►
From YouTube: FE Pairing - 20200918 - Let's Review some MR's
Description
In this pairing session we walk through reviewing some MR's together.
00:00:00 - Start
00:01:07 - Start reviewing ["Remove unused HAML"][1]
00:15:30 - Create [follow-up MR][2]
00:19:30 - Start reviewing ["Add info_line"][3]
00:48:45 - Start reviewing ["test for vscode extension"][4]
[1]: https://gitlab.com/gitlab-org/gitlab/-/merge_requests/42210
[2]: https://gitlab.com/gitlab-org/gitlab/-/merge_requests/42713
[3]: https://gitlab.com/gitlab-org/gitlab/-/merge_requests/42443
[4]: https://gitlab.com/gitlab-org/gitlab-vscode-extension/-/merge_requests/97
A
A
So
let's
do
I
did
just
get
back
on
this
one
and
that
one
is
and
the
pipeline's
running
so
that
one
should
be
fine.
I'm
just
gonna
go
up
and
we'll
check
it
out.
Oh,
this
would
be
really
cool.
This
has
to
do
with
the
vs
code
extension
and
I
haven't
even
checked
out
the
vs
code
extension
at
all,
so
that
would
be
really
cool
I
kind
of
but
I'll.
Let
you
see
I'll.
Let
you
pick
one
of
these.
If
you
think
here's
one
about
removing
unused
handle
templates,
that
sounds.
A
A
B
A
A
A
B
A
A
B
A
A
B
No
idea
what
this
file
is
well,
it's
just
on
the
comment
at
the
top,
but
basically
hamline's
creates
a
bunch
of
to-do
items.
Okay,
so
in
case
you
were
skipping
on
your
hammer.
Linter
hamelin
to
do
basically
keeps
track
of
all
the
things
that
you
skip
rules
on,
so
we
can
come
back
to
it
later
and
fix
it
okay,
but
we
got
rid
of
this
so
yeah,
probably
in
the
next
hamlet,
probably
a
master.
A
B
B
A
C
A
B
D
A
C
A
B
A
What
do
you
think
can
we
can
we
really
get
is
is
searching
for
fa
times
enough.
Do
I
have
to
like
look
for
like
icon,
remove
or
something
I
don't?
I
don't
think
so
because
yeah,
because
we
do
this
fa
icon
exactly
so.
If
I
do
like
icon,
I
don't
I'm
not
I'm
looking
for
mainly
just
like
in
hamel
or
ruby
or
something
here
we
do
the
sprite
icon.
A
D
B
A
A
A
A
A
A
B
A
B
B
D
A
A
A
A
D
A
A
A
A
A
So
I
like
to
keep
my
like
things
that
I'm
waiting
for
I
kind
of
push
them
over
here.
In
my
browser
I
have.
I
have
two
buckets.
I
put
my
tabs
in
and
they're
separated
by
my
email
and
slack.
I
put
all
the
things
I'm
waiting
for
to
the
left
and
the
things
my
active
tabs
over
here
to
the
right.
While
reviewing
this,
I
noticed
you
notice.
There
was
a
helper
that
isn't
being
referenced.
A
B
D
A
A
B
B
A
A
A
A
A
A
D
B
B
A
A
But
we
moved
our
oh
yeah,
we
we
okay,
we're
still
importing
these
messages,
but
we
moved
them.
We
don't
have
to
expose
them
to
the
template
anymore,
which
is
cool
it's
nice.
These
are
still
coming
from
doc,
link
geo
link,
target
blank
content,
href
message,
dot,
link,
yeah
that
that'll
and
it's
expiration
policy,
help
page
path.
Okay,
that
all
sounds
nice
I'd
like
to,
I
think,
also
looking
at
the
tests
and
seeing
how
the
tests
have
changed
would
be
helpful
at
just
confirming
that
that
this
is
done.
But
this
this
makes
sense.
A
B
B
A
A
A
A
A
C
A
A
A
A
A
D
A
A
A
B
A
Away
from
this
guy,
because
I'm
joking,
no,
what
what
I
just
done
here
was
previously
we're
doing
the
this,
this
expectation
to
say
hey.
We
expect
us
to
not
equal
that
we're
containing
this,
and
this
is
like
a
super
permissive
expectation,
because
if
I
had
even
like
typoed
something
here,
it's
not
it's
gonna
pass
like
there's
very
rarely
will
it
not
pass.
We
were
like.
We
did
then
copy
these
lines
down
here.
A
A
If,
if
you
can-
and
I
think
we
can
in
this
place,
I
think
it
makes
it
a
bit
more
more
helpful.
So
that's
the
comment
that
I'm
gonna
leave.
Oh,
it
looks
like
we've
done
it
a
couple
of
other
times
too
over
here,
yeah
and
we'll
just
do
the
same
thing
here
too
yeah.
So
we're
doing
this
not
to
equal
thing
and
I'm
suggesting.
Let's
just
keep
it.
C
A
Let's
just
do
it
when
the
expiration
is
message
does
not
exist,
message
does
not
exist.
This
is
when
it
does
exist,
and
here
he
says
when
there
are
no
yeah.
This
is
the
same
thing
as
messages
not
exist
yeah.
So
let's
do
this
that'd
be
fun.
That
might
be
a
nice
thing
to
leave
fine.
So
the
way
that
I
would
do
this
since
we
do
have
like
some
descriptions
and
stuff
here,
I
I
usually
have
a
description
as
one
of
the
parameters
to
my
parameterized
spec.
A
A
A
B
A
C
B
A
A
B
A
Yeah,
so
what
is
the
case
that
makes
it
true?
This
case
makes
it
where
it
does
show
up.
So
then
I
want
to
just
flip
the
other
bits.
I
want
to
do
one
where
images
count
and
that's
what
this
one
is
checking
over
here.
I
want
to
do
one
where,
when
there
are
no
images,
when
there
are
no
images,
so
I
want
to
do
it
where
it
was
false,
but
image
count
is
zero,
like
that
one
should.
A
Pass
yep
yeah
and
if,
if
I
was
gonna,
do
this
really
well
like?
I
would
also
like
to
put
in
here
like
here's.
The
expectation
like
these
are
the
ones
the
values
I
expect.
I
expect
it
to
sometimes
show
it
up
and
of
expected
not
to,
and
then
it
makes
it
so
obvious
of
here's,
the
one
that
works
and
then
did
we
check
it
out
when
we
flipped
the
other
bits
that
it
doesn't
show
up.
A
One
is
only
helpful
if
I,
if,
if
we've
done
false
and
one
and
checked
this
one-
and
I
think
it's
worth
and
that's
something
that
showed
up
because
we
moved
it
to
parameterize
spec,
I
think
that's
worth
bringing
up
so
I'm
gonna,
I'm
gonna,
bring
it
up,
but
I'm
gonna
delete
this
one
because
we've
captured
that
now
the
way
I
do
this.
Yes,
let
me
check
here
so
I've
left.
A
A
A
A
B
B
A
Yeah,
I
I
find
it
to
be
helpful
of
reducing
some
noise,
but
I
don't
I
don't
do
it
all
the
time.
Sometimes
I
go
really
far
and
I'll
say
and
I'll
add
below
this
and
I'll
say
you
can
apply
this
patch
by
running
like
by
copying
it
to
your
clipboard
and
like,
but
I
I
think
I've
already
sent
patches
to
robert.
If,
if
I'm
not
sure
someone
you
know
has
used
patches
before
then
I'll
just
leave
it
once
I
usually
do
that
for
like
community
contributors
or
something.
A
A
A
A
A
A
A
I
think,
but
here's
the
thing
too,
is
I
don't.
I
don't
review
this
reviewing
process
as
like,
even
though
we
have
approvals
and
stuff
like
that,
like
I
really
want
to
emphasize.
This
is
a
collaboration
effort
like
we're
we're
just
collaborating
on
it,
anyways
cool
yeah.
I
have
some
some
small
comments
here,
but
then
we
should
be
good
to
go,
but
it
looks
like
nico
created
it,
but
would
you
mind
doing
oh,
he
did
create
it.
D
A
A
We
test
that
the
pipeline
status
bar
item
gets
correctly
populated.
It
seems
a
bit
finicky,
but
they
are
reliable,
looks
when
they
start
failing.
Okay,
it's
not
really
easy
right,
but
vs
code
debugging
helps
a
lot.
It's
very
cool,
I'm
gonna,
I'm
gonna!
I
know
I'm
gonna
wanna
review
this
locally.
Do
you
wanna?
Do
it
on
your
screen?
Do
you
wanna?
Do
you
wanna
drive
it.
D
A
A
I
think
that's
hard
to
do.
I
think
it'd
be
hard
to
review
someone
and
like
have
them
be
there
not
just
because
not
because
it's
like
anything
personal,
but
I
think
it's
just
be
so
hard
not
to
create
all
the
discussions
like
I
feel
like
you'd,
have
lots
of
discussions
that
would
be
generated
right
and
not
like
that's
a
bad
thing,
but
it's
just
maybe
maybe
some
of
these
discussions.
It's
really
good
that
they're
asynchronous
yeah.
That's
fair,
anyways,
yeah!
Okay,
so
I
check
this
out.
I'm
gonna
run
my
little
script.
A
A
A
A
A
A
Yeah,
yes,
and
I'm
excited
to
start
learning
more
and
more
about
the
the
vs
code
integration
stuff.
I
really
really
really
want
to
bake
in
my
whole
workflow.
I
do
here
into
the
vs
code
like
that
would
be
amazing
like
if
I
could
see
comments
leave
comments
here.
I
can't
I
can't
wait
for
that.
You
wouldn't
leave
bs
code.
A
Well,
I
think
it
was
already
async,
but
we
we
didn't
yeah.
It
was
already
async,
but
I
guess
we
weren't
actually
waiting
for
it.
Here's
another
interesting
bit
too,
so
we
we're
creating
the
pipeline
status
and
we
also
set
this
interval
for
every
30
seconds,
we're
going
to
try
to
refresh
the
pipeline
status
yep,
but
we're
not
awaiting.
A
Yeah,
are
you
familiar
with
with
with
reactive,
extensions,
reactive
extensions,
not
really
yeah
rxjs
rxjs,
oh
okay,
yeah
that
one
yeah
they
have
baked
down.
So
it's
like
a
whole
another
paradigm,
but
they
have
really
baked
in
doing
like
these
kind
of
non-deterministic
stuff
of
like
we
want
to
pull
something
and
that
thing's
going
to
run
an
asynchronous
thing.
A
I
mean
really
really
well
like
where
I
can
pull
it,
but
don't
do
it
again
until
the
thing
finishes
like
and
that's
the
scary
thing
that
could
happen
here
is
since
we're
not
really
waiting
for
anything
we
trigger
it.
It
takes
forever,
then
we've
triggered
it
again
and
like
and
then
now
there's
a
race
condition
of
what
happens
and
reactive
extensions
helps
just
everything
go
through
one
stream
and
it's
and
there's
a
it's
very
esoteric
because
it
uses
some
very
specific,
like
function
names.
B
A
A
Streams-
and
so
I
would
say,
go
is
a
little
bit
more
like
that,
but
like
rxjs,
and
so
let
me
see
if
I
can
find
rxjs
polling
thing.
Api
or
interval.
Oh
yay,
yeah
yeah,
so
like
all
of
their
methods
like
they
all
have
these
like
marble
diagrams
of
like
how
they
get
hits
or
whatever,
and
so
like
here,
you're,
creating
an
interval
of
whatever,
and
so
none
of
it
actually
is
going
to
take
place
until
you
subscribe
to
that
observable.
A
But
you
could
do
something
like
okay,
we're
going
to
create
an
observer.
We've
we've!
You
know
this.
Marble
is
going
to
drop
every
thousand
milliseconds,
but
then
we're
going
to
map
each
one
of
those
to
an
asynchronous
call.
Then
we're
gonna
like
join
it
somehow
and
like
make
sure
that
we
don't
drop
another
marble
until
our
previous
one
was
done,
and
but
it
just
ends
up.
You
end
up
with
just
this
huge
chain
of
like
it's,
a
very
functional
kind
of
approach
to
things.
A
A
A
B
A
B
A
A
A
A
A
A
B
A
A
A
B
C
B
A
B
A
I
wonder
no
and
that's
what's
interesting
too,
with
projects
like
this
that
are
like
really
fresh
like
it's
it's
hard
finding
a
you
know,
especially
when
you're
just
reviewing
you
know
finding
the
focus
for
yourself,
but
then
also
on
the
mr
itself,
like
what
things
do
I
want
to
highlight,
or
not
so
we've
upgraded
and
now
it
fails
on
unhandled
requests.
Oh,
I
see.
Oh
okay,
that's
great.
A
A
B
B
A
A
A
Oh,
they
want
to
use
typoscript
yeah,
and
that
was
that
was
a
conversation
as
and
I
think
that
kind
of
thing
would
be
super
helpful
because
it
was
concerning
that
we
had
like
calls
that
were
actually
async,
but
we
didn't
await
them.
Using
typescript
would
have
caught
that
right,
but
I
think
what's
a
little
tough
is
it's
already
a
product?
It's
like
already,
it's
not
a
huge
project,
but
it's
still
existing
and
it
has
like
no
tests.