►
From YouTube: PWA Studio Community Meeting Sept. 27, 2019
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
Oh
sweetheart
yeah
there's
a
picture
of
me
in
the
jungle
now
I.
Don't
know
why
yeah
I
was,
we
were
gonna,
go
to
our
stand-up,
but
we
but
Jordan
and
I
were
working
on
well,
I,
guess
a
bunch
of
things
during
my
day
there,
but
we
never
figured
out
how
to
fix
entity
caching
and
the
graph
QL
sigh.
That
was
a
pretty
deep
hole.
A
It's
not
secret
I
mean
the
main
things
that
we
accomplished
during
our
hackathon
are
one-day
hackathon
together,
I'd
say
the
headline
is
that
Jordan
found
a
way
to
bust
the
cash
manually
by
accessing
the
client
and
and
running
the
actual
operation
methods
that
are
on
the
Apollo
cash
object?
And
so
the
question
so
I
matically?
You
got
a
way
to
do
it
manually
and
the
question
is
whether
we
can
abstract
that
out.
I
guess
the
other
headline
is
we
got.
A
B
D
A
A
B
B
C
E
E
Everything
was
in
service
of
this
goal
of
taking
logic
out
of
any
of
components
and
putting
them
in
param
as
we're
talking
about.
Basically
the
idea
there
being
if
we
ever
want
to
add
extension
points
to
Venier
components.
We
don't
want
extension
points
and
logic
to
be
in
the
same
file
because
then
overwriting
us
so
wanna
take
it's
much
logic
out
of
any
ax
and
then
have
kind
of
a
one-to-one
relationship
with
been
a
proponent
and
peregrin
component
hook,
so
that
you
could
easily
create
different
variant
components
without
having
to
re-implement
alter
logic.
E
This
is
for
us,
like
a
pretty
important
way,
to
preserve
flexibility,
as
we
start
to
lock
things
down
with
extension
and
month,
so
what
we
didn't
have
for.
That
was
a
name.
So
if
you
were
paying
attention
to
the
community,
slack
you'll
see
that
our
tentative
name
for
that
was
mix-ins,
because
if
you're
writing,
you
acted
like
you,
don't
write
mix-ins,
you
don't
use
mix-ins
and
so
kind
of
just
went
with
that.
E
But
that
meant
the
wrong
thing
to
a
lot
of
people
and
was
probably
pretty
unintuitive
and,
of
course,
there's
also
an
ancient
concept
and
reactive
mix-ins
with
classes.
So,
instead
we
went
with
this
name.
Talents
suggested
by
James
Cal
common,
because
tones
are
basically
hooks
right
and,
more
importantly,
this
name
can't
collide
with
anything
that
I
know
up,
though
that's
what
this
means.
Talon
specifically
refers
to
a
component
hook,
not
a
hook.
That's
designed
to
be
reusable
like
use
query
or
to
be
a
utility
like
use,
drop
down
or
something.
E
This
is
specifically
for
components
that
already
exist
in
many
ax
and
had
some
logic
in
it.
Whether
it
was
running
a
query
having
a
side-effect,
something
like
that,
that's
what
we're
calling
a
talent,
so
we've
been
going
component
by
component
in
Bennion
when
I
say
we
I'm
talking
about
Stephen,
Drew
and
I,
so
putting
up
PRS
for
each
of
the
components,
inventing
a
one
folder
at
a
time,
so
we're
going
to
walk
through
a
couple
of
pr's
one
of
his
one
of
mine,
I,
guess
that's
also
one
of
mine,
so
we'll
go.
E
So
this
is
what
the
Talon
looks
like
so
here
in
Paragon,
just
number
lib
challenge
off
bar
and
we'll
work
out
the
top
level
exports
later.
You
know
all
the
react:
ports
there's
it
doesn't
pull
in
react
itself,
and
that's
because
we're
not
doing
any
create
elements
in
these
folks,
they're
just
logic,
so
use
use
callback.
We
get
context
the
same
as
of
any
component
would
and
then
it's
just
a
function.
So
it's
gonna
accept
an
object.
E
E
In
this
case
we
just
need
disabled
show.
My
account
show
sign-in,
which
are
going
to
be
functions,
and
then
we
get
some
information
from
context
and
all
we're
really
doing
in
this
case
is
creating
a
couple
of
callbacks
couple
of
handlers
and
then
from
the
disabled
logic
we're
creating
a
more
explanatory
name
of
what
it
does
so
deriving
some
sub-state-
and
this
is
this
return-
this
return
object
shape,
is
basically
exactly
what
the
component
needs.
So
if
we
come
down
to
the
off
bar
component.
E
So
if
there's
logic
around
what
specifically
to
render
that'll
still
exist
in
the
component,
because
the
component
is
in
charge
of
what
gets
returned
with
what
elements
get
created.
But
the
hook
will
handle
any
logic
which
may
be
pretty
simple
or
maybe
pretty
complicated
as
we'll
see
another
in
other
things
like
checkout.
E
You
can
see
that
logic
for
leaf
components
like
you're,
literally
category
leaf
and
category
branch
tends
to
be
pretty
thin,
so
the
talons
are
pretty
simple
or
something
that's
a
little
higher
up.
You
know
that
contains
kind
of
a
whole
tree
of
logic.
In
this
case
literally
category
tree,
it's
going
to
be
a
little
more
complicated
may
have
side
effects.
E
So
here
we
do
the
fetching
we'll
do
running
of
queries.
Returning
of
the
data
and
the
state
from
those
queries
updating
of
Redux,
as
you
know,
after
the
query,
has
run
and
then
some
transformation
of
the
data
that
we
got
back
so
here
we're
basically
doing
the
first
operation
that
we
do
on
the
data
in
order
to
shape
it
into
child
category
objects
that
can
easily
be
turned
into
the
category
notes
themselves
on
the
front
of
I.
E
E
That
means
that
we're
probably
going
to
need
to
come
up
with
a
with
a
way
to
indicate
from
the
hook.
You
know
through
documentation
or
comments.
What
kind
of
query
it
needs
right
now.
It's
not
gonna
be
an
issue
because
we're
always
gonna
get
right
query,
but
that's
something
that
will
document
in
a
way
similar
to
prompt
things.
E
Underneath
you
have
the
ability
to
create
totally
different
variants
of
components
without
reinventing
the
logic
which
is
really
important,
I
think
going
forward
so
keep
an
eye
out
for
all
of
these
PRS
that
are
rolling
into
the
repo
or
maybe
a
third
of
the
way.
I'd
done.
Half
way
done,
I
would
say,
with
those
components
we're
not
going
to
go
overboard.
A
E
A
E
Generally
speaking,
if
you
are
making
a
an
event
handler,
at
least
it's
it's,
where
you
would
start
considering
making
one
okay,
that's
like
the
most,
that's
just
the
most
common
operation.
You
know
you're
like
this
is
really
worth
it.
Yeah.
E
Talents
tend
to
be
pretty
small
if
you
don't
have
much
logic
to
move
over
the
more
the
more
tricky
question
of
should
I.
Do
this
or
not,
is
whether
the
logic
is
specific
to
the
rendered
UI
or
whether
its
operating
on
props
and
data
so
like
we
saw
there
if
we
have
a
condition
where
we're
deciding
what
to
render
like
between
two
different
trees
of
elements.
Okay,.
A
E
A
A
Sure
I
mean
like,
since
that
makes
a
lot
of
sense,
and
it
seems
like
probably
it's
gonna
be
demand
driven
because
sometimes
there
are
yeah
data
structures
that
clearly
lend
themselves
to
a
UI
concept
that
everyone
likes
and
shares
and
the
demand
for
that
will
become
clear,
so
fair
enough.
That
make
sense.
That's
a
good
answer.
E
Yeah
I
would
not
like
I
said:
I
would
not
go
overboard
with
them,
so
we're
just
we're
doing
them
where
the
need
is
where
the
need
is
obvious,
and
then
what
will
happen
when
we're
done
here
is
we'll
look
back
on
it
and
we'll
see
opportunities
to
combine
things
that
are
basically
doing
the
same
logically,
where
we
we
could
have
abstract
them.
We
just
never
saw
it
or
things
to
simplify
or
hook
or
we'll
just
say.
E
A
F
F
E
C
C
C
So
I
just
thought:
I'd
get
the
team
a
chance
to
do
a
little
round
table
and
say:
there's
there
there's
a
discovery,
here's
their
performance
that
we're
really
looking
to
try
to
move
forward
and,
if
you're
interested
in
you're
on
this
call
or
your
member
of
the
community
in
this
particular
work
stream.
This
is
how
you
can
get
involved.
So
maybe
it
will
start
there
with
move
on
and
then
go
around
and
then
end.
D
The
call
from
there
yeah
I'm
working
on
the
service
workers
and
I'm
trying
to
see
how
we
can
utilize
the
back
door
ease
up
the
build
process
of
also
how
to
speed
up
the
passion
on
the
plant
side.
So
how
do
you
utilize
the
passion
years
in
service
well
present,
because
it's
just
like
a
separate
thread?
That's
a
little
bit
for
us
to
use
it
for
anything
that
can
be
like
caching,
prefetching,
reloading
everything.
G
G
You'll
start
to
see
some
pretty
small
PRS
right
now,
but
that
are
pretty
impactful
and
then
I
kind
of
as
we
go
on.
You
know
that
will
switch,
but
yeah
the
ones
that
are
coming
out
now
should
be,
should
be
pretty
easy
to
look
at
and
digest
and
should
have
a
big
benefit.
So
I
am
doing
all
of
these
tickets.
C
A
Know
there
are
so
many
good
issues
out
there
for
people
to
take
just
as
a
quick
review.
I
went
to
Munich
for
a
contribution
day
at
magenta
Munich,
with
armed
with
little
prototype
of
the
scaffolding
mechanism
and
p.m.
in
it
Magento
PWA.
That
was
updated
for
the
first
time
since
imagine
actually,
and
it
now
works,
although
until
we
publish
it
for
real,
you
still
have
to
use
the
secret
meet
Magento
Munich
dis
tag,
which
goes
a
little
something
like
this
NPM
in
it
at
Magento
/pwa
at
CDN.
A
A
And
the
point
of
this
again
is
so
that
you
can
begin
a
project
and
have
it
connected
to
the
PWA
studio
dependencies
through
real
semantic
versioning
in
your
package
JSON,
instead
of
having
to
fork
our
project
and
then
try
to
pull
in
the
updates
and
changes
through
merging
on
the
in
the
git
workflow
willy-nilly.
The
way
that
only
really
advanced
users
like
Jordan,
etc
or
the
folks
that
JH
can
really
do
so
so
starting
a
project.
A
This
way
helps
and
as
we
discussed
and
as
we
discovered
in
at
experience
on
Monday,
it's
not
so
hard
to
adapt
an
existing
project
from
a
fork
into
this
system
as
well.
So
we
are
creating
a
project
called
community
sync
and
I've
demoed
this
before
so
we'll
go
through
it
real
fast,
but
this
is
going
to
go
ahead
and
download.
Many
a
concept
at
a
particular
version
from
NPM
and
as
a
secret
feature
buried
inside
this.
A
So
anyway,
that
I'm
gonna
go
ahead
and
take
the
focus
away
from
because
the
other
cool
thing
about
meet,
Magento
or
sorry
at
about
CDM,
you
see
was
that
we
actually
got
some
pull
requests
and
merged
them
and,
like
always,
the
core
team
has
done
a
great
deal
of
cool
work,
but
there's
also
been
a
lot.
Let's
see,
those
are
the
open
pull
requests.
They're
also
been
a
lot
of
cool
little
contributions,
and
this
is
stuff
that
even
things
of
this
size
are
really
valuable
to
us,
for
instance,
Lucas
and
I'm.
A
You
see,
though
he
had
never
done
any
react,
work
or
work
on
a
project
of
this
nature
before
he
went
in
and
corrected
a
little
issue
where
we
logged
a
reminder
to
do
it
and
do
a
custom
origin
because
of
a
little
bug
with
how
we
did.
We
detected
it
same
with
the
sticky
attribute
from
pagination,
which
turned
out
to
be
a
small
fix,
but
it's
still
really
valuable
to
have
to
go
in
and
to
do
the
research
and
the
verification
to
do
that,
which
Lucas
did
very
very
quickly.
A
A
The
repository
and
getting
going
the
biggest
thing
that
we
found
that
I
think
we
need
to
address
is
the
the
issue
of
connecting
the
the
PWA
studio
and
venĂa
instance
that
you
have
to
a
local
instance
of
Magento,
and
we
worked
really
hard
to
make
sure
that
you
don't
have
to
do
that.
So
we
can
welcome
people
into
the
fold
who
don't
necessarily
know
how
or
have
an
easy
way
of
setting
up
a
local,
the
Gentiles
on
their
own.
So
we
let
you
connect
it
to
cloud
instances.
A
However,
that's
sort
of
because
we
didn't
cover
that
issue
very
much
over
the
course
of
the
last
few
iterations
of
development,
since
we
can
avoid
it
here
on
the
core
team.
There
was
some
usability
issues
with
it
that
that
arose,
that
we
started
to
see
when
people
in
in
Munich
who
do
have
the
ability
to
create
a
Magento
instance
and
who
are
actually
the
most
familiar
with
starting
their
work.
That
way,
they
noticed
that
there
are
actually
some
problems
and
trying
to
do
it,
the
main
problems.
A
A
But
we
I
need
to
find
the
issue
where
that
happened.
So
this
is
a
great
example
of
the
benefit
going
to
contribution
days
for
us
as
a
team
that,
even
though
we
we
have
great
advantage
in
speed
in
doing
things,
our
own
way
bringing
in
people
from
the
Magento
world
is
going
to
help
us
notice
things
that
we
wouldn't
otherwise
notice.
A
So
between
that
and
a
fantastic
experience
that
I
had
going
and
sitting
down
with
Jordan
that
experience,
they
were
kind
enough
to
host
me
and
to
blow
some
productivity
on
spending
a
day
showing
me
their
stuff
and
talking
to
me
about
accessibility
concepts.
We
have
a
whole
bunch
of
stuff
in
process
and
in
the
pipeline
as
issues
which
is
going
to
help
with
that
tremendously
and
and
eater
Nova
talked
about
that.
A
Just
for
knowing
that
we
are
doing
the
underlying
data
layer
right
and
in
a
way
that
scales,
so
that
people
can
build
features
without
worrying
about
slowness,
and
not
only
that,
but
dev
has
been
opening
issues
and
Steven
issues
and
of
course
mr.
Calculon
has
always
issues
about
documentation
and
then
revanth
who's
very
new
to
the
project
has
opened
a
whole
bunch
of
issues
about
well
looks
like
these
cover
performance
as
well
as
testability
and
everything
else.
So
take
a
look.
A
If
you
were
on
this
call
from
the
community
and
you've
been
waiting
to
find
something
that
you
can
really
engage
with.
No,
no,
we
haven't
groomed
these
issues
and
assigned
them
Help
Wanted
tags
yet,
and
we're
going
to
do
that.
But
if
you
want
to
take
some
initiative,
get
out
ahead
and
grab
stuff
that
that's
going
to
have
a
big
benefit
without
too
much
work
on
your
part.
Now
is
the
time
it's
more
possible.