►
From YouTube: AsyncAPI SIG meeting 31 (September 15, 2020)
Description
This is the recording for the AsyncAPI Special Interest Group (SIG) meeting #31.
Attendees:
- Collin Sullivan
- Fran Mendez
- Jonas Lagoni
Moderation:
- Ćukasz Gornicki
Agenda & Notes:
https://github.com/asyncapi/asyncapi/issues/435
A
Okay,
recording
started
so
welcome
to
asking
api
open
meeting
it's
tuesday,
7
15
of
september
2020,
4
pm
utc
on
the
call.
It's
me
lukas.
A
Api
self
called
maintainer
from
the
founder
jonas,
the
the
als,
also
old,
school
all-time
contributor,
and
we
have
also
colleen
from
the
nuts
projects
project
that
joins
us
very
often
to
support
us,
especially
jonas,
with
the
work
on
the
template.
A
So
welcome
guys
and
before
I
jump
to
the
topics
I
added
to
the
agenda,
is
there
anything
that
is
not
on
the
agenda?
Something
super
important
that
you
prefer
to
talk
about
before
we
jump
into
the
argent
items.
A
Three
two
one:
okay,
so
let's
jump
the
first
one
is
the
template
for
generator
templates.
So
I
just
wanted
to
share
something.
I'm
sharing
quite
regularly
on
the
on
the
slack
channel
and
also
in
the
status
update.
A
What
was
I
about?
Oh
yeah,
template
for
for
the
template,
so
I'm
working
on
a
on
a
kind
of
template
like
github
template
that
someone
could
use
to
quickly
create
a
repo
with
a
generator,
a
generator
generator
generator
template.
A
A
So
it's
kind
of
presenting
those
features,
plus
it
has
some
extra
stuff
like
proper
s-linked
configuration
and
the
release
pipeline.
So
it's
much
easier
to
kick
off
with
automated
releases
of
the
template.
It's
mimicking
the
the
release
pipeline
that
we
have
for
asking
api
and
it's
what
I
want
to
share
is
maybe
the
status.
So
it's
the
the
work
is
eighty
percent.
A
I
would
say
eighty
percent
done
so
I
managed
to
create
a
template
that
covers,
of
course,
the
hooks
the
filters
partials
all
the
filters
that
not
all
the
filters
like
it
shows
the
usage
of
filters
that
comes
from
nun,
jacks
filters
that
come
from
our
official
package
of
filters
and
also
a
a
custom
filter,
and
it
also
covers
the
yeah.
The
docs
cover
the.
How
to
set
up
the
pipeline,
how
to
get
the
quality
assurance
and
how
to
write
tests
from
the
for
the
template?
A
A
A
It's
a
lot
of
content,
so
I
say
follow
up
after
that's
done.
I'm
gonna
work
on
a
interactive
tutorial
where
you
create
the
same
template
but
step
by
step.
So
you
don't
get
everything
with
one
click,
but
you
actually
go
through
step
by
step
tutorial
where
you
add
a
specific
part
of
the
template,
and
you
end
up
with
the
with
the
final
template.
A
The
outcome
of
this
template
is,
as
you
can
see,
I'm
a
super
pro
css
master,
but
I
covered
here
something
that
was
interesting
for
me.
So
a
generation
of
the
of
the
diagram
of
how
the
how
the
dependencies
between
the
the
schemas
look
like
in
a
visual
manner,
so
something
that
we
can
consider
long
term,
maybe
for
html
templates
behind
the
future
flag
that
you
don't
get
on
the
docs,
but
also
maybe
the
the
diagram.
So
that's
an
interesting
part,
but
I
think
what's
interesting
to
see.
A
A
Okay,
can
you
see,
can
you
see
my
visual
studio
code?
You.
A
A
Better
okay,
so
the
the
interesting
part
I
was
playing
with
the
the
the
last
days
was
how
to
test
the
result
of
the
generation.
A
So
the
the
outcome,
the
yeah,
the
output
of
the
generation,
is
the
index.html
file
where
the
diagram,
the
mermaid,
I'm
using
the
mermaid
library,
the
mermaid
diagram,
is
generated.
So
I
wanted
to
have
a
kind
of
way
of
making
sure
that,
with
my
my
example
as
an
api
file,
I'm
gonna.
A
I
wanted
to
make
sure
that
the
result
that
I'm
gonna
receive
after
generation-
let's
say
I
pump
the
the
mermaid
dependency
or
or
just
modify
something
in
the
style
css.
A
I
just
wanted
to
make
sure
that
the
the
index.html
file
will
not
be
will
not
change
like
during
the
review
of
the
pull
request,
for
example.
So
here
I
used
the
the
snapshot.
A
Snapshot
testing
from
jest,
so
it's
basically
what
I'm
doing
is
I'm
really
I'm
using
generator
generating
the
using
this
template,
I'm
generating
files,
then
I
read:
read
the
file
and
then
match
it
to
the
snapshot.
A
A
A
A
Okay,
I
broke
these
snapshots
with
some
tests,
so
you
can
see
that
the
there
was
also
a
snapshot
created
for
those
tests,
so
the
snapshot
is
basically
for
the
template.
Have
a
look
so
it's
this
html
file
and
every
run
of
the
test.
A
Compares
compares
the
the
snapshot
with
the
with
the
file
that
got
generated
during
the
test
was
run
basically
and,
as
you
can
see,
with
the
second
run,
the
test
was
successful,
but
if
I
would
change.
A
Do
some
change
to
the
index.html.
A
With
the
next
test
run,
it's
gonna,
it's
gonna
fail
because
it's
gonna
detect
a
change
in
the
in
the
snapshot,
and
it's
completely
fine,
sometimes
that
it's
gonna
fail,
because
you
really
changed
something
in
the
in
the
index.html
and
the
change
is
expected.
A
So
in
this
case,
you
have
to,
of
course,
have
a
look
on
the
snapshot.
If
only
the
the
part
that
you
changed
has
changed,
and
then
you
have
to
remove
and
regenerate
the
snapshot.
A
So
so,
basically,
it
means
that
it's
not
always
the
best
solution.
The
snapshot
because,
like
you,
can't
even
find
many
articles
about
it
that
people
write
that
at
some
point
of
time,
people
get
a
habit
that,
because
of
large
changes
in
the
snapshot,
they
simply
regenerate
snapshot
without
investigating.
A
So
I
find
those
snapshots
as
useful
whenever
there's
a
change
in
the
pull
request,
like
a
new
flag
is
added,
for
example,
and
the
the
default
result
should
not
change
after
introduction
of
the
of
the
parameter,
but
only
let's
say
new
tests
should
appear
with
a
new
snapshot
where
we,
where
we
test
the
the
result
after
using
the
parameter
in
the
test.
A
I
I
have
no,
you
know
what
I
mean,
so
that's
that's
basically,
where,
where
those
snapshots
will
be
very
useful
so
to
validate
that
adding
a
new
parameter
or
adding
some
new
file
to
the
output
of
the
generation
that
the
the
old
files
are
not
broken.
A
And
jonas,
I
guess
that
if
we
would
start
using
the
snapshot
approach,
if
we
we
would
like
to
start
using
snapshot
approach
like
for
all
the
templates.
A
A
Yeah,
the
only
thing,
I'm
afraid,
afraid
of
and
for
example,
not
sure
if
you
remember
some
time
ago,
we
agreed
that
if
we
are
working
on
the
template-
and
we
want
to
have
a
some
example
spec-
that
we're
testing
against-
we
should
not
commit
it
to
the
to
the
template,
but
rather
use
this
generic
dummy
one
and,
for
example,
I
I
actually
commit
the
dummy
one
here,
mainly
because
of
those
snapshots,
because
you
can
imagine
that
if
the
dummy
file
would
change
somewhere
remote
and
I
would
still
test
snapshots
against
it,
then
of
course
it
could
be.
A
There
could
be
a
risk
situation
that
okay
on
pr
everything
is
fine,
but
then
between
pr
and
release,
something's
changed
in
the
in
the
remote
file
and
then
the
snapshots
fail.
So
so
that's
the
risk.
If
we
really
will
use
think
about
using
tck
for
such
testing,
because
then
again
those
test
cases
can
change.
I
mean
now
they
are
not
super
stable,
they're,
changing
all
the
time.
So
that's
not
the
time
where
we
can
use
it
now,
but
in
the
future.
It's
still.
I
think
it
would
be
super
risky.
A
If
we
tell
all
the
template
developers
to
use
the
same
testing
approach
and
then
with
one
file
change,
we
break
all
the
snapshot.
Tests.
E
What
isn't
the
tcp?
What's
it
called
tc.
A
D
C
D
Just
saying
so
you
don't
have
it,
you
don't
have
to
maintain
a
duplicate
on
your
reaper
right.
E
A
A
A
But
now
I
will
basically
starting
from
tomorrow,
I'm
I'm
gonna
stop
working
on
the
template,
postpone
it
a
bit
not
because
I'm
tired,
it's
just
because
there's
another
topic,
I'd
like
to
focus
on
and
it's
called
hacktoberfest.
A
So
I'm
not
sure
how
many
of
you
know
what
oktoberfest
is.
So
let
me
first
google
it
for
you
so
oktoberfest,
let's
say
every
every
year
during
the
during
october,
like
october,
1st
october
1st,
you
get
it
so
it's
a
it's
a!
It
was
usually
sponsored
by
by
actually
the
company
that
I
have
t-shirt
on
me
today.
It's
a
digital
ocean
and
they
didn't
pay
me
for
having
it
today.
A
They
sponsored
it
started
it
three
years
ago,
it's
a
one
month,
long
event
for
people
around
the
world
to
create
they
have
to
sign
up
to
oktoberfest
and
they
have
one
month
for
creating
five
pull
requests
to
different
projects
and
as
a
result,
they
get
a
swag
pack,
and
I
love
swagbucks.
That's
why
I
always
took
part
in
the
oktoberfest
and
it's
a
it's
pretty
widely
known
event.
A
A
But
you
can
imagine
it's
third
or
fourth
even
like
this,
so
there
are
probably
hundreds
of
thousands
of
features
labeled
with
october
first
huckleberry
fest.
A
So
it's
not
enough,
and
I'd
like
to
yeah
involve
asking
api
in
the
oktoberfest
in
let's
say
much
more
professional
way,
and
the
the
goal
is,
of
course,
the
like
promote
as
an
api
initiative
through
this
through
this
event,
but
most
important
is
to
promote,
to
like
clearly
show
that
async
api
initiative
is
not
just
the
spec
like
open
api,
for
example,
initiative
where
they
only
focus
on
the
spec
to
show
that
we
are
also
working
on
tooling.
A
Actually,
a
lot
of
tooling
and
the
other
one
is
to
kind
of
have
a
side
effect.
First
of
all
make
it
easy
for
existing
community
to
jump
in
into
contributions
to
asking
api,
but
also
the
materials
that
the
resources
that
I
want
to
work
on
for
the
event,
some
of
them
can
be
reused
in
the
future,
to
make
it
much
easier
for
people
to
jump
into
the
project
and
contribute.
A
I
basically
need
to
do
for
next,
two
weeks
to
prepare
for
oktoberfest.
A
So
unless
you
want
me,
I
don't
want
to
just
take
your
time
and
go
into
details
of
what's
written
in
this
document.
I
think
the
most
important
is
at
least
for
me
and
and
and
from
the
the
last
point
so
some
time
ago,
frontiered
with
me
his
desire
to
publicly
actually
to
my.
D
A
To
have
a
have
a
live
coding
sessions.
A
That
are
not
prepared.
It's
rather
like
everyday
job,
but
online
for
people
available
to
join.
To
see
how
front
is
working
on
some
particular
features
of
asking
api
spec
is
it
did
I
explain
it
well.
D
Yeah,
actually,
maybe
not
just
not
just
a
developing
specific
feature
or
or
change
in
the
spec
or
something,
but
also
whenever
you
have
to
work
on
trying
to
understand
what's
happening.
D
Why
should
we
add
something
to
this
back
or
why
is
resultant?
Do
it
this
previous
work?
That
needs
to
be
done
before
you
make
a
proposal
before
you
even
start
working
on
it
or
like
before
you
even
officially
start
working
on
it.
There
is
a
there
is
a
work
before
this.
This
happens
right
and
it's
usually
happening
behind
the
scenes
like
nobody
sees
it,
and-
and
I
think
it's
useful
to
have
this-
this
kind
of
information
in
public.
D
So
when
someone
asks,
for
instance,
why
don't
you
do
x
and
instead
of
y,
it's
yeah?
I
thought
about
it
like
it's
here
it's
registered
and
you
can
see
why
we
decided
not
to
go
this
way,
but
instead,
this
other
this
other
way.
Of
course,
this
can
also
be
written
afterwards
as
a
document,
but
but
it's
also
cool
to
have
to
enjoy,
to
engage
people
live
and
during
these
live
sessions
because
they
can
people
can
also
jump
in
and
help
us
like.
Hey,
I
see
you're,
not
thinking
about
it.
D
Have
you
think,
have
you
thought
about
it
about
this
other
option
say
it
will
be
like
a
pair
programming
with
lots
of
people
if
they
wish
to
just
jump
in
or
pamper
me
or
pair
thinking.
If
you
want
right,
not
just
for
ramen,
so.
B
Yeah,
that's
a
great
idea.
One
one
thing
we've
done
on
the
gnats
team
that
has
kind
of
you
know
has
been
well
received.
Is
an
ama
ask
me
anything,
and
then
we
have
queued
up
examples
that
we
can
show,
and
this
is
diving
down
into
the
code
showing
prompts
you
know
and
and
showing
how
users
how
to
do
things
that
fills
in
the
gaps
while
we're
waiting
for
other
questions,
and
that's
worked
well
as
well.
D
I
agree
that
this
is
also
work.
I
asked
me
anything
and
and
also
like
you
know,
involving
people
into
into
the
decision
making
discussions
or
previous
decision-making
discussions.
I
think
it
helps
a
lot.
A
A
So
obviously
it
was
not
the
best
time
to
start
it,
and
I
was
thinking
like
okay,
I'd
like
to
have
a
kind
of
support
office
hours
for
people
that
would
be
interested
in
taking
part
in
the
october
first
with
asking
api,
and
why
not
connecting
it
like
with
the
with
this
idea
that
front
had
so
we
we
can
just
jump
online
on
some
schedule
or
me
or
fran
or
both
of
us.
A
If
anyone
else
would
like
to
join
like
like
jonas,
maybe
you'd
like
to
so,
and
then
this
this
office
hours
would
be
okay
if
nobody
joins
from
for
oktoberfest
and
needs
some
help
with
some
pull
requests,
then
we'll
just.
I
will
probably
work
on
some
code
and
show
my
crappy
thinking
process
when
I
code
and
and
fran
will
show
his
probably
better,
probably
better
thinking
process
than
I
have.
C
D
D
D
D
Yeah,
it's
thinking,
feeding
animals,
tweeting
process.
D
Just
to
make
it
clear
for
octoberfest,
I
think
it
helps
to
have
office
hours
and
scheduled
office
hours.
So
people
know
when
are
we
going
to
jump
online,
but
the
other
idea
that
we
was
we
were
talking
about
before.
D
I
think
it
would
be
super
cool
to
do
it
like
non-scheduled,
like
I'm
just
working
on
this
right
now,
I'm
gonna
jump
in
in
in
20
or
13
minutes
or
30
minutes.
I'm
gonna
jump
live
and
and
share
with
you
like
totally
improvised
and
see.
You
know
like
you're
you're
inspired
now,
and
I
I'm
feeling
inspired
now
and
I'm
gonna
share
with
you
right.
I'm
gonna
share
this
with
you
in
case.
You
want
to
to
jump
in
and
enjoy
nothing
really
scheduled
or
anything.
A
Why
not
awesome
21st
century,
my
god
and
one
last
good
thing
I
think
about
this
oktoberfest
will
be
that
we
always
wanted.
I
mean
whenever
we
talk
with
run.
We
always
wish
to
engage
more
diverse
community
into
the
initiative,
and
we
during
the
asking
api
conference
during
the
preparations
for
the
conference
we
with
emeline.
She
helped
us
to
connect
with
the
woman
who
code
that
also
helped
us
to
advertise
as
an
api
conference
in
women
community.
A
So
I
guess
it's
gonna
be
also.
It's
gonna
be
good
to
share
as
well
with
them
give
them
all
these
materials
that
we
prepared.
How
easy
it
is
to
kick
off
with
any
kind
of
contribution,
so
maybe
we'll
somehow
give
back
to
them
and
attract
them
more
into
joining
the
initiative.
A
D
Yeah,
another
thing
that
I
would
like
to
remark
here
is
that
we
we
need
help
making
the
zkpa
community
more
labors.
This
is
this
is
we're
struggling
with
this,
like
I
mean
look
at
the
meeting
today
like.
D
People
say
it's
like
we
still,
which
is
fine.
I'm
not
saying
this
is
bad,
but
we
should
we
should
be
trying
to
make
it
more
diverse
and
not
just
women,
but
you
know
like
diverse
in
terms
of
minorities,
and
you
know
diversity.
D
Basically,
so,
like
I
don't
know
like,
we
try
to
we're
trying
to
reach
some
communities
like
women
who
code,
but
it's
I
think
we
we're
not
probably
doing
enough
or
we're
not
doing
it
well
or
or
both
or
or
we're
not
able
to
attract
women
or
or
more
diverse
groups,
let's
say
or
underrepresented
groups
to
participate
in
these
in
kpa
community,
and
this
is
this
is
something
I
would
love
to
improve.
D
A
Yeah
yeah,
like
my
suspicion,
was
always
that
it's
simply
hard
to
jump
in
into
such
a
big
project
like
asking
api,
and
it's
just
scary.
So
I
hope
that
when
I
will
record
all
those
videos
how
to
like
kick
off
with
the
setup
for
long-term
contributors,
how
to
like
find
out
what's
in
in
the
project
who
can
help
with
what
and
maybe
that's
that's
gonna,
make
things
easier.
B
Yeah,
there's
there's
also
more
ways
to
contribute.
You
know
we're
focused
on
code
because
that's
what
we
do,
but
there's
more
ways
to
contribute
to
a
project
than
code
in
terms
of
hosting
it
right
now
virtual
meetups,
but
we
used
to
really
encourage
people
to
host
meetups
give
them.
You
know
swag
packages
with
stickers
and
stuff
to
hand
out
and
then
also
reward
people,
for
you
know
doing
documentation
and
and
even
just
general
marketing.
A
Yeah
so
like
from
my
point
of
view,
I
I
always
say
that,
of
course,
code
is
not
the
only
way
to
contribute
like
emily
that
I
mentioned
before
she
doesn't
code
but
she's,
one
of
the
greatest
contributors
of
asking
api,
because
she's
she's
writing
blog
posts
about
the
initiative
she
like,
like.
I
said
she
helped
us
with
the
woman
who
code
to
approach
them
so
she's
super
active
and
promoting
the
the
initiative.
So
it's
definitely
a
contribution
that
I
value,
but
it's
just.
A
I
think
that
like
in
like
in
the
engineering
space,
we
simply
should
also
make
it
much
much
easier
to
jump
in
and
and
contribute
not
only
docs.
So
all
those
not
all
those
issues
will
be,
of
course,
related
to
to
pure
coding.
Some
of
them
will
be
ci
related,
some
dogs,
but
what
you
said
about
the
meetups,
I
think
that's
also
something
that
we
long
term
should
have
a
look
at.
A
I
think,
some
time
ago
we
talked
about
having
a
official
slide
deck,
for
example,
about
I
think
api
to
make
it
easier
for
people
to
go
to
meetups
and
talk
about
the
the
spec
and
the
tooling.
So
yeah,
that's
that's
a
good
point.
It's
good
that
you
brought
it
up.
D
There's
a
slide
deck,
but
it's
not
really
well,
it's
not
really
super
visible,
but
it's
on
the
it's
on
on
our
google
drive
yeah.
So
it's
a
google
presentation
or
google
slides
presentation
so
and
it
has
a
lot
of
slides,
already
prepared
slice
in
case.
You
want
to
to
do
a
presentation
about
this
in
kpi,
but.
D
Make
it
more
prominent
and
visible
maybe
to
have
on
the
website
and
the
new
website
that
we're
working
on?
Maybe
we
should
have
something
like
a
resource
center
or
something
like
this,
where
people
can
download
the
logos,
download
presentations-
and
you
know
all
these
things
that
usually
they
ask
for
or
they
end
up
taking
from
the
website
or
from
other
places
or
asking
us
privately.
D
This
will
be
more
prominent
for
sure
and-
and
that's
just
like
the
first
step,
to
try
to
make
it
easier
to
create
like
calling
set
to
create
like
virtual
meetups
and
once
covered
is
gone
or
it's
not
a
problem
anymore,
maybe
in
maybe
live
or
physical
meetups
will
be
great
as
well,
but
but
yeah.
This
is
something
we
will
be
doing.