►
From YouTube: Code Review Weekly Workshop - Aug 19, 2022
Description
In this session we discuss some best practices pertaining to code review. We also mob-pair on a code review.
A
Hey
everybody
thanks
for
showing
up
to
the
code
review
weekly
workshop
yeah.
So
let's,
let's
kick
it
off
with
some
show
and
tell
and
questions
so
I
wanted
to
show
a
couple
of
interesting
things
that
came
up
from
this
one.
Mr,
let
me
go
ahead
and
share
my
screen,
so
we're
looking
at
the
same
thing.
A
A
We
usually
don't
like
to
just
close
them,
we'll
try
to
finish
it
for
them,
if
they've,
if
they
have
stopped
contributing
to
that,
mr,
so
she
opened
up
this
new
mmr
to
finish
it
off.
That
was
all
really
great.
We
added
some
tests,
and
originally,
let
me
see
if
I
can
can
I
started.
A
A
A
It's
pretty
good
to
have
somewhere
in
the
test,
something
that
asserts
either
the
default
state
that
would
have
been
in
just
a
totally
separate
test,
or
it's
even
totally
cool
to
just
put
it
before
the
action
in
the
very
test
itself
and
one
of
the
things
when,
when
we
did
this,
for
instance,
when
we
click
on
this,
we
would
expected
visually.
The
visibility
should
should
change.
A
That
actually
never
changed
because
is
visible,
actually
doesn't
quite
behave.
The
way
that
we
expected
it
to
so
this
test
wasn't
actually
doing
everything
we
thought
it
was
so
that
happens
a
lot
whenever
you're
doing
an
action
that
could
change
state
to
get
really
good
test
coverage,
asserting
the
default
state,
either
in
a
separate
test
or
in
the
same
test.
That's
doing
the
action
having
a
before
and
after
expectations
can
be
really
helpful.
A
And
the
one
other
point
about
that,
mr
was
not
sure
if
you
all
are
aware
that
we
have
a
whole
process
for
finishing
merge
requests,
and
this
is
usually
done
by
mr
coaches,
which
work
closely
with
the
community
to
help
triage
their
mrs
and
help
community
contributions
get
pushed
forward,
but
really,
like
any
engineer,
can
get
ping
to
finish
a
community
contribution
and
probably
the
biggest
task
here
that
might
be
unknown.
Besides,
just
hey
we're
just
going
to
finish.
A
This
thing
is
to
try
to
preserve
some
sort
of
original
commit
that
gives
credit
to
the
author.
I
don't
think
it
can
actually
be
a
git
trailer.
I
don't
think
our
change
log
process
supports
author
as
an
actual
git
trailer.
I
think
it
actually
has
to
be
the
author
of
a
commit,
so
I
need
to
change
this
wording
because
I
don't
think
that's
correct
and
I
hope
someone
can
correct
me
if
I'm
wrong.
A
So
what
I
had
done
here
in
this,
mr
I
just
rebased,
because
I
don't
think
deepika
was-
was
aware
of
that,
because
it's
tucked
away
in
mr
coach
documentation,
but
I
had
just
rebased
the
commit
she
made
and
then
I
chose
to
edit
that
first
one
and
then
I
just
amended
it
but
applied
the
author
of
the
commit
that
we
were
coming
from
because
that
old,
mr
actually
had
like
merge
commits
in
it.
A
A
I'm
pausing
because
I'll
just
talk
for
an
hour.
I
need
to
force
myself
to
stop
all
right
cool
thanks.
Where
are
we
at
sam
yeah.
B
So
I
just
had
an
mr
that
I
reviewed
this
week
where
I
felt
a
bit.
B
I
was
swimming
in
the
deeps
for
a
while,
so
I
had
to
I
felt,
like
I
had
to
read
into
a
lot
of
around
it
in
context
to
figure
out
what
was
actually
going
on
and
I
felt
like
it
took
me
two-ish
days
to
actually
start
actually
reviewing
the
code
and
not
just
trying
to
catch
up
to
where
this
is
I'm
kind
of
wondering
what
it
might
might.
It
have
been
better
to
hand
it
off
to
a
domain
expert.
C
Yeah
that
one
is
something
I
struggle
with,
sometimes
when
the
mrs
have
a
lot
of
like
already
discussion
on
the
mr,
or
also
a
very
long,
I'm
looking
at
them
or
now.
Maybe
like
a
very
long
description,
I
would
say
I
am
encouraged
by
they
have
instructions
on
how
to
set
it
up
locally.
So
I
feel
like
that
would
make
me
maybe
look
at
it
more
than
if
there's
no
instructions
on
how
to
validate
it.
B
There
were
people
already
commenting
on
it,
but
not
as
a
reviewer.
If
I
got
that
correctly.
A
For
me,
I
feel
like
the
non-domain
expert
review
is
like
really
really
valuable
so,
like
I'm,
I'm
not
inclined
to
just.
If
something
is
hard
to
understand,
if
you're
not
familiar
with
it,
that's
a
really
good
signal
of.
A
Maybe
the
thing
is
hard
to
understand
in
general,
so
I
I
I
appreciate
from
time
to
time,
I'll
I'll
put
mrs
and
I'll
get
people
reviewing
it
and
their
perspective
is
just
so
valuable
because
I
realize
I
realize
that
I'm
taking
a
lot
for
granted
of
how
familiar
I
am
with
this
code
base
and
I
really
appreciate
all
the
questions
you're
asking.
So
that's
I
would
say
yeah
at
some
point,
maybe
maybe
it
is
warranted
like
hey.
I
don't
think
I
can
do.
A
Let's
try
to
make
sure
either
the
maintainer
is
a
domain
expert
or
another
domain
expert
looks
at
this,
but
just
doing
a
flurry
of
all
of
your
questions.
That
would
be
great.
I
I
feel
like
so
you
don't
have
to
feel
like
you
have
to
learn
it
all
yourself
from
scratch,
but
just
saying
I'm
unfamiliar
and
I
have
a
bunch
of
questions
about
it
and
sometimes
I'll
even
just
leave
non-blocking
questions
and
that
can
create
follow-up
issues
because
things
yeah
are
legitimately
confusing.
C
Yeah
I
often
find
myself
struggling
with
like
graphql,
mrs
even
like
I'm
a
maintainer,
but
I
don't
do
brexit
well
at
my
team.
Doesn't
they
don't
we
don't
work
in
graphql?
So
every
time
I
look
at
it,
I'm
like?
Oh,
you
know,
and
I
have
to
go
like
refresh
myself,
but
it
is.
C
A
For
me,
that's
why
I
set
a
timer
and
it's
like
if
I
can't
understand
this
on
my
own,
I
move
to
starting
asking
questions
like
if,
if
the
author
was
sitting
right
next
to
me,
what
questions
would
I
ask
to
to
try
to
understand
this
because
yeah,
if
you're,
trying
to
understand
something,
that's
really
outside
your
domain,
you
could
easily
spend
a
whole
day
on
it
and
that's
more
than
you
expect
or
more
than
you've
originally
budgeted
for.
D
Joined
late,
sorry
about
that,
I
I
joined
because
I've
just
become
a
reviewer
and
yeah.
It's
super
intimidating,
like
it's
sort
of
like
I
feel
like
I'm
just
sort
of
still
getting
up
to
speed
with
what
my
team
looks
after
and
even
there
there's
a
bunch
of
questions
I
have
so
yeah
it's
it's
kind
of
a
bit
scary.
I
mean
the
I
I'm
I'm
guessing,
because
I've
just
become
a
reviewer.
D
All
of
the
things
I've
been
asked
to
review
are
very
simple
changes
that
they're
just
very
easy
to
approve
so
far
but
yeah,
I
am
kind
of
dreading
sort
of
you
know
getting
pinned
on
something
where
I'm
just
like.
What,
because
a
lot
of
the
merge
requests,
my
own
team,
I'm
still,
you
know,
sort
of
going.
Okay
what's
happening
here,.
A
And
there's
so
much
history
to
the
gitlab
code
base
that
yeah
so
much
inconsistencies
that
yeah
I
mean
it's
like
it's
like
on
as
a
brand
new
reviewer,
it's
kind
of
like
uncovering
an
ancient
civilization
you're
like
oh
wow.
This
totally
changes
my
perspective
on
the
entire
civilization,
the
yeah.
A
Well,
hey,
that's
that's
awesome,
but
I'm
excited
for
you
to
having
fun
with
code
review,
that's
a
great
way
to
learn
about
the
code
base
and
for
us
to
break
silos
yeah,
and
then
I
would
just
suggest
that
the
main
value
I
feel
like
we're
trying
to
hit
with
that
initial
code
review
is
collaboration
and
so
don't
feel
like
don't
feel
a
whole
lot
of
pressure
with.
D
Yeah
when,
when
you
said
that
earlier,
I
sort
of
felt
a
bit
of
relief,
I
was
like
okay,
good
yeah,
I
mean
the
other
thing
is,
I
think,
like
as
a
developer,
I
don't
think
code
reviews
been
a
strong
part
of
my
you
know.
Like
yeah,
I
don't
think
it's
one
of
my
stronger
skills,
and
so
I'm
also
keen
to
improve
stuff.
A
That's
a
great
that's
a
great
question:
does
you
have
any
or.
C
D
C
D
Terry
was
that
when
you
say
manual
testing,
you
mean
like
kind
of
kind
of
running
yeah.
C
I
check
the
branch
out
get
gdk
up
and
running
and
I
try
if
it's
something
that
can
be
like.
That's
like
a
ui
or
like
an
api
that
I
can
run
also,
especially
with
like
the
graphql
stuff.
I
don't
know
graphql
super
well
so,
like
manually
testing
that,
like
I
come
from,
I
used
to
work
as
a
performance
tester
so
like
I'm
always
like.
I
want
to
break
it.
Yeah.
D
C
C
Oh
yeah
good
suggestion.
I
see
paul's
writing
about
review
apps.
I
I
just
only
recently
figured
out
what
those
were
so.
E
A
Let
me
let
me
maybe
I
have
one
already
open,
so
I
can
show
what
it
looks
like.
Yes,
all
right,
yeah.
Let
me
share
my
screen,
so
I'm
gonna
say
review
apps
yeah,
you
can
get.
A
You
know
you
can
get
locally,
but
or
if
the
pipeline's
passed
and
you
know
cross
your
fingers
and
everything
of
this
thing
deployed
correctly,
there's
usually
a
button
that
says
review
to
view
the
app
and
then
our
login
credentials
are
in
our
our
engineering,
1password
organization,
and
so
this
this
this
could
be
an
easy
way
to
test
some
changes.
Some
changes
need
like
configuration
stuff
before
you've
started
the
jdk,
and
clearly
we
can't
do
that
here.
A
There
is
a
way
to
like
turn
off
and
on
feature
flags
with
review
apps
that
there
should
be
a
way
yeah,
but
it's
through
the
api-
and
I
always
forget
like
how
to
do
it-
it'd
be
so
cool
if
we
could
do
a
rails
console
into
the
review
app.
I
wish
we
could
do
that,
but
I
don't
know
maybe.
F
F
A
I'm
not
really
sure,
I'm
not
entirely
sure.
I
get
nervous
about
adding
keys
to
my
keychain,
because
sometimes
I
drop
and
lose
my
keys.
I
misplace
them
yeah.
You
asked
a
great
question
about
just
general
best
practices
and
I
wanna.
Maybe
we
can
talk
about
it
for
just
a
couple
more
minutes.
I'm
gonna
do
a
a
shameless
plug
of
a
communication,
best
practice
that
has
really
revolutionized
how
I
communicate
on
mrs
called
conventional
comments.
A
So
whenever
I
leave
review
comments,
I
always
preface
it
with
a
label-
and
this
is
actually
more
for
myself
as
much
as
maybe
not
more,
but
it's
as
much
for
myself
as
the
the
person
reading
it,
because
if
I'm
leaving
a
suggestion,
I
should
probably
leave
a
suggestion
me
just
saying
this
isn't
right,
it's
in
a
suggestion
and
then
also
just
being
able
to
tag
a
whole
bunch
of
things
as
non-blocking.
A
So
if
I
notice
user
facing
issues
I'll
leave
these
issue
labels
and
stuff,
so
I
describe
a
handful
of
labels
that
I
find
helpful
and
I've
seen
used
in
the
wild.
So
you
might
see
that
happening
in
mrs,
u
reveal
or
author,
but
this
has
been
really
helpful
for
me
because
I'm
also
don't
like
to
be
thinking
about
how
I'm
going
to
phrase
something
a
lot.
So
this
actually
automates
a
lot
of
that
thought
process
of
how
I
want
to
phrase
something
are.
B
B
Yeah
and
I
just
do
like
double
exclamation
cc
and
then
it'll
pop
up
a
menu.
Give
me
like
a
that's
a
selector
like
what
kind
of
comment
would
I
like
to
leave
and
then
it'll
just
paste
it
in
there
like
pre-formatted
and
everything?
That's
pretty
neat!
That's.
D
A
sort
of
utility
application,
yeah.
B
That
no,
I
need
to
look
where
I
found
it.
I
did
tweak
it
a
little
to
my
preferences,
but
it
was
more
or
less
complete.
A
There's
a
lot
of
like
browser,
extensions
and
stuff
for
conventional
comments,
which
is
if,
if
you
like,
installing
third-party
software
in
your
computer,
you
may
should
go
for
it.
A
G
G
D
A
A
We
have
a
browser
extension,
that's
for
conventional
comments.
We
have
one
like.
I
would
just
suggest
if
you're
gonna
do,
that,
I
think
everyone's
has
their
own.
This
is
you're
getting
to
personal
philosophy,
but
like
we
have
our
own,
that
is
open
source
and
so
that's
a
little
nice.
But
it
is
scary
when
you
install
browser
extensions
and
they
want
all
sorts
of
privileges
like
we'd
like
to
you
know,
we'd
like
to
deposit
to
your
checking
account
like
that's
a
big
red
flag.
D
I
don't
know
saying
that,
but
yeah
yeah
tend
to
avoid
houses.
A
Cool
thanks
for
bringing
that
up
carla.
That
was
a
that
was
a
good
discussion,
good
question,
yeah
good
luck
and
have
fun
code
reviewing.
B
One
thing
I'd
also
like
to
add:
I
like
to
start
with
reading
what
the
issue
is
about
before
I
even
dive
into
the
code,
a
to
know
what
it's
about
and
a
little
bit
more
death
and
b
to
see,
if
the
mr
actually
does
what
the
issue
was
trying
to
ask
for,
because
I
I'll
notice
myself
if
I'm
working
on
an
issue
for
a
longer
time-
and
I
haven't
reread
it
in
a
while-
I
might
have
an
idea
of
what
I
think
I
was
trying
to
solve,
but
it's
actually
different
and
then,
when
I
re-read
the
issue,
I'm
like.
B
A
Cool
does
even
have
anything
else,
they'd
like
to
add
to
the
show
and
tell
their
questions.
Part
of
this
session
coming
up
is
pair,
reviewing
which
I'll
be
eager
to
get
to,
because
we
didn't
really
get
to
have
a
whole
lot
of
time
last
time,
but
if
someone
does
have
something
that
would
be
valuable.
Let's,
let's
talk
about
it,
sure.
A
All
right,
let's,
let's
jump
into
some,
let's
jump
into
some
code
reviews.
Does
anyone
have
any,
mrs
that
they
would
like
to?
Maybe
we
could
all
mob
pair
review
on
it.
So,
if
you're
not
familiar
with
how
we
do
pair
programming
on
our
back
end
or
front-end
pair
programming
sessions,
usually
there's
one
person
driving
and
they
share
their
screen
and
then
the
rest
are
observing
navigating
discussing-
and
it's
usually
surprisingly,
works
really
well.
A
So
it's
kind
of
this
mob
pairing
thing
and
it's
facilitates
some
really
good
collaboration
and
knowledge
sharing
so
yeah
I
have
some
mmrs,
but
I
don't
wanna,
I
don't
wanna
be
so
attention
seeking
which
I
tend
to
be.
Oh,
a
database,
mr
in
the
future,
yeah,
that's
a
great
idea,
terry,
so
yeah.
If
you
get
painted
on
a
database,
mr
on.
F
C
C
A
Okay,
so
I
have
one
here:
I
haven't
reviewed
it
yet
so,
let's,
let's
do
it?
Let's
see
what
happens?
I'm
gonna
share
my
screen.
A
I'm
going
to
open
up
this,
mr,
so
the
goal
here
we're
a
mixed
bag,
a
front
end
and
back
end
right
now.
The
goal
is
to
mainly
facilitate
discussing
the
review
process.
A
So,
let's
check
out
the
context
of
this.
Mr
highlight
runner
name
and
move
locked
icon.
This
is
miguel.
Writes
great
mrs
updates
run
name,
so
it
appears
bold
and
move
the
locked
icon.
Next
to
it,
some
ui
polish
stuff,
here's
a
screenshot,
it's
bold
and
the
locked
icon
has
moved
over.
A
So
this
is
like
one
of
this
seems
and
there's
only
one
file
change.
I
love,
mrs
that
have
like
a
really
simple
scope
and
there's
only
one
file
change.
It.
F
A
To
have
a
really
simple
scope
and
it's
like
12
files
changed
make
me
like
so
nervous.
Okay,
so
I
see
we
have
this
slot,
which
we
don't
have
to
worry
about
the
specifics
of,
but
it
looks
like
we
are
simply
wrapping
this
slot.
That
was
previously
there.
Where
that
runner
name,
I
guess,
will
get
injected
around
with
a
strong
element,
and
we
have
moved
this
runner
badge
over.
A
So
that
the
icon
is
before
it
and
we
added
these
classes,
so
this
all
looks
fine.
This
is
sticking
out
to
me
a
little
bit
because
so
there's
a
technical
specific
here
when
it
comes
to
strong
elements,
they
can
only
contain
inline.
A
So
I
always
feel
funny
about
this
from
the
front
end
because
browsers
are
pretty
forgiving,
but
there
is
like
an
html
specification
that
certain
elements
can
only
have
certain
kinds
of
children
and
strong
elements
can
only
have
inline
children
and
given
that
this
is
a
slot
previously,
that
slot
could
be
anything
could
be
a
div
span
whatever,
which
is
part
of
this
component's
interface,
but
now
we're
kind
of
limiting
the
scope
of
what
this.
What
this
slot
can
do,
because
technically
it
should
only
be
probably
a
span
or
even
just
raw
text.
A
On
top
of
that,
this
happens
so
often,
especially
with
backhanders
the
right
front.
End
code,
I'm
sorry,
backhanders,
I'm
picking
on
you
all
strong,
so
there's
a
whole
semantic
to.
I
guess
I'm
not!
I
don't
know
if
this
is
still
a
thing,
but
like
semantic
html
was
we
tried.
We
tried
to
prioritize
it
sometime.
Strong
also
means
like
urgent,
like
your.
Your
child
is
sick
urgent.
A
So
we
use
it
all
the
time.
Just
a
bold
text,
it's
not
really
what
it's
meant
for
so
this
is.
This
is
great.
I
I
feel
like
it's
all
going
to
work.
It
clearly
showed
screenshots
that
it's
going
to
work,
but
it's
like
there's
other
ways
to
do
this
and,
like
maybe
we're
maybe
strong,
wasn't
the
right
element
to
use
here.
A
So
my
question
I'm
asking
myself
right
now
is:
do
I
ask
a
non-blocking
comment?
Do
I
not
say
anything?
Do
I
just
leave
a
small
suggestion
of
like?
Can
we
do
this?
What
do
you
where
you
all
land
on
this.
A
That's
a
good,
that's
a
good
question.
I
thought
about
that.
Let
me
let
me
check
this
out
locally
so
that
I
can
navigate
not
mucks.
Oh,
I
can't
nest
my
I'm
already
in
tmuxland.
Let
me
check
this
out
locally,
so
we
can
dive
into
it.
G
That
was
a
really
really
good
observation,
anna,
because
if
it
is
only
used,
one
place
you're,
keeping
it
more
cohesive
and
less
coupled
to
this
component
and
then
in
the
future.
If
this
perhaps
is
used
somewhere
else,
it
will
and
it
will
still
be
bold
like
does
the
bowl
belong
with
whatever
is
in
this
slot
or
does
the
ball
belong
here?.
A
Yes,
where
does
it
belong?
Is
a
really
good
question
yeah
all
right,
so
I
love
to
just
curl
the
diff
and
apply
it
rather
than
checking
out
the
branch.
F
A
My
life,
especially
when
you
have
the
gdk
running
cool
so
yeah
here
we're
wrapping
a
strong,
let's
check
out
runner
name
and
see
where
that's
used,
so
runner
name,
seeing
where
this
is
all
imported.
So
it
comes
from
runner
summary
cell
and
admin
runners,
app
and
group
runners
app,
so
admin
runners,
app.
A
A
That's
I
like
I
like
that
thought,
though,
so
I'm
gonna
ask
question
non-blocking.
A
Yeah,
so
for
me
personally,
I
would
leave
this
as
since
there's
nothing
urgent
here
and
I'm
like,
let's
not
use,
misused,
strong,
I'm
leaving
a
polish
comment
of
like
hey,
let's
just
polish
it
up.
This
is
the
first
mr
cycle.
This
is
the
only
small
thing
I
see
it's
like
this
yeah
we
like
to
merge
things,
but
let's
just
polish
it
up
a
little
bit.
I'm
saying
that
strong.
G
G
Where
does
it
say
like
bowls
should
be
used
if
you're
not
talking
about
your
child
being
lost,
like
is
that?
Can
you
give
some
authority
that
so
that's.
A
G
Like
a
reference
to
they're,
just
like
a
link
to
that
concept,
semantic
styling
like
to
provide
some
education,
yeah.
A
We
actually
have
that
in
our
docs.
Okay,
if
we
do,
if
you
do
a
search
for
html
semantic,
I
bet
we'll
find
it
yep
there
we
go
and.
G
A
Yeah
prefer
semantic
classes
used
accurately,
probably
just
regular
class.
I
don't
know
what
semantic
classes
means,
but
yeah.
A
Yep,
so
strong
element
isn't
quite
appropriate
here.
Also,
I'm
pretty
sure
this
just
limits
the
children.
F
Yeah
cool
that's
reveal
great
yeah,
I'm
gonna.
I.
A
Like
whoa
paul's,
just
really
spazzing
out
after
leaving
that
comment
yeah
and
so
then.
G
A
Like
a
guess,
that
is
a
good
point
too.
I
didn't
think
about
that,
but
yeah,
anticipating
the
context
of
how
these
are
going
to
be
read
is
helpful.
Thanks.
F
A
Much
for
this,
mr
mcgill
left
some
comments
for
your
consideration.
A
E
G
G
The
review
group
is
in
the
process
of
redesigning
this
process
and
soliciting
feedback
and
there's
some
issues
open
in
several
epics
about
splitting
this
into
different
states
how
to
represent
it
in
the
ui.
C
Yeah,
I
thought
you
were
supposed
to
unassign
yourself
because
it
kind
of
went
one
way
and
then
it
flipped
back.
But
now.
G
Added
the
ui
element
with
it
in
the
right
with
the
the
treatment,
but
then
they
abandoned
that
feature
and
they're
not
doing
that.
So
now
I
think
they're
iterating
on
other
aspects
of
the
design
of
more
like
the
workflow
in
different
stages
to
you
know
whether
you're
requested
or
re-requested,
and
who
has
the
responsibility
for
it.
C
Yeah,
but
it's
really
not,
I
I
couldn't
find
that's
like
it's
in
the
link
under
dog
fooding,
the
reviewers
feature,
but
I
couldn't
quickly
find
where,
because
it
says,
oh,
here's,
a
summary
of
the
changes,
also
reflected
in
this
section
above
I
can't
find
in
the
section
above
quickly.
C
C
C
C
A
G
That's
the
feedback
I
was
giving
to
them
yesterday
as
part
of
that
their
work
on
this
I
was
like.
I
have
an
email,
centric
workflow
and
I
have
a
like
an
automated
job
that
gives
it
a
big
red
label
that
says
reviewer
when,
when
that
special
header
is
in
the
email
contents
down
which
works
most
of
the
time,
but
like
that
doesn't
work
well
for
unassigning
and
reassigning
myself,
then
I
have
to
go
in
and
watch
this,
which
is
that's
two
places
to
look
yeah.
A
A
This
one
is
interesting.
This
is
working
with
an
interesting
part
of
our
app.
Is
this
the
one
that
I'm
thinking
of?
Oh?
No,
I
think
this
no.
I
was
thinking
a
different
one.
That
one
is
interesting
all
all
mrs!
I
get
are
interesting
that
wasn't
the
one
I
was
thinking
of,
though
this
one
is
about
adding
some
tracking
to
our
pipeline
wizard,
so
we're
having
a.
I
think,
it's
behind
a
feature
flag,
but
it's
a
it's
a
wizard
for
creating
the
yaml
needed
for
paige's
pipeline.
A
I'd
already
done
a
review
on
this,
but
giannis
was
out,
and
so
this
is
kind
of
getting
resurrected
from
a
handful
of
weeks
of
inactivity.
Some
of
the
comments
I
left
was
one
of
them
was.
A
A
Whenever
I
see
events
like
this
from
experience,
I
get
nervous
that,
like
how
often
are
we
doing
this
like
how
often
is
touch
happening
and
it
was
happening
like
ev,
every
keystroke
like
it
was
it
was.
It
was
emitting.
A
So
I
think
it
seems
like
giannis
addressed
this,
so
I'm
interested
to
see
how
that
happened.
I
think
there's
he
probably
used
once
to
see
what
happened
there.
It
looks
like
I
praised
them
for
some
nice
specs.
A
And
then
I
identified
there
was
a
missing
assertion
in
the
unit
test
so
like
this
line
wasn't
covered.
F
A
I'm
so
glad
you
asked
this
question:
oh
yes,
I'm
so
glad
you
asked
this
question.
This
is
like
by
far
one
of
my
favorite
things
to
do
during
review
such
a
good
question.
So,
like
all
reviews,
I
checked
them
out
locally,
so
I'm
back
clean
master
branch.
A
I
add
all
the
changes,
so
what
I
do
to
test,
because
some
of
these
test
files
like
really
big
too,
but
I
will
run
the
test
locally
without
even
looking
at
it.
I'll
just
run
this
locally.
A
Oh,
come
on,
these
first
runs
are
going
to
take
longer
than
the
subsequent.
A
Yeah
I
got,
I
got
a
fancy.
I
got
a
fancy
computer
though
right
my
fancy
computer
told
me
it
can
it
has.
I
could
multitask,
unlike
myself,
so
what
I
do
then.
So
this
is
the
unit
test.
I'm
running
and
I
will
just
straight
up
like
change
I'll
see
what
lines
we're
changing
and
I'll
actually
just
introduce
bugs
like.
A
Yeah,
so
this
is
actually
called.
This
is
called
mutation
testing
and
what
you're
doing
is
you're
introducing
mutants
into
the
subject
or
the
production
code,
and
if
the
tests
pass,
your
coverage
fails.
So
if
I
can
introduce
bugs
and
the
tests
still
pass,
I
have
not
adequate
tests.
So
I
will
do
that.
That's
my
favorite
way
to
test
our
test
coverage
is
I'll,
admit,
looking
at
the
lines
that
are
changed.
A
A
Just
at
the
unit
test
level
of
me
knowing
what's
covered
and
what's
not
so
you
could
do
the
same
thing
in
the
back
end
code
of
changing
a
function
to
throw
an
error
or
be
erroneous
values
or
something,
and
if
this
would
cause
a
bug
and
our
tests
aren't
failing.
That's
a
really
good
signal.
If
our
coverage
isn't
good
enough.
G
And
so
wouldn't
be
a
call
with
me
and
paul
on
it
if
I
did
not
say
this,
but
the
jetbrains
ide
in
ruby
mine,
which
I
use
when
you
run
your
test.
If
you
right
click,
there's
an
option
to
run
with
coverage
and
then
it
will
automatically
highlight
red
and
green
the
lines
which
are
covered
and
which
are
not
that's
fancy
yeah.
F
C
C
G
It's
like
there's
the
under
coverage
thing
that
will
fail,
but
it-
and
I
just
had
this
happen
because
we
try
to
minimize
the
test
run
time
and
run
like
only
the
subset
of
specs
which
are
needed
by
the
code
that
you
changed
sometimes
like
it
can
actually
be
covered,
but
not
by
a
spec
that
got
run
as
part
of
that,
mr
so
it
thinks
it
wasn't
covered
because
it
didn't
run
the
you
know:
100
000,
specs,
that
we
have
in
the
entire
suite.
B
A
So
I'm
trying
to
which.
G
A
So
going
back
to
the
scope
of
this
change,
I'm
noticing
some.
A
A
F
D
One
is
there
a
do
you
have
to
you
have
to
kind
of
you
have
to
commit
the
changes.
Don't
you
in
the
ide,
so
would
it
be
more
useful
to
track
that
particular
so.
A
A
A
G
A
D
I
guess
I'd
look
at
the
issue
and
see
what
what
the
requirements
were
on
the
product
level.
A
Now
this
I
I
believe
giannis
is
in
a
single
engineer
group,
so
I
I
think
I
I
believe
so,
and
so
I
think.
G
D
A
We
measure
something
and
misinterpret
what
it
means,
but
that's
not
really
that
bad
and
that's
just
for
a
small
window.
So
I
feel
like
this
is
a
non-blocking,
because
I
feel
like
the
worst
case
scenario
of
this
name
not
being
totally
right
is
not
that
bad.
A
So
let
me
ask
a
non-blocking
question:
maybe
we
should
consider
changing
the
label
to
reflect.
G
F
A
That
sounds
good.
That
makes
sense
the
I
do
I
do
want
to
I.
I
really
do
want
to
verify
this
once
behavior.
A
I
do
want.
I
really
do
want
to
verify
that
for
my
own
sake,
because
we
don't
do
these
onces
a
lot
in
view
and.
A
I
am
compelled
to
want
to
just
double
check
it.
It
looks
like
the
other
significant
change
to
this.
Mr
is
just
these
doc
changes
where
it's
hard
to
tell
what's
going
on
in
this
kind
of
diff.
But
if
I
go
here,
we're
just
adding
a
lot
of
white
space,
but
adding
this
documentation
to
these
yaml
interfaces
that
he's
had
before.
A
I
think
I
think
that's
really
the
only
to
do
left
for
me
on
this.
Mr,
is.
I
want
to
verify
that
once
behavior,
I'm
looking
now,
I'm
just
doing
a
read-through
of
seeing
if
anything
is
really
jumping
out
of
me,
but
we're
doing
we
have
testing
for-
and
I
remember
verifying
this
earlier
and
praising
it.
I
don't
think
anything's
changed
if
I
was
really
nervous
about
something
that
I
reviewed
previously
changing.
A
I
would
visit
the
commits,
and
yes
so
like
these
were
rather
than
just
overriding
the
same,
commit
be
honest:
did
the
really
helpful
approach
of
just
adding
new
commits
for
for
me
to
look
at
to
see?
What's
actually
changed
since
the
last
time
I've
seen
this.
A
G
A
A
We've
got
lines
of
code
that
we're
not
necessarily
testing,
maybe
they're,
being
tested
elsewhere.
Let
me
run,
let
me
run
it
for
the
whole
folder,
let's
double
check
and
then
it'll
be
time
to
be
time
to
bid
a
deal.
A
G
And
next
time
I
can
show
how
to
do
the
coverage
in
the
ide
yeah.
That's
good.
G
That
will
determine
how
many
concurrent
reviews
you
have
assigned
to
you.
So
the
unassigning
yourself
interacts
with
that.
Like
paul
loves
reviews,
he
would
happily
do
like
50
a
day.
I
personally,
I'm,
like
you
know
couple
couple
at
a
time
is
my
max,
maybe
one
if
it's
a
bad
week.
So
if
I
remove
myself,
I'm
gonna
get
another
one.