►
From YouTube: PWA Community Meeting February 15, 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
C
C
That's
a
little
introduction,
I'm
Lewis
I
work
at
experienced
here
in
the
Netherlands,
and
we
acknowledged
in
Magento
one
that
we
always
wanted
to
generate
modules
because,
as
we
were,
writing
the
same
code
over
and
over
again,
the
mates
gen
was
born
in
2016
and
they're,
currently
still
developing
on
it,
and
one
cool
feature
that
we
added
is
a
source
package
file,
support,
/
2.3
because
of
PV
studio.
So
if
you
switch
to
2.3,
you
will
be
have
the
opportunity
to
use
safe
and
download
package
files.
C
For
a
product
image
carousel
in
P,
V
W,
we
have
to
add
a
custom
component
like
a
carousel,
you
can
select
the
style,
type
s
CSS
or
CSS.
The
reason
I
have
added,
that
is
that
Jordan
is
working
with
as
CSS
my
colleague
in
default,
always
CSS.
Then
it
generates
the
folder
structure.
I
have
added
in
this
case
also
vendor
and
module
structure,
actually
kind
of
like
Magento
module
structure.
So
you
have
also
offender
opportunity
within
your
components.
C
I
know
that's
currently
not
the
the
way
components
are
added
to
the
to
the
team,
but
I
think
that
if
you're
going
to
install
third-party
components,
you
need
to
have
like
a
reference
of
the
vendor,
but
that's
just
my
opinion,
so
it
generates
the
code
for
you.
What
does
it?
It
adds
default
classes
which
you
need
with
the
CSS
files
and,
of
course,
the
basic
index
file.
C
This
one
can
be
downloaded
separately
from
the
module
because,
if
you
have
also
a
Magento
module
really
did
to
this
students
component,
you
want
to
download
it
to
your
Magento
installation
to
your
app
code.
But
in
this
case
you
want
to
edit
the
package
files
for
for
your
veneer
concept,
and
then
you
almost
have
like
the
same
download
functionality.
The
difference
is
that
there
is
a
type
specified.
C
It
makes
a
difference
between
module
and
package
files,
which
is
also
supported.
This
is
a
graphic
QL
endpoint
I've
already
demonstrated
that
earlier,
the
magenta
part
in
the
graphic
well
meeting,
but
now
it
also
adds
an
example.
Query
you
can
implement
within
your
Pina
concept.
So
in
the
queries
it
already
sets
an
example
query
for
you,
with
the
given
input
you've
put
in
the
parameters,
a
question
from
James
that
Lin
came
forward.
Is
that
okay,
that's
good?
But
what,
if
you
have
front
ends,
will
only
develop
for
PVA?
Then
you
need.
C
You
need
to
give
them
like
a
console
command
which
is
used.
Npm
makes
gen
already
gives
the
opportunity
to
do
it
through
income
online.
That
is
written
in
Putin
and
you
have
to
install
some
packages,
so
you
can
eventually
use
it
to
what
I've
been
working
on,
but
I'm
also
having
a
meeting
next
week.
I
think
next
week
with
James
himself
and
some
of
his
colleagues
is
the
following.
If
I
switch
green
now
one
second.
C
C
Then
you
get
an
identifier
back
from
which
you
can
download,
but
of
course,
maybe
in
the
future,
it's
like
it's
already
directly.
How
do
you
say
it
installing
it
within
your
PBA,
but
that
is
something
that
is
work
in
progress.
This
is
just
proof
of
concept
and
will
be
later
talking
about
that
with
james
zetlen.
This
is
the
other
screenshot
where
you
see
okay,
these
are
snippets
which
are
already
available,
from
which
you
can
currently
only
use
two
for
PPA,
but
that's
it
shortly.
Are
there
any
questions.
E
F
James
I'm
sorry,
I'm,
troubleshooting,
some
deployment
stuff,
so
I
need
all
the
internet.
I
can
get
my
videos
off,
but
Louis.
This
is
fantastic
yeah.
Thank
you
guys
for
waiting
on
me
and
thank
you
Louis.
We
did
speak
about
this.
I
was
in
adobe
tech
summit
at
the
time
and
we
haven't
set
up
that
meeting,
but
but
I
will
today.
Okay,
so
just
for
context
for
those
of
you
who
haven't
done
of
a
Magento
modular
extension,
development
mage
to
gen
is
a
really
popular,
really
beloved
tool
and
it's
a
scaffolding
generator
with
graphical
user
interface.
F
You
saw
what
Louis
and
I
discussed
was
that
it's
part
of
the
philosophy
of
pwace
to
do
is
to
enable
you
to
do
developments
just
in
your
area
of
expertise
and
in
your
area
of
interest.
We
wanted
to
see
if
the
sort
of
stuff
specific
to
the
PWA
studio
ecosystem
could
work
independently
and
I,
just
tossed
it
out
as
a
suggestion
like
what,
if
this
was
a
no
generator
and
you
could
use
it
through
NPM
and
then
mage
Jen
connected
to
that,
and
that
was
what
like
six
days
ago
anyway,
Luis
did
it.
It's
done.
D
B
B
F
You
probably
know
you
probably
can
the
one
that's
alias
to
venía
PW,
a
comm
is
still
mid,
deploy
kind
of
troubleshooting.
Why
the
the
last
one
didn't
work,
but
you
can
to
play
it
if
you
have
an
open
pull
request
or
even
if
you
have
a
closed
pull
request,
you
can
fetch
that
individual
deployment,
URL
that'll,
probably
work.
D
B
On
the
board,
you
can
see
those
items
that
are
still
in
progress
and
that
are
still
in
the
review,
a
column
that
are
tagged
for
two
2.0
release.
Moving
those
items
across
the
board
we're
targeting
early
next
week
to
get
the
release
out.
I,
don't
know
if
James
you
have
anything
else
to
say
on
just
B
to
release,
but
in
terms
of
the
details
of
it
or
preview
of
it,
but
that
is
kind
of
the
state
that
we're
in
for
release.
B
F
F
This
product
is
I
think
that
the
way
you
can
think
of
this
is
that
this
release
is
our
private
guarantee
to
the
developer
community
that
we're
going
to
begin
following
semantic
versioning
with
our
changes,
and
that
means
specifically
that
once
two
point
0.0
comes
out
that
that
means
that
any
additional
API
changes
if
they're
additive
will
result
in
two
point
1.0.
If
they're
breaking
it
will
be
3.0
point
out,
so
our
major
versioning
we're
not
sentimental
about
it.
We
make
breaking
changes
so
that
you
can
keep
doing
cool
new
things.
F
That
will
mean
pwace
to
do
four
at
five
and
six
in
short
order,
and
so
we
are
still
working
on
a
plan
or
on
a
sort
of
an
agreement
for
what
versions,
major
and
minor,
we
will
try
to
support
or
accept
bug
fixes
for
all
at
the
same
time,
and
so
we
we
still
need
to
to
get
that
completely
settled.
But
this
is
a
community
project
and
we
want
sort
of
community
input
and
understanding
about
it.
F
Semantic
versioning
is,
on
the
one
hand,
really
clear
and
on
the
one
hand,
can
cause
some
confusion.
So
let
me
just
say
from
the
top
of
my
head
as
I've
been
thinking
for
for
a
while
that
we
all
want
to
have
PWA
Studio,
2.0,
point
O
and
as
soon
as
that's
out,
we
want
to
start
working
on
two
things
at
once.
We
want
to
begin
working
in
one
consolidated
branch
on
the
work
that
incorporates
graph
QL
additions
from
Magento.
F
Two
point:
three
point:
one:
we
would
call
that
release
three
and
then
we
also
want
to
do
a
fast
follow
to
incorporate
things
that
didn't
make
this
release
deadline,
but
it
will
add
quality,
and/or
fix
known
issues,
and
we
would
call
that
hot
fix
2.0
point
one,
and
that
means
that
we
will
be
maintaining
two
deployable
branches
at
that
time.
At
the
same
time,
that
makes
sense.
I
think
we
all
know
that
at
the
very
minimum
that
that's
something
that
we're
going
to
do.
F
What
remains
to
me
is
I'm.
Just
opening
this
to
the
community
floor.
We
want
to
do
this
transparently,
is
to
understand
at
any
point
in
the
future.
How
many
branches
of
code
that
could
potentially
be
released
will
be
operating
at
once.
We
know
that
at
least
there
will
be
a
major
release,
a
hotfix
release
and
potentially
a
minor
release.
But
if
you
look
at
other
JavaScript
projects
or
projects
that
interact
with
the
NPM
ecosystem,
like
say,
react
or
angular
or
view
major
versions
can
be
patched,
but
usually
it's
only
the
most
recent
major
version.
F
That
gets
any
active
iteration
and
we
want
to
make
sure
the
people
in
the
community
are
kind
of
prepared
for
the
way
that
that
will
operate.
So
I
want
to
open
the
floor
about
that
any
concerns
or
any
questions
about
what
the
maintenance
plans
will
be
and
what
the
released
cadence
will
mean
for
your
projects
in
your
work.
F
Is
to
say,
I,
guess
what
would
people
worry
about
with
our
release?
What
are
people's
first
burning
questions
about
the
release?
How
it's
going
to
operate?
I.
Think
if
I
were
one
of
you
guys
that
would
ask
how
is
the
support
between
Magento
2.3
versions
and
PWA
studio
versions
going
to
work?
There's
a
there's,
a
good
question.
I
can
pretend,
when
one
of
you
guys
asked
that
right.
F
So
the
answer
is
fairly
simple:
if
Magento
introduces
a
breaking
change
in
its
API,
then
a
version
of
PWA
studio
and
vania,
which
supports
or
works
with
that
breaking
change
is
also
a
breaking
change,
and
so
that
means
it's
a
major
version
in
Magento.
Two
point
three
point:
one:
there
are
a
couple
of
small
changes
to
the
graph
QL
schema,
which
are
not
purely
added.
It
yeah
an
example
that
a
lot
of
us
have
already
experienced
is
in
the
category
specification.
F
The
two
point
three
point:
one
and
latest
versions
of
Magento
master
that
graphical
schema
has
changed
so
that
small
underscore
image
is
now
its
own
complex
type.
It's
an
object
with
a
string
property
for
path
and
that's
a
breaking
change
in
graph
QL,
because
a
query,
that's
written,
to
expect
a
string
going
to
cause
a
real
error
message,
which
is
sometimes
inconvenient
feeling,
but
the
more
you
get
into
these
strong
types,
the
more
it's
great
to
feel
like.
Oh,
it
failed
fast.
F
It
told
me
something
was
wrong
rather
than
just
have
your
abs
start
behaving
in
mysterious
way,
so
it's
loud
and
it
complains,
but
we
like
it
and
a
breaking
change
like
that
is
believe
it
or
not,
something
that
we
would
call
a
full
major
version.
Just
one
name,
one
change,
and
that's
just
in
order
to
be
completely
clear
about
support,
though
the
version
of
PWA
studio,
whose
graph
QL
query
includes
an
object
for
small
image.
F
F
E
E
F
It's
an
older
project
there
is,
there
is
a
joke
in
the
community
about
Magento
3,
you
say:
Reitsma
makes
t-shirts
that
say
Magento
3
on
them,
but
the
underlying
fact
is
that
Magento
2
point
3
point:
X
is
which
called
marketing
version
a
lot
like
how
Windows
like
deep
down
has
its
own,
like
versions
that
are
like
9
and
10,
and
11
and
12.
F
The
Magento
versions
that
are
used
for
like
semantic
versioning,
are
in
the
composer
JSON
and
it's
like
a
hundred
something
dot,
Magento
versions,
and
when
we
actually
declare
describe
rules,
we
want
to
add,
for
instance,
site
info
schema
information
so
that
when
the
project
builds
against
the
Magento
instance
that
it
queries
the
Magento
instance.
For
that
compatibility,
it
would
be
asking
for
that
version.
F
The
choice
to
call
it
to
point
to
point
one,
even
though
there's
breaking
changes,
that's
more
of
a
marketing
choice
and
it's
a
choice
that
reflects
that
it
is
release
which
enhances
functionality
or
fixes
gaps,
but
does
not
like
add
a
banner
feature:
Sara
Lee,
okay,
yeah,
that
make
sense.
Thanks.
No
problem.
C
One
question
about
versioning:
does
it,
how
do
you
say
the
main
version
of
the
module
which
provides
the
graphical
endpoint?
Does
that
one
go
up,
or
is
that
also
like
a
bug
fixing
them
more?
Like
the
two
point,
three
point
one,
or
is
that
going
to
be
like
what
two
point?
Four
point:
zero
take
a
look
at
module
version
number
well,.
F
First
of
all,
let
me
say
a
big
caveat
that
I
am
NOT,
probably
the
biggest
expert,
on
the
way
that
Magento
modules
and
graph
QL
extensions
are
versioned
even
on
this
call.
But
I
can
tell
you
that
the
way
that
our
graph
QL
system
is
set
up,
there
is
an
underlying
graph,
QL
module
that
provides
the
functionality
and
then
individual
modules,
such
as
catalog
and
cards,
use
that
new
functionality
to
extend
and
describe
additions
to
the
graph
QL
schema.
F
So
if
there
is
a
change
across
the
board
to
the
way
graph
GL
functions
like,
for
instance,
if
it
adds
a
feature
or
fixes
a
bug
in
the
parsing
of
syntax,
then
that
will
be
a
revision
of
the
graph
QL
module
itself
magenta
graph
QL.
But
the
change
in
a
small
image
property,
for
instance,
is
that
is
version.
Changing
the
catalog
module.
F
So
we're
going
to
have
to
be
careful
about
this,
so
the
the
policy
in
the
process
we
all
decide
on,
but
my
hope
is
that
we
will
arrive
at
a
system
where
the
versioning
is
always
clear
that
it's
very
rare
that
you
need
to
make
a
sort
of
a
decision
that
might
be
disputed
with
api's.
For
instance,
if
a
signature
changes
or
a
behavior
changes
or
any
sort
of
return,
type
changes.
It's
breaking
every
time
and
there's
very
little
need
for
discussion
unless
you're
me
and
I
would
argue
about
it
pointlessly.
F
But
I
try
not
to
do
that
and
then,
if
you
know,
if
it's
an
additive
change
to
schema
that
wouldn't
break
anything
else
that
does
duck
typing,
then
that's
a
minor
change
and
if
it's
something
that
fixes
functionality
but
doesn't
change
any
types
or
any
guaranteed
behaviors,
then
it's
a
patch
in
the
real
world,
sometimes
that's
hard.
But
ultimately
you
know
what
the
core
maintainer
x'
would
make
a
call.
F
B
F
Are
a
couple
of
things
we
could
discuss?
I
think
it's
probably.
The
discussion
is
significant
enough
that
it
ought
to
be
recorded
or
textual,
but
we
can
call.
This
out
is
Jimmy
on
the
line
he's
not
well,
let's,
let's
do
kind
of
two
pieces,
but
there's
there's
two
distinct
subjects
to
discuss.
One
of
them
is
about
server-side
rendering
and
we've
been
describing
our
approach
to
this
for
a
while.
It's
not
as
clear
and
simple
as
just
do
react,
server-side
rendering.
F
So
it
can
be
hard
to
describe
we're
committed
to
restating
it
as
much
as
we
need
to,
but
our
approach
has
always
been
that
we
don't
want
to
lock
anyone
in
from
the
largest
merchants,
the
most
wealthy
partner
or
si
down
to
the
smallest,
don't
want
to
make
them
use
a
text
stack
in
addition
to
the
magento
that
they
already
know.
So
the
the
SSR
story
was
always
constrained
by
the
fact
that
we
we
just
don't
want
to
demand
that
you
run
a
node.js
server
on
your
back-end
period.
F
Now,
most
probably
the
larger
deployments
and
those
deployments
that
use
docker
and
other
architectures
kind
of
yourselves.
You
probably
will
it
makes
sense
to
run
a
back-end
for
front-end
and
know
that's
fairly
fast.
You
wrap
it
in
some
caching
and
then
you
can
do
good
server-side
render
that
doesn't
fully
answer
the
question
of
server-side
rendering
a
app
that
loads
dynamically,
but
that's
a
different
story
so
yeah
it
makes
sense
to
run
node,
but
we
worked
very
hard
in
the
PWA
studio
architecture
to
guarantee
that
the
PWA
can
run
even
with
without
a
node
back-end.
F
The
way
that
that's
done
is
a
PWA
has
to
understand
exactly.
What's
the
server
is
supposed
to
do
and
that's
what
the
upward
file
is
there
for
it
to
declare
the
behaviors
of
the
origin
server
that
it
requires
in
order
to
run.
You
know:
here's
where
my
JavaScript
needs
to
load
from
here's,
how
images
need
to
resolve
here's
where
I
expect
graph,
QL
and
rest
endpoints
to
be,
and
one
of
those
rules
is
for
a
given
page.
Is
it
age
it
searched
out
HTML
or
is
it
the
root?
F
F
That's
why
there's
a
property
name
and
the
template
resolver,
and
it's
not
just
mustache,
that
mustache
is
the
default,
so
you
would
write
a
new
renderer,
very
simple
API
for
the
JavaScript
client,
just
string
and
data
end
string
out
and
then
you
would
output
react
and
then
upward
doesn't
really
know
or
care
how
to
do
react.
Ssr,
that's
just
a
mode
of
its
template
resolving
and
that's
our
server-side
vendor.
F
So
we've
talked
about
this
because
at
the
end
of
the
day,
I've
just
spewed
a
lot
of
technical
garbage,
but
people
want
SSR
for
search
engine,
optimization
and
marketing
optimization,
and
it's
it's
not
as
easy
to
do,
server-side
rendering
on
a
reactive
as
some
other
apps.
It's
an
ongoing
question,
but
we
want
to
continue
making
that
better.
We
have
some
early
ideas
about
pre-rendering,
a
lot
of
your
landing
pages,
which
is
something
that
projects
like
Gatsby
j/s.
Do
the
react.
F
Team
is
going
to
add
better
server-side
rendering
for
the
asynchronous
systems
that
we
use
for
progressive
load
like
react,
suspense
and
dynamic
imports.
So
I
think
we
we
will.
We
have
no
rule
against
adding
the
react,
SSR
systems
inside
the
venía
concept,
but
we
don't
want
them
to
be
absolutely
required.
We
would
call
it
a
progressive
enhancement
and
that's
our
policy.
F
That's
our
story
with
server-side,
rendering
it's
open
to
review,
it's
open
to
discussion
and
challenge
as
always,
and
we've
gotten
some
interesting
challenges
about
it,
but
we
we
are
committed
to
making
sure
that
this
whole
technology
stack
is
available
for
all.
So
that's
the
SSR
thing
and
that's
just
kind
of
my
my
five
minutes
on
it
and
the
other
question
is
about
graph
QL
queries
where
a.
H
F
Marketing
so
when
I
say
marketing,
I'm
talking
about
the
term
demo,
social
marketing,
social
media,
so
SEO
and
SML
were
dependent
on
the
outter
systems
that
are
doing
the
inspection
like
that's.
Why
everyone's
always
doing
this
criminology
and
like
guessing
whether
Google
is
really
rendering
your
whole
page?
Are
they
doing
it?
And
it's
GaN
Dex
doing
that
you
know
so
for
SEO,
it's
like,
let's
just
make
sure
that
this
page
tells
Google
or
whoever
else
is
the
authority
on
its
relevance
like
what
it
has
in
it.
F
You
know
metadata
schema.
This
is
a
product
it
can
be
bought
and
that's
a
moving
target.
But
obviously
you
need
to
do
some
managed
server-side
render
and
the
recommendations
from
Google
and
everyone
else
keep
changing
yeah.
But
then
SM,
oh,
like
for
marketing,
is
another
slightly
simpler
subject.
When
you
take
a
link
from
a
store
and
you
paste
it
into
slack
or
Facebook
or
an
iMessage,
then
the
link
in
these
modern
systems,
it
hydrates
it
comes
alive
and
it
has
a
media.
F
It
has
a
little
like
media
object
with
the
picture
of
the
product
and
the
title,
maybe
the
beginning
of
the
description
and,
if
you're
lucky
some
actions
to
take.
So
that's
the
media
object
concept
in
social
media.
That's
in
Facebook,
for
instance,
relies
on
Open
Graph
and
that
also
I
mean
those
systems.
Definitely
don't
hydrate
your
whole
page,
that's
just
slack.
It
is
looking
at
the
HTML.
So
that's
another
reason
to
have
enough
SSR
to
populate
those
objects.
So
it
is.
H
F
And
I'm
being
cheeky
about
most
marketing,
yeah
I
mean
SEO
and
SM
o
is
like
it's
the
phrase
that
we
use
to
roll
up
our
general
obligation
for
many
reasons
to
do
server-side,
rendering
the
page
represents.
Okay,
cool
things
been
sorry,
just
rough,
no
problem,
I
guess
the
other
thing
is
graphical
queries
which
I
wish
Jimmy
was
here
because
he's
the
guardian
of
that
idea
in
that
best
practice,
but
I
understand
it
and
I
can
represent
his
point
of
view.
F
So
we
have
graph
QL
queries
and
separate
files
and
our
projects,
though
not
all
of
them.
There
are
still
files
that
can
include
graph
QL
queries
constructed
in
line
via
the
gql
tag,
and
it's
easy
to
think
that
it's
good
to
encapsulate
it's
good
to
have
them
be
separate
files
that
allows
for
code,
reuse,
and
that
is
indeed
our
thinking
upward-
can
read
graph
QL
files.
So
when
they're
separate
files,
you
can
use
the
same,
get
product,
detail,
query
and
upward
to
resolve
the
and
preload
the
data
that
the
products
root
component
page
also
needs.
F
You
have
to
get
the
whole
product
you
have
to
describe
that
it
needs
the
whole
product
and
that
might
be
like
a
you
know:
40
kilobyte
Chaisson
entity
and
in
graph
QL.
The
idea
is
that
right,
in
line
with
writing
the
react
itself,
you
just
open
a
little
string
literal
to
graph
QL.
You
type
a
query
with
three
things
in
it,
and
so
the
principle
is
that
graph
QL
queries
are
a
part
of
components.
Declaration.
F
That's
the
flexibility
of
graph
QL
is
designed
so
that
you
can
tune
a
query
for
your
particular
views
needs
now.
That's
an
important
thing
to
respect,
so
we
want
to
allow
for
that.
We
we
want
to
clarify
that
our
best
practice
does
not
demand
that
every
graph
QL
query
is
a
separate
document
or
a
separate
file.
In
fact,
we
are
interested
in
future
optimizations
where,
for
instance,
during
the
webpack
bill,
we
could
identify
in
line
graph,
QL
queries
done
with
the
graph
QL
template
literal
and
extract
them
into
files
or
things
like
upward
reuse.
F
But
the
point
is
that
it's
it's
important
for
us
to
remember
that
graph
QL
is
most
important
belief
or
the
convenience
of
the
developer
tying
their
specific
visual
view,
implementation
to
data.
So
that's
something
we
don't
want
to
lose
sight
of,
and
that's
a
summary
of
that
discussion
are
there
others
that
I
have
missed
or
that
we
need
to
talk
about.
I
know
like
at
least
one
thing:
we
have
five
more
minutes
that
the
docker
compose
solution
that
Kayden's
been
working
on
is
really
exciting.
I
Yeah
I
mean
it's
still
a
work
in
progress
in
PR,
so
people
can
check
it
out
if
you
want
to
and
yeah
so
right
now,
I
just
got
something
up
to
have
it
also
support
Windows,
but
I
wasn't
able
to
get
the
automation
with
windows
for
the
cert
and
SSL.
So
if
anybody
is
a
Windows
automation
with
certs
and
SSL
guru
out,
there
happy
to
see
any
kind
of
comments
on
my
request,
but
otherwise
there
are
manual
instructions.
It's
still
in
the
grips
to
set
up,
so
you
don't
have
to
run
anything
extra
it'll.