►
From YouTube: 2022-12-06 Product Analytics Group Sync
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
B
Yeah
so
hello,
everyone,
it
doesn't
look
like
like
Christmas
we
have
to.
We
have
to
put
some
decorations
in
the
meetings.
So
let's
take
a
look
at
the
board.
We
have
the
Milestone
planning
issue
still
open.
You
know,
maybe
maybe
we
should
close
it.
We
have
the
investigation
of
pre-aggregation
of
anonymized
user
ID.
This
is
from
Max.
B
Max
is
not
in
the
meeting,
but
this
is
basically
a
for
optimization
of
our
queries
and
I
believe
this
is
also
blocking
one
of
the
yeah
one
of
the
the
issues
we
have
further
about
adding
support
for
the
pre-aggregations.
This
issue
is
more
of
an
investigation
than
actual
than
an
actual
feature,
but
but
yeah
we
also
have
another
investigations
like
a
slash
Spike
about
it.
A
Axel
yes,
I
was
I,
was
working
with
Max
on
this
last
week
and
so
hopefully
trying
to
make
some
progress
today,
but
we're
we've
got
the
sessions
Cube
joined
with
the
the
user
or
the
event
table,
which
is
working
quite
well.
We
were
just
having
a
bit
of
trouble
to
track
the
session
and
time.
Those
are.
A
Those
values
aren't
looking
correct
right
now,
so
we're
in
the
process
of
just
like
trying
to
work
with
those
queries
to
basically
be
able
to
tell
when
a
session
ends,
which
will
be
useful
for
actually
being
able
to
determine
and
display
the
duration
of
a
given
session
on
a
dashboard.
So
right
now
my
priority
is
the
internal
handbook
and
getting
that
instrumentation
merge,
request
updated,
which
we
can
speak
to
later
today
or
later
in
the
agenda.
A
B
B
Our
analytics
on
the
internal
handbook,
so
I
I,
don't
think
we
have
much
to
say
they're
dead
and
we
have
a
a
bunch
of
issues
related
to
the
onboarding
and
is
doing
a
good
progress
with
the
with
the
issue
that
is
blocking
this
in
this
task.
And
if
you
have
any
comments.
A
Yeah
so
last
week
we
discussed
having
a
shared
Get
Loud
instance,
as
well
as
a
new
cluster
to
kind
of
test
things
out
with,
and
so
both
of
those
are
set
up
and
actually
kind
of
going
back
to
the
previous
blocked
issue
for
tracking
the
internal
handbook,
I'm
going
to
go
ahead
and
get
it
connected
to
the
shared
instance
and
the
new,
not
the
new
cluster
or
existing
cluster.
A
Just
so,
we
can
move
forward
with
the
internal
handbook
or
Mr
it
does
look
like
infrastructure
is
ready
to
implement
the
changes,
but
there
are
some
details.
I
need
to
know
on
my
end,
to
actually
know
how
they're
going
to
connect
to
the
cluster,
but
until
then
I
think
we
can
just
unblock
ourselves
at
this
point
and
start
collecting
data
officially
on
the
internal
handbook
and
then
swap
things
over
once.
Staging
and
production
are
connected.
D
A
I
need
to
send
over
some
credentials
for
the
production
change
request,
but
I
also
need
to
make
sure
that
there's
right
now,
the
stack
is
under
a
security
policy
to
request
or
to
block
any
requests
so
that
are
they're
coming
from
unknown
sources.
So
that's
what
I
need
to
figure
out
with
John
Travis
in
terms
of
if
there's
a
set
of
Ip
ranges
that
I
can
white
list
so
that
we
can
allow
this
request
specifically
or
if
I,
just
need
to
expose
the
cluster
globally.
A
For
now,
until
the
staging
and
production
are
connected,
so
I'm
trying
to
figure
that
out
with
infrastructure,
so
I'll
update,
hopefully
next
week.
Thank.
C
Yeah
so
from
last
week,
the
code
for
this
got
quite
unwieldy
from
a
reviewer
perspective,
so
I
ended
up
needing
to
split
it
into
two
separate
Mrs
that
first
of
all's
been
merged.
This
morning.
The
second
number
is
approved:
we're
just
waiting
on
someone
from
test
to
approve
the
feature
spec
I've
created
for
the
second
Mr.
Since
that's
done,
then
that
will
get
merged.
C
Hopefully,
tomorrow
there
I
believe
there
are
plus
eight
hours
on
me,
so
they
will
be
well
awake
before
I
am
and
we
can
then
unblock
all
the
other
onboarding
work
that
needs
to
take
place.
Hopefully,
tomorrow-ish
fingers
crossed.
B
D
It's
rendering,
which
is
good.
A
D
Axel
and
Rob
for
the
great
review
I
think
we're
starting
to
get
into
a
territory
where
yeah
it's
it's
getting
like.
There's
a
lot
of
parsers.
So
Rob
made
some
good
comments
about,
like
rethinking
that
so
I'm
taking
some
time
to
think
about
how
we
can
best
handle
that,
but
it
should
be
merged
pretty
soon.
B
Yes,
yes
and
I'm
next
and
talking
about
parsers,
that's
the
the
I
just
borrowed
your
code,
GM
from
with.
B
Work
just
fine
and
because
for
my
single
stat,
I
needed
only
a
single
value
from
from
a
query,
so
we
are
working
on
the
transformation
layer
and
I'm
also
demoing
this
for
during
the
in
the
meeting.
So
so
yeah
I
also
had
a
pretty
nice
review
from
GM.
Thank
you
so
much
so
what
we
close,
we
I
believe
we
only
have
the
top
two
if
I'm
not
wrong.
My.
A
A
A
couple,
so
thanks
Axel
for
the
billboard
review,
there's
a
read-only
item
that
I'll
just
quickly
touch
on.
If
you
want
to
catch
up
on
what
Sam
and
I
were
discussing
as
far
as
sessions,
please
feel
free
to
check
out
that
recording,
Sam
and
I
also
had
a
good
session
together
to
kind
of
discuss
what
we
wanted
to
get
out
of
the
dashboards
and
really
put
in
some
concrete
definition
and
requirements
in
there.
So
please
review
the
issues
at
this
point.
E
Yes,
so
looking
at
the
example,
dashboards
I'm
not
seeing
anything
that
we
would
be
missing
as
soon
as
we
could
take
the
table
and
to
see
instead
merged
or
am
I
mistaken,
do
we
need
area,
charts
bar
charts?
Anything,
no,
not
really
I.
E
Then
the
the
only
question
that
I
saw
was
the
percentage
calculation,
how
we
would
do
that
in
Cube
Etc.
This
is
something
we
would
need
to
figure
out
how
to
calculate
something
yeah
a
cube
query,
but
there
must
be
an
example
for
it.
A
A
The
problem
is
that
we're
missing
the
end
date
to
denote
like
a
completed
session,
but
there
is
a
case
where,
basically,
if
there
is
no
previous
end
time,
that
is
effectively
a
new
session,
so
we
so
what
I'm
trying
to
say
is
we
do
track
new
and
returning
sessions,
but
we
can't
return
either
of
those
right
now
until
we
fix
that
query,
so
I
would
say
we.
A
E
We
can
we
can
anyhow
enhance
step
by
step.
That's
a
nice
part
of
it.
Yeah
the
the
percentage
might
be
also
something
so
that
we
have
two
numbers
and
can
calculate
in
the
single
state
a
percentage
and
the
question
would
be
if
we
would
rather
do
this
on
a
query
level
or
if
we
would
do
this
on
on
a
component
level,
so
that
the
visualization
we
basically
say,
here's
100
number
and
here's
the
actual
number
and
go
ahead
and
go
in
percentage
mode
right.
What
do
we
do?
A
A
Yeah
I
wanted
to
see
where
the
single
stat
looks
up
like
post
review,
because
I'm
quite
curious
about
that
I
think
it'll
be
quite
valuable.
You
might
have.
You
might
have
seen
my
screenshot
in
the
slide
camera
previously
about
providing
Delta
values
and
percentages,
and
things
like
that
or
percent
change
yeah.
A
So
I
think
it
is
a
off.
The
top
of
my
head,
like
I,
would
feel
like
it's
a
pretty
simple
calculation
to
perform
on
the
front
end.
Nothing
too,
you
know
intensive
for
as
far
as
front
end
performance,
so
I
would
love
to
see
if
there's
like
an
option
and
I
would
love
also
like
to
learn
more
about
these
like
parsers.
A
If
there's
a
way,
we
can
add
an
option
to
say:
hey
here,
two
values
crunch
these
and
then
provide
a
single
value
for
the
Delta
change
or
as
the
main
Stat
or
things
like
that.
But
we
can.
We
can
explore
that
once
the
component
is
complete
in
its
first
iteration,
but
thanks
for
bringing
that
up,
Tim
cool.
C
I
I
only
just
wanted
to
say
worst
case
scenario
performance
wise.
We
can
always
use
cubes
pre-aggregation
to
pre-query
all
this
data
for
us
yeah
yeah.
A
I'm,
so
I
I
am
quite
curious.
How
pregregations
will
work
at
scale
and
like
how
that
will
affect
the
Cuban
citizen,
Cube
instance,
and
how
many
of
those
instances
will
need
to
support
pre-aggregations
depending
on
the
refresh
interval,
but
that
is
definitely
something
we
could
theoretically
use
I'm
just
curious:
how
that
will
work
in
a
more
curious
how
to
perform
basically
over
time.
B
B
A
Cool
then
I
think
my
last
topic
is:
there
are
ongoing
discussions
with
optimize
as
they're,
trying
to
figure
out
how
to
build
dashboards
for
the
group
level
and
where
to
store
them.
I've
been
just
kind
of
resurfacing
previous
discussions
and
maybe
previous
ideas
we've
had
when
we
were
first
exploring
this
some
months
ago
so
effectively.
A
You
know
we
our
solution
right
now
works
at
the
project
level
and
I,
just
kind
of
wanted
to
ask
or
open
the
discussion
up
to
us
here
in
this
group
about
how
we,
what
ideas
we
might
have
in
terms
of
storing
dashboard
configuration
data
for
a
group
level
or
anything
beyond
the
project
level.
Sam.
You
have
a
comments
here.
F
Yeah,
so
so
my
thinking
on
this
is
that
if
we
store
configuration
files
and
options
in
a
project,
we
can
then
point
the
group
at
that
given
project
using
the
setting.
This
is
pretty
similar
to
what
we
did
with
compliance
pipeline
definition,
files
and
I
believe
it's
also
similar
to
the
group
level
project
template
suggestion
you
made
Dennis
I'm,
hesitant
to
go
down
the
route
of
putting
an
actual
repository
inside
of
a
group
I
think
that's
very
confusing
from
a
user
perspective,
since
many
of
them
mentally
math
project
code
repository.
F
If
we
break
that
mental
model,
that's
going
to
cause
a
lot
of
confusion,
I
didn't
see
a
ux
person
from
the
group
involved,
I'm,
also
wondering
if
their
group
has
a
ux
person,
maybe
they
would
have
a.
A
They
they
do
yeah
I
I
should
have
qualified
that
yeah.
The
wiki
doesn't
give
the
impression
that
you're
operating
on
a
repo
and
I
would
imagine
if
we
were
to
follow
some
other
concept
that
it
shouldn't
just
be
like.
You
shouldn't
treat
it
exactly
as
a
repo.
It's
just
more
of
like
here's,
an
editor.
You
can
save
changes
to
it
and
that's
that
kind
of
thing
like
not
giving
it
the
full
like
same
capabilities
as
you
would
with.
F
A
Yeah
and
a
big
appeal
is
the
Version
Control
aspect
of
it
too.
Right.
That's
like
one
of
the
major
pain
points.
We're
trying
to
solve
so
I
would
say:
I
would
like
the
the
compliance
pipeline.
Slash
group
project,
template
idea,
a
little
bit
more.
That
doesn't
I
think
change
too
much
technically
from
like
a
that
whole
aspect
of
Version
Control
and
all
the
features
you
get
with
that
and
then
that's
a
repository.
We
can
point
to
that
works
at
a
for
organizations
across
the
board.
A
So
but
I
think
there
are
other
comments
here
as
well.
Oh
and
Rob
you're,
just
plus
one.
You
want
to
vocalize
that.
C
Yeah,
just
I
agree
entirely
with
this
approach.
It's
how
security
policies
work
as
well,
where
you
have
to
select
a
security
project
to
then
add
policies
to
and
yeah
I
think
Tim
pointed
out
the
biggest
benefit
to
all
of
this
would
just
be,
and,
as
you
said,
having
those
different
merge,
rights
and
access
controls
to
your
dashboards,
just
like
we
do
with
sizeense,
where
I
knew
some
people
were
able
to
edit
those
yeah
it'd
be
about
I.
C
Think
it
also,
then,
lends
towards
us
being
able
to
control
access
to
individual
dashboards
as
well
having
the
ability
to
Define
which
dashboards
are
accessible
by
which
groups
of
people
or
which
projects
be.
E
Good
at
10
and
hence
to
set
a
later
point
also
that
it's
not
just
a
project
that
you're
pointing
to,
but
rather
like
a
project
plus
a
directories,
for
example,
but
I
think
the
looking
and
having
been
involved
into
Wiki's.
It's
a
pain
and
development
to
call
it
like
this,
it's
it
it's
a
constant
test,
it
never
just
from
listing
Pages.
This
was
never
performed
so
far,
yeah
and
so
everyone
hates
it
to
be.
A
Honest
yeah
we're
effectively
in
a
random
repository
again
right.
So
it's
just
more
more
of
a
headache:
Yeah
the
more
we
discuss
it.
The
more
I
like
the
idea
of
just
pointing
to
a
group
or
a
project
I
wonder
if
a
group
makes
sense
as
well,
if
we're
thinking
or
I
don't
know
potentially
iterating
to
all
of
those
capabilities
because
rob
you
had
a
good
point
about
access
level
requests.
You
know
you
could
have
a
group
with
a
multiple
projects
containing
multiple
dashboards.
A
I
guess
really
following
that
group
project
level,
template
concept
but
I
think
if
we
start
around
that
direction
and
figure
out
like
a
good
first
decoration
or
whatever
is
like
the
path
of
least
resistance,
similar
to
security
policies,
compliance
pipelines
and
group
project
templates,
then
we
can
go
from
there
and
just
expand
from
expand
further,
so
I
guess
as
far
as
action
items
I
will
I.
A
Don't
think
we
have
an
issue,
we
had
like
an
initial
investigation,
but
I'll
open
up
a
more
formal
issue,
at
least
for
us
to
look
at
and
then
potentially
Loop
and
optimize
as
well
to
see
if
we
can
kind
of
converge
on
a
on
a
solution
together.
Yeah.
C
We
have
that
Spike
with
Jeremy
and
and
Max
and
myself,
where
we
discussed
this
in
the
deep
in
some
detail,
but
actually
so
that
was
only
an
investigation
that
wasn't
yeah
proof
of
concept
or
anything
like.
A
F
A
I
would
just
say
it's
more
of
a
coordination
effort,
but
I
think
they'll
be
on
the
execution
side
of
it
and
we
can
stay
focused
where
we're
at
currently
until
we
we
have
no,
we
will
at
some
point,
but
we
don't
have
a
current
need
for
it.
So
just
want
to
make
sure
that
that
all
the
goods
are
on
the
same
page
from.
C
A
technical
level,
the
code
wise
on
the
back
end.
It's
we
have
a
hard-coded
path
that
we
go
to
on
a
project
for
them
point
of
view.
It
would
just
be
changing
that
to
check
to
see
if
there's
a
group
setting
available
or
not
and
use
that
instead,
it's
not
from
an
implementation.
It's
not
going
to
be
that
difficult.
Apart
from
setting
up
a
UI
sure.
E
Maybe
we
could
say
we
also
allowed
to
point
to
itself,
but,
for
example,
gitlab
has
very
rigorous
merch
requests
called
stuff.
That
is
super
complex.
If
you
would
have,
then
someone
editing
dashboards,
because
the
problem
is,
if
you
are
dating
a
dashboard
and
you
save
it,
that
you
would
be
running
into
the
problem
that
you
can't
commit
that.
B
E
A
A
Like
something
potentially
like
a
cascading
framework
where
we
can
override
or
or
even
lock
it
depending
on
their
organization's
needs,
but
I
do
believe
we
have
a
framework
for
that
where
we
can
then
add
that
flexibility
to
Define
it
at
the
project,
level
or
log
in
a
certain
level.
So
that's
definitely
something
I'll
include
in
the
issue.
F
F
So
I
had
the
next
one
I
just
wanted
to
check
in
with
how
the
transition
to
using
our
our
testing
cluster
rather
than
the
infrastructure
managed
one
was
going
from
last
week,
I
wrote
this
question
Dennis
before
your
update.
You
just
did.
A
No
worries
yeah,
just
to
recap:
we're
we're
good
to
go
as
soon
as
I
wrap
up
my
initial
set
of
meetings
today,
I
will
be
updating
the
internal
handbook
Mr
so
that
we
can
get
that
merge
and
start
collecting
data.
So
we're.
F
A
If
you
go
to
the
build
I,
don't
have
it
offhand
if
you
go
to
the
billboard
and
blocked
the
top
issue
that
is
assigned
to
rob.
That
is
the
internal
handbook
issue
change
and
that
will
have
the
the
link
to
merge
request.
A
But
I
will
be
updating
that
with
a
new
app
too,
as
well
as
the
endpoint,
and
then
that
will
point
to
our
cluster,
because
that's
honestly,
what
we've
only
needed-
we've
just
we've
just
been
trying
to
do
the
proper
integration
for
a
bit,
but
we'll
we'll
get
to
it
and
then
swap
over
I'll
retain
the
app
key
and
then
basically
behind
the
scenes.
Nothing
will
have
changed,
but
yeah.
D
Let
me
quickly
share
my
screen.
This
will
be
a
quick
demo
for
me.
I
think
Axel
is
going
to
demo
most
of
the
stuff.
That
is
super
interesting.
So
this
is
the
new
product
analytics
dashboard.
This
was
made
from
Rob's,
nice
and
sign
here
and
to
show
you
the
Delta,
because
it
really
seems
like
we're
starting
to
build
an
actual
product
that
people
would
want
to
use
or
see,
especially
for
the
internal
demo,
and
this
is
how
it
started
out-
and
this
is
where
we
are
with
this
Mr.
D
B
Thank
you
so
yeah,
as
at
the
end,
it
is
going
to
be
a
quick
demo.
This
is
basically
all
of
the
visualization
that
we
have
right
now
in
a
single
place.
We
have
the
single
stat
one
with
a
static
icon
and
and
a
meta
text.
We
have
the
the
table
visualization
that
Gian
is
currently
working
on.
This
is
me
fetching
the
changes
from
gncmr
that
is
in
progress
and
and
rebasing
it
locally
and
the
first
one
we
finished
the
line,
one
so
yeah
I
just
wanted
to
show
you
all
of
these.
B
In
the
same
place,
we
don't
have
much
on
on
the
audience.
Dashboard
and
I
also
wanted
to
take
a
look
real,
quick
to
the
single
stat
options
that
Dennis
was
was
talking
about
so
yeah
right.
B
A
Second,
good
and
yeah:
we
can
explore
that
later
on,
whether
that
happens
in
Cube
or
in
this
component.
But
I
don't
know
if
we
pass
like
an
object
or
an
array
of
values
in
it
and
we
tell
it:
hey:
dude
perform
an
X
calculation
on
it,
but
we
can
that's
a
that's
a
spike
for
another
time,
but
visualization's
a
good
good
job.
Everybody
yeah.
A
8's
got
all
those
dashboards
lined
up
so
yeah.
Please
check
out
those
issues,
make
sure
they
sound
good
to
you
and
it
seems
like
we
have
all
the
pieces
in
place.
Otherwise,
assuming
Max
and
I
can
get
session,
query
fully
tuned
up.
B
Yeah,
this
is
exciting,
so
Dean
take
it
away.
F
E
I
cut
the
Mr,
the
widget
designer
the
first
version.
The
Mr
is
ready
and
I'm
waiting
for
green
pipelines.
All
the
specs
are
written,
which
was
super
tedious
because
there
are
a
lot
of
buttons
thinking
that
can
be
clicked.
This
is
now
the
final
version
of
the
widget
designer.
There
are
mainly
two
more
changes,
so
one
of
them
is
that
you
can
do
page
views,
that's
something
you
already
saw.
E
It
will
always
as
soon
as
you
click
add
to
dashboard.
You
will
see
always
the
code,
because
we
can't
save
it
anywhere.
At
the
moment,
the
most
interesting
thing
that
I
would
say
is
on
one
hand
that
you
can
that
the
grouping
now
works
nicely.
E
So
it
says
events
groups
by
day
and
the
other
thing
is
that
you
can
do
multi-dimension,
adding
it's
pretty
interesting,
because
you
can
say:
okay,
what
are
the
the
hosts
that
we
are
getting
data
and
then
you
can
say
I
want
to
add
another
dimension
for
the
page
path
and
then
you
will
basically
now
we
just
need
a
thing
to
sort,
but
you
will
basically
get
not
only
the
host
but
also
the
path
and
could
say
which
ones
are
coming
from
where.
E
But
you
could
can
also
combine
this,
that
you
say
what,
where
the
you
page
languages
that
were
used
in
a
different
host
and
you
see,
the
offender
is
page.git.com.
That
is
using
Windows
1252,
and
this
is
basically,
you
can
add
as
many
as
you
want,
and
it's
adding
more
and
more
to
that
query
and
is
transpiling
directly
into
the
query
for
Cube,
and
this
is
something
you
can
take
one
to
one
and
it
works
so
yeah.
Hopefully
all
the
specs
get
green.
E
The
only
thing
that,
as
this
is
using,
then
the
data
table
from
yarn,
which
is
also
working
nicely,
I'm,
simply
waiting
until
this
is
merged.
This
looks
already
very
far,
then
I
will
go
replace
and
then
we
can
get
this
into
review
so
getting
there,
and
the
next
version
will
then
be
having
filters
and
date
ranges
on
top
of
it.
So.
E
A
E
A
Yeah,
let's
keep
it,
let's
keep
an
eye
on
how
things
are
going
and
we'll
extending
these
sessions.