►
From YouTube: Community Standup: 12/3/19
Description
Community Standup: Dec 3, 2019
A
A
A
A
The
and
there's
some
more
cuttings
identified
there.
The
second
stage
is,
once
you
have
like
a
cue
column,
inch
from
that
build
or
a
cue
column,
and
they
just
provided
to
you
by
a
vendor.
You
know
what
what
are
the
next
things
you
have
to
do
and
and
that's
like
created,
you
know
the
antidote
user
account,
possibly
installing
packages
or
running
scripts
or
putting
like
a
license
file
on
the
image.
A
A
The
food,
which
is
something
we'd
like
to
maintain,
as
you
know,
is
it
as
a
principle,
no
internet
access
for
the
images
inside
the
lessons,
and
so
there's
three
ways
to
do
that.
There's
when
they're
spun
up
for
a
lesson
via
Napalm
config
by
napalm,
pushing
a
or
using
Python
just
running
a
Python
script,
and
then
the
last
thing
is
using
an
ansible,
playbook
and
there's
gonna
be
things
that
are
pretty
common
to
each
of
those
methods
like,
for
instance,
with
napalm.
A
It's
not
a
config
merge,
it's
a
big
just
that
you
know
config
right
so
there's
things
that
will
you'll
have
to
maintain
inside
of
that
configuration
like
the
antidote,
username
and
password,
and
you
know
things
things
like
that.
So
that
could
probably
be
you
know
automated
in
some
way.
The
same
with
you
know
the
Python
module.
If
the
purpose
of
the
Python
module
is
to
copy
files
over
you
know
we
could
probably,
instead
of
having
everyone,
write
from
scratch
their
own
way
of
connecting
to
the
box
and
then
copying
the
files
over.
A
You
could
probably
come
up
with
a
helper
module
in
Python
that
would
allow
them
to
do
it.
That
kind
of
thing
and
probably
reusable
ansible
play
books
for
similar
tasks.
So
there's
plenty
of
work
items
in
that
mp3
doc.
If
anyone
wants
to
it's,
it's
basically
gonna
be
the
thing
I
spend
most
of
my
time
on
it's
it's
it's
not
too
complicated.
It's
not
nearly
as
complicated
as
empty
one.
B
A
Yes,
so
well
since,
or
the
platform
is
Linux
right
in
production,
you
cow
to
a
queue
kale
to
image
is
functionally
equivalent
to
like
a
VMware
VMDK
file,
correct
and
in
it
schema
that
runs.
The
kiki
mu
is
a
hypervisor
key
move.
/
kb
m
is
the
hypervisor
that
we
run,
which
is
in
production,
which
is
you
know,
equivalent
to
something
like
VMware,
or
is
that
or
hyper-v
so
and
in
q?
Co
2
is
the
native
format
of
of
Kaimuki
vm,
so
we
will
have
to
connect
if
the
image
comes
to
us
from
a
vendor.
A
In
fact,
there's
a
couple
competing
projects,
Qbert
and
firecracker
and
I-
think
there's
a
third
one
that
does
these.
It's
that
the
whole
concept
behind
them
is
the
same.
There's
a
project
called
firecracker
that
yep
that
they
call
it
a
microvia
and
it's
and
it's
basically
they're
running
VMs
and
containers,
but
they're
very
because
you
only
need
verdi.
Oh
they
do
a
custom
kernel
build,
which
is
what
our
goal
should
be
is
to
do
that
and
that
ends
up
with
instead
of
a
3d
image
file
that
you're
putting
inside
of
a
docker
file.
A
Don't
know
if
we'll
actually
use
Qbert
or
firecracker,
mainly
because
some
of
the
plumbing
underneath
is
we've.
We
were
already
running
VMs
inside
of
inside
of
containers,
and
we
already
have
a
pattern
for
stitching
all
the
networking
together,
which
is
really
the
hard
part
and
I
so
I
don't
know
moving
to
Cooper,
it
makes
sense,
I
mean
maybe
we
will
in
the
future,
but
if
we
can
get
to
the
point
of
having
micro
VMs
with
our
current
platform,
it'll
benefit
us
two
ways.
B
B
C
Self-Medicate
should
actually
become
way
better,
because
the
number
of
big-ass
heavy
VMs
that
we
should
be
required
to
run
inside
of
it
should
be
standardized
and
hopefully
reduced.
One
thing:
that's
not
part
of
mp3
is
Derek
and
I
have
been
having
internal
conversations,
putting
the
Juniper
hat
on
a
little
bit,
because
we
contributed
the
V
key
effects
image
it's
possible
that
we
can
actually
replace
nearly
all
of
the
existing
lessons
that
use
the
V
key
effects
with
something
that's
even
more
lightweight
that
still
does
the
job.
C
We
don't
have
to
compromise
on
functionality
at
all,
but
it's
not
running
like
a
full-blown
operating
system.
There's
there's,
there's
a
bunch
of
different
ways
that
we
can
make
the
experience
not
only
for
users
of
the
site,
but
also
for
people
that
are
building
lessons
a
lot
easier
and
mp3
is
one
of
them
and
then
there's
a
few
other
things
that
we've
got
going
on,
that
we
want
to
improve
so
good.
B
Ok,
it
was
working
on
I've
been
trying
to
do.
Packer
builds
for
self-medicate
and
I
haven't
decided
on
putting
those
VMs,
the
docker
pre
loading,
the
images
in
to
self-medicate
or
not,
but
that's,
but
the
task
that
takes
the
longest
amount
of
time
is
those
600,
Meg,
I,
think
they're,
each
600
V
qfx
images,
maybe
they're
800.
They
take
forever
to
to
dr.
pol.
B
Okay
in
the
sticks,
I
have
Suddenlink,
which
is
a
cable
provider,
so
I
have
foreign-made
download,
but
I
and
I
can
speed
test
at
three
four
hundred
on
a
good
day,
but
for
whatever
reason,
I
think
I
may
have
downloaded
too
many
docker
images.
However,
the
case
it
is
docker
either
Dockers
throttling
me
or
suddenly
throttling
me
I
get
about
one
make
a
minute.
A
A
A
They
don't
have
great
internet
dead.
People
could
down
download
the
images
you
know
once
in
a
while
and
and
then,
if
they
just
long
as
they
don't
put,
that
prefix
part
in
they
can
have
a
default
configured
somewhere
like
in
an
environment
variable
and
then
that
image
could
be
pulled
locally.
Possibly.
C
C
C
C
That
is
in
the
lesson
definition,
but
before
it
actually
calls
the
pole,
it
will
append
a
tag
in
code
to
the
actual
to
the
actual
request,
and
so,
if
you're
running
this
is
you're
running
any
lesson
and
self
medicate,
for
instance,
and
you're,
and
hopefully
I
believe
that
this
is
in
the
manifest.
If
it's
not
absolutely
should
be
that
we're
deploying
syringe
inside
of
self
medicate,
with
that
same
environment,
variable
saying
we're
deploying
version
1.0
of
the
curriculum,
then
you
should
only
ever
see
that
image
pulled
like
you
should
never
see.
C
B
C
Not
quite
happy
with
that
just
yet,
but
the
reason
for
what
you
just
said:
if
memory
serves
it's
been
a
while
is
the
Jupiter
notebook
image
is
actually
we
consider
that
part
of
the
platform,
because
that's
sort
of
like
a
feature
of
the
platform.
Yes,
so
it's
not
like
it's
not
like
tightly
bound
to
a
or
set
of
lessons.
It's
it's
more.
Just
like
hey
here's,
some
infrastructure
that
I
need
to
spin
up,
because
this
lesson
requires
the
use
of
a
jupiter
notebook.
C
B
C
And
if
they're
I'm,
looking
at
the
I'm
looking
at
the
syringe
manifest-
and
it
doesn't
have
that
environment
variable,
so
that's
an
that-
that's
that's
the
fix.
That's
what
we
need
to
do
is
tell
ceramic
what
version
of
the
curriculum
we
were
planning
to
load.
I
could
have
sworn
I
did
this
a
while
ago,
but
I
just
must
be
mistaken,
but
any
in
any
case.
What
that
will
do
is
it'll.
Tell
syringe
you.
Don't
you
know,
don't
just
pull
the
lake
as
it
will
pull
me.
It
will
pull
the
latest
version
of
it
without
this.
C
C
C
You
know
that
the
you
know
that
we
had
things
the
right
environment
variable
in
here
properly,
but
it's
not
in
there,
because
what
again,
what
that'll
do
is
it'll
is
it'll
only
pull
that
image
once
because
there
will
never
be
they'll
note
that
image
will
never
change,
and
so,
even
if
the
image
pool
policy
is
set
to
always
so
it'll
probably
still
do
a
poll,
but
it
won't
download
anything
because
it's
not
there
literally.
The
checksum
is
identical.
B
Good,
in
that
case,
I
can
go
ahead
and
preload
the
big
images
for
now
in
the
boxes
and
maybe
I,
could
provide
a
box
with
the
preloaded
and
maybe
a
box
without
the
preloaded.
If
people
are
concerned
about
that,
and
then
we're
really
seeing
a
specific
curriculum,
a
box
for
a
specific
curriculum
version
with
a
specific
mini
cue
version
and
a
specific
key,
a
version
that
all
is
working
and
tested
on
every
machine.
You
know.
C
A
B
Trying
to
get
hyper-v
to
work
inside
a
virtual
box
and
they
basically
said
we're
not
going
to
support
this
at
this
time.
It's
not
one
is
the
first
release
that
introduces
Intel
support
and
even
then
it's
on
select
CPS,
so
they
call
out
the
Broadwell
version
of
the
core.
The
core
series
Broadwell
was
high-fived,
I,
can't
remember
which
one
it
was,
but
there's
newer
versions
than
Broadwell
and
they
say
and
the
issue
notes
they
say
added
additional
support
for
other
versions,
but
they
don't
specify
versions
so
it
it
could
be
an
uphill
battle.
B
C
B
C
A
C
I
mean
no,
yes,
that's
totally
fine,
and
actually
it's
a
byproduct
of
the
fact
that
everything
is
kind
of
related
in
this
project.
There's
no
there's
no
like
isolated,
like
oh
I've,
been
working
on
this
thing.
It's
like
nobody,
you
know
nobody's
relying
on
it
and
nobody
might
be
interested
that,
like
that,
doesn't
exist
like
every
work.
This
every
every
single
piece
of
work.
That's
going
on
right
now
is
the
dependency
of
somebody
else's
work.
A
B
And
its
its
64-bit,
the
issue
is
64-bit
within
the
QAM.
You
may
not
be
supported,
so
you
can
run
32-bit
images.
A
32-bit
bm's
inside
of
a
hypervisor
inside
of
a
hypervisor
but
running
64-bit
requires
the
nested.
Vtx
and
VirtualBox
has
not
supported
that
until
version,
but
v
dot,
v
dot.
Is
it
and
the
event
support
around
until
until
6.1,
which
isn't
out
yet?
A
B
C
Yeah,
nothing
to
show
beyond
what
was
in
last
week.
I
mean
I,
think
I've
I
think
last
week,
given
that
it
was
a
holiday
week,
I
might
have
like
added
some
to
it,
but
it
it's
mostly
unchanged,
I
think
probably
a
probably
next
week
or
maybe
maybe
the
week
after
I'm
doing
a
little
traveling
soon
we'll
see.
But
sometime
in
the
next
few
weeks,
I'll
probably
have
something
to
show
the
VA
the,
but
the
the
same.
The
same
plan
is
there.
Basically
it's
a
get
booked
site.
C
You
guys
all
saw
it
stocks
in
re
labs
done
I/o,
it's
back
ended,
buy
a
new,
a
new
repo.
So
it's
a
you
know:
gonna
be
effect
its
marked
down.
So
it's
you
know
basically
I'll
totally
new
files,
which
I
think
I'll
be
good.
I
think
most
most
folks
last
week,
we're
in
agreement
that
restructured
text
is
just
kind
of
like
May,
which
is
which
I
would
agree
with
its
very
capable
it's
actually
I
I
equate
it
to
like
Emacs
a
little
bit,
but
no
it's
it's
going!
C
Well,
so
basically,
yeah
I
mean
in
the
next
few
weeks.
What
I'd
like
to
do
is
get
to
the
point
where
it's
basically
done
and
and
effectively
once
we
switch
the
site
over,
which
is
still
a
work
in
progress.
We
got
to
figure
out
this
week
to
see
where
things
are
at
with
that,
but
basically,
at
any
time
we
will
be
ready
to.
C
You
know,
start
directing
folks
to
it,
but
hopefully
not
before
we
we
go
live
with
with
the
new
domain,
because
there
they're
all
kind
of
bundled
but
regardless
I
think
there
will
be
a
there
will
be
a
point
where
everybody
can
see.
You
know
everybody
here
on.
The
call
can
you
know
can
can
provide
feedback
as
to
whether
or
not
the
the
the
new
documentation
is.
Is
you
know
up
to
snuff
and
whatnot
I
mean
you
can
technically
do
that
now
it's
all
live
and
it's
all
being
done.
C
You
know
in
its
own
repo,
but
I'll.
Do
it
I'll
do
a
more
or
less
formal
presentation
of
it
I
think
in
the
next
few
weeks,
and
then
we
can
do
some
more
or
less
formal
feedback,
but
same
rule
applies
it's
you
know
it's
public,
but
it's
also
not
super
like
formally
launched,
so
just
keep
it
keep
it
to
yourself
for
now
and
if
you
have
feedback
just
you
know,
send
it
to
me
or
tow
it
or
somebody
else
in
the
forum's.
C
Sure
I'm
there's
a
part,
in
fact,
I
think
another
point
in
the
agenda
that
I
added
I
am
asking
for
guac
feedback
on
the
replacement
plan
for
guacamole.
I
should
really
not
have
a
question
mark
there,
because
people
have
Cabana
provided
feedback,
which
is
good.
Both
Stephen
and
Olivia
have
have
chimed
in
that's
part
of
it.
That's
a
sort
of
a
subset
of
the
work
that
needs
done
so
zooming
out
a
little
bit.
C
Basically,
you
know
the
the
the
the
structure
of
the
new
site
is
there,
but
right
now
the
energy
labs
application
is
still,
you
know,
sort
of
separate.
So
the
idea
is
that,
at
the
end
of
this
effort,
what
we'll
have
is
will
still
have
an
adult
web.
Obviously,
that'll
stick
around,
but
it'll
be
different.
The
way
that
it's
presented
will
be
a
little
bit
different.
C
The
the
the
way
that
the
application
is
changing
a
little
bit
will
be
that
all
the
pay
all
of
the
different
elements
within
a
meta
web
we'll
be
like
widget
eyes,
diz,
the
it's
the
scientific
term.
So
basically,
he
imagined
the
different
components
that
make
up
the
site
that
make
up
like
the
platform
actually
will
be
broken
up
into
they're,
not
quite
iframes,
but
similar
similarly
portable
components.
C
So,
like
the
you
know,
the
lesson
interface
like
how
do
you
interact
with
a
lesson?
The
lesson
catalog,
the
adviser,
the
you
know,
things
like
that
all
of
those
will
be
broken
up
into
sort
of
modular
components
and
effectively
installed
directly
on
the
the
new
site,
which
is
actually
running
in
hugo
now
that
in
order
it
was
very
important,
I've
and
I'm
continuing
to
reinforce
this
with
the
developers
that
we're
working
with.
Obviously,
we
can't
do
that
to
the
point
where
people
that
are
not
running
the
energy
labs
site
can't
use
an
it
at
web.
C
C
We're
doing
in
production,
but
you'll
still
have
a
usable
antidote
web
at
the
end
of
it
by
you
know,
on
its
own
you'll
still
be
able
to
use
it
on
its
own.
It
may
just
look
a
little,
you
know
less
pretty
just
because
it
won't
have
any
styling
around
it.
It'll
just
be
a
very,
very
simple:
it
basically
just
be
a
husk.
C
You
know
like
a
like
a
very
simple
holder
for
those
widgets
if
you're,
just
if
you're
just
working
with
self-medicate,
for
instance,
that's
that's
what
you'll
get
so
that's
the
goal,
the
the
first
part
of
that
which
is
just
building
effectively.
You
know,
rebuilding
the
site
that
we
currently
have
in
Hugo
is
done,
which
is
awesome.
We
you
know
we
want
to.
We
want
to
integrate
the
whole
thing,
there's
actually
a
non-trivial
amount
of
work
that
it's
required
to
sort
of
do
it
halfway.
So
rather
than
you
know,
just
you
know,
launch
the
new
site.
C
It's
I
think
it's
important
to
make
the
changes
to
enter
in
it
web
that
we
need
and
then
and
then
you
know,
go
forward
that
way.
So
that's
what
we're
currently
waiting
on
that,
in
addition
to
paperwork
on
our
side,
to
facilitate
those
changes.
Another
one
of
the
big
decisions
was
how
to
implement
the
new
terminal,
which
is
what
I
covered
in
the
replacing
guacamole
design
dock.
It
sounds
like
there's,
fairly
overwhelming
support
for
the
changes,
which
is
good.
C
Please
keep
the
feedback
coming,
though,
if
you
have
any
obviously
update
update
everybody
if
I,
you
know
if
we,
when
we
do
proceed
with
with
some
of
these
changes,
and
we
do
some
testing
we'll
probably
get
some
of
you
folks
involved
with
that
I'm
sure
but
yeah,
it's
it.
The
the
the
I
wasn't
expecting
necessarily
any
controversy,
but
it
was
good.
It
was
just
good
to
see
that
there's
that
there's
you
know
that
everybody
pretty
much
everybody
supports
replacing
guacamole,
which
I
think
is
good
I.
Think
it's
I,
think
it's
a
good
thing.
F
C
Yeah
we've
been,
we've
been
talking
a
little
bit
on
the
on
the
forums,
especially
since
last
week,
I
mean
I.
Actually,
I
was
watching
some
old
videos
and
I.
Remember
you
popping
up
before
on
on
earlier
stand-ups,
but
last
week
yeah
you,
we
basically
started
talking
about
mp1,
so
just
for
those
that
aren't
super
familiar
with
with
them
with
mp1,
with
with
the
changes
that
we're
gonna
be
doing
and
syringe.
The
very
first
step
is
to
sort
of
re-implemented
database
package.
C
C
Currently,
the
the
way
that
we
report
statistics
is
is
just
basically
we
found
a
you
know:
I
found
a
corner
of
the
API
package
to
throw
some
in
flux,
DB
code
in
and
there
it
is
obviously
that's
not
great
so
going
forward.
What
we're
gonna
do
is
have
a
totally
separate
service,
as
with
most
things
that
are
being
broken
out,
it'll
be
a
totally
separate
service
whose
job
it
is
just
to
simply
look
at
the
database
and
export
statistics
about
it.
So
it's
gonna
look
at
the
database
and
see.
C
Oh
we've
got
five
lessons
loaded
and
there
and
they're
running
these.
You
know,
or
we've
got
five
running
lessons
and
they're
running
these
lesson.
Ids
with
you
know
these
names
and
whatever
other
whatever
other
data
we
want
to
export,
probably
building
on
the
work
that's
being
done
today,
but
also
probably
I,
would
imagine,
rethinking
the
the
metrics
that
we
do.
Export
cuz
I
didn't
really
put
a
ton
of
thought
into
that
it
just
kind
of
works.
C
The
problem
with
that,
of
course,
is
that
you
need
to
be
able
to
communicate
with
the
database
in
order
to
get
that
that
data
so
yeah.
What
Johannes
is
referencing
is
correct.
We
we
in
order
to
paralyze
this
work
effectively
everything
other
than
the
database
calls,
will
need
to
be
done
or
can
be
done,
but
the
database
calls
themselves.
C
I'm
not
sure
if
anybody,
if
anybody
saw
the
details
there
but
I
the
web
ssh-2
project
which
we're
gonna
be
cannibalizing,
is
not
the
right
word,
but
we're
not
gonna
be
using
the
whole
thing.
We're
just
gonna
using
a
single
part
of
it,
but
I
submitted
a
PR
to
enhance
the
security
of
that
project
and
it
was
accepted
within
like
a
day,
so
I'm
I
haven't
tried
contributing
to
guacamole,
but
I
can't
imagine
that
it
would
be
that
easy,
given
minutes
an
Apache
project.
So.
C
C
C
C
C
It's
just
it's
a
it's
a
very
bare-bones
reference
implementation
for
how
this
could
work,
because
you
know
extern
j/s
on
its
own,
doesn't
have
an
equivalent
to
what
we
currently
rely
on
for
for
proxying
connections,
which
is
actually
a
combination
of
guac
tea
and
the
guacamole
Tomcats
servlet,
which
is
in
itself
fairly
complex,
because
it
introduces
this
new
thing
called
the
guacamole
protocol,
which
I
never
really
understood.
Why
they
did
that.
But
that's
ok
anyway,
X
term
J
s
doesn't
have
anything
like
that
built
in.
So
what
webOS
this
h2
does.
C
Is
it
implements
its
own
as
a
node.js
application
using
a
library
called
socket
IO,
which
is
extremely
popular
for
real-time
applications,
especially
like
chat
applications,
pretty
sure
one
of
the
really
popular
electron
based
kind
of
doodads
uses.
It
is
I
mean
I'm,
I'm
thinking
flak,
but
that
might
be
too
much
I'm,
not
sure,
and
they
may
have
replaced
it
since
then
they
may
use
to
have
been
on
it.
C
In
any
case,
it's
just
a
really
fast,
real-time
service
kind
of
library
thing
and
what
they're
doing
is
they're
using
it
to
provide
WebSocket
connectivity
to
x-term
j/s
in
in
much
the
same
way
that
we're
doing
that
with
it
Wacom.
Only
today,
now
the
cool
thing
about
this
there's
a
few
there's,
actually
a
few
cool
cool
things
about
this,
the
current
now
this
isn't.
This
isn't
part.
This
isn't
mandatory
in
guacamole.
C
I
do
believe,
there's
a
way
to
use
WebSockets,
but
we're
currently
not
using
WebSockets,
which
means
that
literally
every
keystroke
is
its
own
HTTP
request,
and
it's-
and
you
can
imagine
that
it's
not
that
as
performance
of
that
is
not
ideal,
especially
on
mobile
connections,
which
we
have
had
demonstrated
to
us
by
myself.
I've
demonstrated
it
to
myself
him
and
our
web
developer
was
like
hey.
You
guys
know
this
really
sucks
right
now
right,
and
so
that
was
one
of
the
things
that
kicked
this
off
anyway.
C
In
general,
moving
to
WebSockets
is:
it
is
a
good
direction,
technically,
that's
possible
in
guacamole
but
I.
To
be
honest,
the
documentation
for
guacamole
in
that
area
is
a
little
bit
of
a
train
wreck
and
also
this
new
architecture
with
with
socket
IO,
makes
it
effectively
mandatory.
It's
just
the
way
that
it
works.
So,
if
you're
not
familiar
with
WebSockets,
you
kind
of
think
of
it
as
like,
a
as
like
a
stream
instead
of
instead
of
a
request,
you
know
an
RPC
kind
of
thing,
so
it's
a
kata.
C
C
It's
a
lot
more
efficient,
socket
IO
is
also
written
in
JavaScript
and
actually
the
whole
back-end
proxy
function
is
written
in
JavaScript
I
have
night
sweats
about
nodejs,
but
I.
Think
in
this
case
it's
actually
a
really
good
call,
because
it
makes
basically
the
end-to-end
solution.
One
language
which
I
think
is
really
good
for
maintainability
we've
actually
got
folks
that
are
that
are
interested
in
contributing
the
antidote
web.
C
That
today
you
the
whole
guacamole
stack
is
a
little
bit
of
a
black
box
and
that's
not
good,
because
that's
a
core
component
to
the
web
app,
so
we've
we've
got
to
make
it
more
accessible
to
folks
that
are
otherwise
just
pure
play
web
developers
anyway,
I'm
rambling.
Obviously,
the
whole
point
of
this
blueprint
was
to
put
all
of
my
rambling
words
into
coherence.
C
You
know
things
and
I
appreciate
the
feedback
that
I
have
gotten
so
Stephen
and
Olivia
both
commented,
like
I,
said
it
sounds
like
there's
pretty
good
support
for
the
change
in
general,
so
I'll,
you
know
assume
that
we're
we're
good
in
general
to
move
forward,
but
I,
don't
know
that
we're
gonna
be
able
to
do
that
in
the
next
week
or
two
in
terms
of
getting
our.
You
know
the
the
web
company
that
we're
working
with
so
I'll
keep
you
updated.
C
If
we
do
make
progress
there
and,
of
course,
if
there's
any
specific
questions
or
concerns
on
any
of
the
details
in
the
blueprint
you
know
you
can
you
can
you
can
provide
that
anytime?
You
know
now
that
none
of
this
is
set
in
stone,
even
if
we
do
start
the
work.
So
just
you
know
chime
in
and
let
us
know
anything.
A
E
I
was
looking
at
accepted
Jesus
the
other
day
it
looked
like
people
trying
to
use
that
library
and
I'm
afraid
there
could
be
another
gang
of
people
trying
to
reinvent
the
same
platform
as
we
are
doing
correctly.
But
I
guess
that's
our
fate
that
maybe
someone
is
doing
another
kind
of
cloud
based
learning
labs.
C
C
Yeah
I'm
sure
they
are
and
I
am
from
and
for
me
I
think
this.
This
change
is
actually
just
about
aligning
with
them.
I
think
I,
think
what
this
says
is
is
the
value
while,
while
having
a
web
browser
or
well
while
having
a
terminal
in
the
browser,
is
obviously
very
cool
and
certainly
helps
facilitate
the
you
know
the
goal
of
a
project
to
make
to
make
this
learning
a
lot
easier.
C
I,
don't
think
that's
our
unique
value,
proposition
I,
don't
think
it
ever
was
so
I
think
moving
to
X
term
j/s
just
makes
that
part
a
little
bit
more
of
a
commodity.
Because
again
the
thing
I
think
we're
optimizing
for
here.
I
think
the
thing
that
we
need
to
optimize
for
is
bringing
more
folks
on
to
be
able
to
maintain
the
platform.
We're
not
gonna
get
there.
If
we
have,
you
know
for
if
we're
effectively
doing
our
own
thing
for
very
little
reason,.
D
E
C
So
that
means,
if
you
change,
if
you,
if
there's
a
file
in
the
lesson
directory
and
you
change
it
in,
you
know
one
of
your
endpoints
while
you're
running
the
lesson,
then,
if
there's
another
lesson
that
has
the
the
same
lesson
or
if
there's
another
image
that
has
this
endpoint,
that
has
the
same
lesson
directory
cloned.
It
won't
actually
have
that
change,
because
they're
totally
isolated.
C
So
one
thing
that
we
were
talking
about
a
while
back
and
I
like
I,
said
I
think
this
is
I
think
this
will
be
the
thing
that
makes
it
much
more
palatable
for
us
to
have
an
IDE
within
each
lesson
is
building
effectively
a
scratch
directory.
So,
instead
of
instead
of
having
init
containers
that
clone
the
git
repo,
you
know
once
every
single
endpoint.
C
We
can
just
do
that
once
for
the
entire
lesson
and
they
all
just
mount
into
that
all
the
endpoints
just
mount
into
that,
and
that
way
you
can
and
that
way
you
can,
you
can
actually
have
you
know
an
IDE
as
it
as
a
tab
as
if
it
was
its
own
endpoint
and
then
and
then,
when
you
edit
files
within
that
environment.
It's
that's
live
for
all
of
the
other
endpoints,
so
any
like
utility
image.
It's
looking
at
that,
like
they're
all
they're
all
gonna
have
a
view
into
that
scratch.
Directory
I.
C
Can't
remember
I
think
I
drew
up
a
blueprint
on
that,
but
I
can
I
know
that
I
haven't
touched
it
in
probably
months,
but
that's
I
in
terms
of
when
you
use
when
you
brought
that
up.
I
was
like
oh
man.
We
definitely
need
to
get
that
scratch
directory
in
place.
Cuz
that
that's
gonna,
be
a
make-or-break
kind
of
thing.
I
think.
B
B
C
So
if
you're
playing
around
with
it,
you
run
into
issues
just
you
know,
please
let
me
know,
because
I'm
kind
of
hungry
for
folks
to
use
that
and
and
and
make
it
make
it
more
palatable.
It
could
be
that
so
it
could
be
that
the
way
I'm
doing
passing
is
just
not
great.
You
know:
I
only
had
a
few
use
cases
in
mind
when
I
built
that.
B
C
C
Maybe
yeah,
like
I,
said
it.
You
know
yeah
my
some
of
my
my
goal.
My
goal
with
that
is:
is
there
a
way?
Is
there
a
way
that
we
can
make
it
as
easy
as
possible
on
the
end
point
builder,
meaning
the
people
building
the
images
so
that
they
don't
have
to
do
what
you
did
with
you
know
doing
in
Gen,
X
and
doing
your
own
configuration?
Obviously
that
will
work,
but
it
would.
It
would
be
nice
if
that
wasn't
necessary.
C
On
the
other
hand,
I
also
don't
really
want
to
I,
like
the
fact
that
the
presentation
type
is
just
set
to
HTTP
and
everything
else
is
is
intuitive
like
I,
don't
I
don't
want
to
I,
don't
want
to
pass
a
bunch
of
you
know.
Ingenix
rewrite
rules
into
you
know
leak
it
into
the
lesson.
Definition:
abstraction
I,
don't
think
that
it
will
be
fun
for
anybody,
so
I
think
we've
just
got
to
figure
out
a
good,
a
good
UX.
For
that
you
know,
if
that's
possible,
if
we
can
keep
those
two
things
in
balance.
C
That
would
be.
That
would
be
ideal.
You
know
cuz,
then
then,
the
user.
The
lesson
the
person
that's
defining
the
lesson
doesn't
have
to
worry
about.
You
know
effectively
redirect
rules,
enemy,
layer,
I,
think
it's
equally
as
equally
as
cumbersome
if
they
have
to
manage
and
genex
rules
in
the
lesson
definition
as
it
is,
if
they
may
have
to
manage
it
in
their
own
image.
You
know
neither
neither
are
ideal
right,
so
yeah
I
think.
C
B
Provisioning
layer
using
ansible
and
I
was
curious
if
there
was
a
way-
and
we
don't
need
this
answered.
But
one
thing
I
was
trying
to
do
was
have
nginx
as
its
own
instance,
its
own
VM,
and
then
see
it
could
be
a
different
VM,
for
instance,
and
then
it
you
make
simple
instable
playbooks
to
change
the
variables.
Instead
of
writing
the
complex
rewrite
rule,
I
just
say:
I
want
to
access
yeah,
be
it
yes,
Ford's
idea.
D
D
D
D
So
at
this
point
our
game
plan
is
to
talk
with
the
folks
that
we
were
originally
working
with
in
terms
of
moving
forward
with
becoming
joint
sponsors
and
talking
with
them
to
see
what
they
would
need.
There
is
that
there
is
a
a
a
for-profit
option
as
well,
which
is
not
ideal,
but
if
it's,
if
it's
simply
a
question
of
having
you
know
a
place
to
park
money
and
that's
what
needs
to
happen
for
folks
to
feel
comfortable
with
with
fighting
on,
that
might
be
what
we
do
in
the
near
term.
A
Saw
some
of
their
wording
and
the
email
and
I
I
guess
I,
don't
understand
what
the
purpose
of
that
Conservancy
is
right.
It's
they're
becoming
what
I,
what
I
like
to
call
a
self
licking
ice
cream
cone.
They
don't
want
to
help
open-source
projects
unless
the
open-source
projects
are
serving
this
conservancies
interests.
That's.
D
E
B
A
I
I
understand
that
concept,
but
the
way
it
was
worded
was
was
not
framed
legally.
What's
like
you
know,
how
was
your
project
gonna
make
us
better?
Was
there
and
the
question
is:
how
does
any
project
make
the
Conservancy
better
its?
It
seems
arbitrary
and
subjective
what
that
means,
and
it's
I
I
was
just
not
a
fan
of
their
response.
That's
all
I
can't
imagine
how
we
would
detract
from
their
purpose.
A
That
I
guess
that's
it,
it's
all
I
got
you
know
they
they
get
funding
we
and
then
they
how
they
help
us
right.
I
mean
beyond
that
I'm
trying
to
understand
what
what
would
be
in
there,
how
how
we
could
not
be
in
their
interest,
barring
obvious
things
like
you
know,
we're
the
software
we're
making
is
unethical
or
something
so.