►
From YouTube: Cloud Foundry Community Advisory Board [August 2017]
Description
Video from the monthly Cloud Foundry Community Advisory Board (CAB) meeting. Learn more at https://www.cloudfoundry.org/.
A
B
A
C
D
A
A
B
A
B
A
A
A
If
we
disappear
from
SF,
you
know
what
happened
is
just
electricity,
maybe
smoking
whole
scenario
yeah
good,
but
let's
open
okay,
so
welcome
everybody.
This
is
August
cab
call,
there's
quite
a
bit
of
things
to
cover.
We
always
start
with
cff
update
foundation,
update
I
know.
Chip
should
be
on
a
call.
Alright,
yeah.
B
No
problem,
two
things
I
want
to
point
out
for
those
of
you
that
are
users
of
Cloud
Foundry.
We
have
a
user
survey
running
through
almost
to
the
end
of
the
month.
Take
a
look
in
the
notice.
I
put
a
link
in
there
at
the
awesomes.
You
could
fill
it
out.
We
want
to
know
who
all
is
using
the
technology
and
then
be
able
to
share
some
interesting
information
back
to
everybody.
You
know
in
aggregate
about
all
of
the
widespread
use
that
we
have
the
other
thing
of
course,
summit
in
Basel
Switzerland,
coming
up.
B
If
you're
on
CF
dev
or
the
CF
flash
mailing
list,
you
should
have
seen
a
contributor
code.
If
you
contributed
anything
to
the
projects,
as
usual
feel
free
to
use
that
to
register
the
first
wave
of
talk,
announcements
of
artists
had
gone
out.
I
put
that
link
in
there
as
well.
There
are
more
coming
so
you
know
really
really
amazing
talk
submissions.
Yet
again
it's
going
to
be
a
great
program,
we're
also
going
to
change
up
the
the
mainstage
quite
a
bit.
It's
going
to
be
a
lot
more
technical,
a
lot
more
demos.
B
A
B
A
A
The
other
thing
we're
going
to
cover
our
sort
of
project
highlights
the
corporate
the
core
PMC's
you.
We
can't
make
it
to
lets
cover
boss
first
because
Dimitri
here
and
he
has
a
contract.
So
DK
give
us
any
I.
Like
you,
wanna
mention
for
Bosch
I
know,
there's
a
DNS
thingy
going
on
hey
yeah!
We
can
rock
so.
C
Yeah
we've
just
recently
shipped
boards
263.
We
had
to
fill
in
the
release
notes,
but
we've
gotten
one
of
the
you
know
long-awaited
DNS
features
that
we
needed
to
do.
This
was
a
DNS
feature
for
being
able
to
point
to
an
entire
link
right
so
covers
all
of
the
instances
inside,
as
opposed
to
what
was
previously
possible
was
just
to
have
a
DNS
entry
for
for
individual
instances.
So
that's
a
one
of
the
last
remaining
steps
for
DNS
to
kind
of
be
a
consumable
fully
by
by
CF.
C
Aside
from
that,
we'll
continue
to
work
on
director
having
ability
to
run
co-located
errands.
So
that's
not
something
that
the
thing
is.
He
have
documented
young
working
on
that
right
now,
but
now
with
two
six
three,
you
should
be
able
to
place
an
errand
on
to
any
existing
service
lifecycle
instance
and
should
be
able
to
run
the
errand
without
having
to
pay
the
cost
for
spinning
up
and
spinning
down
the
machine.
And,
of
course
this
is,
this
kind
of
a
functionality
is
meant
to
support.
C
Maybe
your
errant
can
trigger
some
kind
of
a
you
know,
API
call-
or
maybe
it
can.
You
know,
write
down
some
debug
information
to
walk
files
or
do
something
else
like
promote
a
master
from
you
know
sleep,
someone
right,
so
it's
a
hopefully
opens
up
a
few
use
cases
that
were
impossible
before
them
should
be
a
lot
faster.
Now,
of
course,
the
previous
air
and
functionality
still
available
where
it's
still
life
cycle,
air
and
machines
will
continue
to
be
available
for
upcoming
features.
We
have
some
functionality
that
we
call
internally
hot
swap
over
here.
C
This
is
to
allow
to
quickly
migrate
from
one
machine
to
another
while
doing
a
stem-cell
updates
or
just
recreate.
So
this
is
each
instance,
may
potentially
have
a
second
VM
during
the
update
scenario
such
that
the
downtime
window
for
a
service
to
when
it's
stopped
and
then,
when
it
started
on
the
net
on
the
new
machine
is
smaller
and
hopefully
that
will
also
improve
some
of
the
speed
of
the
deploys,
because
we
would
be
able
to
create
some
bulk
VMs
upfront
and
of
course
we
can
do
a
deletion
later.
A
A
B
B
One
thing
that
we're
doing
on
diego
is
starting
to
work
with
kathy
to
support
rolling
app
updates.
So
we've
been
doing
some
spiking
and
exploration
on
the
diego
side
about
what
api
is.
We
would
want
to
introduce
to
support
that
to
take
as
much
of
the
burden
as
possible
off
of
cloud
controller
in
terms
of
managing
the
instance,
by
instance,
details
of
a
rolling
update
Thrax.
So
that's
that's
pretty
exciting.
That
should
be
fun.
Work
for
the
Diego
team.
Preserver
see
some
other
runtime
folks
on
the
line
too.
D
The
latest
capi
release
I,
believe,
has
a
full
support
for
multi,
go
packs
in
the
v3
API.
There's
no
way
to
use
the
C
up
CLI
to
push
on
app
and
specify
multiple
impacts
that
you
can
do
it.
You
know
in
JSON,
CSCL
I
work
because
I
think
on
the
way
to
talk
to
that
team,
but
I
think
it's
been
well
when
I
hear
and.
D
B
A
I
want
to
thank
the
anonymous
Wolverine.
That's
keeping
really
that's
fantastic!
Thank
you
all
right.
Remember
you
can
post
questions
on
the
cab
cause
you
can't
see,
or
you
want
to.
You
know
like
just
have
a
record
of
it
and
we
can
address
it
now
or
later,
depending
on
what
the
question
is.
Okay,
so
I
guess
this
is
this:
is
it
for
run
time?
Eric
usually
stays.
So
if
you
have
more
questions
later
on,
you
can
ask
him
and
Stephen
will
be
presenting.
A
So
you
can
definitely
ask
him
also
so
for
for
extensions,
because
today's
call
is
going
to
be
covering
pretty
much
two
things
of
extensions.
Then
you
just
mentioned
one
thing
and
then
we'll
just
get
to
the
presentation
is
that
the
bbr
and
I
think
he
raises
here
accepted
it's
an
incubating
project.
Now,
I
would
highly
suggest
you
to
take
a
look
at
it.
It's
it's
called
Bosh,
backup
and
restore,
and
it
allows
you
to
backup
your
entire
Cloud
Foundry
and
also
restore
it.
So
it's
pretty
cool,
give
it
give
it
a
I,
guess
a
try.
B
B
The
world
must
be
curbing
or
it
can
it
be
that
it
can
support
backing
up
for
real
databases
like
PostgreSQL,
for
example,
where
you
can
do
it
well
for
consistency,
for
example,
you
have
to
stop
the
CCTV
from
mutating,
while
you're
taking
a
backup
of
other
databases,
so
there
will
be.
There
is
a
requirement
for
downtime,
but
those
I
know
those
teams
are
working
hard
to
minimize
that
downtime.
Looking
at
things
like
a
read-only
mode.
A
Cool
all
right,
any
other
questions
with
heroes
on
Bush
backup/restore,
no,
okay,
I,
don't
know
if
anybody
else
wants
to
mention
about
extensions-
will
eventually
start
having
more
these
sort
of
updates,
since
there
are
a
bunch
of
projects
incubating,
but
we'll
talk
about
this
one.
So
unless
you
have
any
other
questions,
let
me
just
pass
the
mic
to
Steven
Devine
he's
a
pivot,
I
guess
a
new
pivot
relatively
in
Michigan,
and
he's
gonna
talk
to
us
about
no
he's
been
there
forever.
Sorry,
my
the
first
time
I
meet
him
anyway.
B
D
Yeah,
so
this
is
pretty
quick
demo.
Cf
local
is
a
CF
CLI
plugin
that
sort
of
acts
as
a
docker
translation
layer,
so
it
can
take
applications
in
CS
and
pull
them
down
and
run
them
locally
in
darker
complexions
that
you
know
locally
or
whatever,
backup
to
cloud
foundry
as
droplet,
and
it
can
stage
applications
locally
to
so
let
me
share
my
screen.
I
don't
run
through
a
quick
demo.
A
D
D
It's
pulling
the
droplet
in
the
environment,
so
you
can
see.
I
now
have
a
local
channel
and
a
spring
music
that
droplet
fun.
So
if
I
check
out
that
local
you
know
file,
you
can
see
it
pulled
down
the
environment
of
the
app
at
Cod,
the
app
name
pull
down
a
droplet.
If
you
have
a
custom,
start
command
set,
it'll
pull
that
down.
You
have
staging
environment,
variable
groups
or
running
environment,
variable
groups.
It
will
pull
those
down
in
separate
places,
so
you
can
really
replicate
an
environment
locally
and
then
I
can
run
my
app.
D
So
this
is
a
spring
app.
What
I'm
going
to
do
is
specify
the
name
of
the
remote
application
you
saw
in
CI,
basketball,
so
spring
music.
It
starts
with
the
death
flag
and
what
that
will
do
is
this
will
actually
make
CF
local
use
an
SSH
tunnel
to
connect
up
to
the
application
and
then
do
port
forwarding
to
forward,
through
the
SSH
tile,
to
forward
all
of
the
services
connected
to
the
remote
application
at
the
local
application.
D
D
D
See
I
have
my
jobs,
bill
tech
app
running
in
there
and
they
have
this
SSH
tile,
going
up
to
the
remote
application
and
forwarding
with
local
services,
and
even
if
I
take
a
look
at
the
environment
side
of
the
app.
You
can
see
that
my
V
cap
services
has
been
rewritten
to
use
the
SSH
tunnel
instead
of
the
remote
SSH
tunnel.
There's
a
lot
of
logic,
rewriting
every
possible
kind
of
decap
services
to
point
back
into
the
local
phone,
and
so
they
go
back
here
to
the
applications
running
I'm,
gonna
switch
screens.
D
D
Okay,
one
second,
alright,
and
you
see
spring
music
there
yep.
So
you
can
see
this
is
spring
music,
random,
on
PWS.
No,
here
I'm
going
to
access
spring
music
running
in
my
local
docker
container,
and
you
can
see
it's
actually
talking
to
the
same
database.
I
had
it.
You
know
if
I
change,
rumors
here
to
rumors
changed
and
then
I,
look
at
my
local
app
and
fresh
changed
over
there
too.
D
So
I
have
a
local
app
in
a
remote
app
talking
to
the
same
database
and
that's
most
of
what
I
wanted
mo
I
think
some
other
features
yep
local
has
is
it
can
watch
a
local
directory
that
can
use
a
doctor
volume
mounting
to
mount
a
direct,
the
app
directory
in
the
container
locally
and
then
watch
that
directory?
So
as
you
change
things,
I'll
constantly
be
loading,
and
so
you
can
use
that
to
iterate
really
quickly.
I
couldn't
come
up
with
a
demo
that
showed
that
great
at
size.
We
got
to
talk
about
it.
D
The
this
is
the
github
page.
It
has
a
ability
to
push
apps
to
so
you
can
stage
an
application
locally
and
replace
your
Cloud
Foundry
chocolate
with
the
droplet.
This
all
uses
the
v2
API,
so
it
works
against
any
in
sugar
because
the
WS
or
bluemix
or
whatever
you
know,
working
Australian
foundations
have
been
a
support
soon
too,
and
that
is
basically
you
can
export
applications.
You
staged
locally
as
a
docker
image.
A
D
B
A
C
D
So
that
the
next
interesting
thing
is
a
use
case,
I've
identified
talking
to
a
lot
of
spring
people
is
being
able
to
do
local
development
on
one
micro
service
of
a
large
micro-service
cluster
that
you
couldn't
run.
Look
like
entirely
locally,
so
it
does.
The
service
forwarding
thing
now,
where
you
can
act,
can
connect
to
remote
services,
but
it
no
apps
running
the
platform
can
connect
it.
So
the
next
feature
is-
and
it's
actually
mostly
done
already-
it
will
lay
push
in
it.
D
It'll
push
an
application
automatically
that
doesn't
have
a
anything
running,
but
an
SSH
tunnel
from
port
8080
back
to
the
local
amp.
So
you
can
sort
of
have
a
shim
at
beginning,
Cloud,
Foundry,
that's
secretly
all
forwarding
traffic
back
and
forth
between
your
local
doctor
and
the
you
know
remote
application
as
something
I
didn't
mention.
Is
this
also
works
with
a
remote
dr.,
Damon?
There's
nothing.
You
know
special
about
the
fact
that
my
doctor
is
running
locally,
so
I
can
I
can
so
use
it.
I
can't
even
run
doctor
locally.
D
C
D
So
I
think
I
think
it'll
be
pretty
easy,
because
once
it
can
push
an
application
that
can
be
configured
to
do
container
networking
to
figure
that
application
do
that,
you
know
it'll
be
able
to
integrate
the
same
way
by
forwarding
ports.
You
know
whatever
ports
that
need
to
talk
to
whatever
applications
from
cloud9
be
back
to
local
instance.
A
A
C
Feature
in
Bosch
allow
it
possible
to
run
multiple
errands
in
parallel,
so
there's
a
we've
kind
of
discussed
it
we
haven't,
we
haven't
relaxed
the
locking
just
yet
we're
planning
to,
though
one
interesting
feature
that
we
did
get
more
or
less
for
free
with
this
release
is
that
this
will
run
multiple
errands
in
parallel,
as
long
as
so
the
concept
of
an
errand
really
kind
of
a
shifted
from
a
concept
of
an
instance
group
to
a
concept
of
a
job
right,
because
it
was
always
more
or
less
an
errand
was
a
job.
It
was
just.
C
We
were
referencing
it
by
an
instance
group
name.
Now
what
that
means
is,
if
you
have
five
instances
of
let's
say:
zookeeper
right
in
the
same
instance
group
or
maybe
in
two
different
groups,
and
each
one
of
them
have
maybe
an
arrant
called
I,
don't
know
restart
or
smoke
test,
or
something
like
that
right.
So
it's
the
same
errand,
but
it
just
co-located
on
those
all
of
those
five
instances
right.
C
If
you
run
errant
smoke
instances
or
smoke
tests,
it
will
actually
run
the
air
and
copy
on
all
of
the
instances
in
parallel
and
we'll
give
you,
let's
say,
five
different
results
and
you
can
also
select
which
instances
to
run
it
on
if
you
want
to
be
a
bit
more
selective
right.
So,
even
though
we
currently
don't
have
purely
run
random
errands
in
parallel
on
a
deployment,
we
do
support
current
a
feature
that
has
ability
to
run
multiple
copies
of
the
same
eret.
A
No
he
said
good
information,
all
right,
Thank,
You,
Dimitri,
okay,
so
next
presentation
is
by
me:
I
noticed
that
my
colleague
Simon
Moser,
just
joined
and
he's
in
bed.
So
we're
happy
to
join
you
Simon.
Will
you
join
us
I'm
body,
bed?
Okay,
something
anyway
all
right.
So
let
me
share
my
screen
and
then
we'll
get
started.
Okay,.
A
Okay,
so
the
questions
that
a
lot
of
people
have
been
asking
me
since
I
took
on
CF
extensions.
Is
you
know
where
can
I
find
those
extensions,
and
you
know
essentially
what
are
the
get
up
URL
for
them
and
which
one
of
officials
in
which
or
not
like?
For
instance,
you
just
saw
that
Steven
is
adding
this
local,
but
it's
not
necessarily
official
yet.
A
So
how
do
you
answer
that
question?
Well
right
now,
the
answer
to
this
question
the
is
to
go
to
a
Google
Docs
and
essentially
find
the
extensions
here
and
then
hoping
that
it's
got
its
updated
and
all
the
links
are
correct
and
so
on.
Right
and
obviously
this
is
not
necessarily
a
very
workable
solution,
especially
if
we
wanted
to
like
if
I
wanted,
to
see
okay,
what
are
the
ones
that
are
incubating
versus
the
one
that
opposed
and
so
on
right.
A
So
there's
tons
of
different
views
of
this
data
and
then
what
if
we
needed
to
add
more
data
to
it,
then
I
have
to
kind
of
ping
everybody.
So
it's
it's
kind
of
a
pain
and
if
we
get
to
a
hundred
extensions
and
it's
gonna
become
a
real
nightmare,
so
to
fix
this
I
created
a
github,
but
so,
instead
of
like
going
through
the
details,
let
me
give
you
an
example
about
it
works
so
there's
a
repo
for
it.
A
Obviously
it's
proposed
so
I'm
looking
for
your
feedback
and
this
list
is
going
to
become
big,
so
we'll
have
different
views
of
it
and
we'll
cut
and
slice
it
differently.
So
how
is
this
thing
generated?
Well,
there
is
like
a
JSON
database
that
is
kept
up
to
date
and
I'll
show
you
how
that
works.
But
how
do
you
add
yourself
to
this
list?
Well,
it's
very
simple.
Let's
assume,
for
instance,
that
you
have
a
new
extension
that
you're
working
on
sorry.
A
Okay,
so
let's
say,
for
instance,
you
have
an
extension
called
test
and
you
wanted
to
add
it
to
to
the
list
of
official
extension,
because
you're
going
to
propose
it
and
so
on.
You
have
to
do
one
thing.
Well,
really,
two,
but
the
first
thing
you
do
is
you
say
you
add
a
topic
called
CF
extensions
and
that's
it
so
once
you
do
that
eventually
have
a
bot
that
will
run
periodically
I'm
gonna,
just
gonna
run
it
manually
here
and
when
it
runs,
it
discovers
all
projects
that
have
CF
extensions.
A
Now,
not
all
projects
and
all
organizations
to
a
certain
list
of
organizations
that
we
will
just
curate
I'm
thinking.
You
know:
incubator,
community
Cloud
Foundry,
as
well
as
all
the
different
member
companies
and
then
once
you
do,
that
it
will
automatically
discover
that
there's
an
extension
and
because
this
project
is
not
a
tract
extension.
Yet
what
it
does
is
it
submits
an
issue
to
you
right
well
should
have
submitted
an
issue.
I
guess
I
already
had
an
issue,
so
it
noticed
that
I
had
an
issue.
A
So
that's
why
I
didn't
resubmit
it
and
that
in
that
issue
you
have
to
do
one
thing,
which
is
to
add
yourself
a
metadata
file
that
contains
this
information.
So
basically
owner
company
contact,
email
description,
the
tracker
URL,
the
proposal
URL
eventually
and
then
the
date
that
you're
proposing
once
you
do
that
so
I'm
going
to
go
ahead
to
test
so
I'm
going
to
close
this
issue
and
add
that
file.
A
A
Now
it's
gonna
move
that
project
to
contract
and
the
reason
we
do
this
is
so
that
I
can
control
the
projects
that
are
being
tracked
and
projects
that
are
not
being
tracked
right.
So,
for
instance,
if
you
go
here,
it
should
have
submitted
that
there
is
a
new
track
project.
Eventually
it
should
refresh
it
there
we
go,
so
you
can
see
now
that
the
project
is
on
track
and
then,
let's
say,
for
instance,
the
project
gets
proposed.
A
A
So
if
people
start
putting
some
extensions
everywhere,
they'll
just
go
to
one
track,
which
is
fine
and
I
will
add
an
entry
here
that
says
basically
that
the
test
project
in
the
Cloud
Foundry
incubator
is,
let's
say
unknown
for
this
point
and
then
save
it,
and
then
once
the
the
bot
runs
again,
it
runs
again,
as
I
said
couple
times
a
day.
It
will
move
the
project
now
to
two
tracked
and
then,
as
things
change
and
the
information
changes
and
so
on,
it
will
be
correct.
A
B
A
B
A
B
A
Well,
I'll
figure
out
what
this
is,
but
basically
it
should
update
the
project
list
to
now
include
tests
eventually
here
and
then,
as
new
projects
get
added.
That's
basically
the
idea
so
to
keep
track
of
all
projects,
and
then,
if
somebody
wants
to
remove
themselves
from
any
extensions
from
extensions,
all
they
have
to
do
is
get
rid
of
their
test
project
as
well
all
right.
So
so,
basically,
it's
not
just
you
know,
keeping
track
one
way
to
add
more
projects.
It
also
keeps
track
the
other
way,
which
is
to
remove
projects.
A
So
what
I'm
hoping
for
is
to
eventually
get
to
a
point
where
all
projects
are
added
to
this
process
right.
So
what
you
will
have
to
do
now
fix
that.
One
tiny
bug
here
to
add
yourself
to
extensions
is
to
start
tagging
yourself,
ICF
extensions.
So
it's
not
really
a
tag.
It's
yep
topic,
it's
a
topic
in
in
in
github.
So
if
you
add
your,
you
know,
you
add
CF
extensions
to
your
github.
A
Project
then
it
will
show
up
the
one
thing:
I
haven't
done
so
there's
a
few
more
to
do's
to
make
it
happen
so
this
week,
I
want
to
start
rolling
it
out.
I'll
figure
out
that
one
issue
here,
but
I
want
to
add
all
different
projects
to
this.
The
current
projects,
so,
basically
I
will
ask
you
if
you're
a
community,
if
you
are
an
extensions
PM
to
start
adding
the
topic,
CF
extension
to
your
project
and
then
you'll
get
an
issue
automatically
created.
A
A
It's
this
view
right,
but
I
want
to
also
give
you
views
like,
for
instance,
show
me
all
incubating
projects
versus
project
set
of
proposed
versus
projects
that
are
retracted
and
so
on.
So
different
views
like
that.
I
also
want
to
add
a
little
bit
more
detail
on
the
projects.
For
instance,
there
are
projects
that
are
peer.
A
You
know
that
positive
peer
commit
model
and
in
their
projects
that
follow
the
the
distributed
commit.
You
could
do
that
because
I
have
access
to
the
github
project
and
the
API
there's
tons
of
other
things.
I
can
do
so
like,
for
instance,
here
I'm
showing
like
Watchers
and
stars
and
forks
so
that
you
can
get
an
idea
of
of
which
extension
is
more
popular
or
which
extension
is.
Maybe
you
know
in
active
development,
so
in
other
words,
give
you
a
view
of
all
of
the
extensions
in
a
way
that
is
useful
to
you.
A
So
what
I'm?
Looking
for
you
know,
folks
if
you're
interested
in
this
so
first
is,
if
you
are
an
extension,
TM
tag
yourself
and
if
you
interested
in
this,
you
know
we
welcome
PRS.
The
project
itself
is
an
extension,
but
it's
being
proposed
and
it's
distributed
commits.
So
if
you
submit
commits
that
PRS
that
would
be
nice,
I
have
a
blog
post
that
will
be
pushed
soon.
A
That
discusses
the
but
and
that's
a
summary
of
what
happens
right
so
basically,
when
it
runs,
finds
all
projects
with
this
yep
extensions
and
it
looks
to
see
if
you
obviously
have
extensions
JSON
file
in
your
project.
If
you
don't
have
that,
then
it
submits
it
adds
you
to
the
on
track
and
then
submits
an
issue
only
if
you
don't
have
an
issue.
A
Obviously,
if
you
have
an
issue,
it
doesn't
keep
adding
more
issues
and
then
once
it
runs
and
notice
that
you
have
tagged
yourself
and
that
you
have
a
file,
then
it
starts
looking
to
see
if
it
should
track
you
or
not,
and
it
checks
whether
or
not
your
status
and
then
it
generates
all
those
documentation
files.
So
it's
pretty
straightforward,
not
not
too
hard.
It's
written
and
go
and
I
look
forward
to
any
feedback.
B
A
I
guess
it's
it's
an
interesting
thing.
Maybe
maybe
I
could
change
it
to
populist
company
the
company
that
purple
is
there
or
something
like
this.
The
only
reason
I
wanted
this
is
to
get
a
feel
for
where
that
project
came
from.
So
maybe
like
instigating
company
something
like
that.
So
you're
right,
the
owner
is
probably
a
long
term.
I'll.
Take
that
that's.
B
It
for
what
it's
worth
I
think
I'd
be
much
happier
if
it
if
it
reflected
like
the
composition
of
the
team,
as
opposed
to
like
one
company
like
garden,
wasn't
proposed
by
I,
mean
I,
guess
Gardens
been
there
for,
since
it
started
Diego
and
but
I.
I
spending
don't
feel
there's
any
company
that
owns
it
for
all
time.
Right
we've
got
a
lot
of
companies
on
it.
Yeah
no.
A
That's
fair,
that's
fair!
Actually,
maybe
what
I'll
do
is
I'll
modify
that
but
I
think
getting
an
idea
of
the
team
distribution
which
is
not
hard
for
me
to
do
because,
again,
by
scraping
stuff
from
github
I
actually
have
access
to
all
of
your
commits
and
I
mean.
Obviously,
your
project
has
to
be
public
right
if
it's
not
public,
nothing
works,
but
if
it's
public
I
have
I
can
access.
Like
friends,
you
know
this
year.
What
are
the
commits
and
and
just
like
github
shows
these
graphs.
The
distribution
graph
I
can
do
all
that
right.
A
So
so
in
time
you
know,
as
people
ask
for
these
things
we
could
add
it
and
part
of
the
idea
is
to
have
a
simple
location
and
then
give
enough
information
about
the
projects
to
help
people
make
decisions.
So
so,
thanks
for
that
feedback,
George
and
Jules,
okay,
any
other
question
I
saw
somebody
posted
something.
A
B
B
B
A
By
the
way,
the
tester
should
it's
just
that
github
took
a
while
right,
so
you
can
see
it's
showing
now.
It
just
took
a
while
for
the
data
that
I
committed
from
the
data
that
gets
returned
from
the
API,
so
I
don't
have
a
book,
it's
just
I
knew
I
tested
it
before
anyways
and
actually
that's
a
good
point,
sometimes
github
crashes,
which
is
interesting
because
then
the
book
fails
and
you're
like
what's
going
on
anyways
Brett.
Do
you
mind
typing
your
question
because
the
whole
time
you
were
talking
you
were
grating.
A
B
A
That's
a
that's
a
fair
point:
let's,
if
you
don't
mind
Brett
like
started
discussion
once
you
have
their
vote
directly
with
me.
Cuz
I
want
to
understand
what
you
mean
by,
for
instance,
scanning
policy,
and
things
like
this.
Are
you
thinking
because
there's
a
series
of
bots
on
github
that
do
these
things
automatically?
What
are
you
thinking
that
the
person
that's
instigating
the
project?
Does
it
itself?
You
know,
and
things
like
that.
Let's,
let's
have
a
discussion
about
that.
A
A
Ok!
So
if
you're,
a
extensions,
PM
just
basically
add
a
topic,
you
have
extensions
and
then
let's
get
started.
Obviously
it's
new,
so
we'll
learn
from
this.
If
you
have
suggestions
on
how
to
make
the
process
easier,
I'm
happy
to
discuss
it,
but
again,
I
know
Josie's
laughing
because
he
saw
it
yesterday
and
he
had
something
bad
I.
I
think
you
know
a
couple
back
and
forth
is
easy
and
I
love
the
fact
that
the
project's
themself
manage
extensions
metadata.
A
But
if
you
have
different
opinion
on
this,
that's
the
discussion
started,
but
lets
us
see
start
and
then
and
then,
of
course
we
can
always
change
it.
We
don't
have
too
many
exchanges
right
now,
right
like
we
have
maybe
a
12,
so
whatever
we
do
should
be
shouldn't
be
hard
to
keep
that
database.
Even
like
you
added
manually,
the
problem
is
become,
of
course,
when
we
start
having
a
hundred
of
extension.
A
Okay,
there
is
one
question
here
on
slack
for
moogle
I'm,
not
sure
what
company,
but
he
said
he
or
she
said
my
dream-
is
that
when
you
lost
yourself
as
an
extension,
cff
insurers
scanning
happens
and
submits
issues
PRS
to
you
for
security
problems.
Yes,
yes,
I
see
your
point
and,
as
a
matter
of
fact,
this
whole
idea
of
like
sort
of
forcing
people
to
do
things.
So
when
I
submit
the
issue
and
I
asked
you
to
add
the
CF
extensions
I
could
change
that
to
submitting
a
PR.
The
only
reason
I
didn't
do.
A
This
is
because
submitting
PRS
vs
issue
is
is
like
order
of
magnitude
more
complicated,
because
I
have
to
fork
the
repo
and
then
obviously
I
start
having
a
bunch
of
repos
that
I
forked.
So
where
do
I
delete
them
or
when
do
I
delete
them
and
so
on,
but
definitely
that's
that's
a
possibility.
So
thank
you
for
that.
Okay,
unless
there
are
any
questions
this
is
it
I,
guess,
there's
one
more
chat,
no
I
think
that's
it!
Okay!
Everybody!
If
you
have
no
questions,
follow
up
on
email,
we'll
chat
in
September
take
care
Cheers.