►
From YouTube: Eclipse Che Internals | Cloud Tool Time | Mario Loriedo
Description
Eclipse Che is well known as a web IDE but developers are unfamiliar with its architecture. What are the different components of a Che workspace? How does Che relate to Eclipse Theia? How many resources does a Che workspace require? What's the startup sequence of a Che workspace? These are some popular questions about Che internals that we would like to answer in this webinar.
Presenter: Mario Loriedo - Jan 21, 2021
Get Social!
Website: https://ecdtools.eclipse.org/
Subscribe: https://www.youtube.com/user/EclipseFdn
Follow us on Twitter: https://twitter.com/ECDTools
Like us on Facebook: https://www.facebook.com/eclipse.org/
Join us on LinkedIn: https://www.linkedin.com/showcase/ecd...
Tag us: #EclipseFdn #ECDTools #ToolTime
A
Hello,
everybody
and
welcome
to
another
cloud
tool
time.
This
is
our
third
edition
of
the
cloud
tool,
time
series,
and
today
our
topic
is
eclipse.
J
internals
presented
by
mario
lorietto.
Mario,
is
the
senior
principal
software
engineer
at
red
hat
and
the
eclipse
che
project
lead,
he's
also
an
architect
for
red
hat
code,
ready
code,
ready
workspaces,
the
red
hat
distribution
of
eclipse,
che
and
che.openshift.io.
A
The
online
chase
service
hosted
by
red
hat,
as
he
will
tell
you,
he
is
an
italian,
but
his
living
has
been
living
in
france
for
the
last
10
years,
which
is
long
enough
to
get
a
french
accent
when
speaking
english,
but
not
long
enough
to
prefer
french
cheese.
Just
yet
our
session
today
will
show
you
the
different
components
of
a
che
the
resources
it
requires,
the
startup
sequence
and
answer
other
popular
questions,
hopefully
leaving
you
with
us
a
more
a
better
familiarity
with
the
che
architecture.
A
B
Yeah,
thank
you.
Thank
you
very
much.
Thank
you
for
for
having
me.
I
hope
that
you're,
you
can
hear
me
well,
so
I
will
we'll
talk
about
eclipse
che
and,
in
particular
we're
gonna
see
the
different
components
of
eclipse
eclipse
che
and
if
you,
if
you
don't
know
here
we'll
ever,
I
have
a
couple
of
slides
to
summarize
what
eclipse
is
and
why
is
it
useful?
B
So
as
basically
it
eclipse
che
is
a
server
that
allows
you
to
run
ides
on
kubernetes
and
that
actually
has
a
has
a
lot
of
benefits.
The
first
one
is
that
you,
the
developers
when,
when
the
their
ide
is
running
on
kubernetes
and
is
accessible
through
a
web
browser,
they
will.
It
will
be
easy
to
onboard
new
developers
and
to
start
a
development
workspace,
so
starting
kodi
code
coding
with
zero
effort.
That's
the
first
benefit.
B
B
It
will
never
transit
your
your
your
laptop
or
your
developer's
laptop.
It
doesn't
prevent
developers
from
doing
that,
but
it
allows
workflow
that
avoids
downloading
source
code
on
local
laptops
and,
finally,
what
it.
What
che
does
is
that
it
allows
to
define
your
development
environments
as
code.
So
you
can
check
in
the
definition
of
your
ide,
your
level,
development
tools
and
your
git
repository
along
with
the
source
code
of
your
application.
B
And
that
cyclops
js,
as
I
said,
run
on
kubernetes
on
kubernetes
cluster
and
it's
easy
to
try
it
on
chairs
of
service,
the
osa
osted
version
of
che
by
red
hat
that
is
available
on
che
openshift
io.
That's
the
easiest
way
to
try
it
if
you
want
more
resources
or
if
you
have
want
to
play
more
with
the
server
side
of
j,
you
probably
want
to
try
it
locally
and
we
have
many
options
with
local
clusters.
Kubernetes
cluster,
like
mini
cube,
kind,
docker,
desktop
we've.
B
B
B
So
here
I'm
going
to
go
a
little
bit
more
in
deep
on
what
are
the
different
server
components
and
that
those
those
components
that
you
see
here
correspond
to
different
pods
that
you
have
once
you
install
j
on
your
kubernetes
cluster.
So
I'm
going
to
have
a
look
at
it.
So
here
I
have
a
mini
cube
that
is
running
I'm
in.
B
B
So
this
one
that
is
here
on
my
slide
as
the
this
that
file
registry
here
on
the
bottom
left.
That
is
actually
a
catalog
of
the
different
stacks
that
you
could
use
as
a
sample
stacks
that
we
provide
to
users.
You
can
of
course,
start
developing.
You
do
build
your
custom
stack
with
any
programming
language
that
support
that
is
supported
by
our
plug-in
system
and
with
any
container
that
you
would
like
to
include
in
your
development
environment,
but
we
provide
a
catalog
of
existing
stacks
with
the
most
popular
technologies.
B
So
this
the
second
one
is
the.
The
second
part
is
the
plug-in
registry.
Plugin
registry
contains
the
definition
of
some
sidecar
containers
that
implement
some
extra
functionality
of
workspaces,
so
the
ide
is
functionality
of
your
of
your
of
our
workspace.
For
example,
the
editor
part:
we
you
you
can
we
we
we
support
multiple
editors,
the
default
one
is
eclipse.
Tia
is
one
one
id
based
on
eclipse
that
we
call
chetia.
B
That's
the
plugin
registry,
so
the
other
part
is
the
server
we
have
already
seen
it.
Then
we
have
an
identity
provider
that
is
key
clock
by
default.
That
allows
to
authenticate
and
authorize
users
that
open
the
id
in
their
in
their
browser.
If
you're
not
an
indicator,
you're
not
able
to
access
or
start
or
call
the
api
of
the
chase
server
and
and
then
there
is
postgres.
B
So
we
use
a
database
to
store
the
metadata
of
the
workspaces
of
the
different
users,
and
then
we
have
an
operator
or
an
elm
chart
depends
on
what
you
prefer
to
provision
all
all
those
pods.
So
the
the
operator
is
a
little
bit
more
powerful,
but
the
uncharted
is
is
well
probably
more
widespread
than
than
the
operator.
So
those
are
the
two
installer
for
for
the
cha
for
che
for
the
chaser
part.
So
that's
once
you
you
run
the
operator
or
the
chart.
B
So
that
is
pretty
useful.
I
haven't
installed
the
image
puller
on
my
mini
cube.
Here.
Don't
really
need
that,
so
I
don't
have
a
big
cluster.
It's
just
a
single
node
cluster.
So
once
I
start
tia,
first,
the
the
first
time
when
I
start
chatia,
that
container
will
be
pre-pulled
and
I
will
have
it
here
locally.
So
I
don't
have
the
the,
but
I
have
I've
I've
installed
it
using
the
operator
and
the
operator
as
then
installed
all
the
older.
B
That's
the
server
components
now
we're
gonna
go
and
see
what
actually
are
what
what
are
the
components
of
the
shea
workspaces
and
why?
Where
we,
what
are
their
role
so,
first
of
all.
Well,
let's
look
at
the
definition
of
a
chair
workspace,
it's
yeah
ml.
It
can
be.
It
can
be
even
like
a
few
lines
here.
I
have
a
kind
of
complicated
one
so
just
because
I
wanted
to
show
the
different
things
that
you
can
add
in
the
definition
of
a
chair
workspace.
B
Then
the
the
second
section
will
be
the
component
section
where
you
actually
define
the
different
component
of
your
workspace.
Here
we
have
a
couple
of
components:
one
is
a
development
tooling
container
that
it
actually
will
contain
java
and
maven,
and
then
we
have
a
second
one
that
actually
is
mysql.
B
So
those
two
container
will
be
started
when
you
start
your
ide
and
will
be
started
along
with
the
editor
itself
and
with
the
plugins
of
the
editor.
The
last
part
is
the
comments
part.
So
these
are
commons
that
the
developer
will
will
see
will
will
be
able
to
to
see
in
the
user
interface.
So
you
won't
need
to
do
maven
clean,
install
every
time
you
will
be
able
it
will
be.
B
So
when
we
start
when
the
workspaces
started,
what
happens
so
now
things
get
more
interesting,
so
what
I
will
do
is
that
I
will
actually
so
I've
already
saw
this.
This
is
this
is
my
sample
example
that
I
started
before
I'm
going
to
start
another
another
workspace
here,
the
the
workspaces
that
I've
started
here.
I'd
started
on
chopin
shift
io.
Now
I'm
starting
something
on
my
local
machine.
I
will
actually
use
a
link,
so
it's
the
easiest
way
to
start
a
container
so
I've.
B
B
It
will
start
an
instance
of
a
chair
workspace
that
will
be
using
the
the
definition
that
is
here
in
this
dev
file
here.
So
what
I
will
do
is
that
I
will
click
here
and
that
will
start
much
ease,
and
so
he
had
found
my
dive
file
and
is
starting
the
workspace.
So
what
what
actually
will
happen?
So,
let's
go
back
here
and
let's.
B
B
That
is
actually
going
to
look
at
the
metadata
of
of
the
plugins
that
are
listed
in
the
dev
file,
so
it
will
get
the
the
metadata
with
the
definition
of
the
container,
the
siders
that
are
needed
to
run
those
plugins
and
we'll
get
them
back
to
the
chase
server.
Once
that
is
done,
the
chase
server
will
be
able
to
call
the
kubernetes
api
with
to
create
a
deployment
that
actually
will
define
the
workspace
pod
that
will
first
of
all
run
in
it
as
a
an
init
container.
B
The
plugin
broker
that
time
will
that
will
that
in
that
second
time
will
actually
run
to
get
the
visual
code
extensions.
In
that
case,
this
is
circle
extension
because
we're
using
chatia
that
is
compatible
with
business
record
extension.
So
it's
going
to
get
the
circle
extension
that
we
want
to
include
in
in
the
workspace
and
it
will
download
them
in
a
plugins
volume
where
the
id
or
the
containers
that
will
then
be
started.
B
B
The
git
repository-
and
so
I
see
the
the
git
repository
the
github
repository
where
I
started
from,
has
been
cloned
here,
so
I
will
be
able,
so
I
can
see
that
the
dev
file,
where
that
is
actually
the
definition
of
the
work
workspace
where
that
is
running,
and
there
is
the
the
java
source
code.
So
let's
just
open
one
file
and
we
have
the
the
java
vs4
code
extension
that
is
starting.
B
Let's
do
that.
We
we're
not
going
to
look
too
much
here,
but
so
because
that's
not
the
goal
is
not
I'm
not
going
to
look
at
the
feature
of
the
chat
workspace,
but
I'm
more
interested
in
what's
happening
behind
so
now.
The
the
visual
code
extension
that
that
supports,
so
the
javascript
extension
is
running
I
can
see
that
is
that
it
there
is
an
outline.
So
what
happened
here
is
that.
B
B
In
that
pod,
here
I've
so
on
in
my
browser,
first
of
all,
the
id
has
been
loaded,
so
there
is
a
a
part
of
the
id
that
is
running
on
my
browser.
So
and
even
if
I
go
offline,
I
will
be
able
to
continue
editing
in
my
browser
the
the
source
code.
Then
there
are
some
features
that
are
actually
require
some
server
side
services
that
won't
be
working,
but
the
part
that
is
in
my
browser
as
well.
The
editor
that
is
in
my
browser
will
continue
working.
B
B
B
B
The
of
the
http
request
that
comes
to
the
to
ear
to
the
services
that
are
exposed
in
the
workspace.
So
we
don't
want
that
anybody,
anybody
to
be
able
to
use
the
ide
or
execute
commands
in
the
workspace
of
somebody
else.
So
we
have
one
proxy
that
takes
all
the
connections.
So
all
the
connections
need
to
go
through
that
proxy
and
then
the
requests
are
routed
to
the
right
service
that
is
running
in
one
of
those
containers
and
then
these
are
these
four
containers
are
id
tooling.
B
So
these
are
not
something
that
I've
defined
in
my
in
in
my
that
file,
or
that
I
will
want
to
define
as
a
user.
So
as
a
user,
I
would
like
more
to
define
so
my
development,
tooling,
so
a
version
of
maven
or
java
and
all
the
tools
that
I
need,
like
git
or
any
other
tool
that
I
would
like
to
have
in
my
command
line.
So
that's
the
development
unique
container,
so
I
I
will
have.
B
I
will
want
to
have
all
the
tools
that
I
will
need
to
build
test,
publish
my
my
application,
so
those
tools
are
in
this
container.
Then
there
will
be
another
container.
That
is
my
the
my
sql
container.
That
is
actually
will
be
useful
when
I
test
my
application
to
you,
so
I
will
be
able
to
use
my
application
with
that
version
of
of
my
sql.
Those
are
just
example.
You
can
change
them
with
whatever
you.
You
would
like
to
customize
your
workspace.
B
B
Okay,
so
I
have
if
we,
if
we
look
at
the
pause
here-
that's
so
we
we
have
looked
at
it
before,
so
we
can
see
now
that
there
is
another
pot,
so
this
spot
is
actually
that
body
here
that
has
been
started.
So
we
have
the
chaser
pod
right
here
and
then
we
have
the
the
workspace
part.
That
is
this
last
part.
Here
I
have
a
command
here
to
get.
B
Container
name
so
workspace
container
name,
so
I
will
actually
that's
a
just.
If
I
look
what
what
it
does
it's
it's
it's
it
does
a
cube,
ctl
get
pod
and
it's
filter
out
for
the
the
workspace
the
workspace
part
I'm
looking
for
and
then
it
gets
for
all
the
all
the
different.
The
different
containers
that
are
definitely
in
the
body
gets
the
name,
so
these
are
actually
the
different
containers
that
are
in
the
pot.
B
B
Let's
say
that
there
is
something
that
is
always
we.
We
got
always
the
the
question
about,
so
so
people
want
to
try
try
out
che.
They
want
to
know
how
how
many
resources,
what
what's?
What
are
the
resources
that
a
che
workspace
will
will
require
and
that's
a
different,
difficult
question,
because
that
depends
on
a
lot
on
what
application
you're
going
to
run,
what
languages,
what
framework
you're
going
to
run
on
your
id.
B
We
see
that
the
editor
is
taking
as
as
as
limit
500
megabyte,
so
alpha
gig
and
one
gigabyte
dot,
five
in
1.5
gigabyte
for
the
the
java,
obviously
extension
that
is
required.
I
think
it's
that's
that's.
That
is
the
the
default
of
the
of
the
java
extension.
We
are
on
the
safe
side,
so
we
that's
the
limit,
so
we
don't
want
the
java.
We
also,
we
are
so
called
extension
to
be
killed
when
we're
indexing,
the
the
source
code
of
our
of
projects.
B
B
My
sequel
is
300
megabits,
so
I
think
my
slides
are
bad,
so
I
haven't
put
the
the
right
number
but
yeah,
then
vso
code
is
1.5
gigabytes
and
the
exact
container
is
125.
These
are
like
the
default.
125
megabytes
is
the
default
of
of
plugins
that
don't
define
a
specific
memory
limit
and
then
the
last
is
the
chat
here.
The
the
editor
one
that
is,
that
is
taking
512.
B
so
that
again,
that
common
is.
I
get
that,
as
with
a
cube,
ctl
comment.
So
as
just
cube,
ctrl
get
powered,
I
get
the
the
parts
and
the
part
and
then
with
no
way
it's
not
okay,
it
should
be
better
now.
So
I
get
the
output
in
json
and
then
I
do.
I
use
jq
to
get
the
containers
and
to
get
the
name
of
the
container
and
the
resource
resources
limit
memory.
B
B
We'll
see
a
couple
of
variations
for
a
couple
of
different
things
and
the
first
one
is
a
different
way
to
containerize
the
ide
so
I've,
just
as
we
have
seen
we
have
in
the
previous
example
and
the
one
that
is
running
in
my
in
my
browser.
Right
now
we
have
split
the
different
processors
in
different
containers.
B
That
has
some
limits
and
some
benefits,
so
the
main
limits
is
are
first
of
all.
Is
that
you're
not
isolating
your
ide
from
your
from
the
developer,
the
development
application,
the
application
that
you're
testing,
so
that
is
from
resources.
Point
of
view
so
will
be,
will
make
your
ide
or
your
application
more
unstable.
B
So
you
you
don't
have
the
containerization.
So
second,
second
problem
is
that
that
container
that
is
defined
here
is
a
result
of
the
build
of
of
an
image.
So
if
you
want
to
add
something
so
extend
that
add
other
tools,
you
will
probably
need
to
do
another,
build
and
off
of
the
the
this
image,
and
this
is
actually
the
image
that
contains
the
editor
and
contains
a
lot
of
dependencies.
B
So
that
may
be
three
key,
because
you
may
be
ending
with
a
custom
image
where
your
editor
will
be
an
editor
that
you,
you
got
a
few
months
ago,
that
you
haven't
updated
and
then
you
need
to
maintain
yourself
your
image.
There
is
not
an
is
an
isolated
image
that
gets
updated
by
the
upstream
community
and
your
tool
that
you
manage,
and
you
upgrade
so
it's
you're,
not
decoupling
the
editor
from
your
your
custom
tool.
B
Those
are
the
limits,
but
on
the
other
end
you
have
some
benefits
because
you
will
have
everything
in
the
same
container
and
those
technologies
are
actually
being
designed
to
run
on
the
desktop.
So
those
are
are
not
new
technology,
so
those
technologies
like
video,
soco
extensions,
are
processes
that
that
are
being
designed
to
run
in
a
laptop
where
actually,
the
everything
is
is
running
in
the
same
in
the
same
host.
B
B
So
that
means
that
that,
if
you
put
everything
in
one
container,
it
won't
be
for
a
user
easy
to
maintain
the
customized,
the
their
customized
workspace
image,
but
they
won't
occur
in
problems
that
are
typical
of
extension
that
are
separated
in
in
different
containers.
So
that's
that's
the
benefit
of
of
this
approach.
B
So
what
we
are
doing
is
that
we
are
trying
to
to
find
something
in
the
middle
of
the
two
approaches,
because
having
too
many
too
many
containers
may
be
complicated,
may
also
add
some
time
when
we
start
one
workspace,
but
we
want
to
have
to
make
it
repeatable
and
flexible
as
it
is
today.
So
we
would
like
to
keep
this
architecture
of
satire
architecture
still.
B
So
that's
one
different
way
of
doing
counterization
so
another
way
where
we
could
do
some
things
different
is
that
in
the
example
before
the
development
tooling,
where
may
where
we
had
maven
in
java,
we
actually
were
running
the
application.
The
pet
clinic
application
there
and
my
sequel
was
running
in
that
same
pod
as
well.
That
means
that
we
are
somehow
modifying
the
production
container,
the
production
image
so
because
we
we
are
adding
container
to
the
to
the
image
and
we're
not.
B
We
don't
have
a
parity
between
production
and
development,
so
that
can
be
an
issue,
so
maybe
another
another
way
to
do.
That
would
be
to
run
the
production
containers
without
any
modification
and
running
the
kubernetes
pod
of
so
the
work.
The
workspace
kubernetes
part
as
a
sidecar
to
that
production
containers
and
doing
the
build
so
using
maven
and
the
development
tooling
in
the
workspace
pod.
But
then,
once
we
have
the
result,
we
we
could
update
the
the
production.
B
The
production
pod
either
copying
the
artifact
that
is
result
of
the
build
in
from
the
workspace
pod
to
another
pod
to
the
the
production
pod.
Doing
that
or
even
rebuilding
the
image
and
doing
a
rolling
update
of
the
production,
the
the
production
image.
So
that's
another
approach.
The
problem
is
that
copying
the
artifact
or
rebuilding
the
image,
our
operation
task.
That
takes
time
the
take
time,
so
it
may
add
to
the
inner
loop.
B
B
B
There
is
a
chase
server
that
is
a
manager
of
workspaces
and
a
che.
Workspace
is
actually
powered
by
a
few
pods
one
pod
with
a
few
containers
in
it,
and
we
have
seen
that
from
a
membrane
point
of
view,
the
the
resources
that
are
used
by
that
workspace
are
similar
to
what
a
desktop
id
will
all
actually
use.
B
So
last
here
are
some
links
to
to
learn
more
about
che.
You
can
go
on
the
che
website
or
you
can
contribute
to
to
our
source,
to
our
repository
git
repository,
but
just
opening
issues
or
asking
questions
providing
feedback
or
helping
with
the
documentation
or
even
doing
opening
up
requests
to
add
some
some
new
features
you
can
reach
us
on
twitter
as
well
and
on
matter
most
most
of
our
engineers
will
will
answer
any
questions
we
are
always.
B
A
B
Yeah,
that's
that's
a
pretty
pretty
good
question,
so
we
have
we've
been
adding
the
support
for
restricted
environments
in
the
last
year,
so
we've
been
working.
A
lot
on
that
so
jay
today
is
is
mature.
So
on
that,
from
that
point
of
view,
so
we
have
instructions
to
to
deploy
chain
restricted
network
well
with
openshift,
because
openshift,
so
the
red
hat
kubernetes
distribution,
as
provides
some
tools
that
some
comments
that
will
allow
to
mirror
all
the
images
that
will
be
required
by
you.
B
B
So
that's
for
the
container
part,
as
just
a
matter
of
mirroring
again
open
shift
for
that
open
red
hat,
open
shift.
Does
that
makes
that
easy
really
easy?
So
it's
not
painful,
then
I
don't
know
other
distribution
how
how
they
manage
that.
But
that's
done
with
a
few
comments
on
on
openshift,
so
you
provide
the
list
of
images
that
are
required
by
che.
B
So
for
that
we
we
have
a
mechanism
to
inject
the
configuration
for
the
maven
settings
xml
for
example,
or
for
other
languages
like
so
python,
with
pip
or
golang,
et
cetera,
all
the
dependencies
of
all
the
different
languages.
So
you
we
have
a
mechanism
to
inject
the
configuration
that
are
needed
to
provide
the
url
of
the
repository
on
the
restricted
network
where
the
dependency
should
be
looked
for
and
once
that
is
injected
it
will
be
possible.
B
Then,
to
start
the
pod,
the
the
workspace
spot
will
start
and
you
will
be
able
to
do
maven,
clean,
install
and
all
the
dependencies
will
be
fetched
by
from
from
that
repository.
So
I
don't
think
I've.
So
the
other
thing
is
the
source
code.
Of
course,
the
the
git
repository
need
to
be
in
the
restricted
network,
otherwise
che
won't
be
able
to
clone
the
git
repository
and
our
example.
B
We
are
actually
packaging,
the
the
source
code
and
zip
files,
so
that
when
you
deploy
che
in
a
restricted
environment,
you
will
be
able
to
use
the
example
without
having
to
go
to
to
clone
the
the
example
source
code
from
a
github
repository
that
is
not
available
as
not
reachable.
You
will
just
so
che
will
just
get
the
source
code
from
the
zip
file
and
copy
it
locally,
so
that
you
will
be
able
to
start.
A
I
think
that
was
a
fairly
thorough
answer.
Rob
does.
I
hope
that
answers
your
question.
He
says
thank
you.
So
it
seems
like
it
has
another
question
from
andreas.
You
may
have
already
touched
on
it,
but
how
does
che
compare
to
if
you
have
touched
a
lot
already,
maybe
just
give
like
the
brief.
The
brief
overview,
the
key
points.
B
No,
I
mean,
I
think,
it's
a
it's
a
it's
a
really
good
question.
I
haven't
touched
that
much
so
I've
talked
a
little
bit
about
it,
but
it
so
the
eclipse
tia
is
a
framework
for
to
build
ids
and
we're
using
that
framework.
We
are
one
of
the
main
contributors
to
to
eclipse
tia
project
and
we're
using
that
to
build
chetia.
That
is
it's
our
default
editor,
it's
the
one
that
we
use.
So
we
we
have.
B
We
have
contributed
a
lot
of
of
features
like,
for
example.
The
support
for
video
circle
extension
is
something
that
our
so
the
reda
team
has
has
contributed
yeah
at
the
beginning
of
of
the
project,
and
we
are
continuously
investing
in
that
we're.
Investing
a
lot
of
time
and
effort
in
that
so
tia
is
really
chetia
is
really
part
of
of
che
today.
B
On
the
other
end,
we
see
customers
that
are
asking
for
other
the
using
other
editors
in
particular
for
intellij,
so
we
we
have
also,
as
a
user
you
can.
We
can
select
to
start
your
workspace
using
as
the
I
on
the
id
part
intellij.
So
as
it
will
be
intelligent,
it
will
run
in
a
container
and
then
that
you
will
be
able
to
and
you
will
be
able
to
access
the
ui
through
vnc
with
from
your
from
your
web
browser.
B
So
that's
how
it
works,
and
then
we
have
also
some
community
some
community
support.
So
there
there's
been
some
great
contribution
in
our
community,
so
not
something
that
is
comes
from
our
teams
that
have
contributed
a
coder
support.
So
it's
another
cloud
id
that
that
we
that
you
can
run
in
your
browser
and
it's
actually
it's
just
another,
a
build
of
vs4
code.
So
it's
vs
record
that
runs
in
your
in
your
browser.
So
you
have
options,
but
we
have.
A
B
So
we
had
jupiter
support
at
some
point
and
dirigible,
so
the
yeah,
the
dirigible
jupiter
jupiter,
is,
I
think
we
have
removed
that
because
today
you
can
run
jupiter
from
within
tia,
so
yeah
jupiter
notebooks
these
are
really
popular
and
so,
but
that
that
now
works
from
using
a
video
circle
extension
so
using
the
python
video
circle
extension.
So
we
don't
need
to
maintain
a
separate
editor
for
that.
We
can
just
use
usgt
so
but
yeah
you.
B
There
is
also
the
the
eclipse
desktop
eclipse
ide
that
you
can
run
in
your
browser,
but
that's
something
that
we
we
haven't,
invest
a
lot.
We
don't
so
those
are
technologies
that
are
not
being
designed
to
run
in
your
browser,
so
the
the
the
the
the
final
feedback
user
experience
of
a
user.
We
think
that
would
probably
be
better
for
us
to
invest
on
web
technologies
rather
than
adapting
existing
desktop
technologies
to
run
in
the
browser.
B
Well,
I
think
that
we
are
getting
more
and
more
closer,
so
as
as
everything
so
the
we
have
more
and
more
application
that
we
do
not
install
anymore
on
our
laptop.
If
I,
if
I
need
to
if
I
start,
counting
the
number
of
application
that
I've
installed
locally
so
graphical
application,
these
are
like
really
less
and
less
every
year,
so
most
of
the
application
that
we're
using
are
starting
using
web
technologies
and
use
that
so
for
ides.
B
So
there
are
others
competitors
that
we
have
on
on
the
space
that
we
are
doing
exactly
the
same
thing
so
and
more
and
more
there
are
big
companies
that
are
investing
in
that.
We
think
that
that's
that's
inevitable.
The
the
thing
that
is,
that
changes,
the
challenge
that
we
have
when,
when
you
run
in
a
browser
there
are
a
few
challenges,
so
the
first
one
is
from
a
security
point
of
view.
B
Is
that
your
application
is,
is
going
to
run
in
a
platform
that
is
shared
with
other
applications,
so
you
need
to
be
really
careful
about
security.
You
cannot.
There
are
a
lot
of
things
that
you
you,
you
are
not
allowed
to
do
to
avoid
cross
origin
attacks
so
and
browser
are
more
and
more
strict
about
that.
So
that's
something
that
makes
building
an
application
in
the
running
in
the
browser
more
difficult
than
building
an
application
with
the
same
technologies
that
will
actually
run
in
electron.
B
B
A
B
Well,
because
we
so
we
have
a
lot
of
so
for
for
for
the
different
components
we
have
started,
we
we
we
have
basically
a
different
container,
a
different
repository,
not
sure.
If
we
currently
are
able
to.
We
could
probably
be
able
to
reduce
that
today.
So
we
have
been,
we
have
been
doing
an
effort,
so
we
had
at
some
point.
B
If
you
want
to
run,
go
the
goal
of
your
circle
extension,
you
probably
will
need
to
run
it
in
a
container
that
has
go
and
other
tools
that
you
will
need
when
you
do
go
tooling,
so
go
to
me,
then,
if
you
have
your
java
via
social
extension,
you
will
probably
need
another
container,
so
we
ended
up
with
supporting
tens
of
languages
and
with
tens
of
sidecars,
and
we
had
one
git
repository
for
every
docker
file
and
was
with
the
dockerfile,
came
also
some
script,
etc.
That
would
run
on
the
sidecar.
B
So
I
think
that
most
of
those
are
because
of
that,
but
then
so
we
have
have
you
seen.
We
have
six
component
server
side,
then
we
have
all
the
side,
cars
and
other
six
components
on
the
on
on
the
side
of
the
of
the
workspace.
So
that
means
that
we,
I
don't
think
they
will
never
be
able
to
have
less
than
30
to
the
git
repository.
A
B
Yeah
thanks
for
for
the
question,
that's
yeah,
interesting
quan!
We
are.
There
are
a
few
things
that
we're
working
on.
Of
course,
I
don't
want
to
yeah
to
unveil
all
the
the
things
that
we
so
that
there
are
some
big
changes
that
we
are
working
on,
but
we
don't
know
exactly
when
we
will
be
able
to
deploy
them,
etc.
So,
but
there
are
a
couple
of
things
that
are
coming
that
are
interesting.
So
first
is
suggestions
so
recommendations.
B
So
when
you
start
a
github,
so
we
have
seen
that
you,
you
can
start
your
workspace.
Your
chair
workspace
from
a
git
repository.
You
have
seen
that
we
add
a
dev
file
in
that
git
repository,
but
when
you
don't
have
that
file
in
that
kit
repository
you,
the
you
you
won't.
Actually
you
will
have
the
default
workspace
that
doesn't
have
plugins
that
doesn't
that
only
contains
an
editor
and
the
exact
and
the
authorization
plugin
that
we've
seen
before.
B
So
at
that
point,
when
you
start
your
workspace
and
you
don't
have
plugins
and
we
see
that
there
are
some
java
files
that
will
start
recommending
to
users
if
they
recommending
some
plugins
and
exec
extensions
and
a
stack.
So
we
we
we'll
we'll
introduce
that
in
the
in
the
next
month.
So
that's
that's
something
another
thing
that
more
for
on
the
enterprise
side,
we
want
to
support
multiple
running
chair,
cross,
multiple
kubernetes
clusters.
B
So
we
are
working
on
that
as
well.
So
we'll
have
one
single
entry
point
for
che
and
but
then
the
workload,
so
the
the
workspaces
may
be
provisioned
on
one
of
multiple
kubernetes
clusters
so
that
that
is
required
for
customers
that
have
developers
that
are
distributed
geographically
or
for
big
teams
that
have
that
want
to
have
more
than
android
developers
being
able
to
have
more
than
android
developers
coding.
Concurrently.
A
Great
well,
I
think
that's
something
that
everyone
can
really
look
forward
to,
so
I
think
we're
out
of
questions
for
the
for
the
time
being,
so
we'll
wrap
this
up.
Mario,
thank
you
very
much
for
joining
us
today.
I
think
this
is
an
amazing
session
and
we
will
have
this
on
youtube
in
just
a
few
hours.
If
people
would
like
to
re-watch
it
or
you
can
re-watch
it
any
time
by
using
this
link.
So
thank
you
very
much,
mario.
I
am
going
to
publish
right
now,
a
survey.