►
From YouTube: Octant Community Meeting - January 13th, 2021
Description
Octant community meeting is held weekly. We discuss and talk about the current state and future of Octant, demo upcoming features and releases, and preview new ideas we are considering for Octant.
Meeting agenda: https://hackmd.io/CzaPxtmXT_SW8nEpdwvGzw?view
A
A
A
Yeah,
it
feels
like
it
right
so
yeah
welcome
everyone.
We
are
we.
The
main
thing
is
just.
I
want
to
give
some
updates
on
the
electron
work
that
we've
been
doing
and
and
say
thank
you
to
sam
along
the
way
and
and
we've
been
slowly
moving
this
forward
and
it's
been
a
bit
of
a
process,
but
we
are
it's
getting
close
now.
A
So
one
of
the
things
that
that's
on
here
right
now
is
this
1822
issue,
which
was
merged
in
just
the
other
day
that
came
in
through
a
pr
from
fabrice.
Thank
you.
A
We're
able
to
get
that
merged
in
which
means
that
when
we
do
get
to
having
our
electron
build,
it
will
work
on
m1.
So
well,
allegedly,
I
don't
think
we
actually
have
a
thing
to
test
it
with.
So
I'm
assuming
that
the
breeze
after
this
merged
tested
it
and
it
worked
for
them.
So
hopefully
that
means
it
does
work.
Maybe
maybe
that's
a
that's
a
good
excuse
to
get
the
company
to
send
me
an
m1
laptop
is
the
requirement
for
testing
octan.
A
The
next
thing
up.
There
was
the
auto
updates
yeah.
So
there
was
a
conversation
that
happened
around
this
that
sam
brought
up.
I
believe
it
was
in
our
stand
up
when
we
were
talking
about
what
was
left
to
do
for
electron,
and
I
think
we
just
want
to
call
out
here
that
the
the
first
pass
of
auto
update
is
going
to
be.
A
The
system
is
going
to
download
the
new
binary
and
then
it's
going
to
restart
itself
right,
like,
like
you,
see
with
a
lot
of
other
applications
where
you
won't
just
get
the
banner
that
flashes
and
says:
hey.
Congratulations:
we've
updated
you're
on
a
new
version,
you'll
be
a
prompt
to
say
click
here,
it'll
download
you
restart
and
you'll
be
on
the
latest
version
that
that's
that's
going
to
be
our
our
first
pass
with
auto
update
and
then,
after
that,
we
will
continue
to
explore
more
what
it
means
to
do
partial
updates.
A
Just
the
way
we've
decided
to
package
the
things
together
for
the
first
release
of
the
electron
application
and
also
just
the
time
investments.
We
don't
really
have
the
time
to
fully
explore
like
a
very
a
nicely
crafted
and
well
thought
out
partial
update
functionality,
so
we're
gonna
kind
of
they'll
be
auto
updating,
but
it
will
be
download
the
latest
and
restart,
which
is
still
a
big
step
in
the
right
direction
and
will
be
useful
to
folks.
So
was
there
anything
that
I
you
wanted
to?
C
Yeah
I
covered
it
nicely.
The
default
setting
for
checking
for
auto
updates
is
once
every
20
minutes
and
that's
kind
of
standard
in
the
electron
space.
C
C
So
yes
and
the
partial
updates
they
use.
This
thing
called
a
block
map
file
and
we're
just
excluding
that
from
our
belts
all
together
and
it's
super
nice
too.
It
works
out
on
our
end
because
it
reduces
build
time
significantly
for
whatever
reason-
and
I
guess
on
the
next
line-
interoperability
between
multiple
release
tools
to
a
single
tag.
C
So
essentially,
we
need
to
create
our
own
bucket
and
then
upload
that
all
together
at
once
as
a
single
final
tag
with
our
own
machinery
and
that's
just
going
to
take
a
little
bit
of
work
on
our
part,
it'd
be
kind
of
cool.
If
there
was
some
tooling
already
in
this
space
like
some
sort
of
universal
artifact
publisher,
but
maybe
that's
not
a
real
thing,
it
could
be
fun
for
a
side
project
in
case
anyone
listens
to
this
and
feels
like
you
know.
Yes,
that's
what
open
source
needs
right
now.
A
Yeah
and
just
to
highlight
what
sam's
talking
about
so
sam
did
some
work
that
does
push
up
the
latest
electron
build
as
well
as
the
the
server
binary
you
can
see
here.
It
gets.
It
currently
gets
pushed
up
as
like
two
separate
labels
tags,
whatever
it
releases
releases,
it's
the
same
tag
but
two
well
a
different
tag.
Actually
one's
v
v,
seventeen
one
zero
seventeen.
So
it's
pushed
up.
A
As
you
said,
what
we're
doing
is
working
to
merge
those
together
so
that
it's
all
the
way
people
expect
it
to
be
and
and
all
rolled
up
into
into
one
spot,
and
these
are
the
block
map
files
that
sam
was
referencing.
That'll
go
away.
B
C
Yeah
I'd
assume
it
checks
on
startup.
I
it's
in
the
electron
docs
you
know
like
this
should
be
like
it's
abstracted
away
like
we
don't
even
like
it's
just
like
on
a
constant
update
or
whatever.
A
Yeah
and
then
there's
there's
a
project
board,
which
I
pulled
up
briefly
there
for
a
second,
but
I
can
show
it
again:
it's
our
o17
and
we're
tracking
these
things.
It
looks
very
similar
to
the
last
time
we
showed
it.
These
two
issues,
1479
and
1480-
are
the
ones
that
were
kind
of
sam,
and
I
have
been
circling
around
making
progress
on
so
once
those
are
in
place.
A
That
will
mean
that
the
ability
to
easily
build
and
run
an
electron
application
from
our
build.go
file
will
be
in
place,
as
well
as
the
all
of
the
actions
to
support
releasing
the
electron
binary
via
our
release
pipeline
and
publishing
that
up
into
our
nightly
binary,
so
people
will
be
able
to
start
grabbing
the
nightly
of
the
electron,
build,
that's
kind
of
like
foundation.
A
Work
to
then
allow
the
rest
of
these
to
we've
really
been
just
hyper,
focused
on
getting
that
stuff
done
getting
this
pipeline
and
tooling
work
done
so
that
we
can
then
address
and
burn
down
the
rest
of
the
issues
for
the
o17
sprint.
We've
decided
is
kind
of
a
requirement
for
for
electron
and
we
do
have
a
separate
project
which
is
just
electron
application.
A
I
think
there's
like
three
or
more
three
or
four
stories
in
here
that
are
also
in
o17,
but
that
that
will
likely
come
later,
but
they
are,
they
are
related
to
things
we
want
to
do
once
electrons
in
place,
so
expect
expect
to
see
this
into
the
the
electron
build
in
the
nightlys
very
soon
and
once
that's
there,
then
that's
going
to
be
the
primary
mode
of
operating
octant
for
the
dev
team,
which
will
then
allow
us
to
really
start
to
address
all
of
the
outstanding
work,
so
that
kind
of
updates
where
electron
is
at
and
where
the
project's
at,
and
why
is
why
it
feels
like
not
much
is
going
on
is
because.
A
In
the
forward-facing
part
of
things,
there
isn't
much
going
on
it's
all
back-end
work
and
and
pipeline
and
infrastructure
stuff.
So
there
is,
however,
work
happening
through
the
design
work
that
is
going
on
with
resource
viewer.
We
showed
and
shared
some
of
that
a
couple
meetings
ago
and
I
just
wanted
to.
A
I
don't
know
if
we
have
a
discussion
around
it,
but
I
wanted
to
call
out
that
milan
and
isha
have
been
working
extensively
on
kind
of
redesigning
and
retooling
and
thinking
about
retooling,
redesigning
and
re-uh,
structuring
kind
of
like
how
the
resource
viewer
there's
like
two
pieces,
there's
like
the
look
and
feel
of
resource
viewer,
but
then
there's
the
the
like
the
actual
implementation
and
like
what
is
important,
how
things
are
drawn?
A
What
like
so
there's
two
scopes
of
work,
kind
of
happening
and
one
is
a
much
like
deeper,
heavily
invested
scope
of
work
and
the
other
ones
are
more
of
like
hey,
let's
clean
up
some
things
we
know
about,
so
I
don't
know
if,
if
isha
or
milan
want
to
add
more
to
that,
but
I
do
know,
there's
some
issues
being
created
and
some
some
work
happening
around
that.
B
You
I
I
think
you
know
we
shared
the
current
design
and
the
progress
previously,
so
we're
just
currently
moving
forward
with
that
and
trying
to
finalize
design
and
trying
to
start
working
on
some
prototypes
soon
inside
the
storybooks.
So
that's
pretty
exciting.
I'm
I'm
happy
that
we
get
a
really
good
second
take
on
this.
B
That
will
hopefully
improve
the
user
experience
and
make
a
resource
viewer
more,
not
only
like
a
display
element
but
more
navigation
element
and
show
little
richer
information
for
the
end
user.
D
Yeah,
I
think
what
we're
hoping
to
do
is
like
build
it
in
code
in
storybook
and
then
maybe
get
some
like
better
feedback.
If
your,
if
users
are
able
to
actually
like
use
the
resource,
viewer
and
click
on
it
and
see
how
it
works,
and
then
we
can
test
it
against
different
data
sets
and
see
and
see
how
it
comes
out.
So
we're
in
the
process
of
that.
D
I
think
right
now
we're
still
thinking
about
the
design
that
we
want
to
build
in
the
prototype,
but
hopefully
next
week
we
can
once
we
like
solidify
on
that.
We
can
share
that
out
in
the
community
meeting.
A
E
I
wanted
to
add
a
few
things,
so
I
just
wanted
to
share
that
we're
also
working
on
improving
the
plug-in
author
experience
and
we
have
a
good
set
of
plug-in
authors
to
interview.
But
if
there's
anyone
else
in
the
community,
who'd
like
to
be
interviewed,
please
reach
out
to
me
or
isha
or
any
of
us
in
the
octan
channel.
E
E
So
I'm
excited
to
share
that
and
I'm
going
to
continue
going
to
other
sessions
to
learn
more
from
many
of
the
open
source
teams
within
vmware
such
as
valero
and
robin
and
pew
and
many
others.
D
A
A
All
right,
so
moving
on
to
open
q
a
I
just
added
a
couple
items
that
I
wanted
to
discuss
quickly:
just
kind
of
get
people's,
at
least
like
start
to.
Oh,
actually,
I
see
there's
a
question
from
the
youtube
stream
I'll,
add
it
to
the
list
and
we'll
answer
that
one
first,
so
reginald
reginald
had
a
question,
and
it
is
says
that
the
great
tech
it
chooses
a
lot
of
cpu
when
idle
is
that
being
addressed.
A
So
I
would
say
no
because
I
don't
know
if
that's
a
common
problem
so
reginald.
I
would
encourage
you
to
open
an
issue
with
like
your
platform
and
any
log
output
that
you
have.
That
can
be
really
helpful.
I
do
know
that,
in
the
case
when
octant
is,
is
in
a
rbac,
restricted
environment
or
is
failing
to
look
up
certain
resources
within
the
cluster.
A
What
can
happen
is
that
it
causes
it
to
log
very
verbosely
to
the
console,
which
can
then
result
in
high
cpu
output,
because
the
polar
that's
running
with
octane
is
running
every
second,
and
so
then
the
result
is
that
you
get
either
one
or
mult
like
one
plus
n
log
lines
for
certain
errors
into
the
console
every
second,
and
that
can
result
in
a
high
cpu
when,
when
you're
just
kind
of
not
doing
anything,
so
I
would
I
would
encourage
you
to
open
an
issue
so
that
we
can
take
a
further
look
into
that
and
see
and
see
specifically,
why
you're
running
into.
A
C
Yeah-
and
this
is
just
tricky
in
general-
simply
because
we
don't
really
have
any
type
of
benchmarking
or
even
type
of
like
any
type
of
automation,
that
kind
of
checks
for
this.
So
really
it's
a
lot
of
times
it's
when
we
run
it
and
just
based
off
of
intuition
and
our
devices-
and
I
know
that
sounds
awful
from
a
software
development
standpoint,
but
it's
something
we
can
improve
upon
and
it's
and
there
are
stuff
we
can
also
check
like
at
compile
time
like
to
check
for
go
routine
leaks.
C
I
think
that
was
discussed
a
while
back
ago,
but
if
you
have
a
specific
scenario
of
using
octet
that
causes
a
high
cpu
usage,
we
are
absolutely
interested
in
knowing
what
causes
that.
A
Yeah-
and
I
would
I
would
add,
to
that-
I
would
say
like
if
it
if
it
starts
up
fine
seems
to
be
fine
while
you're
using
it
and
then
it's
idle
and
then
it
spikes,
like
that's
all
like
just
like
how
it
manifests
itself
when
it
seems
to
happen.
All
of
that
is
useful
information
and
helping
us
kind
of
determine
what
might
be
the
root
cause
since
that,
like
like
sam,
was
saying
it's
kind
of
one
of
those
we
don't
know
until
we
know
so
yeah.
A
A
All
right,
moving
on
to
my
question
so
listing
resources
from
a
cluster.
It's
a
that
is
a
very
generic
wide-open
question.
Specifically
this.
This
comes
up.
A
There's
there's
a
couple
things
that
we
are
we're
doing
when
we
pull
and
list
resources
from
a
cluster,
and
one
of
them
is
we
do
a
has
asked,
has
access
check
which
I've
talked
about
before,
and
I
have
an
old
branch
that
is
probably
well
past
its
its
its
ability
to
to
serve
as
anything
useful
but
there's
a
subject:
access
branch
which
kind
of
removes
this,
has
access
and
uses
subject
access
to
list
resources
and
it
is
much
much
faster.
A
A
This
is
work
that
is
happened
independently
of
octan
in
other
projects
and
and
side
efforts,
and
so
that's
something
we
want.
I
I
just
wanted
to
bring
up
and
then
the
other
thing
is
when
we
are
in
our
back
restricted
environments,
we
we.
A
We
have
some
difficult
times
like
listing
resources
in
a
way
that
is
consistent
and
predictable
and
like
the
failure
cases
of
like
when
you
can
and
cannot
see
things
even
with
our
hazard
access
checks,
are,
are
interesting
and
and
aren't
reliable
in
a
way
that
we
can
say
like.
Yes,
this
will
run
in
your
environment
or
no,
it
won't,
or
it
should
degrade
nicely
anyway.
The
so
the
reason
I'm
bringing
all
this
up
is.
A
I
I
I'm
going
to
open
a
discussion
on
our
github
discussions
and
specifically
around
like
kind
of
revisiting
and
rethinking
how
we
do
listing
of
resources
from
a
cluster
like
like
what
we've
done
now
is
just
like:
hey
use,
client
go
discovery,
client,
dynamic,
client,
cluster
client,
go
get
stuff,
bring
it
in
right
and
that
and
that
does
work
and
and
we'll
still
need
to
use
those
apis
at
a
low
level.
But
I
think,
there's
probably
a
better
abstraction
we
can
create
around
once.
A
We
get
that
like
first
off
failing
gracefully
when
we
make
an
api
call
and
it
doesn't
allow
us
to
like
we
get
access
to
night
or
something
like
that.
We
we
ideally
like
instead
of
this
pre-check,
where
we
ask
for
permission,
we
would
have
a
really
graceful
way
to
to
get
forgiveness
and
then
the
other
piece
to
this
is
just
because
the
kubernetes
api
gives
us
these
things
in
a
certain
way.
It
doesn't
mean
that
that
is
the
way
that
we
must
hold
them.
A
Our
object
store.
Our
dynamic
cache
is
already
kind
of
doing
this,
where
we
we
kind
of
store
them
in
a
format
that
we
want,
so
that
we
can
they're
they're,
they're
cached
and
we
can
go
in
and
and
fetch
them
easily
and
pull
them
out
of
the
store
and
and
maybe
like,
and-
and
it's
very
right
now-
that
abstraction
is
kind
of
mapped
one-to-one
to
like
a
gvk
or
or
yeah
to
gvk's.
I
think,
and
we
may
benefit
from
creating
a
data
model
internally,
like
in
memory
where
we
store
this.
A
That
is
more
like
the
model
that
we
generate
for
the
resource
viewer
right.
Like
imagine,
we
went
and
we
collect
all
this
information
about
the
cluster,
and
then
we
put
it
into
a
graph
by
default
right
and
then
now
we
are
exploring
the
graph
and
we're
doing
like
graphql.
A
You
know
to
to
get
objects
and
data
out
of
it
and
then
there's
still
an
informer
sitting
with
a
cache,
that's
kind
of
updating
this
stuff,
but
then
drawing
out
relationships
and
adding
in
extra
context
on
those
relationships
and
doing
all
this
they
become
like
natural
to
the
system.
Instead
of
right
now,
it's
like
a
very
foreign
to
the
system
thing
where
we're
like.
We
take
this
data,
we
pass
it
through
the
resource
viewer,
it
generates
a
graph.
Then
we
have
this
graph.
A
Now
we
go
back
to
kubernetes
to
try
and
add
context
to
the
graph
right.
Well,
what?
If
the
the
root
of
the
data
star
was
already,
the
graph
is
something
I'm
thinking
about
and
and
for
just
kind
of
long
term,
from
like
the
direction
we
want
to
take
resource
viewer
and
like
some
of
the
screens,
we've
shown,
I
think
we'll
we
have
to
kind
of
potentially
think
about
a
different
way
to
store
our
data,
maybe
maybe
not.
B
Yeah,
I
I'd
love
to
see
that
you
know
maybe
doesn't
have
to
be
graph.
You
know,
maybe
it
could
be
just
a
different
data
structures
used
to
represent
the
the
data,
but
you
know
it
would
be
really
good
to
to
kind
of
move
that
one
level
away
from
the
kubernetes
view
things
and
maybe
make
it
a
little
easier,
accessible
and
a
little
easier
to
to
gather
the
information
for
the
resource
part
of
the
parts
of
review
system.
I
know
when
I
was
there
trying
to
add
something
for
the
resource
viewer.
I
mean
it.
A
Yeah
yeah,
so
I
was
just
throwing
it
out
there
as
something
to
think
about
and
and
something
I've
been
thinking
about
and
like.
I
think
I
think,
there's
like
a
there's
a
list
of
problems
that
we
run
up
against
with
our
current
approach.
That
need
to
be
solved,
but
then
there's
also
just
like
how
do
we?
How
do
we
like?
Can
we
do
it
better
right?
Can
we
just
do
it
better,
knowing
what
we
know,
knowing
what
how
we
want
to
display
the
data?
Can
we
do
it
better?
A
B
So
I
just
want
to
share
you
know
the
my
view
of
the
resource
here
is.
I
would,
I
would
definitely
like
to
see
to
be
a
map
of
your
kubernetes
cluster
and
that
map
includes
everything.
So,
if
you
think
about
the
map,
like
traditional
map,
you
can
have
different
levels
of
zoom,
you
can
have
filters
you
can
have
so
as
long
as
we
have
that
map,
it
would
be
pretty
easy
for
us
to
say
okay,
so
I
want
to
focus
on
this
part
of
the
map.
B
I
just
want
to
see
this
deployment
and
maybe
filter
it
by
some
labels
and
things
like
that,
but
at
the
end
of
the
day,
it's
it's
still
just
a
different
view
into
the
same
map.
So
if
we
can
find
a
good
way
to
represent
that
map
to
the
data
structures
on
the
backend
side,
I
think
that
would
be
awesome.
A
Reginald
asked
another
another
question
that
I
wanted
to
answer:
has
anyone
run
octant
as
a
container
and
exposed
it
as
a
service?
Yes,
people
have
done
that
it
isn't
supported.
Octant
is,
is
the
only
official
mechanism.
Procten
is
running
locally
as
an
application,
and
the
reason
this
is
the
case
is
is
due
to
well.
One
of
the
big
reasons
is
is
due
to
the
global
context
and
the
way
we
load
config
files
and
and
the
assumptions
that
we
make
of
where
octant
is
running.
We
can
do
it.
A
We
we
get
a
lot
of
benefit
from
being
able
to
assume
that
octant
is
running
on
a
local
machine
and
not
in
a
deployed
environment
around
like
reliability
and
and
like
what
we
can
do
with
it.
So
that
is
why
we
haven't
really
explored
that
avenue
of
like
making
it
something
that,
like
giving
first
class
support
to
it
running
as
a
service
people
have
done
it.
A
People
have
used
config
maps
to
expose
the
settings
and
and
and
get
it
working,
but
and-
and
that
is
an
issue
that
we
have
open
in
our
backlog
for
potentially
like
long
term
or
down
the
road,
but
in
the
short
term
it's
it's.
We
just
don't
have
the
the
team
size
to
really
do
both.
Well,
we
just
kind
of
have
the
team
to
do
one
well,
so
that's
what
we're
focused
on
is
the
application
side.
A
And
so
the
other
question
that
I
put
in
here
was
the
web
components
and
custom
components.
A
This
has
come
up.
Some
folks
are
writing
a
plug-in
and
we're
asking
about
this,
and
it's
come
up
before
it's
in
our
backlog.
It's
it's
a
thing
that
that
we've
been
asked
about
before,
and
I've
talked
about
before,
and
I
I
like
the
idea,
but
I
am,
I
am
actually
I'm
I'm
struggling
mostly
to
to
understand
exactly
what
like
the.
Why
and
the
benefit
and
the
how
like
less
the,
how
more
the
the
why
and
the
benefit
of
like
what?
What
do
people
who
want
this?
A
Why
do
they
want
it,
and
maybe
this
is
something
that
that
pia
can
find
out
during
the
plug-in
author
interviews
of
like?
Is
that
something
that
would
interest
you?
How
would
you?
How
do
you
see
yourself
using
a
custom
component?
A
What
is
the
like,
why
a
custom
component
instead
of
you
know
making
modifications
or
or
creating
a
new
upstream
component
like
and
one
of
the
things
that
I
think
I
think
all
of
us
have
said
before
is
specifically
around
like
charts
and
graphs.
A
B
Yeah
I
mean
you
know:
user
interface
is
pretty
complex
and
the
the
amount
of
use
cases
is
pretty
much
infinite.
So
I
I
kind
of
understand
why
people
want
to
put
different
things
in
and
I
I
I
feel
that
at
some
point
we
should
provide
a
little
better
support
for
the
web
components,
but
you
know
I
mean
even
now.
I
think,
if
you,
if
you
are
working
with
a
typescript
plug-in,
you
should
be
able
to
add
the
as
long
as
you
you
know,
gather
all
the
data
on
all
all
the
information
you
need.
B
You
should
be
able
to
add
any
component.
You
want
to
the
to
you
to
your
plugin.
If
that's
what
you're
talking
about,
then
I
assume
we
are
so
you
know,
but
at
some
point
maybe
we
should
make
like
a
custom
components
like
a
first
class
citizen
and
and
providing
a
little
better
interface.
A
Yeah
yeah,
I
don't
know
either
I
I
in
my
head,
I
think
of
it
as
being
something
like
there's
some
type
of
of
web
view
component
or
something
like
that
and
it
and
it
you
know,
creates
a
shadow
dom
that
and
then
anything
that's
defined
in
there
gets
loaded
into
it
and
renders
with
its
own,
you
know,
so
you
can
have
script
and
head
tags
and
things
that
go
out
and
fetch
in
these
css
and
javascript
resources
and
just
load
it.
And
it's
like
a
read-only,
isolated
shell
from
from
octant.
A
That's,
I
guess
to
me
that
would
be
like
the
step.
One
would
be
like
to
create
it
as
like
this
read-only
view
and
then
from
there
figure
out,
like
what
do
interactions
like
reactive
patterns.
Look
like
in
that
world.
A
So
I'll
open
a
discussion
around
this,
I
think
I
think
it
would
be.
It
would
be
useful
to
start
collecting
that
information
and
we
can
take
like
the
stuff
that
pia
finds
when
talking
to
plug-in
authors
and
add
it
to
the
discussion
as
well
and
just
have
have
a
good
place
to
maybe
so
when
people
are
searching
for
this
or
it
comes
up,
it's
a
place
to
point
people
to
and
they
can
add
their
opinions
and
thoughts.
D
Yeah,
well
we'll
definitely
ask
about
that
in
our
interviews,
but
I'm
also
wondering
how
much
of
it
is
like
an
awareness
issue
like
milan
was
saying
like
this
can
be
done
in
typescript.
But
how
familiar
are
people
with
like
being
able
to
create
these
components
in
typescript
or
knowing
that,
like?
Oh,
if
I
choose
to
do
a
typescript
plugin
over
a
go,
plugin
I'll,
have
this
flexibility
or
capability?
A
I
think
he
was
saying
that,
ideally,
you
would
be
able
to
because
it's
like
that
same
declarative,
syntax,
that
people
are
used
to,
but
right
now
there
is
no
way
to
put
your
own
custom
stuff
up
there,
except
for
there's
the
there's
the
take
a
text
component
flag.
It
is
marked
down
flag
it
as
secure
content,
and
then
you
can
basically
free
render
html.
A
C
But
if
we
had
web
components,
we
could
implement
or,
like
I'm
wondering,
if
we're
missing
a
certain
class
of
components,
as
opposed
to
just
actually
needing
to
build
out
the
system
of
web
components
and
custom
components
really
well,
and
that
is
something
we
probably
want
to
clarify,
because
it's
much
harder
to
get
web
components
right
and
it's
a
little.
It's
a
lot
less
work
to
implement
a
new
component.
A
Yeah,
I
think
that's
a
good
point
tim.
I
and
I
agree
with
you.
It
is
yeah
doing
doing
a
new
component.
I
I
think
I
think
that
that
would
be
like
we
as
a
team.
We
want
to
research
and
decide
what
graph
library
we
want
to
use
and
what
chart
library
we
want
to
use
right
like,
and
I
don't
know
if
there's
already
one
that's
kind
of
like
the
de
facto
standard
for
open
source
projects
at
vmware
or
not.
A
B
No,
no
there's
no
standard
in
myanmar,
I
mean
the
the
one
we
built
the
one
I
would,
I
think
most
of
you
know
that
I've
been
before
working
knocked
and
I
was
working
on
actually
data,
visualization
library
inside
the
portal
and
that
that
was
used
in
both
healthwatch
and
up
metrics
that
were
beautiful
products,
but
unfortunately,
at
some
point
the
the
observability
team
decided
to
move
towards
using
grafana
internally,
so
that
that
there's
still
open
source
library.
That's
actually
really
good.
B
Fortunately,
in
the
react,
so
I
mean,
maybe
you
can
try
to
find
a
way
to
use
it
and
kind
of
open
source
it
a
little
more
than
it
is
right
now,
because
it's
it's
a
publicly
available,
but
not
completely
open
source,
it's
kind
of
totally
different
issue,
but
I
there
are
some
other
solutions
too,
and
there's
a
lot
of
good
libraries
out
there,
most
of
them
d3
based
what
we
did
in
our
case
is
we
we
didn't
want
to
invent
the
hot
water.
B
We
basically
just
decided
we're
gonna,
find
a
really
good
library
out
there
and
build
on
top
of
it
and
mostly
like
features
that
are
important
for
from
observability
downsampling.
You
know
support
for
trump
ql
and
different
standards
and
making
specific
charts
that
are
interesting
for
monitoring
and
observability.
B
So
not
only
do
we
provide
a
good
charge,
but
also
some
mechanism
and
and
like
recommendations
which
you
know
for,
for
example,
for
latency.
You
don't
want
to
use
bar
chart
for
waste
and
see
right
there
different
ways
of
doing
that.
So
so
it's
a
it's.
B
It's
a
very
you
know
complex
space
and
there's
a
lot
of
variables
there,
but
if
you
feel
that
we
should
develop
on
chart
I'll
definitely
like
to
see
that
happening,
because
I
think
it
would
improve
a
lot
user
experience
in
on,
especially
like
you
know,
small
things
like
a
spark
charts,
you
might
be
interested
when,
when
you
look
at
the
the
parts
and
how
they
behave,
the
cpu
you
can,
you
can
even
display,
like
spark
chart
inside
there.
B
Maybe
it's
even
inside
the
resource
viewer
showed
a
little
chart
showing
you
know
how
the
cpu
was
behaving.
The
reason
things
like
that.
A
A
I
wanted
to
pull
this
up
just
because,
like
I
do
agree
like
I
think,
there's
a
lot
of
things
in
here
that,
like
could
be
made
part
of
that,
could
make
the
octane
ui
better
right,
the
like
the
overall
or
the
ux,
the
overall
user
experience
better
by
having
them
as
components
that
are
part
of
the
user
interface,
and
you
know
the
like
just
like
you're
like
these,
like
sparklines
right,
like
things
that
you
were
saying
exactly:
yeah
like
heat
maps
and
and
and
polar
charts
and
things
like
you
can
you
can
give
a
lot
of
information
in
a
small
space
with
with
some
of
these
charts
more
than
just
like
the
verbosity
of
a
table
and
text
so
yeah.
A
I
think
that's
something
we
want
to
explore
exactly,
maybe
maybe
along
the
line.
I
I
I
do.
I
think
I
like
what
sam
was
saying
about
just
like
you
know,
extending
our
components
so
that
we
have
more
charts
right
and
just
like
just
like
finding
areas
like.
Maybe
we
can
work
with
the
issue
to
find
areas
in
the
you
in
the
current,
like
user
interface,
where,
like
hey,
if
this
was
this
kind
of
chart,
this
is
what
it
might
look
like
and
then
use
that
to
drive
like
okay.
A
What
charts
are
we
going
to
implement
because,
like
even
even
just
the
donut
chart,
which
we've
done
folks
already
use
it
right,
they
already
will
use
it.
They
already
want
to
extend
it.
They
already
want
to
make
it
better,
so
I
think
just
getting
more
in
there
we
would
see.
I
know
that,
like
the
the
jenkins
x,
plug-in,
would
probably
take
advantage
of
a
lot
of
different
charts
if
we
put
them
in
place,
you
know
showing,
builds
and
and
like
stacked,
stacked
information,
and
things
like
that,
so
I
am
yeah.
A
I
like
that
idea,
something
to
think
about.
After
once,
we
start
working
on
stuff
after
an
electron.
B
Yeah
and
then
in
the
same
time,
I
don't
yeah.
In
the
same
time,
I
don't
think
we
should
go
too
deep,
like
you
know,
moving
to
like
prometheus
grafana
territory
or
bayfront,
or
things
like
that.
But
you
know
maybe
we
can
provide
interfaces
for
those
packages,
but
I
think
it
would
be
good
to
provide
just
out
of
the
box
octant
experience
where
you
can
see.
For
example,
you
know
when
you
open
the
octane
in
money.
A
So
yeah,
that's
all
that's
all
I
had
let
me
check
and
make
sure
there's.
No
other
questions
did
anyone
else
have
anything
to
add
before
we
call
it.