►
From YouTube: 🚀IPFS Core Implementations 2019-07-15🛰
Description
Description: https://github.com/ipfs/team-mgmt/issues/992
Notes: https://cryptpad.fr/code/#/2/code/edit/G2OMM7r9aryUS1B92ye1ycnG/
A
B
Cool,
let's
do
it
hello,
everybody
and
welcome
to
the
ISS
Cory
parent
ations
weekly,
sync,
it's
July.
The
15th
is
2019
still
it's
kind
of
summertime
outside
here
anyway,
so
I'm
cool
I
created
the
doc
because
there
was
no
doc
and
and
I
put
some
things
in
it.
So
I
thought
we
might
want
to
talk
about
them.
If
you
are
here,
please
put
your
name
on
the
attendees
list.
Can
I
get
someone
to
take
notes
for
today
or
volunteer
to
take
notes
at
least.
B
Thank
You
Molly,
Brad,
okay,
cool,
so
yeah,
okay,
I'm,
not
sure
I'd,
entirely
exactly
how
it
works,
but
we've
got
a
note-taker.
We've
got
everyone
we're
putting
their
names
on
and
then
so
I'm
just
gonna
kind
of
go
through
the
document
and
see
what
see
what
happens
if
you've
got
initiatives
like
things
that
are
kind
of
big
overall
tasks
or
projects
that
are
happening
then,
and
you
want
to
kind
of
give
an
update
about
them
and
they
feel
it's
important
then
put
that
put
that
down
in
the
dock.
B
And
if
there
are
any
blockers
that
you
have
in
your
weekly
update,
oh
yeah,
you
should
probably
tell
people
about
that
if
you
ever
have
a
weekly
update
of
what
you've
done,
what
you're
blocked
on
and
what
you've
got
next
put
it
on
the
bottom
of
the
nearing
the
bottom
of
the
document.
Just
so
you
have
a
record.
We
won't
talk
over
of
it
over
it
all.
C
C
A
D
B
Okay,
this
is
very
good
news
that
we're
going
to
be
upgrading
our
release
process
and
and
making
it
more
well
tested
doing
Morrissey's
getting
people
to
test
it
before
it
goes
goes
out
generally
having
more
stable,
reliable
releases
and
the
possibility
of
following
Semba
as
well
to
some
extent.
So
that
is
super
good
news.
In
my
opinion,
anyone
have
anything
else
to
say
about
that
before
you
move
on
Oh,
shall
we
get
on
with
it
all
right?
Let's
get
on
with
it.
Upgrade
testing
process
go
ipfs,
yeah,
okay,
Steven,
I,
guess
again
is
that
you.
C
Yes,
everyone
should
read
that
linked
document
yeah.
This
is
the
current
plan.
It's
obviously
up
for
debate
and
change
for
how
to
start,
or
basically,
the
required
steps
to
start
really
seeing
fest
again
and
then,
after
that,
like
the
rest
of
the
plan,
this
is
again
more
focused
on
good
best,
because
that's
really
the
only
people
relying
on
in
a
lot
of
production
systems.
C
C
Once
again,
yeah
yeah
first
for
milestones
done,
which
unfortunately
means
probably
two
to
three
months
of
work,
so
this
also
required
buy-in
from
many
different
teams.
You
may
notice
that
some
people
are
here
listed
as
do
something
to
be
surprised,
but
will
hopefully
be
pulling
help
from
in
frog
from
package
managers
gateway
team
from
just
about
everywhere.
C
A
A
C
B
C
That's
basically
the
plan
it
we
don't
feel
yeah
sure
that
we
won't
introduce
progressions.
It
is
also
partially
the
fact
like
we've
been
saying,
we
want
to
do
this
for
ages
and
we
haven't
been
doing
it.
This
is
a
way
to
to
force
us
to
actually
make
progress
and
that's
actually
like
up
way
to
our
engineering
process.
C
C
B
Got
it
cool?
Thank
you,
alright,
that
is
the
upgrade
testing
process.
So
this
next
section
is
something
I've
put
in
for
J
sake,
FS,
and
there
is
at
the
moment
a
big
kind
of
falling.
This
really
last
quarter
that
all
most
of
this
work
has
been
done,
and
it's
only
this
only
now
that
kind
of
we've
been
able
to
look
into
it
and
and
start
fixing
our
anyway,
but
so
garbage
collection
doesn't
exist
in
JSI
events
and
one
of
the
things
that
we
kind
of
need
if
we're
gonna
store.
B
So
there
is
so
I
put
a
heart,
emoji
and
duck
MC
at
the
bottom
of
there,
because
it
is
basically
all
of
Dirk's
good
work
that
he's
done
on
this
and
he's
done
a
really
good
job
of
looking
into
what
what
go
ipfs
stars
and
to
do
something
similar
and
and
so
yeah
it's
it's
kinda.
It's
coming
along
I've
been
reviewing
the
pull
request.
B
They
are
large,
there's
a
lot
of
stuff
that
it
touches,
but
they're
like
we
are
kind
of
reasonably
close
on
sites
like
having
a
kind
of
first
past
implementation
of
garbage
collection
working
in
a
similar
way
that
doesn't
go
activist
also
working
in
China
succubus.
Oh
man,
that's
cool,
and
so,
like
I
put
there,
it's
kind
of
uncovered
a
few
problems
with
our
existing
pinning
things
like
our
internal
blocks
were
being
kind
of
provided
to
the
DHT.
So
we
need
to
fix
that.
We
realize
that
go.
B
Ipfs
uses
an
in-memory,
IP
o
d
for
those
for
those
nodes,
I
believe
so
that
they
don't
get
put
onto
the
DHD
and
so
yeah
when
you
sort
that
out,
and
we
found
out
a
couple
of
other
problems
with
pin
types
not
being
filtered
properly
for
pin
LS,
and
we
know
that
performance
is
not
great
when
pinning
in
Jessica
fest
at
the
moment.
So
there
is
I.
Think
that's
been
also
been
looking
into
how
we
might
fix
up
that.
So
next,
next
up
for
that
is
just
well
finish
off.
B
G
Have
one
question
very
quick
because
okay
I
don't
remember
this
hundred
percent,
but
I.
Remember
like
at
some
point
in
time
we
had
problems.
We
need
XE,
be
Auto,
garbage
collection
right
now.
So
when
you
started
filling
in
annex
VIII
database
like
the
browser
itself
will
be
the
one
trying
to
garbage
collect
just
ensure.
Like
my
kid.
G
Oh
no,
it
keeps
the
objects
that
are
most
most
used
and
I'm
wondering
if
this
work
with
GC
like
contemplates
that
and
like
test
for
different
browsers,
because
there's
also
problems
in
Firefox
Chrome,
in
which
I
don't
respect
second
flags.
We
we
like
try
to
get
next
to
meet
these
things.
Are
we
destined
for
those
things
or
or
this
is
just
like
first
version
per
situation,
yet.
B
G
B
H
A
C
Mike
wait
there
was
that
for
providing
version
for
re,
providing
we
just
walked
the
data
store,
I,
don't
think
we
even
look
at
or
exclude
anything,
that's
in
the
inset
I
mean
he
should,
but
you
do
for
initial
provides
again
I,
don't
I'm,
not
sure
what
we
do
there.
I
don't
know.
If
we
take
any
explicit,
if
you
do
any
things
mostly
to
avoid
providing
him
think
might,
but
it
would
be
kind
of
a
hack.
Anyways
is
not
very
reliable.
F
C
F
C
I'm
not
sure
if
we
like,
create
a
separates
the
block
service
stack
when
we
initially
write
these
to
the
datastore
I
know
we
definitely
provide
because
we
don't
have
any
of
the
distinction
there,
but
we
initially
write
them.
We
make
quite
separate
stack,
I,
don't
know,
I
mean
I
can
check
that
quickly.
B
Okay,
if
you
examine
anyway,
that'll
be
great
but
yeah
cool,
that
is
garbage
collection
and
it's
coming
to
a
JSI,
be
best
node
near
to
you
and
then
so.
The
second
thing
I
I
wanted
to
talk
about.
Also
I
put
this
down
to
dedicated
routing
I
know
this
is
this
is
something
that's
kind
of
super
important
to
me.
That's
working
in
the
browser.
B
We
don't
have
a
DHT,
even
if
we
did,
we
wouldn't
necessarily
be
able
to
connect
to
half
the
people
on
the
DHT
because
of
same
origin
policy
and
people
with
web
sockets.
That's
that's
fine,
but
a
lot
of
them
are
on
just
HTTP
and
connecting
to
them
from
HTTPS
is
impossible.
So
yeah
we
have
challenges
in
the
browser
we
have
GSI
person.
B
Delegated
routing
is
one
of
those
one
of
those
things
that
can
that
can
really
kind
of
help
that
connectivity
magic
feeling,
at
least
even
if
it
isn't
actually
a
curly
and
I
peer-to-peer
thing,
one
of
the
things
that
is
blocking
this
from
from
being
realized
and
being
released
in
J's.
Okay,
this
is
the
fact
we
need
a
way
of
configuring.
The
go
IDF
s,
demon
gateway
port.
B
So
just
just
to
say.
That
is
what
it
is.
There
are
two
modules
that
I've
been
working
on
this
week:
they've
been
around
for
a
while
I
had
to
sort
out
the
dependencies
they
were,
depending
on
a
really
old
version
of
I.
Can
that's
httpclient
the
old
one
called
HTTP
api
TFS
api,
which
meant
that
we
had
an
extra
200k
of
bundle
size,
put
being
pulled
into
our
bundles,
so
I
fixed
that
up.
So
that's
no
longer
the
case,
so
they
are
ready
to
go.
We
just
need
this.
B
A
I
Through
thing
or
garbage
collection,
tell
your
routing,
don't
work.
Curie's
got
either
I'm
really
interested
in
delegates
at
I
think
because
it's
unlocks
embedded
JSOC
EFS
in
brave
to
be
useful
before
we
implement
actual
peer-to-peer
transport
right
now.
Unless
you
are
connected
to
note
that
already
has
a
content,
the
embedded
note
is
not
that
useful.
You
are
not
able
to
use
it
for
anything
practical.
That's
why
it's
not
like
the
default
in
the
brave,
so
I
played
with
it.
I
have
opinions
and
even
filled
some
shoes.
B
You,
okay,
yeah
I,
mean
and
specifically
I
guess
there
are
only
a
few
options
for
finding
and
connecting
connecting
two
nodes:
finding
peers
and
and
getting
content
for
JSON
APIs
in
the
browser,
and
this
is
adding
another
ones
that
would
be
really
really
reliable
to
the
mix.
So
it
would
be
good
to
get
it
together.
Cool,
okay,
thank
you.
Sip,
just
raise
your
hand,
sorry.
J
C
Person's
content,
okay,
cool
this
could
also
partially
fixed
by
the
DHT
or
getting
a
lot
of
the
D
gene
nodes
offline.
If
they're
not
currently
make
hundreds
of
connections
that
used
to
your
question
at
like
very
worst
case
scenario,
make
it
twenty
connections,
usually
like
five
connections.
Okay,
that's.
F
C
C
F
C
A
I
Like
a
quick
clarification
about,
what's
actually
blocked,
our
preload
nodes
are
set
up
so
that
actually
we
are
able
to
use
them
without
making
those
changes
to
go
ideas.
So
it's
like
not
a
hard
block.
The
problem
is
that
people
who
would
be
interested
in
running
delegated
routing
they
would
have
to
use
our
infrastructure.
So
that's
what's
blocked.
F
I
B
C
B
K
Trust
me
so,
basically,
last
week,
I
finished
most
of
the
remaining
work.
Integrating
those
lips
of
Angelica
P,
including
it
also
interoperability,
test
suite
go
so
we
did
last
Friday
Shane
safe,
guys,
released
a
new
version,
including
the
also
the
fallback
for
a
flood
soap.
So
Jacob
is
on
only
late
this
week,
but
next
week,
I
hope
that
we
will
release
release
candidate
for
Jess,
which
we,
including
gossip,
so
and
meanwhile
I
also
will
start
working
on
carrying
Jessica
Fest
ready
for
this
new
release.
K
B
B
B
G
C
C
K
G
B
I
Yeah,
it's
something
we
talked
before
today.
We
Steven
but
I,
feel
it's
something
worth
mentioning.
That's
happening,
subdomain
gateway.
It
turns
out
to
be
an
interesting
infection
point
between
our
CID
v1
base,
42
initiative
and
also
oh,
it's
a
vessel
to
solving
original
isolation
for
both
of
the
gateways
and
also
local
host
gateway.
So
there
are
multiple
problems
that
it
theoretically
solves,
but
it
could,
depending
on
what
we
will
implement
in
which
order
it
will
solve
different
things
for
different
people.
I
So
I
just
wanted
to
highlight
this
and
give
you
a
brief
overview
where
we
are
and
what's
happening
here
so
subdomain
gateway.
It's
basically
content
identifier
in
sub
domain
instead
of
location
path
and
it's
actually
shipped
in
userland
CloudFlare,
and
also
we
have
the
web
link
public
gateway
which
supports
that
the
problem
is
that
the
way
it's
supported
right
now,
it's
at
the
nginx
level.
It's
like
River
reverse
proxy,
which
takes
care
of
Trane's,
taking
a
CID
from
sub-domain
and
converting
that
CID
to
request
to
regular
old
school
gateway,
which
was
a
addressed.
I
So
the
first
thing
that
subdomain
in
gateway
support,
like
native
support
for
subdomain
gateways.
Would
salt
would
be
the
need
for
having
something
in
front
of
go
ipfs
to
have
something
subdomain
gateway,
so
we
want
to
make
it
easier
for
people
to
run
their
own
public
subdomain
gateways
that
provide
original
isolation
similar
to
cloud
cells
without
the
need
and
the
complexity
of
running
a
reverse
proxy
and
setting
that
up.
So
that's
one
thing
and
there's
like
a
related
issue
on
that,
and
we
identified
some
open
questions.
I
How
maybe
that
in
the
next
step,
so
the
second
step
is:
if
we
have
support
for
subdomain
gateway
and
we
run
go
ipfs
on
localhost,
then
we
have
those
something
something
that
local
host
fake
domains
for
local
host.
That
could,
in
theory,
provide
a
subdomain
gateway
for
your
local
host
machine,
and
this
is
where
it
gets
interesting
because
you
get
a
parity
between
origin,
isolation
on
the
public
gateways
and
you
also
have
originals
relation
of
the
local
gateway.
I
So
the
problem
here
is
this
transition
path,
where
we
have
people
running
local
gateway
which
is
path
address,
and
then
we
suddenly
introduced
subdomain
one.
We
have
some
nodes
in
the
issue
I
linked
about
how
that
should
we
figured
out
the
redirects,
because
we
don't
want
to
support
both
on
the
same
origin
that
would
break
the
isolation.
I
So
that's
like
an
open
effort.
We
will
difficult
out
through
this
week
and
another
thing
when
we
have
that
we
could
also
implement
HTTP
proxy.
On
top
of
that,
that
would
be
like
a
much
smaller
change
and
that
would
give
us
better
UX
in
a
web
browser
that
has
like
companion
or
the
other
web
browser
that
sets
up
go.
I
PF
SS
HTTP
proxy.
That
would
enable
us
to
remove
a
custom
port
and
make
host
names,
look
better
and
picky.
By
going
back
to
the
base
32
initiative.
I
Subdomains
could
support
not
only
ipfs
but
also
IP
NS,
and
we
finally
sort
of
figured
out
how
to
represent
peer
ids
in
subdomains
and
we've
added
multi
codec
for
the
p2p
keys
and
the
remaining
steps
are
wiring
up.
The
CID
support,
instead
of
multi
Hodgkiss
in
multiple
places,
both
in
go
and
enjoy
as
I
believe
we
assumed
pure
ID
to
be
multi
harsh,
effectively,
CID
v0q
em
something
and
we
would
need
to
find
all
those
places
to
our
eat
it
up.
I
F
Sort
of
like
putting
this
directly
in
ipfs,
but
at
some
point
it's
like
you
know
like
at
what
point?
Are
you
now
replicating
nginx?
You
know
and
something
that
is
like
really
the
domain
of
something
that
is
designed
to
be
a
web
server
like
a
professional
web
server
versus
something
that
is,
you
know,
a
basic
HTTP
support.
C
So
basically,
here
we
just
we
want
enough
so
that
we
can
integrate
admits
into
web
browsers,
and
that's
it
I,
don't
think
we
have
any
intention
of
like
working
really
hard
on
before
when
it's
not
that
in
terms
of
like,
like
really
seasoned
Ross,
it's
like
a
shame,
a
lot
of
really
complex
since
we
could
use
it.
We're
probably
not
going
to
do
everybody's
all
the
pricing
shakes
out
in
front
of
it
in
this
case,
it's
more
about
like
we're.
F
C
Same
origin
and
whatnot
so
like
we
could
do
is
say:
hey
users,
you
have
to
install
these
two
things
or
we
could
say
whatever
so,
let's
1k
and
it
just
kind
of
magically
works,
and
if
that's
really,
the
key
part
here
is
Newswatch
was
a
it's
all.
This
one
application
and
you're
done.
I
would
like
the
part
of
this
conversation,
you're
going
to
have
with
utility
vestige
of
two
pieces
of
goods,
the
library
and
then
going
with
the
application.
C
F
F
C
That's
we
already
do
we.
Actually,
we
do
literally
shove
a
single
binary
yeah,
but
yeah
basically
like
this
is
not
for
like
this.
Oh
it's
for
my
server.
Yes,
we
could
say
like
hey,
like
you
install
it,
the
festival,
ain't
used
on
nginx
reference
proxy
and
it's
actually
outputting
to
everyone
once
the
Gateway,
but
we're
friendship
and
usually
the
use
case
here-
is
an
end
user.
I
I
We
we
support
like
HTTP
caching,
semantics
we
put
like
see
IDs
in
attacks
and
other
stuffs
of
the
web
browsers
benefit
from
built-in
HTTP
caching
layer,
that's
already
like
implemented
everywhere,
but
we
we
only
implement
what
makes
sense
in
ipfs
context.
It's
not
like
it's
not
one-to-one
mapping.
We
just
need
to
be
very
careful
about
what
we
map
and
what
we
expose
over
HTTP
sure.
B
Thank
you.
Thank
you.
I
was
really
good
question.
Actually
any
other
questions
before
we
move
on
to
blockers.
B
Alrighty
then,
we've
got
block
as
an
answer
relatively
small,
a
small
ask
to
help
out
on
the
forums
on
the
discuss
forums
on
discuss
that
idea,
I
thought
I,
oh
there's
notes
and
on
the
IRC
channels,
I
have
s
and
ipfs
dev.
Please
hang
out
or
help
out.
If
you
can,
with
things
on
there,
did
you
ever
put
that
down
yep?
C
So,
like
the
the
set
of
people
who
are
doing,
this
is
actually
fairly
small
and
it's
a
lot
of
work
and
very
like
distracting
time-consuming.
So
we
had
more
people.
Jumping
on
this
would
be
really
helpful.
It
also
helps
the
community
like
see
that
hey,
yes,
for
involves
we're
interacting
with
them
kind
of
stuff.
C
I
just
discuss
forms
could
definitely
software
help
IRC
and
generally,
because
if
you
have
more
people
different
time
zones,
even
have
better
responsiveness
and
like
again,
better
community
engagement
right
now
that
we
have
some
great
people
in
the
community
doing
this,
but
they
don't
always
have
visibility
like
into
we're
currently
on
so
more
people
who
are
like
in
the
core
foundations
group
that
understand
what
we're
working
on
going
out.
That
you
mean
like
telling
people
what's
going
on
really
helpful.
B
C
H
C
C
C
B
C
C
Yeah
in
the
current
answer
that
is
like
these.
From
my
perspectives,
we
were
first
working
on
the
testing.
There's
like
we
have
something
cuz.
They
were
not
sure
about.
So
first
gonna
make
the
testing
like
stall
is.
We
can
actually
tell
what
these
fixes
will
do
to
the
network.
That,
like
is
throwing
everything,
then
we're
going
to
release
much
fixes
and
the
timeline
on
this
is
unfortunately,
probably
two
three
months
before
we
start
to
find
these
fixes
the
network
that.
H
C
Commitments
on
the
two
to
three
months
thing,
because
we
don't
know
I,
definitely
actually
don't
do
that
I
do
we
can
give
the
roadmap
of
like
yeah.
Now
this
this
jettisoned
document
is
public
I,
don't
know
how
widely
want
link
to
that,
but
they
asked
me
say:
look
like
we
are
currently
building
out
massive
test:
the
infrastructure
to
try
to
better
resume
networking,
better
person,
change
debate,
and
that
is
the
cart
thing.
Yeah.
H
C
G
Well
I'd
say
there
is
was
get
through
really
without
thing
and
I
feel
very
confident
and
I
feel
that
it's
very
smooth
can
I
get
the
support
from
the
partners,
and
so
on,
and
people
saying
like
every
day,
I
mean
anything
like
quite
intensely
reason,
and
am
I
doing
the
Buccos
person
I
might
like
my
standpoint.
Do
something.
That's
then
we'll
have
more
updates
so.
C
I
think
those
are
two
issues
here
is
like
there's
the
release
process
in
the
current
release,
which
will
fix
some
burning
pyres,
but
it
will
not
fix
any
to
get
patience
when
you
touch
them
yeah.
The
Gateway
issues
are
long-standing
other
issues
that
won't
get
fixed
until
we
are
actually
up.
We
have
testing
of
a
structure
and
starts
at
big
shipping
like
really
large
changes
to
the
network.
C
G
G
C
A
Think
the
thing
we
can
do
for
it
not
to
take
three
months
is:
we
can
significantly
cut
down
the
amount
of
time
it
takes
to
build
large-scale
testing
infra
right
now,
that's
slated
to
be
the
largest
thing.
If
we
can
pair
a
little
ice
that
more,
we
can
get
more
people
on
it
in
such
a
way
that
it
takes
less
time
to
build
out
and
that
will
significantly
decrease
the
time.
That's
the
biggest
Walker.
E
G
B
It
would
be
cool
for
it
to
take
much
shorter
time
than
that.
Then
what
we
think,
okay,
so
look
as
an
asset.
We've
got
no
other
blockers
and
asks
then
we're
on
to
questions.
Does
anyone
have
any
ad
lib
questions
that
have
just
popped
up
in
their
head
that
they
want
to
quickly
ask
we've,
got
six
minutes
left.
G
Quick
question
gonna
be
talk
about
tests.
I
would
like
just
checking
out
familiar
are.
Is
they
go
ipfs
team
on
the
net
benchmarks
that
I,
because
the
team
like
that
I
wanted
me
on
lead
for
as
in
January?
And
if
you
are
all
updates,
are
you
so?
Is
that
for
you
or
these
are
like
yes,
what
all
luck,
useful
I
heard.
C
C
B
B
It's
not
like
thousands
and
thousands
of
nodes
testing,
but
it
is,
it
is
doing
specific
actions
we
have
ipfs
and
tracking
the
time
it
takes
to
like
transfer
a
file
between,
like
ji
per
person,
go,
give
us
or
adding
a
file
to
just
JCPS
or
just
go
ipfs
I'm.
It's
that
sort
of
benchmarks
and
they're
a
space
for
coding,
more
Bureau
yeah.
That's.
C
Exactly
what
we
need?
That's
like
the
the
first.
If
you
look
at
the
the
our
roadmap
I,
the
basically
first
section
is
just
like
get
a
bunch
of
those
benchmarks
running
and
then
track
these
over
time
and
have
like
some
crappy
web
UI,
I,
say
crappy,
because
I
don't
want
to
waste
time
on
it.
Just
like
more
progress.
E
B
Yeah,
it's
it's
pay
for
this.
Alright
I
will
I'll
get
it
up
and
running
again
as
a
priority.
There's
an
absolute
wonderful
that
would
save
us
yeah.
E
B
Alright
I'll
try
and
get
it
running
and
then
maybe
give
it
from
the
presentation
or
record
something
so
that
you
guys
can
watch
and
understand.
What's
there
and
and
yeah,
we
awesome
if
I
could
also
look
at
CI
as
well.
Cuz
I
want
it
as
well
alright!
Well,
it
is
do
we
have
any
other
questions
quickly.
A
The
thing
I
was
gonna
bring
up
was
just
two
beads
spec
maintainer
PR
I.
Think
that
use
more
visibility
and
also
I
think
there
are
more
people
that
have
mentioned
blockers
in
their
personal
notes
than
mentioned
blockers
in
this
meeting.
If
you
have
blockers
that
you're
going
to
need
people's
visibility
into
to
call
them
out,
but
I
thought
and
urges
at
hand.
B
Yeah
I
just
opened
an
issue
on
Friday
I
think
but
hadn't
seen
any
activity
on
it.
I,
don't
know
what
the,
how
kind
of
the
level
of
triage
that
we're
doing,
but
it
would
be
good
for,
like
in
general,
if
people
are
opening
issues
just
to
acknowledge
them.
Even
if
we
can't
look
like
we
know
that
we
don't
have
the
bandwidth
to
do
anything
about
them.
I.