►
From YouTube: Geo: design workshop part1
Description
This is part of the design discussion about the monitoring Geo. Starting from summarized outcome from the sketching activity to the user interview.
A
This
is
Sanjay
and
Fabien
we're
going
to
talk
about
our
interview
experience.
It's
not
our
experience
that
we'd
like
to
wrap
up
those
like
comments
from
the
users
and
we
would
like
to
move
on
based
upon
their
feedback
and
I,
have
set
up
this
mural
board
for
the
Geo
team,
starting
from
our
nicely
designed
Geo
future.
Look
and
I
made
some
notes,
because
we
have
already
some
of
the
like
common
patterns
about
the
very
first
view
of
the
geo,
so
I
think
we
could
start
like
looking
into
this
part
if
you're,
okay.
A
Yes,
let's
do
that.
So,
thanks
to
the
team,
we
have
five
inputs
and
I
realized
was
really
interesting
to
see
that,
like
Mir
titled,
like
more
seeing
the
very
first
UI
for
like
monitoring
tool,
of
course,
like
our
current
UI,
is
also
for
the
monitoring
but
like
how
it
looks
like
looks
quite
different
and
of
course,
like
there
were
a
certain
like
tendency
that
the
okay
we
have
to
show
the
very
first
level
of
overview
and
if
user
click
on
those
overview
that
we
could
show
those
detail
pages.
A
That's
how
I
see
this
drawings,
and
also
there
were
like
some
needs
from
our
team
more
about
this
video
license
data
I,
really
love
this,
like
graph
from
tones
I
think
there
could
be
like
better
way
to
show
this
replication
worse
or
education
works.
I
think
there
could
be
some
nicer
way
to
visualize
this,
so
we
could
put
some
like
better
data
visualization
on
the
very
first
page
and
also
some
some
P
we'll
try
to
mention
the
location
so.
B
B
A
I
think
that's
that
could
be
our
very
quick
summary
and
also
some
question
for
me
is
like
how
how
shall
we
manage
the
other
feature,
like
other
very,
very
big
feature
like
failover,
for
example,
like
upgrading
geo,
is
also
a
big
thing,
but,
like
we
kind
of
like,
haven't
actually
think
about
it
at
the
time,
because
the
task
was
moved
mostly
focusing
on
monitoring
things.
You.
A
A
B
Mean
we
we
have
to
like
see
where
it
makes
sense,
but
I
can
totally
see
for
the
failover
specifically
right
and
also
for
the
installation
maybe
like.
Actually,
this
is
something
where
a
lot
of
interesting
work
and
Happiness
of
like
how
can
we
as
a
separate
sort
of
upgrading,
get
lab,
including
to
you
or
installing,
give
lab,
including
on
like
hardware
versus
like
setup
required
to
actually
couldn't
like
configure
a
geo
like
system
right,
yes
and
I?
Think
it's
especially
for
like
the
set
up
part
right.
B
We
can
do
not
better
in
the
in
the
UI.
It's
also
like
an
open
question
for
me
is
sort
of.
Where
is
the
entry
point
for
a
user
right?
Where
do
we
leave
the
general
sort
of
hey?
You
know
like
you
need
you
need
a
working,
get
lab
installation
right
and
now
you
know,
comes
the
Geo
part.
So
it's
a
it's
interesting
yeah,
but.
A
Yeah,
it's
also
a
bit
tricky
because
I
I
mentioned
in
the
interview
notes
but
like
there
are
just
like
one
person
normally
who
needs
to
take
care
of
the
gillip
version,
yes
and
then
geo
version
at
the
same
time.
So
probably
they.
You
just
think
this
task
as
one
big
task
way
of
grading
things
correct
so
yeah
I
think
this
is
a
really
good
point.
So
I
just
like
mark
the
phone
on
the
mural
yeah.
B
A
B
Many
systems
administrators
have
their
own
homegrown
monitoring
system,
which
we
may
or
may
not
want
to
replace
with
like
get
labs
monitoring
system
I
know,
that's,
there
may
be
a
separate
discussion
to
have,
but
they
the
pattern
that
you
describe
here
I
think
is
really
right.
Right.
It's
like
most
of
the
time
you
would
go
to
the
like
get
lab,
hey
like
give
up
geo
overview.
If
you
had
the
problem
right
and
then
you
know
like
once,
you've
detected
that
problems
ID
live
here,
some
form
of
monitoring
and
you
actually
go
there
right.
B
It's
not
like
people
open
up
the
geo
interface
every
morning
to
look
at
it
and
say
like
wow.
This
is
so
nice
right.
It's
very
sort
of
like
task-driven,
as
in
I've
detected
this
problem
now.
I
know
I
want
to
resolve
it
right,
so
the
user
story
here
is
really.
You
know.
I
want
to
open
that
itu
administrator
interface
to
understand
which
files
were
in
synced
right,
which
repositories
are
not
saying
something
like
that.
Yeah.
A
That
that
was
very
true
like,
and
it
was
kind
of
like
not
my
assumptions
at
the
time.
So
if
you
see
there
like
behaviors
area
like
like
you
just
mentioned
like
they,
they
they
don't
go
to
the
Geo
UI
like
everyday
or,
like
really
regularly
I,
think
only
one
user
group.
They
answer
that
they
go
regularly
mm-hmm
every
week
like
once
a
week.
That
was
because
they
don't
have
like
homegrown
monitoring
system
for
the
geo
replication.
I
think
that
was
the
only
reason
yes,
but
if
they
had
their
homegrown
like
monitoring
system.
A
In
that
case
they
they
always
monitor
those
dashboard
or
data,
and
then,
if
there's
something
wrong
and
then
they
just
like
okay,
there's,
probably
something
wrong
on
the
secondary.
He
knows
where
or
I
don't
know
in
some
thinking
it's
problematic,
and
then
they
decided
to
like
come
to
our
primary
UI
and
coming
to
the
Geo
UI.
So
it's
like
that's
because
they
have
to
come
into
the
admin
area
and
they
have
to
click
once
more.
A
Like
geo
knows
yes,
so
this
is
already
like
second
stats
and
and
what
we
expected
to
see
at
the
time
like.
Okay,
like
probably
they
are
doing
something
on
the
UI,
but
until
now
we
don't
have
like
very
clear
idea
like
what
we
are
actually
doing
on
the
current
UI,
but
I
realize-
and
probably
you
have
already
noticed
that
they
go
to
the
secondary
UI
first
and
check
the
healthy
status
but
like
they
were
all
the
very,
very
professional
geo
users.
So
they
already
know
that
the
healthy
doesn't
mean
actually
it's
healthy.
A
B
I
think
the
healthy
I
think
actually
they've
kind
of
learned
mm-hmm
by
probably
failing
a
little
bit
right,
is
that
the
healthy
indicator
is
really
a
an
indicator
of
like
like
I.
Don't
like
system
Hills
right
as
in
is
there
like
is
gitlab
like
running.
You
know
and
configured
correctly
right.
That
is
what
it
means.
B
It
doesn't
mean
everything
is
synced
I,
don't
everything
is
in
sync
right
or
there
are
no
other
problems,
so
I
think
that
is
really
misleading
right,
because
you
always
also
have
to
sort
of
expand,
whether
like
synchronization
information
in
order
to
see
oh,
it's
healthy,
but
like
two-thirds
of
these
files
are
not
saved
right.
So.
B
A
Could
have
some
like,
probably
at
first
time
they
were
very
confused
and
then
they
they
realize,
because
thankfully
they
are
very
smart
people,
so
they
just
realized.
Okay,
this
is
just
healthy,
but
still
we
need
to
thicken
and
they
they
learn
from
their
experience
and
they
know
that
they
still
need
to
open
the
verification,
information
and
the
sync
information,
and
they
they
check
the
details.
A
A
A
And
also
some
some
users
mentioned
like
the
left,
sidebar
navigation.
We
sometime
just
call
it
like
subpages
the
label
itself,
it's
kind
of
probably
fake
to
them.
Of
course
like
now,
we
know
how
what
it
means,
but
probably
for
them.
It
looks
a
little
bit
confused
and
this
was
also
mentioned
like
twice
as
far
as
I
remember
and.
A
A
B
B
A
That
was
very
true,
I
I
could
agree
on
that
and
then
I
think
we
can
make
some
changes
on
this
and
I
think
the
last
pain
point
that
I
found
was
like
the
separate
information
of
verification
and
sync
information.
This
was
mentioned:
I
think
that
they
understand
the
meaning
of
each
artifacts
and
then
why
we
represented
those
data
like
that
way
because
they
had
seen
the
document
they
have
used
this
product
for
like
some
months
or
they
there
are
really
professional
users.
A
B
A
I
think
we
need
to
think
about
like
better
solution.
Yeah
I
think
I
need
to
think
about
better
solution
how
to
visualize
those
two
different.
It's
actually
on
the
same
level,
but
anyway
it's
a
different
information,
so
I
think
I
need
to
visualize
this
in
a
different
way.
I
think
I
have
shown
you
to
lobby.
B
B
Also
I
think
there's
also
this
I
think
we
are
incomplete
in
this
regard
as
well
right.
It's
like
you,
have
some
pages
for
projects
or
something
you
can
do
stuff,
but
we
don't
have
this
for
some
other
bits
like
LFS
files,
right
and
so
I
think.
There's
also,
you
know,
there's
sort
of
different
user
stories
right.
It's
like
you
go,
you
know
to
the
UI
because
he
is
synced
Anthony.
You
click
on
details
and
you
want
to
see
them.
Then
you
want
to
do
something
about
it
right
and
functionality.
B
A
And
yeah
I
think,
like
the
interview
was
like
very
valuable
to
me
to
understand,
like
their
pain
points
and
their
like
current
status
and
our
current
status
also
and
I
think
basically
like
and
also
surprisingly
like
they
really
wanted
to
solve
some
problems
on
the
UI,
because
my
assumptions
was
because
they're
so
used
to
like
using
the
terminal
and
then
probably
they
can
just
I
use
the
rake
path
and
then
some
of
the
very
familiar
comments
that
they
can
use,
but
still
like.
They
really
wanted
to
solve
the
problem
on
the
UI.
A
B
You're
actually
right
there,
it's
like
I
think
for
like
what
I've
found
or
like
my
impression,
was
if
they
have
to
go
there
anyways,
even
if
they
are
capable
of
figuring
out
what
to
do
next
right,
that's
more
work!
Well,
it's
like
you.
You
have
today
switch
context.
They
have
to
go
like
I
think
we
had
one
instance
where
user
went
into
like
the
Postgres
database
right.
A
B
B
That
makes
you
more
productive
right.
This
is
the
I
think
the
the
difference
is
really
will
come
like
okay.
If,
if
solving
the
problem
means
I
have
to
like
I
have
let's
say,
I
have
a
thousand
LFS
files
that
are
not
replicated
correctly.
I
only
have
a
single
button
right,
like
retry
like
them,
one
by
one
clicking
a
thousand
times.
People
will
not
do
that.
B
You
know
they
then
may
go
to
the
command
line,
but
then
maybe
we
need
something
like
a
batch
processing
right
and
then
that
goes
away
so
I
think
it's
more
of
like
if
the
things
that
they
would
do
on
the
command
line
actually
work
when
the
graphical
user
interface,
where
there
are
already
it
felt
like
they
would
just
do
it
there,
because,
yes,
why
not?
Why.
A
Not
yeah
I
think
this
is
like
what
like
they
have
questions
about
like.
Why?
Not,
of
course
they
they
will
understand
like
it
will
be
hard
for
us
to
implement
everything
at
once,
but
the
seal.
What
they
expected
was
like
I
want
to
solve
this
problem
on
the
UI,
because
I
have
found
the
problem
on
the
UI
and
I
want
to.
Yes,
it's
cleared
up
so
yeah
I
think
this
is
like
what
they
expected,
and
this
is
like
really
like
very,
very
simple
somewhere.
A
B
A
I
agree,
so
what
we
can
do
is
based
upon
their
comments.
So
just
do
the
number
two
exercise
that
it
is
a
little
bit
different,
so
it
seems
like
they
are
having
a
problem.
Of
course
they
already
used
to
it,
but
still
they
want
to
see
this
information
and
they
want
to
explore
this
information
more
in
an
easy
way.
A
A
A
B
A
B
I
think
that
would
be
really
nice
I
think
we
should
for
for
the
next
exercise.
I
think
we
should
maybe
invite
the
team
as
well,
and
we
could
also
just
just
a
thought-
and
we
could
run
a
first
session
with
some
of
the
team
members.
But
leave
is
open
for
like
people
to
like
asynchronously.
Add
some
some
post-its
in
here
as
well.