►
From YouTube: Product Analytics 15.10 frontend group retro
Description
In this call the Product Analytics frontend group retrospects the 15.10 release.
A
Let
me
start
this
off
by
sharing
my
screen
very
kindly
so
that
we
can
see
the
gym
data
I.
Can
everyone
see
my
screen
awesome,
getting
some
thumbs
up,
I,
don't
think
we
need
to
walk
through
the
entire
Milestone
board,
but
I'll
just
quickly
show
it
in
general.
All
I
can
see
here.
That's
like
we.
We
have
some
agenda
items
for
slip
stuff,
but
it
seems
like
we
have
a
feature
spec
that
Alvin
that
you
haven't
started
with.
A
B
A
It's
it's
fine
though
we
can,
we
can
move
it
on
and
other
than
that
the
Milestone
is
looking
pretty
good.
We
have
like
a
lot
enclosed,
a
lot
in
verification
and
most
of
it's
in
review.
We
only
have
like
a
few
items
here
and
not
really
anything's
front-end
related.
Besides
the
work,
that's
Tim's
doing.
B
Not
a
lot
so
first
item
we're
based
on
the
custom,
the
product
and
Stitch
dashboard
listing
with
the
new
shared
one.
We
now
are
at
the
project
level.
So
all
of
the
analytics
code,
except
for
onboarding,
has
been
shifted
over
now
in
its
entirety,
and
the
menu
item
has
been
updated.
I
am
in
the
process
of
verifying
it.
B
It's
only
I
just
got
notification
from
the
bot
to
say
it's
ready
for
me
to
do
the
review
state
that
verification
process
so
I'll
be
doing
that
tomorrow
morning,
so
that
now
means
that
all
of
our
code
now
lives
on
the
under
analytics.
We
have
scheduled
for
the
next
Milestone
which
we'll
talk
about
in
the
kickoff
meeting
the
onboarding
changes
once
the
onboarding
changes
are
done.
We
can
delete
that
product
and
analytics
folder
and
any
related
specs
and
close
off
that
area
entirely.
B
B
B
My
assumption
is
that,
as
optimize
gets
some
board
and
starts
making
use
of
this
listing
as
well
and
others
they
will
be
contributing
to
the
visualization
design
are
what
they
need
to
put
into
it,
because
we
don't
know
the
data
sources
that
they're
going
to
be
working
with.
So
there
is
that,
in
terms
of
the
default
date
range
that
was
super
nice
super
straightforward,
nice
and
easy
just
made
it
so.
B
The
date
range
is
no
longer
limited
to
31
days
defaults
to
no
date
range,
but
because
it's
customers,
dashboard,
is
a
shared
component
that
can
be
used
by
others.
Technically
does
nobody
else
is
using
it
right
now,
but
with
the
working
dashboard
working
group
finishing
up,
hopefully
people
will
start
making
use
of
this.
B
It
does
mean
that
if
someone
needs
a
date
range
like
if
we
wanted
to
use
this
on
audit
events,
if
that
became
a
dashboard
type
structure,
you
could
just
add
one
in
and
then
the
last
thing
was
something
I
did
this
afternoon
that
merged
very
quickly.
So
thank
you,
Epi.
We
now
on
renovate
bot.
So
all
three
of
us
are
linked
to
the
renovate
bot.
We
already
have
a
first
Mr,
which
Owen
and
myself
are
assigned
to
since
I'm.
B
A
maintainer
I
suggest
that
oh
and
if
you
do
the
initial
review,
then
I
will
do
the
maintain
a
review
on
that
one.
If
you
like,
so
we
know
it's
working.
Tim
has
made
this
a
comment.
What
about
grid
stack,
which
was
not
in
the
initial
issue,
so
I
set
a
reminder
for
myself
to
create
an
issue
tomorrow
morning
for
that
to
be
scheduled
in
for
us
to
add
that,
but
it
will
be
super
straightforward,
easy
to
do
one
line
change
within
the
renovate
repo.
So
that's
me.
A
Awesome
yeah
I
had
a
look
at
the
Mr
and
seems
surprisingly
simple,
like
it's
all
configuration
based
based,
which
is
very
nice
and
yeah
thanks
Tim
for
pointing
that
out.
I
think
we
all
just
completely
forgot
about
it.
Yep
onto
my
stuff,
what
I
did
was
improve
the
panel
responsiveness
and
consistency.
A
This
turned
out
to
be
quite
a
rabbit
hole
specifically
for
charts,
because
we
use
the
e-charge
library
in
gitlab
UI,
which
is
then
used
in
gitlab.
That
is
the
component
library
that
we
use
in
the
front
ends
and
key
charts.
Has
this
Behavior
where
you
can
set
High
to
Auto
and
it
will
fall
to
the
parents
containers
Heights,
because
we
were
wrapping
charts
inside
charts
inside
charts
in
gitlab
UI,
pretty
much.
A
That
was
like
broken
right
across
the
board
for
every
single
chart,
so
I
wanted
to
go
in
and
just
fix
it
for
line
chart
and
then
discovered.
Oh,
this
is
actually
broken
on
every
single
chart
and
I
created
a
massive
related
Mr
that
has
I
think
just
been
virtual
in
kitlab
UI,
so
I'm
very
happy
about
that.
But
Axel
also
did
some
preliminary
work
on
this
to
improve
grid
Stacks
responsiveness
so
that
it
doesn't
collapse
into
and
it
has
a
Min
height
so
that
it
doesn't
like
go
into
something.
A
A
C
A
C
Cool
yeah,
so
this
milestone
for
me
has
been
a
lot
of
Handover
work
from
the
previous
team
on
the
product
analytics
front.
It's
mostly
so
far,
just
being
this
column
chart
visualization
that
I've
added
in
which
has
yeah
just
been
a
wrapper
over
the
UI
chart
and
plugging
it
into
our
visualization
designer
and
dashboards.
A
A
A
Yes,
totally,
no,
as
I've
talked
about
the
GitHub,
UI
change
just
got
merged,
thanks
for
reminding
me
Rob,
and
this
will
likely
still
slip,
even
though
it's
already
been
merged,
because
patching
gitlab
usually
takes
a
few
days.
So
I
only
expect
the
height
Auto
fixes
to
be
in
by
the
end
of
the
week
and
then
I
can
create
an
MR
on
gitlab
to
actually
implement
the
property
changes.
I
can't
do
it
yet
because
height
Auto
is
bugged,
every
single
chart
would
go
to
zero
height
and
would
not
render
so.
A
B
B
Where
he's
listed
noted
out,
why
and
more
details
around
his
refinement,
whether
it's
ready
for
development,
what
sort
of
weight
is
going
to
give
it?
Why
he's
giving
it
a
certain
way
how
many
mls
or
issues
he
expects
to
give
whether
it
needs
documentation
or
not,
which
I
really
like
that
little
extra
touch,
because
it
gives
it
reminds
us,
do
we
need
do
we
need
documentation
before
we
even
start
development?
We
need
to
start.
B
It
reminds
us
to
consider
that
during
planning,
which
then
makes
it
easier
for
us
to
let
Lorena
know
ahead
of
time
what
she's
going
to
need
to
be
involved
in,
because,
obviously
she
does.
She
covers
more
than
just
group
product
analytics,
so
I
just
thought
I'd
raise
that
I
I,
probably
I'll,
probably
change
the
format
slightly,
maybe
into
a
table,
maybe
using
our
labeling
system
a
bit
more
or
links
describing
whether
it
is
going
to
be
multiple
Mrs
or
issues
rather
than
combining
the
two
and
explaining.
B
A
Yeah
I
think
this
is
pretty
nice,
like
I,
said
it's
definitely
efficient.
Let's
counterpart,
no,
not
just
technical
writing,
but
I
feel
like
Engineering
in
general
and
also
product,
know
like
what
this
is.
What
this
change
actually
entails,
so
I
like
it,
maybe
if
you,
if
you
could,
could
you
please
create,
like
you
said
you
wanted
to
edit
some
of
this
like
create
like
a
snippet
and
share
it
with
us
and
I'm
thinking.
If
that
snippet
becomes
like
something
that
we
use
a
lot,
maybe
we
could
add
it
to
the
handbook.
B
Yeah
just
again,
I
really
appreciate
both
of
you
putting
in
so
much
time
helping
me
with
these
the
massive
and
very
complicated
miles
that
I
ended
up.
Refusing
you
know,
helping
me
break
them
down
further,
pointing
out
in
potential
problems
and
pitfalls
and
getting
them
over
the
line.
I
wouldn't
have
been
done
it
without
I
review.
So
thank
you
very
much.
A
Great
job
on
that
and
I
have
the
next
one,
a
big
thanks
and
shout
out
to
Alvin
for
already
submitting
a
new
visualization
column,
charts
in
your
first
Milestone
and
doing
it
in
such
a
clean
and
professional
way,
really
yeah
really
great
work
and
really
also
super
happy
to
have
you
on
board
with
the
team
and
it's
good
that
it's
no
longer
just
Rob
and
I
I
feel
like
it's
just
been
the
two
of
us
for
how
many
years
now
so
it's
good
to
have
another
person
on
our
team.
Everyone.
C
A
A
B
Agreed
it's
worth
saying:
Kevin
is
still
transitioning
from
his
previous
team,
so
he's
probably
going
to
be
considered
doing
that
over
the
next
Milestone
as
well.
So
me
and
you
we'll
be
we're
not
the
DIYs,
but
we
are
going
to
be
around
to
support
Kevin
during
that
process
and
then
at
that
point
we'll
be
getting
rid
of
our
figment
licenses.
Yeah.
A
Then
the
next
one
is
we'll
be.
Switching
the
product
analysis
stack
state
from
the
back
end
and
removing
the
or
to
the
back
end
and
removing
the
related
front
into
logic
within
the
next
Milestone
or
two
attach
a
related
issue
there.
It's
just
an
FY.
That's
when
adding
like
business
logic
to
the
front
end
just
know
that
it's
temporary
and
it's
probably
going
to
be
removed
or
hopefully
be
removed
soon.
B
I'm
still
well
yeah
I'm
still
in
conversations
around
some
of
the
SDK
discussions.
There
is
a
issue:
I
will
find
a
link
for
discussing
whether
to
make
the
SDK
typescript
or
not.
So
if
anyone
wants
to
dive
in
on
that
feel
free,
it's
I
don't
mind
either
way,
but
I
just
want
to
make
sure
that
we've
listed
out
the
pros
and
cons
and
make
sure
we
come
to
a
considered
decision
rather
than
a
personal
preference
style.
One.
A
C
B
I
mean
it's
a
really
simple
basic
application.
That's
just
a
sort
of
proxying!
Well
well,
snowplowing
there
are
coming
up,
but
Jitsu
originally
did
so.
We
wrapped
it
up
in
our
own
configuration
to
separate
what
the
user
you
know.
So
this
user
doesn't
have
to
install
Jitsu
it
just
installs
the
gitlab
SDK,
and
they
don't
need
to
worry
about
how
it
works
under
the
hood.
So
it
that's.
It's
a
really
simple
basic
JavaScript
library.