
to
answer
too
like
for
more
like
who
have
more
overarching
view
but
I
think
yeah
as
long
as
there's
no
demand
from
Ghost
site
for
this.
Maybe
they
can
just
hold
off
implementing
this
until
as
long
as
nobody
asks
for
it.
B
But
yeah
I
could
be
also
maybe
a
good
exercise
right.
That's
what
I
was
taking
as
well,
because
I
I
thought
that
it
was
perfectly
spec
that
nothing
will
be
missed
or
that
we
want
to
add
something
new
and
then,
if
something
new
happens,
then
what
what
is
the
solution
again
right
I
mean
maybe
this
one
we
can
do
something
else
like
postpone
it
or.
But
what
about?
If
there's
like
a
real
new
future?
B
A
A
B
A
B
B
Didn't
go
it
would
it
would
break
I
know,
for
example,
because
I
did
some
stuff
in
the
past
and
I
looked
at
this
there's
a
research
paper
from
from
from
the
goal
people
gold
developers
about
when
they
didn't
didn't
pass
the
context
around
like
this.
You
know
this.
This
context
object.
So
there
were.
B
There
were
a
lot
of
each
GP
methods
that
didn't
have
the
context
and
that
they
wanted
to
add
the
context
because
yeah
propagation
and
all
that
stuff
and
then
they
basically
duplicated
the
methods,
and
there
is
a
bunch
of
approaches
how
to
do
it,
they're
using
like
V2
or
or
two
or
something
at
the
end.
There's
like
or
past.
This
very
attic
passage
as
a
very
attic
parameters
or
yeah.
Yes,.
A
B
A
Would
need
to
maintain
kind
of
two
different
versions
of
the
interface
and
I
think
that
is
what
gives
them
a
bit
of
headache.
Yeah.
A
B
B
A
B
A
B
B
It
didn't,
but
now
it
does
yes
now
we
do
support
it.
B
Like
I
said
last
week,
I'm
still
asking
entirely
if
I
can
show,
show
it
up
to
you
in
one
of
the
meetings
but
yeah.
So
we
do
support
it
now,
it
you
have
a
so
you
have
the
the
trace
View
and
then
you
click
on
the
trace.
There's
a
bunch
of
so
there's
a
all
the
details
of
this
pen,
and
we
also
show
the
logs
there
so
there's
like
there's
different
dams
and
then
one
of
the
tabs
now
is
are
the
links
and
then
we
click
on
the
tab
industry.
B
You
see
all
the
links
and
then
you
click.
You
go
to
the
other
similar
to
what
you
you
have
on
on
Jaeger,
for
example,
oh.
A
B
There's
no
like
visual
representation
yet
like
if
you
collect
a
span
and
if
you
see
the
little
links
like
you
have
in
Azure
in
Azure,
monitor
but
I
I
assume
that
they,
it's
not
my
it's
another
team,
I,
don't
work
with
them,
but
I'm
I'm
trying
to
inquire
to
see
what
what
what
is
the
goal
and
how
how
it
will
look
like,
but
now
it's
it
it
is.
It
is
there
if
I
understood
correctly,
it
was
always
somewhat
supported,
not
just
not
being
visually.
A
A
B
B
B
Yeah
yeah,
it
was
good
because
I
I
brought
it
up
actually
that
this
is
something
that
we
are
planning
to
be
supporting
hotel
and
more
and
more
people
will
be
yes
using
it
as
as
soon
as
like
instrumentations
have
it
so
yeah
that
probably
played
played
some
influential
on
them
prioritizing
work
on
it,
yeah
great.
A
Yes,
okay,
but
as
nobody
else
is
showing
up,
I
I
suggest
we
kind
of
yeah
close
early
today,
because
I
think
it
doesn't
have
that
much
sense
and
we
discuss
them.
Then
we
have
kind
of
the
discuss
with
others
again
yeah.
B
Yeah,
no,
no
I
think
that's
a
good
idea.
I
will
definitely
for
for
for
for
true
I
will
look
at
the
hotel.
I
I
have
I
started
a
review,
but
I
didn't
publish
it
because
I
didn't
finish.
Yet
we
are
quite
busy
with
some
some
internal
things
lately
on
Thursday,
oh
I,.
A
Have
I
understand
we
all
are
so
I'm
just
trying
to
kind
of
yeah
close
and
open
commands
on
the
old,
Tab
and
and
then
I'm
enjoy
I,
never
blocked
with
the
adding
Links
of
the
span
creation.
But
hopefully
when
we
can
then
push
this
out
about
and
then
we
can
make
work
on
an
actual
like
more
detailed
roadmap
for
stability
for.