►
From YouTube: Layer5 Websites Weekly Meeting (Feb 14th, 2022)
Description
Layer5 Websites Meeting - February 14th, 2022
Agenda:
- Filters in mobile view #2527
- Loading animation for calendar #2601
- Convert background image to Gatsby background image on homepage. #2596
- Synthesize meshery.io and docs.meshery.io navbars.
- Issues with mobile menu
- Docs Sidebar
- Exposing test results on public dashboard on https://smp-spec.io
Join the community at https://layer5.io/community
Find Layer5 on:
GitHub: https://github.com/layer5io
Twitter: https://twitter.com/layer5
LinkedIn: https://www.linkedin.com/company/layer5
Docker Hub: https://hub.docker.com/u/layer5/
A
B
A
Okay,
just
briefly,
if
I
could
briefly
just
because
I
know
you're
you're
in
the
zone,
which
is
great
but
I'll
also
say,
welcome
everybody
to
the
layer.
Five.
A
I
should
stop
the
youtube
live
stream,
it's
always
so
difficult
because
you
start
the
live
stream.
It's
like
bringing
up
the
browser
and
if
you
close
it
beforehand,
you're
not
sure
if
it's
going
to
actually
make
the
live
stream,
so
so
with
no
further
interruption.
It's
monday
february
14th,
we
do
record
the
calls
thanks.
Everybody
for
coming
priya
is
hosting
today.
A
A
There's
some
folks,
like
abba
jay,
who
it's
becomes
quite
obvious,
based
on
their
back
in
the
background
and
zoomed,
like
he's,
been
on
a
call
before
one
or
two
or
ten
yeah
yeah,
any
anybody
else
so
anshu
and
aditya.
It's
not
obvious
by
your
lack
of
a
last
name,
whether
or
not
you've
you're
an
aditya
or
an
onshoe
that
we
already
know
there's
a
few
of
them
here.
Do
you
guys
want
to
say
hi
aditya?
Aren't
you.
A
All
right
so
that
that
aditya
is
the
mr
subedi.
Okay,
we
we
know
that
riff
raff.
We
know
that
guy
and
then
anshu,
say
hi
in
chat
or
or
on
the
microphone.
If
you
would
otherwise
we'll
just
keep
harassing
you.
E
B
Okay,
so
the
recent
update
which
the
pioneered
to
the
filters.
B
B
So
a
couple
of
things
which
I
have
told
him
to
change
and
you
guys
can
give
your
feedback
to-
is
that
the
first
thing
was
so
in
the
desktop
view.
Right
now,
all
these
filters
are
open
by
default
because,
like
there's
a
lot
of
space,
so
we
can
keep
it
open.
So
I
told
him
to
keep
that
open
in
the
desktop
view
by
default
and
they
keep
it
closed
in
the
mobile
view
when
it
opens
for
the
first
time.
That
is
one
thing.
B
Another
thing
is
this:
one
see
this
filter,
it
has
a
cross
and
filters,
and
this
is
clear
filters
and
across
so
they
kind
of
give
the
same
impression
to
me.
It's
like
you
know
clear
all
filters
that
kind
of
an
impression
to
me.
So
I
told
him
to
change
this
to
say,
apply
filters
and
probably
change
this
into
a
button
as
well,
like
probably
a
clear
filter
button
and
apply
filter
button,
because
otherwise,
if
we
keep
this,
they
have
to
be
aligned
right
now.
The
text
is
not
aligned
this
one
so
yeah.
B
F
I
think
using
different
button
is
not
necessary.
We
can
like
change
the
filter
button
as
well
like
we
can
change
the
states
as
if,
if
user
is
using
the
drop
down
and
then
it
is,
it
will
be
applied
straight
and
when
a
user
is
not
using
drop-down,
then
it
will
be
like
cross
filters.
F
Like
the
drop
down
is
not
visible,
then
the
state
will
be
apply
filters
and
when
the
drop
down
is
visible,
then
state
will
be
cross
filters.
The
same
button
no
need
to
like
use
a
different
button.
Yeah
am
I
audible.
B
G
Okay,
the
thing
is:
actually
we
need
a
button
to
clear
the
filters
like
when
you're
selecting
something
like
articles
or
blogs.
You
need
a
separate
button
for
that.
Okay,
the
thing
is,
we
can't
work
around
that,
but
about
the
applied
filters.
I
don't
think
we
need
it
because,
like
we
are
filtering
all
the
options
with
an
on
click
thing,
so
having
a
separate
apply
button
apply
filters.
Button
doesn't
make
any
sense.
B
I
wasn't
saying
create
a
separate
apply
filter.
I
was
just
saying
I
think
you
weren't
here
before
see
this
one
cross
filters
and
this
one
clear
filters.
Suppose
you
don't
have
this
one,
this
kind
of
gives
an
illusion
that
you,
you
know
you're
removing
all
the
filters.
Instead,
we
can
just
change
this.
Instead
of
the
icon,
then
just
say
apply
filters.
It
doesn't
have
to
do
any
other
functionality.
That
is
what
I
was
saying.
H
H
Yeah,
because
across
will
be
symbolic
to
closing
the
filters.
H
And
clear
filters,
I
don't
think
clear.
Filter
needs
a
cross
button,
a
clear.
G
Yeah
yeah,
I
removed
it
by
mistake.
I
think
I
would
add
it
again.
H
And
one
other
thing:
we
can
decrease
the
font
weight
of
type
and
decrease
the
font
rate
of
type
make
it
make
it
make
it
different
from
filters
main
filters
and
the
circuit
itself
the
keep
the
phone
fit
of
the
filters
and
decrease
it
of
the
type
written.
B
Yeah
that
can
be
done
also
the
button,
if
you
missed
it.
One
more
thing
that
I
mentioned
is
these
filters
should
be
open
in
the
desktop
view
by
default
and
in
the
mobile
view
it
should
be
closed
by
default.
Can
you
do
that.
B
I
I
It
was
like
only
loading
animation
and
the
after
the
changes.
It's
like
we
blur
the
means
the
calendar
is
blood
and
just
showing
that
user.
That
calendar
is
loading.
F
F
It
is
very
fast.
It
is
like
not
even
rotating
in
a
full
circle
too,
so
it.
If
it
has
some
delay,
then
it
will.
It
looks
good.
I
J
I
I
A
So
it's
not
it's
not
the
most
critical
that
we
actually
show
the
calendar
and
full
opacity,
but
in
some
respects,
ideally
or
like
50,
because
it's
it's
only
the
last
like
half
of
the
calendar,
that's
waiting
to
be
loaded
so
give
that
some
thought,
if
you
would,
if,
if
we
don't
get
through
a
full
cycle
of
the
spinner
beautiful,
if
we
never
show
the
spinner
beautiful,
it's
like
this.
A
That
just
means
that,
like
we
didn't
really
need
it
in
the
first
place
or
not
necessarily
or
you
know,
sometimes
it's
good
work
and
and
this
is
desired,
and
it's
a
nice
custom
spinner
you
might,
you
might
double
its
size
or
something
you
know.
A
I
don't
know-
maybe
just
to
you
know
just
so
that
it's
really
apparent
that,
like
something
is
there
good,
good
job
on
the
reducing
the
opacity
it's
like,
even
if
you
leave
it
just
as
is
it
you're
still
indicating
to
people
to
visitors
that,
like
hey,
you
know,
there's
something
here,
it's
not
yet
interactable.
You
know
you
can't
interact
with
it
yet,
but,
like
you
know
it
the
whole.
A
A
So
so
we
generally
don't
have
an
issue
with
it
loading
quickly.
The
calendar
data
that's
coming
from
a
separate
site.
That's
coming
from.
You
know
google
calendar,
I'm
trying
to
think
if
there
are
other
places
on
this
site
or
potentially
other
sites
that
you
know
remote
data
is
being
loaded,
and
this
can
be
reused.
A
Calendar,
it's
actually
the
same
calendar.
It's
a
I
mean,
I'm
sorry,
it's
actually
the
same
data
same
calendar
data
source,
it's
a
different
implementation,
but
food
for
thought.
In
case
you
want
to
capitalize
on
your
investment.
You
know.
A
I
L
Okay,
so
I
don't
know
I
don't
have
to
do.
I
have
to
share
my
screen
because
I'm
not
really
running
locally.
I
just
joined
the
call
right
now.
B
L
No,
no!
It's
fine!
It's
right!
Let
me
check
so
I'll,
probably
like
go
through
the
issue.
I've
not
really
I'll
go
through
the
link
that
is
posted
here.
L
So
basically,.
L
Did
here
was
I
discussed
with
rbj?
I
think
I
think
if
I
pronounced
them
correctly,
so
he
was
working
on
the
optimizing
optimizing,
the
front
the
home
page.
So
I
wanted
to
do
other
contributions
like
aligning
the
the
big
hero,
images
and
other
things.
So
he
said
he
has
he's
working
on
them
and
he
has
created
ppr,
and
so
I
looked
through.
I
saw
that
there
was
a
there
was
see
an
issue
with
the
big
image.
In
the
background,
which
was,
this
is
a
background
image
here.
L
Let
me
go
through
this
so
and
so
when
I
asked
him,
if
he
had
any
plans
in
trying
optimizing
the
image
he
said,
he
found
that
there
was
an
issue
like
he
saw
that
gut
speed
doesn't
really
like
support
doing
it,
but
during
the
background
image,
so
I
I
had
to
do
some
research
and
I
found
that
there
was
a
way
you
could
actually
do
it
with
some
plugins.
That
was
thought,
supporting
background
image
and
like
the
previous
version,
but
then
they
had
some
other
plugins.
They
called
gb
gb
bridge.
L
That's
the
bridge
that
helps
you
like
do
some.
You
connect
the
the
gasb
plugin,
the
the
current
version,
and
you
do
the
conversion
of
the
background
image.
So
I
implemented,
I
implemented
that
after
the
research
and
I
was
able
to
like
turn
this
particular
background
image
to
like
when
it's
loaded
to
a
web
web
image.
So
from
and
from
what
I
see
when
it's
loaded
you
take,
it
goes
from
being
one
mega
one
mb
size
to
around
kilobytes
in
size,
so
that
is
what
I
basically
did.
L
So
it's
also
like
a
really
a
tangible
reduction
in
I
mean
like
in
the
optima
in
size,
so
I
felt
it
was
good.
I
I
created
the
pr
for
that
as
well.
I
intend
to
also
do
this
same
thing
in
all
background,
like
from
background
imaging,
but
in
other
pages,
as
well
as
well
as
implement
the
gatsby
static,
imaging
the
other
ones
that
are
not
done
yet,
so
that
that's.
That
was
literally
what
I
did
so
so
to
like
show.
L
So
if
I
try
to
this
image
right
now
is
basically
like
around
it's
about
seven
kilobytes
so
from
from
one
from
one
megabytes,
so
so
that
that's.
A
Just
kidding
yeah,
no,
that's
amazing!
Isn't
it
that
so
is
this
leonard?
Is
that
webp
format
here
yeah
yeah
yeah,
it's
in
what
people
might
hear?
Yes,
yeah!
Do
we
great
oh-
and
I
think
I
I
was
going
to
ask
a
question,
but
I
think
I
see
the
answer
so
for
should
a
visitor
be
using
a
browser
that
doesn't
support
web
p.
Is
that
second
background
image?
L
Okay,
okay,
so
when
I,
when
I
was
doing
the
the
other
contribution
that
was
really
that
was
later
closed
I
had
to
set,
there
was
some
configuration
I
I
I
was
supposed
to
make
in
the
I
think
in
the
package.json
at
the
gatsby
config.
That
should
have
a
fallback,
so
it
has
like
the
best,
the
first
one,
the
second
one,
the
fallback.
So
I
know
just
because
the
first
that
contribution
wasn't
matched
the.
L
It
is
not
the
quadration
is
not
in
the
is
not
the
gatsby
config
yet
so
so
since,
because
of
the
configuration
is
not
the
this
config,
I
could
actually
also
like
add
that
configuration
so
that,
even
if,
like
if
you're
working
with
the
gatsby
background,
like
the
static
image
that
people
are
like
abby
and
the
rest
has
been
like,
adding
to
the
page
that
we
should
also
have
a
fallback
say
if
you
use
it
in
in
a
browser
that
is
not
supported,
but
then
you
could
actually
try.
L
F
I
think
it
will
work
in
all
browsers
and
you
don't
need
to
config
as
a
nikhil
already
config,
all
the
files
we
have
in
images
folder.
So
maybe
we
don't
need
to
con
config
that
specific.json
file.
F
One
thing
you
can
do
is
you
you
can
inspect
it
and
if
it
shows
two
two
lines
like
a
one
for
web
format
and
one
for
png,
then
it
will
be
fine
and
like
you
can
go
on
that
particular
dip
where
it
is
where
it
is
present
and
if,
if
it
it
is
showing
the
two
files
like
one
for
the
web
format
and
one
for
the
png
format,
then
we
will
like
we
can
go
to
like
it
is
working
fine
for
us.
If
it
shows
the
do
a
format
that
we
want.
A
Yeah,
I
guess
a
couple
of
things
that
that
the
quality
of
the
image
is
I
in
this
case.
I
don't
think
it's
a
concern.
I
don't
know
that
it's
even
to
my
eyes
that
there's
enough,
I
don't
know,
I
don't
know
that
you
can
really
tell
a
difference
in
terms
of
the
quality
of
the
image,
which
is
amazing,
considering
the
savings
and
bandwidth
and
the
savings
of
speed.
So
that's
fanta
the
plug-in
the
the
configuration
that
you
were
speaking
to
inside
a
gatsby
config.
Is
that
plug-in
specific
or
is
that.
L
No,
no,
it
is
not
plug-in
specific
it's,
for
example,
if
if
the
gatsby
static
image
or
start
like
the
gatsby
ways
of
optimizing
images
using
the
page,
the
config
actually
like
it's
more
of
you
implementing
a
fallback.
So
it's
either.
You
do
you
go
from
say
a
webp
to
a
jpeg
or
to
an
author.
So
it's
more
like
it's
a
fallback.
L
So
if
this,
if
this
one
doesn't
load
load,
the
other
one
so
so
for
unsupported
browsers,
yeah,
the
webp
wouldn't
work
so
to
default,
to
maybe
the
jpeg
and
probably
to
use
the
time
to
like
find
the
auto,
probably
like
look
for
the
best,
the
best
ones
that
you
to
to
to
go
like
that.
You
show
in
that
page.
L
L
A
Do
we,
okay
is
the
let's
get
more
speci.
So
since
it's
a
sweeping
change,
this
is
kind
of
a
big,
potentially
a
very
positively
impactful
change,
but
one
that's
large
in
nature.
We
should
dig
in
here
a
little
bit
more,
so
the
for
the
contributors
to
the
website.
How
what's
the
mechanism?
What's
the?
How
do
they
configure
a
fallback?
A
L
Have
to
perform
any
extra
configuration
once
it's
once
you
put
that
line
in
the
in
the
in
the
gatsby
config.
You
don't
have
to
you.
Just
do
your
your
normation
and
maybe
try
to.
If
you
add
the
static
image,
plugin
or
whatever
you
don't
have
you
don't
have
to
configure
anything.
So
far,
is
there
already
in
the
that's
the
config
that
this
is
this
hourglass
we
config
that
has
to
please
okay,.
A
L
A
L
B
Yeah
one
thing
that
we
can
do
in
your
present
pr
is,
since
the
quality
is
reducing
like
a
little
bit,
we
can
probably
put
a
darker
overlay
like
one
shade,
darker,
probably
to
the
image
using
css.
That
is
one
thing
which
we
can.
A
Oh
very
good,
I
was
just
saying
hi
to
ashwin
who's
on
the
call
nice
to
have
you
ashwin,
if
you
don't
mind,
take
a
moment
to
drop
your
last
name
into
zoom
and
that
way,
we'll
know
you
from
the
other
four
ashwins
here,
you'll
be
the
special
ashwin.
A
Oh
nice,
very
good,
yeah,
absolutely
ashwin
poke
poke,
as
you
will
poke
around,
as
you
will
all
right,
good,
so
hit
the
next
topic
up
and
devil
priya.
If,
if
you're,
okay
with
this,
if
you're,
okay
sharing,
I
was
just-
we
had
made
a
recent
comment
in
one
of
our
calls
somewhere,
and
I
think
we
filed
an
issue
on
well
trying
to
make
the
the
difference
between
the
mescheri.o
project
site
and
the
difference.
A
The
fact
that
that's
a
site
and
then
the
documentation
for
measuring
that's
a
different
site,
we're
trying
to
minimize
some
of
the
differences
as
people
switch
back
and
forth.
So
this
is
actually
right
where
you're
sitting
is
a
perfect
use
case
deborah.
If
you
were
to
click
on
to
open
in
a
new
bra
new
tab,
the
docs
page,
I
hope
no
one
else
was
working
on
this,
because
I
I
didn't
quite
check,
but
I
just
made
a
couple
of
quick
updates
to
try
to
pull
the
two
nav
bars
more
into
alignment.
A
And
the
there's
actually
a
gradient.
That
is
super
light.
That's
happening
right
here
in
the
doc's
nav
bar,
it's
very
very
light.
It's
a
little
bit
darker
at
the
top
and
just
a
smidge
lighter
at
the
bottom.
A
A
If
you
go
back
deborah,
if
you
were
to
go
yeah
and
you
go
to
the
very
it's
actually
kind
of
bad,
because
here's
here's
something
that
we
could
use
some
help
on
so
right
here,
there
is
what
I
like
to
think
of
as
a
fairly
pretty
gradual
linear
gradient,
starting
from
a
darker
green
to
a
lighter,
green
vertical
right.
Now
that
same
linear
gradient
has
been
applied
to
the
documentation
nav
bar.
A
So
if
you
switch
over
like
they
should
be,
it's
really
hard
to
tell
like
this,
but
I
I
think
that
these
are
the
same.
I
think
what
I
don't
think
is
the
same
yeah.
I
know
it.
It's
deceiving
like
to
me
right
here.
It
looks
like
the
project
site
actually
has
a
gradient
and
the
documentation
site
doesn't.
A
But
if
someone
chooses
to
dig
in
then
please
prove
me
right
or
proved
me
wrong.
There's
there's!
There
is
something
that,
if
demo
priya
this
is
super
small.
I'm
not
sure
why
we're
talking
about
this
this
long.
But
if
you
go
to
the
project
site,
when
you
start
to
scroll,
we
have
the
nav
bar
follow
and
if
you
notice
it
changes
color.
A
A
Okay,
I
get.
I
don't
know
it
feels
like
it's
a
little
darker,
the
like
yeah
when
you
that
feels
a
little
darker
than
anyway
we're
just
trying
to
minimize
the
differences
between
the
two
sites
so
that
people
don't
really
feel
like
they've
dropped
out
of
one
and
hit
the
other.
Ultimately,
if
they
feel
that
way,
they
know
it's.
Okay,
the
mescheri
logo
is
a
little
bit
smaller
on
the
documentation
site.
That's
just
so
that
it's
the
same
height.
A
I
think,
as
the
the
word
documentation,
but
we
could
change
that
we
could
make
the
logo
a
little
bit
bigger.
So,
whatever
any,
is
this
interest?
Anybody
in
terms
of
trying
to.
A
A
A
Let's
say
that
the
nav
bar
is,
I
don't
know,
150
pixels
tall.
Let's
just
say
you
know
somewhere
in
there
right
now.
If
you
apply
a
gradient,
a
linear
gradient
to
that
you'll
see
that
it
would
go
from
very
dark
to
very
light
in
a
short
little
span.
It's
like!
Well,
that's
not
what
that's
not
the
coloring
that
we
want.
We
want
it
to
be
only
the
first
150
pixels.
A
A
Why?
I
remember
that
I
have
no
idea,
and
so
so
it's
kind
of
an
interesting
challenge
that,
like
okay
well
on
the
doc
site,
what
else
is
you
have
to?
You
have
to.
A
There's
different
ways
to
figure
out
just
how
to
make
that
gradient,
precise
and
so
yeah,
like
I
said,
it's
kind
of.
If
you
challenge
yourself
to
really
make
it
the
same.
It
becomes
an
interesting
challenge
that
type
of
discussion
on
css,
for
some
of
you
probably
feels
like
a
sort
of
ridiculously
small
thing.
Why
are
we
it's
like?
A
Well
actually,
when,
if
you're
really
trying
to
make
good
software,
when
you
can,
when
you
have
the
time
to
like
every
pixel
counts
it
really,
it
really
does
there's
a
nikhil
sharma
who's,
not
on
the
call
today
he's
been
on
the
call.
The
last
couple
of
few
weeks
he's
just
starting
an
internship
with
layer.
Five,
he
is
focused
on
nothing
but
pixels
at
the
moment
and
and
his
impact
is
massive
he'll.
If
he's
here
next
week,
he'll
show
what
he
and
what
he
had
worked
on.
B
B
Okay,
so
the
next
topic
is
issues
with
mobile
menu
in
get
network,
so
I
think
there
was
a
fix
about
the
get
nighthawk
website,
but
there
was
something
with
the
file
change
like
the
changes
were
made
in
the
css
file,
but
they
were
supposed
to
be
made
in
and
sas
fell
or
something.
That
is
why
this
did
not
get
implemented
object.
You
were
coordinating
with
yeah.
B
Yeah,
so
the
changes
and
styles
will
be
the
same,
just
starting
to
move
to
the
other
side.
I
think.
A
Right,
I
don't
right
yeah
right,
I
don't
think
so.
There's
wait
wait!
Actually,
if
you
don't
mind
if
you
go
back
real
quick,
so
we
have
two
levels.
I
think
there's
a
bigger
issue
lurking
here.
We
have
two
levels
of
menus,
so
we
have
the.
We
have
a
nav,
a
site
navigator,
which
is
the
menu
that
you
were
just
showing,
and
then
we
have
a
docs
table
of
contents
or
a
docs
like
sidebar
menu
and
in
mobile
view.
A
A
Pretty
cool
idea,
I
can't
remember
who
worked
on
it.
It
is
a
cool
idea.
It
is
I'll.
Ask
this
everyone,
this
question,
there's
a
couple
questions
one:
does
the
chevron
get
in
your
field
of
view
for
reading
two
or
the
the
chevron
slide
the
button?
Two?
What
would
you
consider
like
when
you
before
deborah
prey
had
clicked
on
the
button?
What
would
you
have
expected
to
happen?
A
Is
it
obvious
to
you
three,
assuming
that
the
answer
to
both
of
those
is
what
I
think
it
is,
which
is
it
kind
of
gets
in
the
way
of
reading
some
stuff
and
you're?
Not
quite
sure
what
would
have
happened
like
well.
Imagine
if
you
take
the
chevron
the
circle,
you
shove
it
all
the
way
over
to
the
left.
So
it's
actually
like
half
a
circle
like
three
quarters
of
a
circle.
You
you
like
cut
off
a
quarter
of
it.
You
shove
it
all.
The
way
to
the
left
cut
off
a
quarter
quarter
of
it.
A
F
At
first
it
is
at
top
like
we
have
a
separate
button
there,
like
I
don't
know
who
might
we
are
regarding
this
change,
but
like
initially,
we
have
a
different
nav
bar,
which,
like
opens
in
a
half
way
on
this
stocks,
page.
F
G
B
F
Like
the
one
we
did
in
the
mystery
side
for
the
side
panel,
yes,
here
she's
working
on
something
similar.
H
Oh
yeah,
I
closed
it,
but
we
changed
the
icon
for
it.
We
can
make
this
the
previous
one,
like
the
previous
one.
We
had.
M
A
Cool
sounds
like
you
might
have
a
partner
in
crime
with
abijah
there
as
well.
If
one
of
you
is
reviewing
the
other
one's,
implementing
or
or
something
yeah
that'd
be
good,
it
makes
you
I'm
really
curious
deborah.
Do
you?
Do
you
mind
if
you
go
to
doc,
docs.mastery.io
like
what?
What
do
we,
what
we
have
going
on
in
mobile
view,
yeah?
What
do
we
so?
Okay?
So
we've
got?
Okay,
do
we
have
a
carousel?
A
A
A
C
So
we
are
trying
to
like
the
issue
yeah,
so
we're
trying
to
make
a
dashboard
for
the
service
mesh
performance
site
in
which
we,
you
know,
display
all
the
results
that
happen
in
the
back
end
of
the
performance
test.
Right
now
we
are
just
showing
like
tests
that
are,
that
does
not
show
the
actual
capabilities
of
the
service
mesh
performance,
so
we're
trying
to
you
know
display
a
more
comprehensive
data
on
the
website
that
for
that
so
today
I
I
asked
I
actually
about
what
are
the
you
know.
A
A
A
C
Yeah
so
there
that
lots
of
data
that
gives
and
start
percentage,
but
it
doesn't
doesn't
give
like
what
the
data
is
about
so
yeah.
I
guess
this
is
the
one
which
we
show
in
the
like
performance
test
like
p50,
p90
p99
max.
These
are
the
values
that
we
show.
I
guess
exactly
yeah
so
yeah
that
that
I
figured
out
this
std
dev
means
I
guess,
standard
aviation
yeah
but
yeah.
This
is
these
things
like
what
are
these
datas,
for?
I
could
not.
You
know,
figure
it
out.
A
So
are
there
60
of
those
or
are
there
count?
Okay,
those
are
the
histograms
themselves,
so
so
it's
yeah.
This
is
not
obvious
and
we
should,
but
on
line
64,
it
says,
duration,
histogram.
You
know
not
that
this
is
obvious,
but
those
are
the
points
on
the
graph
so
of
the
in
this
particular
test.
If
you
scroll
a
little
bit
further
to
the
top,
there
were,
I
think
there
were
60
queries
sent.
I
think.
C
A
Yeah,
if
you
keep
going
to
the
top,
so
runner
results
come
yeah,
so
I'm
pretty
sure
the
return
codes
were.
There
were
60
return.
There
were
60
messages,
responses
with
a
return
code
of
200,
so
all
of
them
were
successful,
so
yeah
there's
60
total
over
a
test
time
of
from
5000
to
5890
milliseconds.
So,
however,
many
seconds
that
totals
up
to
so
anyway,
the
the
array
that
you
were
just
looking
at
for
duration
histogram,
each
of
those
indexes
in
the
array
so
end
count
start
percent.
A
That's
that's
one
of
the
vertical
buckets
that
you're
that
you'll
see
in
the
graph,
where
what
it
says
is
between
the
time
0.04
and
point
zero.
Three,
eight
seven
one
five
between
this
slice
of
time.
There
was
one
request
that
went
that
fast,
like
that
was
the
fastest
request.
Only
one
of
them
came
back
and
out
of
the
sixty,
it's
one
1.6
repeating,
or
the
number
of
requests
that
land
into
that
bucket,
because
it's
not
a
precise
time.
It's
a
range
between
0.387
and
0.04.
A
It's
a
time
range.
How
many
requests
came
back?
Were
that
fast?
Only
one?
What
was
the
percentage
of
the
total
number
of
requests
that
were
sent
well?
Well,
it
was
1.6
of
them,
so
you
go
to
the
next
bucket
tonight
and
this
is
what
a
histogram
is.
It's
like
a
series,
so
what
gets
more
interesting?
Is
you
go
to
like
the
third
or
fourth
bucket?
Okay,
that
that
one,
that's
great,
it's
more
clear!
C
All
right
all
right,
yeah,
I
I
think
I
get
it
now.
Okay,
and
what
are
these
percentage
for,
like
the
success
rate.
A
A
That's
not
exactly
the
right
calculation
there's
something
to
do.
It
might
have
something
to
do
with
standard
deviation
as
well,
but
but
yeah.
Basically,
one
way
to
help
teach
yourself
this
and
begin
to
help
document.
This
spec
is
to
look
at
the
graph,
the
graphed
version
of
it
hover
and
see
the
same
exact
point
you
you'll
be
able
to
map
the
points
all
right.
F
A
Nice,
that
makes
sense,
so
it's
accumulated
it's
like
by
that
point
in
time,
like
we
start
from
the
very
top,
it's
like
from
the
very
top
like
hey
starting
at
the
fastest
responses
and
going
all
the
way
to
the
end
to
the
slowest
responses,
if
you're
looking
at
like
the
fourth
bucket
the
fourth
time
range,
the
what
aditya
is
saying
is
by
the
time
that
you've
arrived
at,
that
fourth
bucket
23,
that
23
fall
were
this
fast
or
faster
right.
A
C
C
All
right,
so,
okay,
so
one
more
thing
like
these
are
these
like:
why
have
you
set
like
50
75
1999
like?
Is
there
any
like
any
particular
reason
for
that.
A
I
think
it's
like
order
of
standard
yeah,
the
in
the
golden
okay
yeah
in
sre
land
for
those
that
are
into
devops
and
those
that
want
to
like
track.
A
The
quickest
way
to
so
that
so
you
can
measure
any
number
of
metrics
in
your
infrastructure
to
tell
whether
or
not
it's
being
performant
and
running
very
well,
there's
a
thousand
there's
like
tons
and
tons
of
which
ones
should
you
measure,
which
ones
are
actually
meaningful.
A
Well,
there's
four
golden
signals
that
people
talk
about
the
ones
that
you
should
measure,
so
some
are
like
some
of
it's
latency,
some
of
it's,
the
latency
that
you
have
on
your
services,
errors
that
you're
seeing
your
overall
request
rate
forget
the
other
one
anyway,
but
for
each
of
those
metrics
those
time
series
what
you
could
do
is
say:
well,
hey,
what's
our
average
latency?
What's
our
50
latency
well,
on
average,
it
takes
150
milliseconds
to
get
a
response.
A
It's
like
hey,
we're
doing
great,
like
that's
really
fast
for
our
website,
you
know,
or
for
whatever
service
it
is,
except
if
you
like,
accept
that
that
average
can
lie
and
hide
a
lot
of
important
info,
because
then,
if
you
look
at
the
well
what's
what's
our
99th
percentile?
What's
of
all
the
ones
that
we're
sending
that,
what's
the
one,
how
slow
are
the
one
percent
and
they
could
be
super
slow,
we
could
have
a
really
long
tale
of
like
well
yeah,
most
of
them
on
average,
when
you
average
them
all
out.
A
It
comes
to
150
milliseconds,
but
some
of
them
take
four
minutes.
It's
like
whoa,
so
so
anyway,
these
percentiles
are
kind
of
commonplace
convention
for
people
monitoring
infrastructure
to
say:
let's
take
a
look
at
our
at
these
percentiles
and
it
just
people
have
kind
of
landed
on
the
convention
that
these
are
the
ones.
You
typically
want
to
look
at,
okay.
A
So
for
the
response
time,
what
we're
doing
here
is
saying:
okay,
for
so
for
latency
of
the
60
that
were
sent.
Let's
take
how
fast
they
all
came
back.
Add
that
all
up
divide
you
know
do
the
anyway.
A
Here's
the
nice
thing,
which
is
some
of
this,
like
it
doesn't
matter
at
all
like
we
didn't
have
you
don't
have
to
understand
any
of
that
in
order
to
have
a
performance
dashboard,
because
all
you
need
to
do
is
like
all
we
want
to
do
is
basically
take
the
same
charts
and
put
them
onto
a
page,
and
then
let
people
switch
between
those
the
first
result.
The
second
result
in
the
third
and
kind
of
starting
there.
F
So
this
d,
would
this
data
be
real
time
or
like
we
would
just
be
using
this
json.
J
C
All
right,
okay,
okay,
then,
so
I
guess
the
graphs
will
be
more
like
same
that
we
have
now
in
the
measuring
ui
or
we
have
to
add
something
else
in
that
it's
a
good
question.
A
C
A
A
A
Measuring
the
performance
of
seo
under
different
scenarios
and
they're
using
actually
the
same
format,
the
same
json
format
that
we're
using
they're
using
they
on
this
site.
What
they're
saying
is:
there's
a
particular
type
of
test
they're.
Basically,
looking
at
this
statistic,
this
set
of
statistics
and
they've
run
it
a
bunch
of
times,
so
you
can
look
at.
A
You
can
go
back
in
history
and
look
at
the
data
from
a
prior
performance
test
and
they
graph
it
out
for
us
starting
out.
We
would
just
yeah
show
the
same
thing
that
we're
showing
in
measure
ui,
probably
I'm
here,
and
we
let
people
switch
between
independent
data
sets
independent
results.
Eventually,
what
we'll
need
to
figure
out?
A
Aditya
is
well
okay,
that
that's
fine,
but
like
not
all
te,
but
sometimes
people
are
trying
to
test
for
how
fast
their
proxy
is
or
for
how
much
memory
it
consumed
or
for
and
that
that's
where
we'll
start
to
break
down
between
those
different
types
of
tests
and
when
you
can,
where
you
can
kind
of
tell
and
we'll
need
to
standardize
these,
but
where
you
can
tell
the
difference,
is
in
the
test
results.
A
You
were
looking
at
the
json
it,
it
has
not
only
test
results
and
histograms
and
standard
deviations
and
stuff,
but
it
has
information
about
how
long
the
test
was
run.
What
type
of
surface
mesh
it
was.
What
the
end
point
was
like
we'll
need
to
start
to
use
some
of
that
data
to
build
out
a
navigation
system.
A
Okay,
so
I
know
we're
spending
okay,
any
thoughts
or
other
so
aditya,
if
you
just,
I
think
for
now,
if
you
just
use
a
static
set
of
json
data
as
your
test
data,
you
crack
open
a
pr
on
this
site
and
you
have
a
relatively
simple
looking
page
that
re
reuses,
one
of
the
existing
layouts
to
let
people
nav
between
result
sets
and
then
ultimately
start
to
nav
between
result
sets
for
one
mesh
versus
the
next.
You
you'll
be
you'll,
have
you'll
be
moving
quite
a
you
know,
you'll
be
doing
good.
A
Now
is
a
good
time
to
signal
to
su
betty,
so
yash's
yeah.
I
also
can
dpr
nice
good,
yeah
they'll
be
so
end
up
being
a
big
deal.
There
were
a
number
of
people
that
reached
out
about
this
last
week
from
some
companies
that
you
all
know
so,
okay,
how
much
okay
so
we're
just
over
time
leonard.
We
should
circle
back
to
you
briefly.
A
So
I
think
you
think
you've
more
or
less
answered
my
question,
which
is
hey.
You've
got
you've
identified
a
particular
plug-in.
The
plug-in
part
of
the
plug-in's
responsibility
is
to
implement
the
fall
black,
the
fallback
strategy
or
implement
so
part
of
what
it
does
is.
It
looks
at
the
pictures
that
are
there.
It
creates
alternative,
more
modern
image
formats
for
the
same
image.
It
leaves,
I
suspect,
the
original
in
place
as
a
fallback
plan.
A
Yes,
and
so,
for
those
that
are
contributing,
there's
no
change
in
the
contribution
flow
like
people,
just
the
plugin
does
its
thing.
L
L
L
A
Okay,
nice,
I
mean
that
when
we
say
it
like
that,
that's
relatively
concise
and
straightforward:
it's
like
hey,
there's
a
new
plug,
there's
a
plug-in
that
will
do
this.
For
us.
Here's
how
it
works.
L
If
so,
that,
if
you
go
to
the
website
channel,
you
see
the
link
and
you
see
what
just
the
formats
you
just
you
you
talk
about
like
you
write
the
full
back,
you
want
the
the
image
the
way
you
want
the
image
to
be
rendered.
So
if
it
doesn't
get
to
render,
I
say
webp,
you
should
go,
it
should
try,
that's
a
jpeg
and
if
it
doesn't
render
as
a
jpeg,
then
it
should
automatically
detects
what
he
wants.
So
this
is
this
is
the
format
the
defaults,
the
particular
defaults.
L
A
This
sounds
like
a
nice
behavior.
A
Devil
priya,
you
might
be
close
to
this.
Like
so
wait,
you
can
customize
it
again.
B
A
L
L
A
Yeah,
I
think,
from
what
I've
seen
from
my
part,
it's
good
to
go
like
it's
good
to
go,
we'll
we'll
we'll
want
to
have
nikhil
lada
check
it
out.
Just
because,
like
demoprio
was
just
saying
that
he
had
made
some
changes
here
recently
and
so
he's
like
he's
in
the
zone.
He's
he's
the
right
one
to
check
this
out.
Let's
make
sure
he's:
okay,
yeah
he's
amazing.
A
Okay,
so
nikhil
is
the
is
the
guy
that
we
want
to
get
a
thumbs
up
from
just
because
he's
been
more
into
this
area
than
I
have
recently.
A
Good,
so
I
just
poked
him
on
slack
to
make
sure
and
then,
in
terms
of
like
other
things
forthcoming,
there
were
so
we've
got
this.
This
initiative
around
surface
mesh
performance,
josh
suberia
on
that
we've,
oh
we're
going
to
talk
about
the
surface
mesh
playground
soon
and.
A
F
A
L
A
Oh
man,
all
right,
very
good.
I
think
that's
a
wrap.