►
From YouTube: AMA with Phil Hughes - Web IDE - Frontend - 2019-07-18
Description
AMA docs for questions:
https://docs.google.com/document/d/1rKw8RE1y1Vq5KM402QzTVCriZPPHNabJhkcYLVrrhRE/edit?usp=sharing
A
So
I
have
my
first
one
and
I'm
gonna
voice
it
and
then
put
it
up
in
the
document.
So,
let's
see,
if
I
can
paste
it
I'm
like
all
right
so
Phil
I,
wonder
if
you
could
give
us
a
few
examples
of
things.
You
thought
we
should
do
with
the
web.
Id
code
base
that
we
never
got
around
to
do
yet
so
stuff
like
which
factors
or
stuff
bad
were
not
exactly
the
in
the
shape
that
you
wanted.
B
Is
the
biggest
issue
tomorrow
is
performance
like
handling
potentially
hundreds
of
thousands
of
files,
the
performance
figures,
good
several
old
man
we
did
have
web
workers
and
I
think
the
delay
between
actually
coming
from
the
web
worker
back
into
the
browser
because
make
it
look
like
the
performance,
but
even
though
technically
wasn't
like
the
browser's
didn't
really
crash.
If
it's
just,
it
was
like
a
delay
which
there
was
gonna
be
coming
back,
and
this
is
probably
the
biggest
thing
we
saw.
I've
always
wanted
to
talk
us
to
performance.
B
All
right,
I
always
always
wants
to
make
it
work
offline,
and
it's
just
like
this.
So
many
complicated
things
until
I
see
making
the
ID
work
offline,
but
I
like
never
really
costume,
even
try
to
do
it
and
I
did
always
bring
go
visit
him.
The
JavaScript
library
that
actually
brings
get
into
the
browser
I
always
wanted
to
look
into
using
that,
instead
of
actually
the
API
and
I
think
that
sort
of
created
more
problems.
B
A
B
Okay,
you
know,
sir,
we
did
have
web
workers
I,
think
you've
being
removed.
Now
the
main
idea
is
the
web
workers,
because,
my
famous
so
when
we
were
gonna
play
the
IIT,
we
request
a
list
of
files
and
it's
just
a
list
all
the
files,
just
a
flat
list
of
an
array
of
strings
the
represent
the
path
of
the
file.
So,
on
the
front
end,
we
then
take
that
path.
B
We
chop
it
into
like
folders
and
stuff
and
generate
that
on
the
front
end
and
the
smaller
projects,
it's
not
so
badly
than
they
play
local
blundered
files.
That
can
happen
quite
quickly.
It's
when
you
started
get
into
Caleb's
EE
style
projects
that
start
to
notice
and
eventually,
before
my
stuff's
getting
wet
so
I
moved
that
into
a
web
where
I
saw
at
like
the
performance
still
needs
to
be
good
in
the
web
worker,
because
you're
still
gonna
get
it
so
technically
slow.
B
It
doesn't
freeze
the
browser
and
see
if
that
was
moving
as
the
web
worker,
which
then
does
all
the
transform
itself
and
then
post
it
back
into
the
browser.
Now
the
problem
comes
with
gadget
performance.
There
is
that,
because
the
web
worker
just
passes
across
a
sense
you
just
Jason
when
it
gets
back
into
the
browser
view
then
has
to
add
it's
been
reactive.
B
If
you
saw
stuff
to
each
of
the
objects,
then
all
the
keys
and
stuff
to
the
silver
violence
problem
there
and
I
guess-
is
that
there's
still
the
delight
coming
back
from
the
way
we're
complaining.
It's
not
massive
issue,
I
think
Paul,
maybe
my
surrounds
it
by
self-improvement
algorithm
a
little
bit.
How
I
am
create
the
array
of
files
again,
this
is
still
eventually
there's
gonna,
be
some
project
I'll,
maybe
like
thousands
of
files
that
there's
still
going
to
be
the
problem.
C
B
Problem
yeah:
well
at
the
moment
we
don't
use
web
workers,
it
was
removed.
The
main
problem
is
the
because
we
need
so
much
data
pair
file
to
actually
run
the
web
IIT.
A
view
needs
to
create
watches
for
each
of
the
keys
in
the
object
and
sage
objects.
There's
like
a
thousand
five
hundred
keys.
You
then
start
to
go
more
and
more
pair
file
technically.
C
Like
removing
data
from
from
the
file
object
is
not
a
real
option
right
because
we'll
eventually,
we
will
still
be
adding
more
data.
So
how
do
you?
How
do
you
see
this
as
a
solution
like
split
the
data
into
chunks
and,
for
example,
preload
stuff
that
is
like
like
first
and
then
lazy
loading,
the
remaining
ones?
You
know
they
fight
the
performance
for
like
how
how
useful
that
and.
B
This
is
like
it's
not
gonna
happen,
but
rewrite
it
in
graft.
You
are.
The
problem
is
because
we're
using
the
rest
EPA
is
we
just
take
all
that
data
that
we
return
and
just
think
okay.
Well,
we
might
need
it
all
eventually
in
leadership,
in
the
object.
Maybe
we
don't
need
that
or
maybe
we
only
need
like
a
couple
of
keys
from
it.
Maybe
we
could
actually
just
check
what
keys
you
guys
if
the
problem
call.
This
is
the
fact
that
there's
so
much
code
to
actually
go
from
right.
How
do
we
determine
that?
D
C
What
do
you
think
that
mission
to
graph
QL
might
improve
the
things
right,
but
then
again
we
need
to
out
what
data.
Actually
we
need
and
it's
my
turn
now.
That's
me
actually
do
needle
back
in
the
same
problem
like
as
the
as
the
simplest
solution.
What
do
you
think
that,
for
example,
like
getting
and
parsing
the
data
in
chunks,
with
lazy,
like
lazy
processing
like
files
offer,
for
example,
one
teeth,
would
soul
the
thing
or
it's
like
it's
impossible
due
to
some
technical
things
and
I.
B
Guess
there
is
like
maybe
could
look
into
the
components
themselves
and
see
how
they
actually
render
stuff.
Well,
we're
not
going
to
get
around
the
memory
problem,
the
more
data
we
actually
store
on
the
front
end,
the
more
memory
unit,
we're
gonna
use,
yeah
yeah
I
mean
we
could
get
a
bit
clever
on
how
we
actually
render
the
components
and
look
ok.
Well
that
completes
my
usual
just
lazy
load
that
in
and
stuff
layer,
ok.
A
So
we
on
the
source
code,
part
on
on
the
dips.
We
did
and
tackle
the
similar
problem,
where
we're
loading
the
whole
dips
and
we're
loading
them
incremental
e.
Then
we'll
just
do
this,
like
the
next
step
of
doing
sort
of
like
a
first
request
to
craft
well
to
get
like
just
a
list
of
the
file.
So
we
can
render
the
tree
in
place
with
the
minimum
metadata,
and
then
we
will
fetch
in
chunks
the
the
full
metadata
that
we
that
we
need
in
the
future,
for
whatever
operations
we
need
to.
B
The
way
we
could
do
it
so
now,
if
our
graphical
TM
print
the
returns,
the
list
of
files
is
fast
because
just
return
the
strings
and
but
then
it
is
it
useful
to
render
that
fast.
Like
do
I
do
the
web
IDs
you're
gonna.
Let
the
file,
if
we
just
render
the
tree
be,
cannot
see
open
a
file.
Is
it
any
news.
A
So
yeah,
that's
that's
the
thing
that
we
have
to
figure
out
exactly
what
a
use
case
were
there
and
then
like,
like
Jill,
was
saying
about
the
removal
of
the
web
worker
that
was
done
in
the
context
of
having
the
fastest
bootstrap
when
web
ID,
so
that
it
could
replace
the
ice
editor.
The
old
editor
file
edit.
So
that
was
the
most
important
thing
is.
If
you
get
a
URL,
then
we
need
to
load
the
file
first
and
then
that
the
tree
can
be
loaded
likeness
as
a
second
class
object.
A
A
E
B
B
E
E
E
E
It's
assumed
a
similar
problem
to
like
the
red
ring
of
video
game
where
you
don't
render
the
entire
world.
You
only
really
thing
the
user
could
see.
I
could
say
that
the
abstract
I
have
no
idea
how
we're
gonna
do
it.
I
just
wanted
to
make
sure
I
understood
it
Bobby.
If
you
could
do
it
open
source
and
that'd
be
kind.
C
C
Yeah
we
discussed
this
I
think
last
week
or
a
couple
of
weeks.
You
know
we
feel
definitely
right
now.
We
have
like
tonight
my
preference,
which
many
erase
stored
on
state
with
with
different
sorts
of
data,
and
the
problem
is
that
not
always
this
data
is
vaguely
updated,
so
we
have
open
files.
We
have
saved
open
files,
state
staged
file,
states
changed
files,
state
entries.
C
This
issue
with
one
of
the
deliverables
when,
when
obeying
entries,
doesn't
abate,
all
this
or
all
those
array
so
I
had
this
idea
of
like
replacing
all
these
open
file
staged
files
changed
files
with
gathers
that
would
get
computed
once
the
state
entries
get
updated
so,
but,
but
you
feel
mentioned
that
you
had
some
ideas
in
this
regard.
So
I
would
like
to
get
those
ideas
right,
yeah.
B
That
was
I
originally
had
this
idea
and
I
might
give
it
quicker
and
I
probably
didn't
like
the
worst
way
possible.
I
just
didn't
like
it
at
all.
The
probably
they'd
still
be
in
Stata,
and
she
would
be
a
flower
checked
with
keys
in
the
actual
file
path
or
the
folds
path
and
then,
for
example,
open
files.
It
will
just
hold
a
string
that
would
then
link
to
steak
dying
trees.
Is
that
that's
all
what
we
were
saying?
Isn't
it,
then
you
would
have
they
compete?
B
C
E
C
The
sort
of
filtering
of
state
entries
he
might
get
slow,
though,
but
those
things
like
open
final
stage
files
and
change
files
performance
for
those
is
not
critical,
really
like.
We
don't
need
them
to
happen
instantly,
but
I
also
thought
about
again
as
well
like
keeping
them
as
flat
flat
arrays
of
just
IDs
of
the
objects
from
state
entries.
So
whenever
we
need
to
retrieve
an
object,
for
example
like
an
open
file,
we
have
the
idea
and
we
get
work
like
path
right.
Yeah.
B
C
Going
to
be
path,
and
then
we
would
retrieve
this
from
my
state
interest
by
the
path,
but
but
again,
I
think
like
it's
one
step
too
many
like
before
would
have
open
files,
for
example,
yet
computed
every
time
we
update
state
entries,
maybe
maybe
this
this
would
work.
What
do
you
think,
like
just
filtering
of
the
off
state
entries.
F
For
example,
somehow
yes
may
I
jump
in
I
when
I
first
saw
bye-bye.
Gee
I
had
exactly
the
same
idea,
so
I
played
with
it
like
some
time
ago,
and
the
problem
is
a
digit
itself.
I
mean
you
can
have
the
file
simultaneously
being
staged
it
and
being
modified,
and
this
is
different
version
of
file
see
since
we
have
like
free
steps
like
preview
original
version
in
each
stage
at
version
and
edited
version
which
is
not
yet
stage
yet
and
having
this
in.
C
For
for
every
l,
dream
state
entries,
I
think
every
entry
there
has
content
and
has
raw
content
parameters.
So
raw
content
contains
the
original
data
that
has
been
like
from
from,
like
the
last
version
story
and
get,
and
the
content
is
the
current
one.
So
technically
the
edited
version
I
think
we
do
have
this
information
stored
on
state.
F
Entries
don't
and
we
have
in
case
we
are
going
for
sorry
for
jump
again.
I
just
play
it
with
this
sound
sound
weekend
and
if,
in
that
case,
you'll
have
the
third
version.
We
need
to
start
it's
a
stage
at
version
because
we
they
could
be
free,
different
versions,
you
mean
staged
version,
can
be
different
from
a
modified
version.
Yeah
it
works,
so
you
can
stage
it
after
that
you
can
change
it
and
you
in
that
case
you
will
have
a
difference
between
stages
version
and.
B
Mean
we
could
still
totally,
we
could
still
do
it
that
way,
so
stay
dying,
cheese
to
be
a
flat
object
with
the
keys
is
the
path
and
and
then
the
open
files.
We
just
add
in
that
path:
change
files.
We
add
in
that
path,
just
as
a
string,
that
line
looks
up
interests
and
cheese,
but
then,
when
we
want
to
move
into
staged
and
the
stage
way,
we'll
just
move
the
actual
object
and
copied.
At
that
point,.
F
Phil
I
have
a
different
question:
I'm
totally
unrelated
to
create
I'm
from
manage
I'm
just
curious,
dissolve,
FIA
and
I
dig
into
the
code.
Did
you
consider
it
some
time
ago
it
early
start
or
maybe
did
you
consider
it
night
now
taking
some
like
existing
JavaScript
implementation
to
maintain
all
the
things?
Yes,
so
that
was
the
best
Princeton
we
raise
them.
Yeah.
F
No
obviously,
we
do
not
want
to
have
the
full
repository
clumped
into
the
memory,
but
I
mean
like
maintaining
the
small
copy
of
this.
Solving
will
have
right
now.
The
code
replicates
ajit,
storing
model,
a
lot
like
this
having
staged
files
and
so
on
and
so
on.
So
it's
not
like,
let's
do
direct
push
via
the
ggs
or
whatever,
but
just
using
the
part
of
their
code
to
replicate
all
this
behavior
which
they
are
trying
to
emulate
and
store
know
as
I
never
looked.
B
A
F
Where
you
could
one
more
question
and
that's
a
question
I'm
just
curious
because,
like
I'm
exploring
the
same
opportunities
for
my
simple
project,
have
you
played
a
heavy
tried
to
look
into
the
web?
Possibly
approach?
Because
right
now
we
are
doing
like
I
said
not
not
emulating
center,
not
cloning,
the
entire
repository,
but
our
other
supporting
get
labs
pretty
modern
and
that
may
injunction
with
indexdb
even
may
help
us
to
have
some
kind
of
a
flying
back.
Yeah.
B
A
daily
part
webassembly
that
looked
into
is
changing
the
syntax
highlighting,
but
the
memory
uses
kind
of
a
rubric
version.
There's
a
webassembly
region.
Is
it
the
c
library
to
do
the
since
exam?
That's
literally
any
part
would
be
somebody
looked
at
or
it
could
be
interest.
The
problem
that
comes
is
how
many
of
those
it
could
actually
write
the
web
assembly
code.
We
be
using
the
library
that
I
guess,
if
I
guess,
when
we
say
well,
if
we
use
a
library,
the
problem
is,
if
we
need
to
extend
it
in
some
way.
F
A
So
thanks
thanks
yeah,
there's,
definitely
more
of
us,
this
guy
and
me
as
well,
but
I
get
your
point
so
I
had
one
question
still
I
wanted
to
see
what
he
thought-
and
this
is
particularly
interesting
because
at
the
moment
we're,
as
you
know,
you
were
considering
fear
and
other
alternatives
for
the
actual
implementation
of
web
IDE
and
I
wonder.
Is
there
anything
that
in
the
past,
when
you
were
I
mean
you,
you
were
right
there
from
the
start
when
you
work
on
the
web
IDE.
A
B
So
from
the
beginning
was
the
right
way
you
started
with
just
like
multi
file
editor.
There
was
a
bit
hockey
a
bit
pull
me
a
bit,
oh
great,
and
let
me
move
it
into
the
web.
Id
moniker
itself
is
fine,
a
thing
in
the
browser.
It's
not
gonna,
be
great
anyway,
and
the
thing
that
always
interests
me
is
to
code
some
hours
from
code.
Some
bugs
have
got
be
asked
in
the
browser
for
code
sandbox.
That
was
always
something
I
thought
we
should
maybe
look
at,
but
if
you
can
fluent
LVS
code
will
interrogate
loved.
B
Carry
work
without
the
ordinate,
because
I
think
code
sandbox
works
in
the
iPod,
some
pretty
sure.
How
does
it
think
they've
done
a
fan
on
top
of
him
yeah.
It
is
something
specific
to
Monaco,
I
think
so.
On
the
fairway
fire.
You
cut
yes,
code,
woman,
but
then
I
said
like
I
steer,
uses
PS
codes,
not
Monaco,
so
maybe
it'd
be
salt.
There's
on
my
problem.
We're
using
theer's,
there's
gonna
be
a
docker
container
running
for
every
session
ID
web
ID.
Is
that
gonna
be
scalable?
I!
Guess!
That's!
That's!
Like
friends,.
B
A
And
there's
more
changes
that
require
that,
but
we're
definitely
taking
very
seriously
for
the
proof-of-concept
into
and
assessing
all
the
work
that
is
involved
in
getting
that
at
the
speed,
but
guys
on
the
calling
is
definitely
doing
to
share
about
anything
about
that.
But
just
to
pick
your
brain
on
that
guy
I.
D
Mean
I'll
turn
my
camera
to
I'm
more
curious
about
that
I.
Think
the
doctor
concerned
is
a
very
real
concern.
I
think
like
implementation
of
running
any
of
these
ideas
that
are
much
more
full-featured
than
the
current
version
of
the
web.
Id
like
hash
I
mean
at
some
point
you
have
to
attach
compute
to
them.
That's
why
there's
been
spending
milestones
and
milestones,
trying
to
like
attached
to
CI
runners
and
I
guess
in
the
way
the
web
ID
works
in
some
super
hacky
way
that
kind
of
just
kids,
it's
there.
D
It
seems
to
be
its
legacy,
so
yeah
I'm
curious.
You
know
like
if
you
think
about
that
or
if
you
think
about
like
the
architecture,
decisions
that
were
made
kind
of
historically
and
then
you
know
I
think
you're
right
like
when
it
was
implemented.
It
was
probably
the
right
path
given
what
it
was
trying
to
accomplish.
But
seeing
where
it's
trying
to
go
now,
like
I
mean,
would
you
rethink
I
mean
I
think
looking
at
it
from
the
lens
of
like
where
it's
trying
to
go
now?
B
I
guess
the
directions
going
to
make
sense
these
somebody's
already
created
idea.
Yes,
their
make
sense.
I
just
just
worried
that
the
touching
docker
containers
know
something
I
know.
So
it
worries
me
a
little
bit.
I
mean
it
might
be
fine.
It's
you
strum
it'll
ends
up
being
okay
and
I.
Let
them
being
great
I,
just
wonder
if
it
could
be
achieved
ourselves
and
try
and
learn
it
for
them.
If
someone's
already
done
it,
why
did
she
stay
away?
Reading
reputable
source.
A
B
Actually
could
be
looked
at
in
a
while,
I
guess,
there's
just
so
much
code
in
there
be
nice
like
Champa,
dad
all
the
way
it
wouldn't
be.
Nice
I
mean
it
looks
nice
mister
through
it
when
it
started.
It
was
just
this
massive
Chara
script
object,
we're
all
every
single
bit
a
was
added
into
it,
and
the
sailors
mutates
and
compared
into
as
me,
take
some
JavaScript
files.
B
A
B
D
B
Yeah
there
was
nothing
really
that
I
came
up
with
the
word
and
was
something
no
Don.
I
always
wanted
to
get
some
time
scheduled
to
get
the
the
syntax
Alex.
It
highlights
an
improved
there's
a
major
crest
open,
but
it
was
just
the
test
would
fail
and
I
just
didn't
really
get
the
time
twisting
go
back
and
fix
them
and
they
sort
of
it
fixed
kind.
A
lot
of
things
that
gave
us
you
support
it,
improve
the
JavaScript
syntax
highlight
some
pretty
shame,
prove
that
Ruby
highlight
them.
B
D
B
Yeah
yeah,
but
the
reason
M
website,
if
I
bring
good
name
escapes
me,
doesn't
make
us
from
with
it
actually
doesn't,
and
that
is
the
C
library
him
yeah
I
guess
matching
that
toward
the
chronic
get
lab
since
X
has
always
been
an
issue
because
they
get
lab
code,
renderer
uses
Rouge
or
is
Monica.
Doesn't
so
they
get
your
color
differences?
Don't
match,
try
to
get
them
similar,
but
it's
just
it's
never
gonna
match.
Unless
you
use
the
C
library.
A
So
so
I
don't
have
a
question,
particularly
written,
but
I'll
just
wing
it.
So
one
of
the
things
that
were
looking
into
the
future
of
the
web
IDE
is
a
possibility
or
possible
scenario
that
we're
gonna
have
to
re-implement
the
integration
with
kit
lab
using.
You
know
a
plug-in
system,
bonding
architecture
using
the
sidebars
that
sort
of
thing
tweaking
the
UI
as
much
as
possible
as
a
plugin
can
do
given
that
you've
worked
on
most
of
the
basic
initial
integration
there.
A
B
B
A
B
Didn't
take
away
that
any
tricky
ones
Dave
Farra
they
existed
like
it,
wasn't
something
that
was
new,
the
API
to
commitments,
multiple
files
or
the
existed,
so
we're
just
hooking
into
that.
Not
really
any
complicated
like
the
easiest
of
them
all
was
integrating
with
code
sandbox.
It
was
just
like
I
bind
to
code.
Maybe
then
we
had
JavaScript
projects
within
the
browser,
well,
I'd,
like
Terence
and
how
it
works.
B
A
Just
a
genius
kid:
so
can
you
talk
a
little
bit
more
about
that
collaboration,
especially
about
integrating
a
third-party
word
tool
thing
from
outside
into
the
github
world?
What
was
the
lesson
well,
should
we
have
done
differently,
especially
about
collaborating
with
someone
who's,
not
full-time
working
for
us,
and
sometimes
we
get
lost
in
generally.
B
We
had
a
lot
of
time
when
we
were
doing
it.
We
always
I
call
him
every
week.
So
that
means
it
would
be
nice
try
to
discuss
these
problems
that
we
had
because
he
went
scheduled.
We
didn't
really
come
up
with
the
problems
that
we
do
now.
I've
host,
know
ourselves
and
so
I
guess,
because
we're
not
booking
them
every
week.
It's
not
a
priority
for
him.
So
you're
saying
that
having
weekly
calls
will
benefit
the
we
used
to
we're
fast,
but
I.
B
B
D
Is
that
something
like
if,
if
say,
the
IDE
changes
entirely
right
to
the
some
other
BS
code
based
IDE
is
like
that
a
is
that
an
integration
or
a
or
a
feature
that
you
see
like
staying
or
having
to
be
rebuilt
or
like
I,
mean
right
in
a
more
full-featured
IDE
with
compute
in
theories
like
that
could
run
differently.
I
guess
right,
because
we're
I.
B
Guess
the
compute
side
of
things
anyway,
to
be
better
for
the
backend
side
of
things
that
you've
written
the
actual
act
for
JavaScript
projects.
It
can
just
be
run
in
the
browser
free
code.
Sandbox.
Do
we
need
to
talk
to
a
docker
containers
right,
see
you
on
the
JavaScript
okay?
So
if
you
run
in
the
browser
more
than
stuff
like
a
rails,
app
that
makes
sense
in
the
backend
run,
something
like
Knox
right.
It
means
to
say
ever
running
that
makes
sense
and
I
think
that's
how
I
cope
somewhat
to
do
it
now
any
said.
B
B
Let's
feel
that
we
used
with
work
for
client-side
JavaScript
wiki
to
build
the
backend
stuff
ourselves
so
well,
III
didn't
get
that
bill.
Can
you
yeah
the
code?
The
library
is
called
smoosh
pack
or
Sam
pack
or
some
clever,
and
that
can
compile
the
front-end
client
code
in
the
browser,
but
it
doesn't
handle
any
back-end
stuff,
which
I
mean
I
wouldn't
meeting.
So
we
need
to
tell
that
ourselves.
Yeah.
A
B
I
mean
the
Kyrie
Tati,
it's
great.
It
is
like
super
easy.
It
just
needs
to
create
this
new
class
and
then
it
just
it
just
asks
for
the
files
that
it
needs.
So
every
time
it
sees
an
import
it'll
just
ask
folks:
hey
listen,
can
I
just
have
this
index
dot
JSON
and
we
can
get
the
content
to
it.
There's
not
complicated
sounds.
A
A
A
A
So
since
you
were
involved,
I'm
asking
this
out
of
the
blue,
but
do
you
feel
like
making
the
web
idea
stand
alone
app
itself
so
that
it
could
be
embedded
wherever
whether
to
run
iframe
or
something
like
that
is
something
pushing
too
much?
Was
it
would
be
doable?
What
would
be
the
your
concerns
there?
A
B
Don't
know
I,
guess
it
the
current
way.
It's
just
not
gonna
work.
At
the
Cohen
case,
that's
been
rewritten,
hey,
I,
think
when
I
said,
Tim
originally
moved
on
to
attend
Ruth,
there's
a
hot
dog,
some
/id,
whoever
is
in
the
idea
that
eventually
would
have
like
personal
and
projects.
So
you
get
instead
of
not
being
projects
to
be
like
order
like
piss
to
my
area,
I
guess
like
snippets
of
way,
you
could
just
write
code
and
save
it
and
I
never
came
about.
So
it
seems
to
the
way
that
we
have
this
like
stupid.
B
A
Yeah
definitely
I
think
we
got
around
back
with
the
code,
suggestions
and
stuff
like
that
or
merge
requests
where
you
can
just
like
embed
those
suggestions
there,
but
not
the
issues
themselves.
I
I
mean
I
could
see
some
some
benefits
of
having
something
like
that
for
educational
purposes
or
articles
where
you
have
full
code
pen
time
that
you
why?
A
But
it's
actually
web
IDE
running
and
it's
got
a
completely
different
use
case,
because
there
wouldn't
be
much
integration
with
get
lab
or
or
if
they
would
probably
be
a
messy
login
workflow,
but
I'm
yeah
I
think
we'll
probably
get
around
that
we'd
snippets
and
they've
been
probably
like
embeddable
snippets,
and
then
we
can
just
have
the
snippets
be
more
interactive
and
and
inherit
some
of
the
web
IDE
features.
No,
let
me
Carrie
I
heard.
A
So
alright
I'll
put
that
on
the
chopper
shopping
list
and
then
get
them
no,
but
but
definitely
there's
a
lot
of
use
cases
that
we
considered
so
one
of
course,
the
use
case
for
the
main
use
case
for
the
web
IDE.
Just
like
you
want
to
do
a
quick
change.
You
don't
check
out
the
code,
but
there's
a
lot
of
other
use
cases
that
we
can
definitely
benefit
from
from
those
advanced
features,
but
not
really
specific.
That
I
want
to
pick
your
brain
on
at
this
point.
A
C
A
A
A
Mean
we
don't
know,
we
don't
have
to
extend
this
on
purpose,
just
just
to
make
the
time
and
then
she
will
be
around
for
a
couple
more
days
yet
so
we
can
definitely
if
we
need
to
just
reach
out
to
him
or
me,
and
we
can
hop
on
a
call
and
record
that
as
well
and
make
that
available.
So
if
you
think
about
anything
else
that
you
haven't
asked
here,
and
he
thought
about
still
and
still
a
couple
of
days
to
do
that,
so
still
do
you
wanna
these
work.
A
C
A
A
B
Stuff,
this
multi
file
that
Stan
and
I
got
caught
Jacob
shot.
His
original
content
manager
lead
person.
It
was
his
almighty
to
allow
people
to
edit
multiple
files
in
the
browser.
So
if
I
happen
to
clear
in
the
project
they
found
that
the
Quelaag
there
was
maybe
four
people
who
win
that
tech
savvy.
He
didn't
really
know
how
to
get
clear
and
stuff
like
that
and
create
commit.
It
was
nice,
the
browser's,
a
pathetic
multiple
files.
They
could
commit
them.
You
don't
have
to
worry
about
anything
and
so
start
of
this
most
foul
editor
decodes.
B
It
was
done
in
a
month.
There
was
the
major
curse
was
huge.
There
was
so
many
comments
on
it.
It
was
like
impossible
I'm,
pretty
sure
you
can't
even
load
the
page
right
now
yeah
it
was
kind
of
buggy,
but
it
awake
and
it's
all
showed
there
was
possible
to
do
and
make
quite
a
lot
of
people
so
found
the
project.
Interesting
I
didn't
think
at
that
point.
It
got
picked
up
by
any
project
manager.
It
was
still
just
Jacobs,
little
Fink
and
then
I
sort
of
came
in
and
decided.
Okay
well
I'll.
B
Just
rewrite
all
the
way,
the
stages
miles
into
the
UX
and
it'll
change,
some
of
the
design
and
I
think
at
that
point.
Eventually,
the
product
mines
picked
it
up,
but
it
became
the
IDE,
but
it
was
named.
The
idea
is
so
originally
it
was
gonna
take
over
the
fire.
Listen,
it
was
all
gonna,
be
the
fire.
Listen
the
won't
be
like
this
separate
boot
to
edit
you
just
click
a
file,
it
would
load.
You
could
change
it
there.
You
could
commit
it
and
yeah.
So
then
the
project
rise
to
go.
B
We
got
some
insiders
in
to
him.
Previously
was
just
taken.
One
of
the
designers
on
the
side
just
come
up
with
some
design
where
it
should
look
like
and
someone's
designers
and
programmers
came
so
I
get
the
time
schedule
to
it.
It
starts
to
look
a
bit.
Nice
start
to
work.
A
bit
nicer,
I
went
to
became
on
it
to
its
separate
Ruth.
At
some
point
in
all
of
this,
it
was
decided
it
would
be
an
ear
and
you
feature
so
for,
like
maybe
a
release
off.
B
It
was
mutants
the
any
code,
and
that
was
mentioned
decisive.
We
were
fighting
to
see,
I
mean
it
kind
of
made.
My
stuff
look
good,
because
my
stats
went
from
Lake
added,
like
1
mm
or
whatever,
like
10,000
lines
into
EA,
and
then
it
went
back.
It's
like
added
10,000
lines
into
C,
so
it
made
like
my
contribution
graphical
kind
of
good,
which
I
was
okay
with
that
and
it's
not
a
competition
but
yeah
I
get
it
yeah.
I
really
feel
like
a
thing.
B
Opposition
says,
because
the
competitor,
but
guess
I'll
give
my
stats
on
who
ID
how
many
lines
in
two
files
and
stuff.
So
it
makes
my
stuff
that
kinda
goes
like
plus
200,000
me
lines
of
code
and
if,
thanks
to
me
with
the
ID
from
C
into
e
backing
from
E
to
seeing
and
yeah.
So
a
came.
This
idea,
the
that
chrome
point
you
could
only
have
changed
files.
There
was
no
such
thing
as
stage
trials,
so
I
think
it
was
maybe
to
meet.
B
E
B
Don't
think
that
gets
us
to
where
we
are
now
we
got
code
sandbox
the
terminal.
The
terminal
was
something
we
always
wanted
to
do
and
we
had
designed
the
way
back
when
it
should
the
multi
file
Edison.
Obviously,
because
we
had
no
back
and
engineers
assigned
to
him.
There
was
no
way
we
could
get
worked
on
front
end
early
Jim.
B
A
B
A
B
Monica
itself
in
can
get
kind
of
our
hands
every
time
you
give
it
file.
So
you
know
it's
piece
of
the
ID
Monica.
Let's
see
you
give
it
a
file
and
British
that
content
and
pretty
sure
you
give
it
a
path
based
on
that
path.
It
will
work
out
the
actuals
since
X
highlight
them
based
on
the
file
extension
and
every
time
you
pass.
B
This
Monica
creates
a
new
model
in
itself
and
was
holding
that
in
storage,
and
so
even
if
you
opened
too
close
to
file,
it
would
create
a
new
model
for
that
same
file.
So
we
ended
up
creating
this
model
manager
type
thing
that
would
cash
it
on
our
side
and
we
could
just
patch
it
into
Monica.
If
we
would
you
open
the
file
yeah
what
has
been
missing
to
the
original
day
or.
A
A
You,
okay,
I,
think
that
basically
does
it
so
again.
I'll
repeat:
if
you
have
any
questions
reach
out
fill
this
last
day,
it's
gonna
be
July,
31st,
so
definitely
reach
out
to
him.
Until
then,
you
need
to
hop
on
a
call
to
record
it
and
so
that
we
can
capture
those
conversations.
I
can
I
can
get
that
set
up
and
for
those
that
are
interested
on
the
topics
of
source
code
will
be
having
one
other
one.