►
From YouTube: Magento GraphQL Community Meeting, 8 November 2018
Description
Agenda:
-) GraphQL-60: Manage Stored Payment Methods (@Pablo Fantini)
-) GraphQL-43: Added new products to WishList by SKUs (@aleksliwork)
-) GraphQL-142: [Mutations] Cart Operations - Add configurable product to Cart (@Yevhen Sh.)
-) GraphQL-214: Incorrect returned values of product_count (@Yaroslav Rogoza)
- Mage2Gen GraphQL demo (@LewisVoncken)
GraphQL Community Project Wiki: https://github.com/magento/graphql-ce/wiki
Mind Map: https://www.mindmeister.com/1094880388?t=a5LVAQ71Vq
Backlog (with Zenhub): https://github.com/magento/graphql-ce#boards?repos=128075669&showPRs=false
Backlog (without Zenhub): https://github.com/magento/graphql-ce/issues
Zenhub Chrome Add-on: https://chrome.google.com/webstore/detail/zenhub-for-github/ogcgkffhplmphkaahpmffcafajaocjbd
A
A
Some
some
topics
here
around
stored
payment
methods.
First
then,
we'll
move
on
to
adding
new
products
to
wishlist
theis
queue.
Next
one
is
adding
configurable
products
to
cart.
Our
contributor,
unfortunately,
is
not
here
today
for
that
one,
but
will
he
recorded
a
little
video
that
will
that
will
share
with
everyone
and
then
the
last
one
is
we
had
a
bug
come
through
about
incorrect
to
return
values
for
the
product
count
yield.
A
B
B
B
A
A
My
question
is
so
normally
you
can
so
a
shopper
can
add
a
card
in
the
current
luma
theme
when
they,
when
they
place
an
order
right
when
they
pay
with
a
card,
they
have
an
option
to
save
it
for
later
and
that's
how
they
that's,
how
they
add
the
card
to
their
accounts.
But
with
with
this
mutation,
are
you
able
to
add
a
card
to
your
account
without
running
a
payment
protection?
Yes,.
B
A
A
B
A
Yeah
and
I
think
I
was
gonna.
Yeah
I
was
gonna
point
this
out,
because
so
my
my
assumption
is
that
we
use
we
store
tokens
that
are
not
visible,
visible,
I'm,
I,
I'm,
assuming
means
just
whether
it's
showing
on
storefront
to
the
chopper
is
that
right?
Yes,
okay,
so
we
I
mean
is
that
we
use
non
visible
tokens
to
manage
things
like
basically
like
order
management
order,
management
actions
by
the
admin
rights.
Things
like
partial
capture
or
partial
refund.
A
A
Exactly
so,
it
should
only
it
should
only
always
return
yeah,
what's
visible
to
the
customer
and
not
show
any.
You
know
it
shouldn't
have
that
property
basically
exposed
I,
think
that
that
makes
sense
to
me
unless
we,
you
know,
unless
we
actually
do
use
non
visible
tokens
on
the
storefront
for
some
operations,
but
I,
don't
think
so.
C
B
A
There
was
a
little
bit
of
there's
work
going
on
basically
independent
lead
to
on
the
on
deleting
of
tokens,
and
because
currently
we
don't
remove
the
the
token
from
the
payment
processor
when
we
call
delete
tokens
and
there
was
there-
was
a
request
basically
to
for
us
to
start
doing
that
so
I,
don't
know
if
that
work
ever
got
complete,
but
let
me
actually
I
found
it.
Let
me
let
me
just
send
this
to
you
Pablo
in
case
you
want
to
take
a
look.
A
I,
don't
know
if
it's
100%
relevant,
but
something
that
I'll
put
in
the
chat
right
here
in
the
zoom,
so
that
anyway,
they
may
not
have
an
impact,
but
just
just
in
case
you're.
Just
so
you're
aware
it's
it's
I
think
I
think
I
was
deleting
tokens
is
probably
okay,
so
cool.
Thank
you.
Let's.
Let
me
ask
if
anybody
else
on
the
call
has
any
questions
about
the
token
payment.
Token
management.
A
D
E
D
So
pretty
it's
pretty
simple,
because
it's
almost
obvious
for
it
to
load
us
it's
adding
new
products
to
lists
by
Eskimos.
So
what
I'm
going
to
show
is
just
process
is
pretty
simple
too.
So,
as
you
can
see,
I'm
created
some
new
customer,
and
here
is
my
wishlist
and
it's
empty,
and
what
I'm
going
to
do
is
just
to
add
some
new
product
I
created
an
admin
panel,
so
simulation
is
pretty
simple
and
everything
is
correct.
D
I've
based
my
poor
quests,
basically
on
the
functionality
created
before
by
some
other
guy,
so
I've
included
his
all
his
code
into
my
pool
request
and
he
actually
approved
it
so
and
I,
based
on
my
logic,
basically
on
my
own
wish
list
items
and
wish
list.
Basically
here
is
how
it
looked
on
the
basically
nothing
so
complicated,
I
just
add
in
again
this
new
notation
and
everything
is
based
on
the
wish
list
and
will
return
this
list.
So
I
wish
list
item.
A
D
Actually,
I'm
not
sure,
basically
using
the
wish
list
was
created
before
I
think
we
can
maybe
put
it
some
somewhere
in
the
common
method
or
in
the
common
class
I'm,
not
sure.
Now
it's
on
the
progress
in
the
progress
and
I'm
going
to
cover
everything
with
API
functional
tests
so
and
I'm
on
the
protists.
Now
on
this
process,
so
I
haven't
finished
it
and
basically.
D
D
I
think
I'll
cover
it
in
the
future.
So
if
I
started
this
I
think
it
would
be,
it
would
be
a
right
thing
to
to
do
everything
according
to
this
epoch,
I
mean
so
what
I
am
I'd
wanted
to
do
is
just
cover,
maybe
function,
functional
tests
and
I'm
going
to
create
maybe
delete,
update
methods
mutations.
So
that's
my
plan.
Okay,.
A
Great
yeah,
that
sounds
that
sounds
really
good.
I
think
that
would
be
so
be
awesome
if
we
can
get
update
functionality.
Working
to
remember
the
tricky
thing
with
wish
lists
is
that
for
EE
or
commerce,
now
right,
we
we
can
have
multiple
wish
lists.
So
I,
don't
remember
what
we
decided
in
terms
of
developing
for
open-source
versus
seat
versus
commerce.
A
But
that's
that's
something
like
you
know
could
be
I,
don't
know
if
we
ever
came
up
with
a
way
of
developing
graphic
you
all
for
commerce,
but
for
now
for
open-source
I
think
everything
is
in
good
shape,
because
you've
got
one
wish
list
there
that
your
that
your
creative,
basically
managing
or
even
adding
products
to
so
just
need
to
future-proof
it
for
to
make
sure
that
it
will
work
with
multiple
wish
lists,
anyways,
just
just
a
comic
okay,
any
anyone
else.
Any
other
comments
or
questions
about
wish
lists.
A
Yeah,
so
the
question
is
about
developing
general
development
of
open
source
versus
commerce
I.
So
our
intentions
for
now
to
keep
developing
for
open
source
only
I
think
since
we
do
have
partners
involved
that
do
have
access
to
commerce
as
well.
We
want
to.
We
want
to
make
sure
to
enable
those
guys
to
develop
for
for
commerce
if
they
want
to
I
I'm,
actually
a
little
bit
unclear
on
what
what's
needed
there
on
our
end,
if
anything,
but
well
let
us
regroup
internally
here
and
we'll
we'll
provide
an
answer
that.
A
F
It's
adding
configurable
product
to
the
cart.
Let's
start
so,
I
want
to
show
you
this
scheme
of
configurable
product
I'm,
adding
the
notation
of
the
name.
Is
it
configurable
product
to
the
cart,
to
cart
and
summer's
over
it's
pretty
simple
I
will
fool
you
yeah,
that's
it!
So,
let's
turn
to
our
test.
F
Put
this
hash
to
our
cart,
ID
copy,
this
copy,
sorry,
and
what
we
have
we
have
our
methyl.
We
have
our
cut
ID.
We
have
our
carte
items,
data
and
SKU
wish.
We
want
to
add
the
way.
Let's
call
this,
it
is
the
same
so
and
what
we
want
to
have
it
as
a
result:
our
ID
quantity
and
product
name
and
school.
So,
let's
start
and
run
it
this
it.
F
C
D
A
A
G
So,
basically,
in
current
implementation,
this
field
represents
the
corresponding
field
in
the
catalog
category
entity
database
table
and
actually
it's
not
the
most
efficient
way
for
getting
the
category
product,
because
first
of
all,
this
value
is
not
very
useful
on
the
store
phone
because
it
represents
the
number
of
all
all
products
in
the
category.
No
matter
are
they
this
product
sellable
or
not,
and
the
second
issue
is
this
value
in
the
database.
For
some
reason
it's
not
always
correct.
G
For
example,
in
my
instance,
you
can
see
even
some
negative
values
here
so
yeah
and
we
decided
to
and
also
regarding
this
issue.
We
have
some
related
issues
created
by
by
some
users,
for
example.
It
here
is
the
first
issue
that
says
that
products
count
should
really
be
equal.
The
number
of
fetched
products,
as
you
can
see,
we
like
help
only
four
products
here,
but
products
count
46
and
the
second
issue
it's
already
closed,
but
it's
like
it
shows
that
this
value
might
be
a
little
bit
confusing
for
users.
G
For
example,
if
user
requests
products
he
sees
that
he
has
one
product
in
the
category,
but
no
items
in
return,
so
my
have
created
a
separate
resolver
on
this
matter
to
to
resort
to
fetch
like
amount
of
profits
that
are
separable
and
that
will
be
fetched
by
using
product
items
queries.
So
if
we
have
only
two
products,
visible
and
so
visible,
I'm,
invisible
in
catalog
and
the
products
are
in
stock
and
have
positive
quality
and
just
just
to
show
the
results.
So
that
was
the
previous
result.
G
For
example,
we
have
only
one
item
return,
but
this
filled
product
count
shows
that
we
have
four
items
and,
after
implementing
a
separate
resolver
this
like
well,
you
represent
the
real
amount
of
products
that
will
be
returned
in
products
items
so
queer.
Also,
this
resolver
does
not
affect
that
performance
should
not
affect
the
perform
a
significant
way,
because
in
a
scope
of
this
resolver
we
don't
laught
at
the
products
collection.
We
only
get
like
this.
The
size
of
this
collection
implementing
some
additional
joins
to
the
products
collection,
but
we
don't
load
that
collection.
G
A
Cool,
thank
you.
So
one
thing
that
comes
to
mind
from
at
least
from
my
perspective
is
that
we
want
to
make
sure
that
this
works
with
whenever
the
whatever
the
store
displays
products
that
are
out
of
stock,
or
you
know,
there's
a
configuration
to
basically
whether
you
display
those
products
or
whether
you
don't
display
those
products
but
in
in
both
cases.
It
needs
to
be
reflected
accurately
right
if
I'm,
if
I'm
a
merchant
that
I
choose
to
still
display
those
items.
A
G
That
make
sense,
yeah
yeah,
of
course,
so
I
took
into
account
disappointment
also
and
I
use
it
to
to
determine
the
product
accounts
are
used
the
same
like
a
collection
processor
that
is
used
for
the
product
section,
so
they
will
be
the
same.
If
the
store
is
configured
to
show
the
out
of
stock
products,
they
will
be
shown
here
and
they
will
be
counted
in
the
product
counts.
Oh
cool.
A
C
Maybe
also
need
to
make
sure
that,
depending
on
current
store,
we
have
also
correct
numbers,
because
it's
possible
that
some
products
on
special
stores,
because
they
just
belong
to
different
root
categories,
so
I'm,
not
sure
you
I,
would
just
test
this
and
make
sure
that
it
works
correctly.
In
that
scenario,
as
well.
G
A
A
So
Tom's
yeah
Tom's
been
for
those
of
you
on
the
call
who
haven't
met.
Tom
Tom
is
gonna,
be
it
was
brought
on
to
the
community
hearing
team
to
help
with
our
testing
efforts
so
he's
he
has.
Basically,
the
original
developers
are
MF
TF
framework,
and
so
he
has
deep
expertise
in
that
field,
and
so
we
want
to
try
to
share,
share
the
knowledge
and
share
the
wealth
that
he
possesses
in
terms
to
to
bring
the
testing
kind
of
robustness
right
to
our
community
projects.
H
A
Awesome
yeah,
we've
noticed
a
lot
of
you
know:
new
people
keep
joining
the
project
and
contributing
which
is
has
been
really
awesome
because
you
know
I
feel
like
this
is
great
great
thing
to
work
on
and
we
we're
always
always
happy
to
to
welcome
new
contributors
and
new
developers
to
the
project.
So
it's
been
really
nice.
So
thank
you
guys
for
your
hard
work.
I
know:
we've
got
veteran
contributors
that
have
been
with
us
for
a
long
time
now
and
not
to
not
to
forget
those
guys
that
appreciate
your
work.
Obviously,
too.
So
sorry.
H
A
A
So
I'll
wait
for
Lewis
for
a
little
bit
longer
here,
but
it
it'd
also
be
good
to
maybe
give
us
a
little
bit
of
context.
I
thought
I
have
seen
you
in
the
like
PWA
channels
mostly,
but
if
you
can
give
us
some
background
of
where
you
what
you've
been
working
on
or
maybe
you
know
what
your
company's
been
working
on
and
where
your
interests
lie,
I'd
be
great
to.
I
E
I
A
J
A
I
Right,
okay-
and
this
is
my
local
environment,
in
which
you
know
as
much
as
you
give
the
Thunder
name
from
its
experience,
a
module
name,
and
then
you
have
all
kinds
of
parts
which
can
and
yes,
for
example,
you
can
create
the
block,
just
simple
block:
you
give
it
a
class
name
and
maybe
a
function
you
want
in
it,
you
can
define
for
which
cope
it
has
to
go.
You
give
a
layout
handle
with
it,
and
then
it
generates
the
code
exactly
what
you
need
for
that,
then
you
can
even
download
it
you
can.
I
There
is
just
food
come
online,
you
get
it.
You
can
also
create.
New
modules
can
sign
a
and
whole
stuff,
but
where
is
where
it's
about?
Is
we
have
no
like
a
switch
version
currently
on
the
life?
It's
two
point.
One
entry
point
two
and
a
new
one:
it's
going
to
be
the
highest
two
point
three,
but
in
default
we
start
with
two
point:
two,
because
everybody
is
going
on
the
developing,
of
course
frontier.
So
if
you
now
go
to
two
point
three,
what
is
the
difference?
Currently?
I
Currently,
it's
only
the
graph
QL
snippet,
which
you
can
simply
add.
You
give
the
gloss,
because
I
thought
take
a
look
at
the
magenta
strip
here.
It
has
always
separate
module
for
every
graphic,
well
API,
for
example
the
CMS
SCMS
graph
QL.
So
if
you
take
a
look
and
you
make
work,
that
is
for
the
gloss
experience
example
work.
I
I
I
I
Probably
and
for
example,
if
you
do
based
on
a
model
you
have
so
if
you
have
previously
generated
block
content,
we
will
admit
gray,
open
for
and
a
Web
API,
but
if
you
guys
going
to
Africa,
of
course,
and
then
it
generates
the
same
one,
but
in
case
of
block
you,
but
probably
one
both
cuz
I'm,
you
want
the
overview
page
and
a
single
page.
But
what
is
your
opinion
about
it?
C
Is
clear
like
why
do
we
have
a
repository
interface
requirement
when
we
create
graphical
models
so
in
general
and
graph
here?
What
hitting
it
like
schema
and
if
arbitrary
resolver
and
there
is
even
no
need
to
have
data
provider
like
if
it's
available
in
CMS,
it's
just
how
Sheamus
is
implemented,
but
it's
not
any
requirement
for
graphical
in
general,
so
no
I
understand
but
I
face
it.
I
C
I
J
I
C
But
we
don't
need
any
repository
to
fetch
data,
for
it
can
be
real.
Collection
can
be
like
anything
like
we
do
not
restrict
like
in
current
public
IP
is
tricky
that
it
must
be
on
top
of
existing
service
contracts,
which
is
usually
like
repository
for
corporations.
Yes,
Tokyo,
you
don't
have
such
restriction,
it
can
anything
and
usually
in
most
cases
we
will
probably
even
not
rely
on
existing
service
contracts
just
for
performance
reasons.
Just.
I
C
J
E
C
Since
you
usually
do
not
depend
on
the
positive,
like
I'm
suggesting
that
you
will
generate
this
advice
to
deposit
errands
and
developers
should
go
and
remove
repository
that
a
provider
because
most
likely
he
won't
in
it.
Thank
you.
Why
don't?
We
just
generate
the
basic
template
which
is
common
anywhere,
so
for
that
we
need
to
have
like
a
module
as
you
as
you
did,
then
we
schema
is
probably
one
query
and
I'm
sure
is
over
and
that's
it.
Okay.
C
Think
we
have
because
it
will
be
basically
example
without
any
functionality,
but
we
have
some
simple
method:
it
already
has
functionality
so
yeah
we
can
have
it
go
yeah.
We
can
have
this
and
discuss
exactly
what
can
be
included
in
this
default
template
which
makes
sense
for
everybody
yeah,
and
my
name
is
Alex
Polaris
by
the
way.
So
you
can
just
reach
me
out
on
slack
on
community.
I
A
I
A
Yeah,
this
is
yeah
thanks
thanks
thanks
for
sure
Mesa
devil.
This
is
a
definitely
a
cool
idea.
I
I
wasn't
aware
of
this
tool.
I,
don't
know
if
I've
seen
it
but
I
yeah.
So
the
point
yeah,
the
point
I
understand-
is
to
kind
of
automatically
generate
this
like
boilerplate
code
right
and
then
for
developers
to
to
kind
of
fill
in
fill
in
what
they
need
to
fill
in
yeah.
I
A
E
A
Definitely
willing
to
work
with
you
on
that
Thanks
thanks
for
sharing
Lewis
Andy.
All
the
feedback
is
appreciated.
Thank
you.
Thanks
awesome.
Well
with
that.
Unless
anybody
I'll
give
I'll
give
people
in
the
room,
maybe
one
more
one
last
chance
to
bring
up
any
topics
that
they
feel
they
want
to
discuss
this
week.