►
A
All
right
here
we
are
so
you
know
what
I
just
noticed
check
this
out.
So
if
you
open
github.dev,
they
now
have
enabled
vs
code
for
oh
okay.
You
have
to
be
signed
it,
but
they've
enabled
vs
code
or
all
that
basically
in
browser
right.
So
the
question
that
I
have-
or
rather
that
we
have
is,
can
we
leverage
the
in-browser
environment
and
bridge
that
to
whatever
machines
we
have
right.
A
So
the
first
question
is,
you
know
they
have
existing
SSH,
connectors
and
and
container
connectors
for
remote
development
with
BS
code.
So
what
we're
looking
into
is
how
to
leverage
the
existing
remote
development
stuff
within
vs
code
and
potentially
patch
in
support
for
connection
to
workspaces,
directly
managed
via
coder,
which
is
something
that
manages
workspaces.
It's
like
vs
code
on
your
own
infrastructure,
so
right.
A
So
what
we're
trying
to
do
is
basically
take
the
vs
code
on
our
own
infrastructure,
which
is
the
dfl
community,
Community
infrastructure
and
we're
trying
to
which
is,
you
know
as
we're
using
to
develop
our
second
right
plug-in
ecosystem
and
make
it
accessible
for
maintainers
or
for
for
contributors,
and
so
so
what
we
can
do,
ideally,
is
you
can
we
will
need
to
spend
out
a
derp
server
and
we'll
need
we'll
we'll?
What
we're
going
to
do?
A
First
is
we're
going
to
roll
some
containers,
so
we're
going
to
have
a
container
registry
and
then
we're
going
to
deploy
this
coder
thing
I'm
on
top
of
kubernetes
and
then
finally,
we
will
well
we'll
first.
Actually
this
will
be
the
deployment.
But
the
first
thing
we're
going
to
do
is
patch
vs
code
to
take
advantage
of
that.
Well,
actually,
okay,
we're
gonna
have
to
so
we're.
Gonna
have
to
run
we're.
Gonna
have
to
do
that.
A
We
should
be
able
to
deploy
code
or
using
this
group
stuff
and
then
okay,
we're
forwarding
fantastic
okay,
and
this
will
let
us
manage
our
workspaces
and
and
infrastructure
with
terraform
so
and
with
so
so
this
way
we
can,
we
can
provide
compute,
you
know
for
specific
use
cases
or
we
can
enable
the
original
thing
that
we
wanted
to
do.
You
know
with
this
whole
thing
is
basically,
you
know
as
we're
all
working
on
different
things.
A
We
need
to
be
able
to
like
jump
into
different
environments,
so
so
that
we
can
review
pull
request
and
so
forth
right.
So
this
has
come
up
with
now,
I
think
if
you
go
back
about
like
a
year
and
a
half
in
the
weekly
sinks
or
more
okay,
so
remote
agencies.
So
let's
dig
in
here
so
we
we
did.
A
We
went,
we
followed
this
Rabbit
Hole
and
that's
documented
in
the
engineering
log,
basically
search
Source
graph,
we
weren't
able
to
find
the
actual
implementation
via
the
documentation
link
for
vs
code.
If
you
go
to
the
plugin
documentation
link,
it's
it's,
it
doesn't
link
you
actually
to
the
code.
It
links
you
to
a
stub
which
is
just
docs.
The
code
it
actually
looks
like
is
contained
within
the
vs
code,
repo
itself,
perhaps
under
this
srcbs
platform,
remote
browser-
and
maybe
this
browser
socket
Factory.
A
So
if
we
can
figure
out
how
to
connect
these
websockets
or
if
we
can
figure
out
what
what's
going
on
with
these
websockets
right,
so
references
new
browser,
websocket,
okay,
that's
here,
that's
where
we
came
from
default,
websocket,
Factory,
so
browser
socket
Factory
is
probably
what
we
want
to
look
for
there,
foreign.
A
A
Where
does
this
thing
get
instantiated
right
and
configured,
because
we
had
originally
found
this
path
by
looking
for
the
show
terminal
login,
which
is
one
of
the
config
options
right,
so
we
started
looking
for
config
options
because
we
figured
okay
well
you're,
going
to
consume
them
somewhere
right,
okay,
and
then
you
know
we
should
probably
it
would
be
good
for
us
to
actually
just
look
at
what
is
the
remote
provider
being
used
here?.
A
A
A
A
A
A
A
A
A
A
Okay,
so
we
want
to
we'll
load
the
data
into
the
postgres
database
and
then
you
know
once
again
we
already
have
the
digitalocean
API
automation,
yeah.
So
we'll
just
add
this
database
stuff
for
that.
So
because
what
we'll
do
is
you
know
we'll
Define
and
obviously
we're
just
doing
spin
up
spin
down
here,
but
you
know
we
want
to
enable
us
to
do
whatever
so
I've,
just
just
picking
Digital
Arts,
because
it's
extremely
easy
so
for
no
reason
other
than
that
everything
that's
just
too
complicated.
So.
A
Okay,
it's
a
little
past
10
cents
an
hour,
very
careful,
okay!
Oh
my
God
wait
a
minute!
Wait!
A
minute!
Wait!
A
minute!
Wait!
A
minute!
Wait
a
minute!
What.
A
A
All
right,
okay,
you
might
as
well
request
it
right
so
because
what
we'll
do
is
okay,
so,
basically,
okay,
so
for
for
four
dollars.
A
The
thing
is,
we
don't
need
that
long.
We
can
just
spin
this
stuff
up
and
spin
it
down.
So
this
is
actually
incredibly
cost
effective.
It's
free
these
things
can
really
crank
at
this
feed.
That
I
mean
eight
CPUs.
A
A
Basic
and
professional
Finance,
okay,
I
just
want
them
to
increase
the
limit
in
case
we
need
it,
add
another
node
pool.
Oh
you
can't.
You
can't
have
even
more
pools.
Okay,
oh
they
don't
want
you
to
just
hang
there
and
structure
I,
see
excellent.
Okay,
a
dollar
an
hour;
okay,
oh
wow!
We
can
go
crazy.
Okay.
A
So
I
found
out
we
found
out
about
this
chaos,
risk
coordinate
group
today
which
pretty
cool
stuff
that
they've
got
going
on
there
and
you
know
chaos
metrics
of
chaos,
metrics
for
the
chaos
God,
so
they've
also
has
a
risk
working
group,
which
is
which
is
interesting
so,
and
this
is
around
way
to
go
yeah.
So
this
is
interesting
need
to
dig
into
this
more
dependency
risk
assessment.
Obviously
we
have
touch
points
there.
So,
where
two
points
are
quite
a
few
of
these
places,
it's
like
all
of
them.
A
Not
you
know
we
don't
really
we're
not
really
dealing
with
business
risk
so
much,
but
you
know
we
want
to
enable
people
too
right
so
via
context
awareness,
so
I'm
going
to
join
these
calls.
So
you
know,
if
anybody
else
wants
to
join
me,
that's
I
will
I
will
be
joining,
see,
see
if
we
can
understand
what's
going
on
there
and
contribute
and
get
involved.
A
Beyond
that,
what's
going
on
so
yeah
I
I
would
like
to
and
then
and
then
after
seeing
the
risk
death
saw
this
auger
auger,
so
I,
basically
very
similar
right,
so
very
similar
to
what
we're
doing
so.
What
I
was
kind
of
thinking
is
that
we
could
just
spin
this
up
run
some
scans
and
spin
it
down
right.
A
So
they
have
this
database
backend
ooh
do
they
have
containers
they
do
they
have
containers
published
to
GitHub?
Okay,
this
solves
our
problem,
largely
Okay.
So
this
is
fantastic
okay,
so
we
can
just
wow.
This
is
great
okay.
This
is
fantastic
okay,
so
we
can
just
start
doing
this
right
now.
A
A
A
A
A
So
what
did
we
want?
We
wanted
the
service
container.
A
Okay,
services,
redis,
okay,
so
foreign.
A
A
I
know
and
then
we'll
add
the
engineering
log
in
the
commit
trailer
right
so
should
I
contribute
okay,
so
yeah.
This
is
fine.
You
know,
I,
think
repo
URLs.
A
A
Or
we'll
just
do
repo
URL
for
now
this
is
workflow
dispatch,
so
yeah,
okay,
so
let's
say
on
push
as
well.
A
A
A
All
right,
so
we
should
be
able
to
spin
up
the
whole
thing
within
the
job.
Is
the
goal
right
and
then
we
can
see
about
expanding
the
different.
You
know
if
we
need
to
run
into
solution
or
something.
A
A
A
A
A
A
A
That
should
be
Docker
compose
a
bad
stalker
compose
and
then
it
looks
like
they
have
separate
Docker
compose
files,
they're
not
actually
using
they're
like
have.
They
have
separate
Docker
composed
cells,
so
they
have
database
composing.
So
just
pretty
nickel
thing
I
think
just
for
the
second
trade
required
so
supports.
A
A
A
A
A
A
I
think
we
can
safely
just
grow
all
these
environment
variables
in
here.
A
A
Okay,
so
foreign.
A
A
A
A
That
could
be
interesting,
so
we
could
do
workflow
dispatch
to
start
other
instances.
So
this
workflow
dispatch
would
start
the
primary
one.
That
also
has
the
database
and
then
we
could
do
like
a.
We
could
add.
We
could
do
the
Derp
server
and
then
we
could
do
a
matrix,
maybe
to
start
more
jobs.
And
then
we
could.
You
know
the
scanning
that
way.
So
I
think
we'll
just
put
this
so.
A
A
A
A
A
A
A
A
A
A
A
A
A
Or
could
get
integration
what's
this
th?
Okay?
So
that's
nice
worker
base.
So
this
is
very
similar
to
the
initial
interface
that
the
neutral
metric
collection
that
can
go
way
back
in
the
code
base.
If
you
go
back
to
like
when
the
first
open
source
release
of
dfml
and
very
similar
concept,
only
the
workers
are
called
the
metrics.