►
From YouTube: Magento MSI Open Demo. January 18, 2019
Description
Today's agenda:
- Current week update. `1.1.0` preparation. Test Coverage report and Test Plan for Distance-Based algorithm by @Slava Moskalyuk
- Distance-Based algorithm API stabilization by @TheRick / @iminiailo
- Magento 2.3 PWA + GraphQL + MSI for POS application by @Alex from Magestore
- MSI upgrade notes by @Lori Krell
B
Will
mute
Issa,
who
is
typing
very
loudly
so
by
the
way,
it's
nice
nice
to
meet
you
nice,
to
have
you
in
our
meeting
Caesar?
So
it
is,
it
is
just
at
this
time
you
decided
to
join
the
MSI
community,
but
we
it's
better
later
than
never.
So
we
help
you
welcome
on
board
and
we
we.
We
really
expect
that
you
will
be
exactly
the
same
eager
and
as
on
PWA
slack,
so
we're
waiting
for
your
contribution,
ideas
and
so
on.
So
let's
follow
our
agenda
and
we
will
start
so.
B
As
you
know,
we
are
working
on
the
on
the
small
release.
Actually,
it's
not
really
small,
so
we
introduced
three
quite
a
big
story
there.
So
we're
going
to
use
a
support
of
elastic
search
for
custom
stocks.
We
introduced
because
you
know
currently
MSI
and
especially
custom
stocks
are
supported
just
on
a
default
just
under
my
sickle
adapter
and
regarding
the
elastic
search
we
have
the
we
have
the
you
know.
B
B
So
we
are
trying
to
make
this
API
as
it's
more
official
efficient
as
possible,
so
that
now
so
Allah
will
tell
you
a
little
bit
more
regarding
the
how
we
are
preparing
for
the
smoke
testing,
which
we
have
to
do
before
before
the
release,
one
dot,
one
dot
all
of
the
MSI
package.
So
so
our
feel
free
to
share
your
screen
and
share
with
your
date.
C
142
of
them
is
automated
and,
according
our
last
manual
testing
cycle
on
that
moment,
we
have
500,
550,
Mario
and
the
result
of
I
think
was
at
468
was
past.
It
well
well
fail.
20
was
skipped
in
5,
a
blocked
now
automated
dating
our
optimization.
Now
on
the
25%
of
all
testing
area,
what
the
ownership
test
and
by
severity
it
splitted
that
57%
of
the
have
0
0
to
reality.
0
1,
have
statistics,
pretend
and
7
second
severity,
60
of
them
about
catalog
40,
about
configuration
32
about
sales.
C
1325
of
them
passing,
which
is
87%,
30
escaped
and
test
creation
trends.
So,
since
7
of
December
of
last
year,
we
have
70
data
and
I
renew
added
to
hit
test,
and
they
all
not
automated
it,
because
formation
is
still
on
the
same
level,
but
seven
of
them
they
on
PR
way
pending
to
mainline
and
I.
Think
that's
all
for
report
from
Tom
and
now
I
will
tell
you
about.
Today.
C
We
created
first
way,
of
course,
settings
and
configurations
that
distance
algorithm
working.
We
enter
a
PK
and
said
other
configuration
settings.
Then
we
create
free
in
my
test
scenario,
three
different
sources,
two
of
them
in
United
States
and
one
in
Kiev
in
stock
I,
said
that
key
of
stores
have
no
priority
and
then,
in
this
scenario,
place.
C
Order
from
New
York
one
source
in
Culver
City
another
in
Austin,
so
according
to
distance
selection,
algorithm
system
have
to
choose
the
closest
city.
What
is
Austin
for
New
York
if
I
am
no
ghagra
vehicle
and
also
another
testing
area
cover
this
case
when
you
are
not
have
enough
quantity
on
one
of
the
source,
so
source
selection,
the
winds
supposed
to
propose
you
to
deduct
on
one
source
and
the
rest
from
another
source.
B
B
Yeah,
it's
pretty
important
for
us
to
have
all
this
use
cases
to
be
to
be
documented,
because
that
this
new
type
of
algorithms
into
use
is
quite
a
lot
of
quite
quite
a
lot
of
business
cases.
Because,
as
Allah
mentioned,
we
support
different
different
mods.
So
we
support
both
offline
mode.
When
there
is
no
any
online
provider
provider
is
involved,
as
well
as
online
mode,
where
we
use
Google,
as
as
a
provider
for
the
for
the
geo
coordinates
and
they
calculate
and
the
calculator
of
the
distances
between
two
points.
B
So
that
also
like
we
have
different
computation
mode.
So
we
will
provide
the
ability
to
deliver
for
merchants
to
make
a
logistic
based
on
driving
mode,
whether
a
cycling
mode
or
just
walking
mode.
Depending
on
how
particular
particular
orders
are
delivered,
so
that
all
of
these
different
types
of
use
cases
supposed
to
be
covered
with
corresponding
with
corresponding
use
cases,
because
there
are
Ave
a
lot
of
way.
A
lot
of
cases
need
to
be
tested
even
in
the
scope
of
the
smoke
testing,
so
that
we
want
to.
B
So
now,
that's
probably
all
regarding
the
the
preparation
for
for
update
for
msi
for
a
messiah
new
release,
and
you
see
that
how
we
prepare
and
like
all
the
testing
stuff
for
this
upcoming
release,
and
now,
let's
proceed,
probably
for
the
for
this
for
the
most
brilliant
feature
in
the
scope
of
this
release.
So
the
feature
contributed
by
ricardo,
the
distance
base
algorithm
and
the
ricardo
will
shed
some
light
on
on
on
refinement.
B
A
A
E
B
I,
so
let
me
let
me
share
with
you
a
little
bit
more
technical
degree
background
on
this
topic,
so
the
main
part
was
so
record
introduced
this
distance
based
algorithm,
an
actual
distance
based
algorithm,
as,
according
to
this
modularity
new
modularity
privileges.
We
are
way
following
following
in
magenta
2
2,
3
4.
All
the
newly
created
features,
so
the
distant
base
algorithm
is
actually
represented
by
three
modules.
One
module
actually
introduced
the
eight
guys
for
the
distant
pace
of
Gerizim
implementation,
part
and
the
admin
UI
part.
B
So
for
all
of
these
functionality
we
have
a
dedicated
module
and
regarding
the
API
part,
we
have
we
have
the
distance
provider
and
we
have
the
service
which
which
calculates
the
distance
by
by
address
and
the
most
interesting
part,
and
the
part
which
we
discussed.
The
most
was
because
this
newly
created
algorithm
should
be
injected
into
our
existing
system
and
extend
the
our
existing
extension
point
so
that
we
have
the
inventories
of
source
selection
API.
So
this
is
module
and
the
invented
resource
selection.
B
Api
already
provides
one
out
of
the
box
implementation,
which
we
released
in
the
scope
of
our
one.
You
know
in
the
scope
of
our
source
to
release
of
MSI,
so
the
priority
based
algorithm,
where
we
actually
did
not
use
any
inbound
data,
but
we
just
use
the
configuration
data
so
the
priority
your
sources
in
the
scope
of
in
the
scope
of
this
talk
to
which
the
sources
are
assigned
to,
but
taking
into
account
that
we
got
this
newly
created
algorithm.
The
distance
based
algorithm
and
distant
based
algorithm
need
additional
inbound
data
to
be
provided.
B
You
know,
in
our
case,
this
data
actually
is
represented
by
the
address.
So
you
can
see
here
on
this
precise
listing
that
we
had
built
in
our
inventory,
request,
object
and
the
before
that.
In
our
previous
version
of
algorithm,
it
was
a
key
just
to
specify
the
stock
ID
and
all
the
items
requested
in
the
scope
of
in
the
scope
of
the
request.
B
So
we
so
the
items
actually
just
a
pairs
of
the
SKU
and
quantity,
but
since
since
we
need
to
provide
the
the
most
nearest
sources,
so
the
most
nearest,
where
warehouses
for
the
to
the
shipping
address-
and
here
we
need
to
provide
the
shipping
address.
That's
why
here
you
see
that
this
is
implemented.
We
are
the
extension
attribute
and
we
said
this
extension
attributes
to
our
inventory
request
entity,
and
why
do
why
do
we
use
extension
attributes
because,
potentially
so
inventory
request
interface?
B
So
potentially
this
object
could
be
extended
with
any
other
data,
so
that
potentially,
is
there
any
father
or
the
algorithm
appear,
which
should
apply,
for
example,
gender
segregation.
It
would
be
easy
to
edit
with
additional
extension
attribute,
for
example,
if
they're
gonna
be
sources
which
provide
just
boots
for
men
or
just
for
women,
it
would
be
easily
to
lose
it
to
use
this
extension
attribute.
B
B
B
And
this
is
how
the
inventory
source
selection
API
looks
like
also,
it's
pretty
interesting
part
how
we
like
releasing
the
new
version
of
algorithm.
We
introduced
some
like
we
had
to
apply
some
changes
to
already
released
files
to
release
files
and
taking
into
account
that
in
magenta
we
we
used
to
proceed
with
non
non
with
non
backwards.
Incompatible
changes
in
the
scope
of
the
patch
releases
I
will
show
you
how
we
try,
try
to
mitigate
all
the
all
the
possible
backwards.
Incompatibilities
and
I
did
this,
probably
not
in
the
scope
of
this
pull
request.
B
This
is
just
a
model,
and
it
happened
that
after
some
refactoring
we
we
no
need
anymore
two
parameters
which
are
injected
to
the
constructor,
and
these
parameters
were
typed
parameters,
so
the
precise
object
represented
was
here,
but
keeping
this
type
is
not
really
good,
even
if
you
don't
use
this
parameter
anymore.
So
it's
not.
It's
not
sufficient,
just
to
remove
the
private
variable
here,
because
what
will
happen
in
this
case,
the
object
manager
will
still
create
this
object
and
will
inject
this
to
the
constructor.
B
B
But
what
we
are
doing,
we
are
just
duplicating
them
clearly
and
say
that
it
should
be
new
here.
We
just
remove
the
type
hinting
removal
type
hinting
happens
just
to
prevent
object
manager
to
create
this
dependency,
so
this
will
increase.
So
this
will
boost
performance
a
little
bit
and
also
we
will
just
no
need
to
create
a
needed
dependency,
but
in
any
case
object
manager
will
try
to
create
something
so
to
prevent
object
manager
to
create
something
we
have
to
specify
the
GI
and
the
in
the
GI.
B
We
have
the
exercise
new
so
that
actually
it
means
that
GI,
our
object
manager
will
just
inject
inject
new
in
this
case,
doing
all
of
this
stuff,
if
actually
help
us
to
eliminate
dependency,
which
we
no
need
anymore
and
by
the
way
I
checked
in
our
in
our
backward
compatibility
guide.
That
removal
of
the
deprecated
code.
B
Regarding
the
removal
yeah
so
the
to
preserves
the
constructor
signature,
remove
type
hinting
for
unused
parameters
to
remove
dependency
on
their
type
and
add
suppress
volume.
But
here
we
don't
have
we
don't
have
statement
mentioning
that
the
GI
configuration
should
be
added
if
you
find
the
excess
I
type
new
for
this
parameter,
because
if
somebody
will
proceed
just
with
two
of
these
items,
actually
he
will
have
the
fatal
error
at
the
time
of
the
object
instantiation,
because
the
object
manager
is
not
able
to
instantiate
required
parameter
without
typing.
B
So
so
that's
probably
a
good
point
for
our
backward
compatibility
guide
to
be
updated,
and
this
is
how
you
see
that
on
a
messiah,
we
preserve
the
backward
compatibility,
updating,
updating,
updating,
even
entities
which
are
not
marked
with
an
API
HPD
block.
So
this
is
pretty
pretty
interesting
part
of
what
I
wanted
to
highlight
here
in
this
demo
meeting
and
as
you
seen
record,
the
just
merge
my
pool
request
into
his
one
and
I
believe
will
process
historic,
west
and
north
into
to
the
tree
develop
today.
B
So
that's
all
regarding
update
on
distance
based,
algorithm
I
believe
we,
you
would
be
able
to
reach
the
distance
based
algorithm
in
our
mainline
branch,
even
in
the
really
near
infusion,
but
now
I
want
to
stop
sharing
my
screen
and
I
want
to
provide
a
possibility
to
colleagues
from
my
store
to
make
a
presentation
of
the
really
interesting
stuff
he
he
will
present
the
post
application,
which
actually
uses
all
the
newly
new
features
provided
by
magenta
to
the
tree.
Stop
so
I'm
stopped
stop
talking
and
just
the
stage
is
yours,
Alex.
F
F
F
F
F
Here
you
can
see
is
a
POS
location.
It
is,
maybe
is
a
physical
store
observation
and
we
consider
it's
like
a
new
search
channel
for
Magento
and
it
connected
with
stock
amazing
tool
and
this
stock
connect
with
only
one
sauce
and
our
pipe
sale
will
read
the
data
from
stop
and
write
two
updates
even
to
reduce
sauce.
So
we
follow
the
rule
of
M
CSI
that
we
reach
a
resonator
from
stock
and
Verizon
data
to
sauce.
F
F
G
F
F
F
D
F
F
When
CMS
III
see
that
basically
I
think
these
architecture
of
CMS
I
see
is
a
really
good
for
for
the
luncheon
with
have
many
physical
store
and
when
they
want
to
do
too
many,
the
inventory
separate
from
test
or
the
MSA
is
very
easy
to
too
many
it,
and
when
we
music
us,
we
we
choose
a
concept
you
to
use
a
new
sale
channel
from
us
in
tow.
Besides
about
the
website,
and
it's
a
really
easy
to
integrate
yeah.
F
B
By
the
way,
maybe
there
are
some
some
pitfalls
or
drawbacks,
which
you
consider
should
be
fixed
in
the
scope
of
MSI.
So
when
we
are
talking
before
you
mentioned,
the
sales
channel
possibility
extension
of
the
sales
channel
and
because,
for
example,
as
I
understood
for
your
purposes,
it
definitely
makes
sense
to
use
the
pause
as
a
sales
channel.
So
not
the
magenta
website,
but
the
post
at
the
terminal
to
be
used
as
an
independent
sales
channel
mm-hmm.
F
Yeah,
you
can
see
here's
a
stock
of
the
Bands
OMSI
and
in
Sochi,
though
we
you
can
see
the
website
mundo
here
and
we
add,
the
PS
looks
in
here
is
considered
like
a
new
search,
enormous
intro,
mm-hm
yeah.
So
when,
when
we
go
to
to
the
stock
of
mass
into
here,
you
can
see
the
website
search.
You
know
and
previous
search
a
new
here.
F
C
F
B
So,
but
actually
I
can
I
can
show
you
how
to
customize
this
part
because
you're
exactly
correct,
so
how
we
return.
We
under
the
hood,
we're
on
the
source
selection
algorithm,
choose
the
the
most
priority
source
and
a
return
to
this
source.
This
is
how
how
its
implemented,
but
if
you
want
to
specify
money
early
only
which
source
you
want
to
make
the
return
you
have
to
customize
this
part.
Actually,
there
is
no
no
more
customization,
not
a
lot
of
customization.
We
did.
You
need
just
to
cuss
one
or
two
classes
to
make
that
possible.
B
C
F
D
Alex,
if
you
have
trouble,
locating
the
W
studio
announcements,
any
materials
or
resources
that
are
associated
with
it
feel
free
to
ping
me
in
slack
I'll
either
provide
you
the
resources
that
you
need
or
connect
you
to
the
right
people
to
the
PW
818
yeah,
because
it's
thinking
you
know
thing
for
us
to
see
what
early
adopters
of
the
PW
a
studio
can
do
and
get
getting
your
feedback.
They
shoes
that
you
encounter
or
things
that
you
like
would
be.
B
So
think
your
Alex
for
your
presentation
and
feel
free
to
continue
provided
feedback
on
whether
MSI
or
PW
a
studio,
or
you
also
mentioned
that
you
use
graph
kill
so
like
this-
are
all
project
where,
like
we're,
both
community
engineering
team
is
involved
either
this
project
very
interested
for
our
poor
team,
because
this
a
very
big
our
features.
This
is
just
a
beginning
so
that
many
new
updates
would
be
rolling
out
in
the
scope
of
each
of
this
tracks.
B
So
keep
us
posted
on
your
feedback
and
we
will
provide
you
our
feedbacks
on
improvements
we
doing
in
the
scope
of
issues
districts
so
that
the
next
one
gonna
be
Laurie
last
but
not
least,
definitely
not
least
because
documentation
is
the
highest
priority.
Sometimes
it's
even
higher
than
in
fact
writing
code.
So
worried
a
lot
more
about
about
what
sure.
G
All
right,
hello,
everyone,
so
a
really
quick
but
very
important
documentation
update,
is
in
our
dev
Docs.
We
do
you
have
a
page
that
walks
through
how
you
want
to
manage,
install
and
now
update
or
upgrade
MSI.
Now,
if
you
have
a
fresh
install
of
Magento
2.3,
either
open
source
or
commerce,
you
will
automatically
receive
MSI.
It
will
be
replacing
all
inventory
management,
as
we've
talked
before,
but
say
you're
on
a
2.1
or
2.2
release
line,
and
you
are
really
eager
to
get
upgraded
to
2.3
and
use
msi.
G
These
modules
will
all
be
disabled
by
default.
This
is
to
ensure
that
any
custom
code
you
already
have
in
place.
It
doesn't
get
you
knocked
out.
You
don't
have
issues
if
you
say
you
use
order
management
from
Magento
things
of
that
nature,
but
enabling
them
is
very
easy
and
I've
added.
Some
information
here
in
this
guide.
G
I'll
also
provide
a
link
in
slack
and
there's
a
link
in
Twitter
I'll
go
ahead
and
get
moving
again,
that
kind
of
walks
through
what
you
need
to
do,
how
to
enable
these
and
including
a
lot
of
other
documentation
links
also
in
the
dev
docks.
Now,
when
we
get
to
the
new
package
for
a110
of
MSI,
which
is
going
to
include
all
that
great
things,
we've
talked
about,
including
that
distance
based
algorithm
and
all
the
performance
enhancements
which
I
have
to
I've
been
using
them
they're
amazing.
G
We
will
have
a
couple
extra
steps
documented
if
needed,
but
we're
hoping
that
this
will
be
a
seamless,
easy
update,
but
just
in
case,
if
you
do
have
to
take
an
extra
step,
we'll
get
it
documented
and
we'll
let
you
know.
We
also
have
release
notes
in
draft
that
you
can
find
in
the
wiki
for
MSI
as
well
as
those
are
now
getting
prepared
to
go
live
as
soon
as
the
package
goes.
Live
for
additional
documentation,
updates
that
are
in
the
pipeline.
G
Kevin
here
in
Devdas,
is
also
working
on
some
API
documentation
updates
to
support
the
new
source
selection
algorithm
for
distance
based.
So
we
should
start
seeing
a
PR
moving
around
that
you
guys
can
kind
of
get
a
quick
peek
into
those
as
well.
That's
all
I
have
this
time.
Unless
you
have
questions
or
feedback
happy
to
take
those.
B
G
B
D
D
Our
meeting
with
the
entire
community
engineering
team,
where
we're
going
to
present
a
quick
status
update
where
we
are
and
goals
for
2019
everybody's
going
to
be
presenting
Igor
is
gonna,
be
there
I'm
gonna,
be
there
everything's
gonna,
be
there
we're
gonna
be
having
some
fun
talking
about
what
we
have
for
you
and
what
kind
of
help
who
they
need
and
talking
a
little
bit
about
how
you
can
join
us
and
help
us.
So
please
check
out
our
community
calendar
for
that.