►
From YouTube: Octant Community Meeting - March 18, 2020
Description
March 18, 2020
Agenda
Last PR in review for 0.11
Terminal PR
Upcoming prorities for Octant
Electron/Application
Editing
Plugin API v1
Future work
In-Cluster
Better networking support
Demo
New Terminal flow
Memory improvements
Open Q&A - add your questions here
Sam - Do we want to move the Namespace picker?
We’ll leave it here.
A
All
right
so
welcome
everyone
to
the
March
18th
octant
community
meeting.
We've
got
just
a
couple:
we're
gonna
talk
about
what's
coming
up
in
2
at
11,
which
we
talked
about
previously,
but
we're
there's
only
one
PR,
that's
left
out
standing
before
we're
gonna
look
at
starting
to
do
that.
Release
we're
going
to
talk
about
our
the
kind
of
next
set
of
upcoming
priorities
for
octant.
A
Well,
what
the
focus
will
be
then
we'll
be
touching
on
kind
of
the
future
work
so
stuff,
that's
more
and
like
the
six
months
from
now
window
and
then
I
think,
depending
on,
if
they're
up
for
it,
Milan
and
Sam
might
have
some
demos
of
some
some
things.
So
I
I'll
just
go
ahead
and
get
started
now.
So
the
the
last
PR
that's
currently
up
for
review
is
the
terminal
component
and
it
is
almost
done
we're
just
doing
some
revisions
on
some
of
the
feedback.
A
So
there's
a
cup,
probably
just
a
couple
more
passes
on
on
some
reviews,
for
it
and
and
then
we'll
be
ready
to
release
ODOT
11,
which
has
a
lot
you'll
see
in
the
in
the
release,
notes,
there's
a
lot
a
lot
of
good
stuff
in
there
around
performance
and
just
some
user
enhancements,
as
well
as
just
some
other,
like
checklist
items
for
user
experience
that
we've
been
able
to
get
done
with
so
that
that
should
be
coming.
You
know
hopefully
soon
well
well,
yeah
I,
don't.
A
Soon,
as
for
the
after,
we
released
all
about
eleven
kind
of
the
upcoming
priorities
for
octamer
are
going
to
be
getting
the
feature
parity
for
the
electron
application,
starting
to
look
at
editing,
specifically
just
we
have
the
Monaco
viewer
in
place
for
our
demo
right
now.
So
I
want
to
look
at
starting
to
extend
some
of
the
editing
capabilities
within
octave.
Some
of
that
might
be
llamo
based.
Some
of
that
might
be
guided
similar
to
what
we
have
now
with
like.
A
Okay,
so
the
the
future
work
which
we've
touched
on
before
this
is
stuff.
That's
kind
of
gonna
be
coming
just
in
I,
just
kind
of
want
to
give
everyone
an
idea
of
what
to
expect
in
the
upcoming.
You
know
these
upcoming
nine
months
to
close
out
this
year
and
that's
going
to
be
a
workaround
in
cluster,
as
well
as
workaround,
better
networking
support.
A
Anyone
we're
gonna
call
it
workflows,
creating
our
workloads,
sorry,
creating
a
workload
via
that
is
a
kind
of
integration
of
our
existing
workload
viewer,
as
well
as
the
resource
viewer,
and
making
some
improvements
around
guided
navigation
of
your
workloads
and
and
getting
good
status
information,
and
then
combining
that,
with
the
editing
support
to
allow
you
to
fix
and
correct
errors.
This
is
all
just
kind
of
like
big
picture
for
octant
and
all
of
these
kind
of
tie
in
together,
but
specifically
we're
calling
out
some
of
these
larger
items.
A
That's
still,
the
driving
motivation
of
octant
is
to
help
people
understand
what's
in
their
cluster
and
then
by
understanding
it
allow
you
to
fix
any
issues
that
you
see
with
things
in
your
cluster
as
you're,
developing
and
deploying
new
things.
So
that's
still
our
main
goal.
That's
still
our
focus
and
all
of
these
things
are
kind
of
just
working
towards
that.
So
these
are.
These
are
the
these
are
kind
of
like
the.
A
If
you
look
at
our
Zen
hub,
which
I
can
see
yeah
good
ours
n
hub
here,
we
have
these
epics
in
place
in
this
column,
and
these
are
these
are
all
in
line
with
kind
of.
What
we're
going
to
be
doing
is
just
getting
the
application
in
place,
the
object
that
we're
editing
and
the
plug-in
system.
You
can
see
these
first
three
kind
of
align
with
the
first
three
that
we're
going
to
be
talking
about
here
or
working
on
here
and
then
feature
stuff
for
in
cluster
and
better
networking
support
is
stuff.
A
C
Yeah
there
was
one
more
that
I
wanted
to
just
bring
out
in
the
open
as
an
issue
to
look
at
before
dot
11.
So
we
moved
the
namespace
bar
into
the
left
navigation
menu
on
the
premise
that
it
was
part
of
a
design
change
for
a
new
navbar
at
one
point,
and
since
we've
kept
the
navigation
pretty
much
the
same
for
now,
do
we
want
to
move
that
back
before
we
release
that
Albin.
B
Yeah
we
did
spend
quite
a
bit
of
time
on
that.
Ross
has
put
a
lot
of
effort
into
doing
different
mock-ups
with
different
solutions
and
I.
Think
we
making
quite
a
bit
of
progress
here
and
probably
the
next
thing
we're
gonna
do
is
I
will
do
a
little
bit
of
prototyping
with
the
side
now
to
make
sure
that
works.
B
The
way
we
expected,
because
we
kind
of
came
back
to
that
as
a
solution
now
thinking
that
the
side
now
with
them,
that's
a
little
smarter
than
the
one
that
clarity
provides
out
of
the
box
may
solve
most
of
the
problems.
So
yeah
I
agree
with
you.
You
should
probably
address
that
right.
After
the
new
release
and.
A
B
C
A
A
B
B
So
yeah
I've
been
working
on
I'm
moving
the
terminal
to
the
tab
and
it's
it's
been
pretty
interesting,
definitely
way
more
than
I
bargained
for
when
I
want
here
to
do
that.
But
it's
been
a
great
learning,
experience
and
I
think
it
works
really
good.
The
way
it
is
right
now
so
basically
terminal
is
now
press,
there's
a
separate
tab
on
the
pod
view
and
if
you
click
on
it,
it's
gonna
up,
I'm.
B
Sorry
things
a
little
slow
on
my
machine,
I'm
running
some
bills
in
the
background,
so
terminal
basically
opens
kind
of
similar
to
our
viewer
and
logs
inside
its
own
tab.
And
you
can
do
you
know
anything
you
you
want
there
inside
the
terminal
and
it's
gonna
be
present
as
long
as
you
have
a
connection
or
you
exit
the
terminal.
So
whenever
you
go
back
to
this
pod,
you
can
go
I,
don't
know
cron
jobs
and
come
back.
B
B
Yeah
things
are
really
slow
on
my
machine
right
now
and
come
back.
It's
it
will
just
create
a
new
terminal
and
start
sharing.
So
I
did
put
a
little
bit
of
water
to
make
this
session
specific.
Basically
trying
to
make
the
each
desktop
window
in
your
browser
has
a
different
session,
but
there's
very
few
problems
with
that.
B
Quite
a
bit,
I'm,
sorry,
quite
a
few
problems
with
that
and
so
definitely
require
a
little
bit
bigger
surgery
inside
our
server-side
code.
So
me
may
address
that
at
some
point,
but
I
think
the
way
it
works.
Now,
it's
it's
pretty
good.
So,
basically,
if
you
have
another
window
opening
the
same
pod,
it
will
show
exactly
the
same
copy
with
terminal
which,
depending
on
your
application,
I'd
be
good.
B
A
And
I
think
that's
that's.
That
would
be
expected
for
us
right
now,
given
that
we
hold
a
global
session
state
and
we're
designed
to
work
from
the
desktop
as
we
as
we
get
into
the
in
cluster
work,
we're
actually
going
to
need
to
create
user
based
sessions
where
it's
actually
each
obviously
each
you
know,
browser
tab
is
its
own
unique
session
to
octant,
and
so
that
work
will
be
like
will
have
to
re-architect,
and
we
engineer
some
things
and
that
work
will
be
handled
there.
So
this
is
as
it
is
now.
This
is
perfect.
B
D
C
Alright,
so
I
feel
like
this
is
the
sort
of
thing
that
would
end
up
in
a
blog
post
at
some
point,
but
this
is
a.
This
is
a
very
interesting
problem
that
happened
when
we
upgraded
to
angular
9,
and
we
use
this.
We
used
to
use
this
library
called
ng
rx
taken
to
destroy
and
for
those
unfamiliar
with
the
angular
space.
There's
the
idea
that
you
have
these
things
called
subscriptions
and
angular,
and
they
just
watch
a
specific
variable
for
data
changes,
and
it
knows
how
to
smartly.
B
C
Itself
as
it
changes
in
real
time
now,
the
problem
with
using
a
library
like
this
is
that
it's
nice
in
the
sense
that
we
had
it
so
then
it
would
just
sort
of
magically
unsubscribe
from
them.
So
when
the
component
gets
destroyed
these
these
Watchers
that
look
for
changes
in
variables
would
just
magically
go
away,
but
what
we
upgraded
to
angular
9,
this
particular
library
just
stopped
working
silently
and
as
a
result,
it
became
this
be
capable.
C
We
created
a
very
interesting
by-product,
which
I'll
show
here
in
our
content
component,
and
we
can
actually
see
it
in
a
very
simple
way
through
logging.
So
an
example
would
be
something
like
this:
not
content
service
dot
current
subscribe.
So
let's
just
put
like
a
print
statement
here,
and
this
is
gated
and
give
it
a
second
to
load
up,
and
hopefully
it
should
move
in
a
second
here.
B
C
B
C
We
just
change
back
again
and
it's
going
to
increase
by
another
unit,
so
it's
not
wasted
by
three,
and
this
is
going
to
continuously
increment
every
time
a
user
clicks
a
link
all
because
we
use
a
third
party
at
subscribing
library,
so
I
guess
the
moral
here
is
know
what
your
dependencies
are
doing.
It
is
great
to
a
different
engine,
just
sort
of
be
aware
that
they
can
have
unwanted
side
effects.
That's.
C
Well,
this
was
more
of
an
observation
right,
so
I
was
looking
for
performance
gains
elsewhere,
because
I
was
trying
to
shave
off
some
response
times
when
clicked
links
and
then
I
realized
it
got
slower
and
slower
as
I
use,
doctrine
and
that's
II
didn't
seem
quite
right,
either
so
yeah.
So
this
was.
This
was
kind
of
a
fun
little
tidbit
on
the
behind
the
scenes.
Evocative
very.
B
A
There's
so
many
but
yeah
the
JavaScript
ecosystems
is
interesting
in
that
way
where
it's
just
I
will
all
could
all
ecosystems
can
end
up
like
that,
but
specifically
I
feel
like
the
JavaScript
one
is
just
kind
of
like
for
me.
It's
like.
Oh,
that
work,
keep
using
it
and
then
but
yeah.
You
make
a
good
point
like
having
to
go
back
and
check
that
it
might
not
be
working
now.
So
that's
silent
failure,
it's
a
killer.
A
D
B
D
B
D
And
I
found
it
by
the
way
and
I
realized
why
I
couldn't
find
it?
We
moved
our
docks,
so
all
of
my
links
are
dying,
but
this
is
what
it
looks
like
right.
So
this
is
a
landscape
for
harbor.
Basically,
we
went
and
found
four
or
five
different
products
that
have
similar
scope,
and
then
we
created
a
set
of
things
that
what
they
do
or
cannot
do
like,
for
example,
like
sink
Brockton
right.
Can
you
edit
our
values
for
deploy
workloads?
Can
you
view
this
type
of
things?
Can
you
view
logs?
Can
you
pods?
D
Can
you
scale
the
parts?
Can
you
troubleshoot?
Can
you
like?
Those
will
be
the
some
of
the
categories
for
rockton
and
I
know
you
already
have
some
of
those,
so
it'd
be
good
for
us
to
have
it.
We
can
just
publish
it
out
as
well,
so
people,
if
they
want
to
know
hey
I,
want
to
pick
octant
versus
X.
Why
should
I
and
build
on
the
value
proposition
and
one
of
the
links
there
will
be
a
road
map
and
it'll
can
tell
them
about
hey,
even
though
we
don't
have
XYZ.