►
From YouTube: 2022-12-12 Product Analytics 15.7 frontend group retro
Description
In this call we discuss our experiences in the 15.7 milestone.
A
A
The
two
main
things
that
we
had
to
do
from
the
front
inside
was
the
onboarding
app
and
then
also
the
charts.
So
the
good
news
is
that
we've
made
I
feel
like
good
progress
there
on
the
chart
side
or
the
visualizations.
We
have
the
line
chart
the
table
and
the
single
stat.
A
Thanks
for
all
your
help
and
contributions
there,
especially
axle
for
helping
out
on
the
single
stack
one
and
yeah
they're
all
merged.
So
those
are
like
the
three
main
visualizations
that
we
need
to
start
building
dashboards.
So
now
everyone's
very
excited
about
that
then
also
Rob.
Congratulations
for
getting
the
onboarding
app
merch!
I
know
that
was
a
giant
one
I'm
sure
it
caused
many
many
a
headache.
B
A
You
feel
like
there's
anything
that
we
could
follow
up
on
to
make
their
process
better,
or
was
this
like
an
individual,
one-off
case.
B
No,
it's
a
you
know
you
have
to
our
SLA
is
48
hours
after
the
request
is
made
and
you've
got
to
give
everyone's
in
different
time
zones.
So
you've
got
to
give
everyone
that
time
that
they
need,
and
it's
just
unfortunate
that
when
a
Reviewer
is
non-responsive
in
that
time
period,
you
have
to
you,
you
can't
proactively.
Just
swap
you
have
to
give
them
an
opportunity
and
then,
when
you
actually
do
get
a
review
that
starts
to
respond.
B
C
B
Yep,
so
sharing
what
I've
done
you
did
well,
we
got
the
onboarding
designs
all
tidied
up
and
the
issues
were
already
set
and
ready
to
go.
That
was
the
beginning
of
the
Milestone
and
I
also
had
some
time.
I
was
fixing
the
cube
proxy,
which
didn't
work
for
some
people.
B
Yarn's
line,
chart
Tim's,
query
designer
name
to
be
determined
at
this
point,
and
we
also
in
one
of
the
Mrs
Young,
raised
a
valid
point
around
getting
a
more
specific
error
in
place
for
when
we
have
no
table
or
Dash
database
returned
from
Cube
It's.
It's
not
yet
generated
on
the
onboarding
process.
So
adding
a
new
404
specific
message
for
that
scenario.
So
we
can
differentiate
between
known,
expected
errors
and
unknown
unexpected
errors.
A
C
Yes,
let
me
vocalize
that
so
yeah.
This
is
a
bit
from
from
the
previous
Milestone,
but
we
got
nicer
ducks
and
we
also
have
a
couple
of
I
wanted
to
mention
it,
because
we
have
these
follow-up
about
that.
I
did
on
LinkedIn
link
it
on
the
on
the
agenda,
but
we
have
this
follow-up
about
adding
information
about
our
stack
into
the
into
the
docs,
and
this
was
one
of
the
suggestions
and
I
do
believe.
This
is
valuable,
so
I'm
going
to
try
to
push
that
forward
is
in
this
next
milestone.
A
Awesome
yeah
I
think
as
we
move
towards
the
customer
preview.
The
importance
of
that
like
is
going
to
raise
right
now.
It's
it's
not
like
our
priority
number
one,
but
the
more
eyes
we
get
on
this
I
guess
the
the
better
it
is
for
us
to
explain.
You
know
the
context
around
like
what
it
is
that
we
use
and
and
why
certain
things
need
to
be
said.
C
Yeah
totally
because
one
of
the
questions
we
got
a
lot
when
we
started
working
on
this
was
what
is
product
analytics
and
what
what
are
we
doing
here
and
what
is
your
vision
and
and
all
this
stuff
and
having
this
tagging?
The
stack
and
description
is
a
good
way
to
to
to
move
that
conversation
forward
as
well.
B
B
The
honor
boarding
issues
are
only
now
realistically
able
to
be
developed
and
we've
got
what
a
week
left
in
the
Milestone
which
we're
doing
this
a
bit
early,
because
next
week,
I'm
not
here
at
least
so
next
Milestone.
Having
a
look
at
the
15-8
Milestone
planning.
The
number
one
item
for
us
is
the
onboarding
flow.
B
Getting
that
finished,
followed
by
any
getting
those
dashboards
set
up
the
audience
the
behavior
ones,
although
I
believe
they
want
to
do
a
shared
session
for
that,
if
I
recall
correctly
so
yeah
from
our
perspective,
that's
our
biggest
thing.
I've
already
got
out
the
empty
State
into
review,
which
I've
shared
with
Axel
for
initial
review.
B
A
Cool
and
I
think
that's
I
think
like
overall
we've
made
good
progress
and
I
think
it's
really
understandable
why
some
of
those
items
May
slip
and
luckily
it's
not
our
top
priority,
which
is
the
you
know,
building
those
dashboards,
so
I
think
we're
we're
in
a
good
space
still
there.
So
that's
nice,
any
any
other
things.
That's
yeah!
Anyone
would
like
to
mention
there
I
can't
think
of
anything.
B
Oh,
that
is
one
thing
the
it's
not
something
we
are
doing.
It's
something
Dennis
is
handling,
but
we've
got
the
the
English
words.
Rob
speak
the
internal
handbook
Mr
that
I
wrote
ages
ago
got
merged
about
testing.
B
We
were
asking
what
we
need
to
do
to
get
immersion.
He
just
got
merged,
not
petite
had
a
problem,
but
it
does
in
the
time
that
I
wrote
it
and
it
got
merged.
They'd
moved
the
doxy
theme
out
of
the
repo,
so
I
added
a
theme
hook
into
the
repo
and
the
theme
no
longer
existed
existed
there.
So
it
didn't
work.
So
we
heard
Dennis,
went
and
created
separate
issues,
one
to
remove
that
from
the
handbook
and
two
to
add
it
to
the
right
location.
B
But
there
is
no.
There
is
now
an
ongoing
discussion
which,
if
I
can
find
the
I
will
link
in
the
document
to
bumping
that
no,
the
handbook
version,
number
I,
don't
know
if
it's
properly
deployed
as
a
result,
it
might
be
deployed.
I
felt
I
thought
it
was
deployed,
but
from
what
I,
but
there's
still
an
ongoing
debate
around
cookies
and
stuff
that
people
are
pushing
back
on
the
internal
handbook.
A
Cool
thanks
for
erasing
that
I
think
that's
valuable
to
know
yeah
and
then
some
ideas
and
action
items
for
improvements
in
the
following
Milestone
open
question
that
I
have
that
I
thought
would
be
cool
to
hear
some
perspectives
on.
This
is
because
it's
something
that
I'm
also
like
I,
wouldn't
say,
struggling
but
still
figuring
out
like
how
to
balance
or
high
standards
for
code
quality
that
we
have
at
gitlab
for
the
application
overall
versus
specifically
what
what
level
of
quality
do
we
need
for
Alpha
releases
that
especially
for
features.
A
B
Yeah,
just
a
general
feeling
that
we're
starting
to
move
away
from
the
sort
of
get
out
quick
phase
where
you
just
need
to
get
things
out
into
a
more
sort
of
into
the
morgue
starting
to
think
about
the
next
step
after
the
after
that
into
internal
preview
and
thinking
about
polish,
not
necessarily
at
the
point
where
everything
needs
to
be
done
and
dusted
and
100
foolproof.
But
we
also
need
to
be
aware
that
we
shouldn't
be
adding
unnecessary
of
a
ux,
technical
or
other
kind
of
DX
any
kind
of
technical
debt.
B
If
we
don't
need
to
I,
think
I'm
I'm
sort
of
Leaning,
it
Axel
wouldn't
know
her
really
much
about
this.
But
do
you
remember
the
first
compliance
dashboard?
We
did
and
I.
B
B
Was
functional
in
a
way,
but
not
you,
but
the
user
experience
was
in
great
and
it
didn't
really
help
much
with
what
the
Persona
was
looking
to.
What
do
with
that
screen,
so
I'm
I'm,
starting
to
think
along
those
kind
of
lines,
is
this
doing?
Are
we
reaching
the
goals
that
we're
looking
to
reach
with
each
iteration
is
deterioration,
progressing
it
forward
in
a
positive
step,
yeah
I'm
waffling
now.
A
Cool
I
have
some
thoughts
on
that,
but
I
would
first
like
to
also
hear
Axel's
response
on
that.
C
Also,
because
what
Rob
said
that
we
are
starting
to
get
a
bit
away
from
that
from
the
priority
on
speed
that
we
we
did
have
before,
and
it
was
pretty
apparent
in
which
basically,
we
were
trying
to
to
move
away
from
the
POC
as
as
fast
as
we
could
and
Chip
something
and
and
I,
for
example,
and
Rob
can
can
remember
this
that
when
he
joined
what
I
told
him
was
hey,
I
I
might
need
help
to
push
forward
this
to
actually
to
actually
got
it
merged
and
and
I'm
talking
about
the
the
the
first
base
for
the
dashboard
and-
and
that
was
because
I
was
a
bit
concerned
about
a
the
feedback
that
I
was
going
to
get
with.
C
C
This
for
me
have
been
like
not
that
significant,
because
I
believe
we
are
doing
a
good
job
on
that,
but
yeah.
Now
the
thing
is
that
now
we
do
know
and
or
our
vision
for
the
work
that
we
are
going
to
that
we
are
doing
is
more
clear,
at
least
for
me,
if
the
Nyan
day
in
indifference,
so
so
I
do
believe
that
in
in
like
the
next
step,
we
are
taking
right
now,
our
our
balance
again
I'm.
C
A
Think,
like
as
Engineers,
that
is
what
we
want
like.
We
want
good
code,
stable
code,
something
that's
readable,
maintainable.
We
don't
want
to
go
back
and
throw
the
crash
like
the
compliance
dashboard,
for
example.
You
know,
even
though,
like
the
the
code
itself
is
actually
great
like,
like
I,
said
Rock,
the
the
business
value
of
it
was
questionable
and
it
was
essentially
just
an
experiment
to
see
like
what
can
we
do
and
what
do
we
need
to
still
bold
to?
A
Actually,
you
know
create
something
usable
that
said:
I
do
have
like
sympathy
for
products
wanting
us
to,
or
or
like
the
the
eyes
that
or
on
our
group
to
to
want
to
see
like
a
quick
progress,
because
we're
now
getting
at
that
stage,
where
it'd
be
great,
if
we
can
have
that
internal
preview
ready
and
basically
show
hey,
this
is
what
we're
thinking
of
building,
but
everything
is
still
also
subject
to
change.
A
This
isn't
like
production
quality
and
for
for
me,
like
those
are
like
two
different
sides
of
like
two
sensible
things
that
I
can
understand
from
both
sides,
and
I
can
also
see
through
time
how
that,
like
bar,
we
have
like
in
Romance
almost
like
a
percentage
chart
like
a
bar
chart
where,
at
the
start,
we're
say
focused
on
like
80
speed,
20
quality
an
hour
shifting
it
over,
but
I
do
Wonder
like
with
each
Milestone.
A
If
I
feel
like
we
need
to
worry
more
and
more
and
more
about
the
quality
of
it
and
less
about
the
speed,
but
for
experiments.
Do
we
throw
that
away?
For
example,
like
with
Widget
designer
should
should
be
just
merge.
You
know
and
experiment
in
and
be
like.
You
know,
we're
probably
going
to
redo
this
entire
thing,
or
do
we
aim
for
for
Quality
code,
and
this
is
kind
of
like
a
conundrum,
because
I
feel
like
I'm
I'm,
not
certain.
B
I
I
agree
with
everyone
else's
sentiment.
I
think
we've
been
striking
a
nice
balance.
We've
been
there's
been
a
minor
tension
in
all
of
our
discussions
around.
Should
we
be
doing
this
fast
or
should
we
do
doing
really
thinking
deeply
around
each
element
that
we've
been
building,
which
is
good.
That
means
that
we
are
thinking
about
the
quality
as
well
as
the
our
group
skulls
and
the
that
tension
of
time
has
Johan
pointed
out.
B
So
as
long
as
we're
keeping
that
in
the
back
of
my
Minds
at
all
the
time,
and
even
if
that
does
cause
a
bit
of
tension,
that's
always
a
good
thing.
That's
never
wrong
or
bad.
B
On
the
specific
point
about
query
designer
we've
not
done
any
kind
of
user
experience,
conversation
discussion,
debate
products
not
even
thought
about
it
in
any
intrinsic
detail
on
a
Persona
level
and
how
it
how
it's
going
to
work
is
very
much
right
now,
a
can
we
do
this
experiment,
that's
sort
of
being
morphed
into
something
that
might
end
up
in
the
product,
but
again
it's
it's
not
yet
had
that
thought
process
put
behind
it.
B
So
I
would
argue
that
in
the
in
terms
of
any
area
that
we're
experimenting
in-
and
that
includes
things
like
sessions
like
what
Dennis
and
Max
have
been
doing
with
the
session
scheme
of
data
and
all
that
all
that
stuff
it's
around
functionally
is
it
of
high-
is
the
code
of
high
quality?
Does
it
fun?
Would
we
want
to
continue
using
that
code
in
future
and
is
it
bug
slash
security
compliant?
You
know,
following
all
our
best
practices
from
a
UI
point
of
view,
I
I'm
a
bit
more
loose
there.
B
If
something
is
blatantly
sort
of
off,
then
I
would
question
it
or
raise
it
as
a
as
a
Improvement,
and
so
they
can
always
push
back
and
say
no
we'll
do
it
later
or
whatever
it
might
be.
Have
that
discussion,
but
I
would
but
from
a
UI
point
of
view,
we're
gonna
have
to
wipe
it
clean
and
start
again,
because
we've
not
had
any
kind
of
discussion
around
it,
but
it
doesn't
mean
that
we
have
to
get
rid
of
all
the
functionality
and
the
code
behind
it
necessarily.
A
Yeah
I'm
in
I
I
feel
like
there's,
there's
additional
tensions
for
me
specifically
on
stuff,
like
the
query
designer
or
adding
like
these
large
features
or
pieces
of
code
where,
on
the
one
hand,
I'm
very
Pro,
let's
push
experiments
and
let's
use
it
and,
let's
figure
out
what
is
the
value?
What
is
it
that
we
want
to
use?
Is
this
usable?
Is
it
something
that
we
want
to
go
forward
on?
On
the
other
hand,
there's
the
all,
unlike
because
we're
adding
it
to
gitlab's
main
code
base.
A
It
isn't
like
the
dev
kit,
where
you
know
it's
much
safer
to
experiment
there,
because
I
feel
like
it's
well,
it's
it's
safer
in
the
sense
that
it's
there
isn't
a
cost
to
someone
some
other
group
coming
and
like
reusing
our
code
and
gitlab's
code
meshes
very
quickly
where
even
in
our
own
group,
we
start.
You
know
we're
aware
of
what
what
what
the
others
are.
Writing
and
we
reuse
as
much
as
possible
and
that's
great,
but
it
also
means
like
if
we
had
to
pull
something
out
to
redo
it.
A
The
cost
of
that
raises,
and
are
especially
aware
of
the
fact
that,
like
we're
gonna
have
groups
like
optimize
also
come
on
board
and
start
reusing
some
of
our
code,
so
I
really
want
us
or
I'm
I'm
starting
to
think,
and
it
would
be
cool
if,
like
I,
think,
everyone
starts
to
think
about
like
a
framework
for
how
to
redefine
or
like
marking
a
file
or
saying
like
this
is
an
experiment.
Please
don't
reuse
anything
here
or
like
starting
to
add
comments
or
adding
some
system
where
we
can
both
deliver
stuff.
A
C
Yeah,
that's,
that's
fine!
That's
a
good
call!
I
I,
totally
I
totally
see.
Where
are
you
coming
from
I,
although
I
do
believe
we
might
not
be
at
risk
of
of
actually
like
like
putting
a
technical
depth?
If
that's,
what
we're
talking
about
yeah,
I,
yeah
and
especially
especially
with
is,
is
the
query.
This
is
the
the
query
designer
the
the
the
next
big
thing
we
have
in
in
yeah.
You
know,
or
are
we
going
to
have
like
more
No.
A
B
The
next
big
thing
is
funnels
and
funnel
data
visualization
of
funnels
and
funnel
analysis.
Career
designer
is
like
five
six
Milestones
out
yeah
from
our
from
a
sort
of
development.
On
our
side
perspective
Tim
has
just
found
it
really
useful
to
get
going
on
this
to
help
with
visualizations
proving
what
we
can
do,
and
also
building
dashboards
and
seeing
that
functionality.
C
A
So
I've
I've
linked
to
the
widget
designer
it's
adding
almost
a
thousand
eight
hundred
lines
of
code.
Second
mayor
Walter
reveal,
but.
A
Yeah
it
it
has
already
been,
and
even
then
it's
there's
also
additional
cognitive
blow
too
it
like
the
smaller
you
make
it
the
more
difficult
it
becomes
to
actually
do
like
integration.
So
it's
basically
honest.
A
But
I
do
think
this
raises
like
a
question
for
me
like
how
can
we
better
push
stuff
like
this
in
a
very
work
in
progress
States,
because
I
think
that's
fine
and
then
keep
it
isolated
and
keep
our
technical
debt
low?
For
that
and
I?
Think
for
me,
it's
like
the
question
is:
should
we
even
do
that
and
to
how
can
we
do
that.
B
This
sounds
like
and
feels
like
an
RFC
to
me.
It
feels
like
we
need
to
come
up
with
a
I
mean
we
can
push
it,
but
it
feels
like
we
need
a
convention
that
the
whole
front
end
would
get
behind,
because
we're
gonna
have
to
do
some
kind
of
annotation
because
there's
a
lot
of
files
here,
it's
not
like
just
one
entry
point
and
then
everything
Cascades
out
of
it
necessarily
there's
views
shared.
B
There's
changes
to
customizable
dashboard,
which
is
a
different
problem,
because
that
is
going
to
be
shared
with
optimize
and
other
groups
that
want
to
use
customizations,
so
yeah,
I
I,
know
we're
pick
we're
not
picking
on
Career
designer
but
we're
using
it
as
a
as
a
as
a
as
an
example
of
the
Cyber
scenarios.
B
We're
going
to
face-
and
it's
not
just
going
to
be
this-
we
have
other
things
that
nobody
else
has
done
so
far
in
our
code
base
around
analytical
data
structure
and
Analysis
that
we
gonna
have
to
experiment
on
and
I
think
yeah
the
problem.
Half
the
problem
here
is,
we
don't
have
any
kind
of.
B
We
don't
have
a
description
of
what
the
crew
design
is
supposed
to
be
to
help
make
that
decision
and
that
choice
too,
which
makes
it
even
harder
to
pick
what
should
or
shouldn't
be
happening
at
any
given
point
from
a
from
a
you
know
from
a
product
point
of
view,.
A
Yeah
I
was
thanks
for
letting
me
like
pick
your
your
brains
here,
because
this
is
like
really
an
early
question
and
I
was
just
you
know,
trying
to
explore.
What
is
it
that
your
thoughts
are
on?
Is
this
a
common
problem
and
I
do
think
like
the
right,
the
right
vehicle
to
take
this
forward,
and
that's
when
we
mentioning
it?
Rob
is
like
an
RFC
and
talking
to
like
the
maintainers
group
and
then
the
full
front-end
group,
and
seeing
seeing
what?
B
A
So
I
don't
want
to
take
more
of
our
a
lot
of
time
because
I
see
we're
already
close
to
the
end
and
I
don't
want
to
or
I
want
to
leave
some
time
for
praises.
So
please
go
ahead.
B
Yeah,
just
before
I
should
probably
just
before
we
get
to
Praises
just
to
read
only
more
or
less
reminder
to
add
your
holiday
allocation,
not
holiday
allocation,
but
your
percentage
of
can't
type
English
enter.
No.
That
is
not
my
cursor.
Where
is
my
cursor?
Thank
you.
Ignore
me
right
colonian,
but
update
your
availability
in
the
15a.
B
My
Milestone
obviously
I'm
expecting
us
all
to
be
off
at
some
point
all
combined,
but
just
so
we've
got
an
idea
if
we
need
to
push
anything
out
of
the
15-8
Milestone
be
good
to
make
sure
that
we
all
know
where
we're
going
to
be
like
I'm
not
going
to
be
here,
for
the
majority
of
it.
I've
got
like
five
days
worth
of
work
and
then
I'm
gone,
so
just
thought.
B
I'd
raise
that
in
terms
of
Praises
I
thought
I
would
say
a
thank
you
to
Jan
for
providing
that
piece
of
feedback.
B
On
my
maintainering,
my
maintainer
review
I
gave
him
I
made
a
made
a
non-blocking
suggestion
and
I
wasn't
very
clear
around
being
more
when
it
started
to
get
a
bit
more
into
the
sort
of
blocking
final
decision
phase
for
me.
But
it
was
great
for
someone
for
him
to
give
me
that
feedback
and
allow
me
to
see
what
I'd
be
aware
of
other
people's
perceptions.
To
help
me
make
myself
a
lot
more
clearer.
As
a
reviewer
and
yeah
I
yeah,
it
was
a
I'm,
not
gonna
lie.
B
C
A
A
B
A
Appreciate
that
you,
you
are
championing,
like
the
user
scores
and
especially
focusing
on
our
ux
and
I,
think
that
is
not
something
that
you
should
stray
from
at
all.
So
yeah,
don't
don't
mistake,
my
feedback
on
the
feedback,
for
we
shouldn't
do
this.
It's
yeah,
just
like
the
blocking
version
is
not
blocking.
Then
I
would
like
to
say
thank
you
both
for
the
efforts
you
put
in
this
Milestone
I
feel
like
we've
made
a
lot
of
progress.
A
I
know
we
didn't
manage
to
get
everything
done,
but
I
feel
like
if
we
look
at
all
the
Mrs
merged,
all
the
black
stuff
that
we've
unblocked
it's
it's
amazing
work,
then
I
also
want
to
call
out
like
something
awesome
that
I
saw
is
thank
you
both
for
the
amazing
coach
suggestions
that
you
made
on
the
cube
analytics
Library
with
regards
to
making
things
more
usable,
exporting
it
thinking
about
different
parsers,
how
we
evolve
that
it's
really
given
me
a
lot
of
fruitful
thoughts
in
terms
of
how
we
take
this
forward,
how
we
use
it
for
different
libraries
and
I
think
it's.
A
It's
made
our
architecture
all
that
much
better,
so
yeah
awesome
work
there
right
unless
anyone
has
any
last
minute
items.
I
think
we're
good
to
to
say
goodbye.