►
A
It's
recording
alright.
Let's
do
this
hello,
everybody
welcome
to
the
ipfs
core
implementations
weekly
sync,
it
is
October.
The
31st
2019
still
chugging
through
Jacob,
has
volunteered
to
be
a
note-taker.
Thank
you
very
much,
Jacob,
that's
very
kind,
all
right,
so
the
hackpad
with
all
of
the
notes-
and
things
is
here.
Let
me
put
it
in
the
chat.
I
chat
capture.
If
you
have
not
already,
please
would
you
fill
in
your
a
sink
update?
It's
at
the.
You
should
put
that
at
the
bottom
of
the
document.
A
We
won't
go
through
this
during
the
meeting,
but
it's
for
people
to
review
asynchronously
as
they
please
and
follow
up
if
they
need
to
all
right
checklist
continue.
What
have
we
got?
So
if
you
are
here,
please
put
your
name
in
the
attendees
bit
so
that
we
know
you're
here
for
for
the
record,
and
we
will
continue
on
with
the
initiatives
where
we
talk
about.
What's
what's
going
on
in
the
world
of
the
J,
what
not
just
yes
go
and
Jess
IVF
escort
implementations
cool.
So
first
up
we
have
upgraded
release
process.
A
B
Actually,
actually
testing
third
party
repo,
so
we
added
support
this
ratio
a
while
ago
and-
and
we
are
actually
now
testings
they're-
probably
repos
in
the
bill-
every
build,
so
not
just
internal
ones,
but
some
set
up.
Oh
he
wants
to.
The
problem-
is
that
some
of
the
third-party
repairs
were
testing.
Although
they
signed
up
for
the
early
release,
custard
cake,
they
don't
actually
depend
on
choice
of
FS
or
the
HTTP
client.
B
Some
of
them
depend
on
you
know:
IP
FSD
CTL
instead
and
don't
pass
a
version
of
it
in
so
it's
like
well,
that's
cool,
but
there's
nothing
there
that
we
can
swap
out
to
test
for
an
update
in
one.
So
there's
some
issues
further
down
the
weather.
When
I
talk
to
them
and
further
down
the
document,
they
explore
that
in
a
little
more
detail,
but
fundamentally
I
mean
we
can
like
it
would
be
lovely
to
test.
B
You
know
dependencies
of
dependencies,
but
it
wouldn't
be
lovely
at
all.
It
would
be
madness
and
chaos
because
we
then
don't
know
like
what's
breaking
whose
fall
there
is
like
what
needs
to
be
done
to
fix
it.
So
we
shouldn't
do
that
and
instead
we
should
ask
these
people
to
instead
depend
on
a
version,
because
that
way
they
guarantee
that
they
work
with
that
version,
and
then
we
can
guarantee
that
they
will
work
with
the
version
that
we
want
it
to
work
with,
and
otherwise
everyone
will
just
go
nuts.
So
that
is
that
any
questions.
A
I
think
that
makes
absolute
sense
I
think
we
should
where
people
are
using
IP,
FSD
CTL
and
not
passing
in
an
implementation
of
ipfs.
We
should
definitely
tell
them
that
that
is
the
best
practice
and
make
that
front
and
center
on
the
IP
FSD
CTL
repo,
which
Hugo
I
think
you
said
you're
already
gonna
do
so.
We.
A
B
A
A
A
C
A
D
So
I
only
got
this
working
yesterday,
so
I'm
gonna
run
it.
So
it's
building
a
bunker
image
with
ipfs
in
it
spinning
up
to
dr.
instances
on
the
same
machine.
It
is
Trent
putting
a
file
in
one
peer
and
sending
it
to
the
others
here.
So
it's
all
synchronized
and
that's
done
so
that's
the
demo
and
kicks
out
JSON
and
with
the
JSON
we
can
do.
We
can.
We
can
make
really
get
that
to
run
on
multiple
machines
now
and
still
arbitrarily
big,
that's,
and
that
will
be
what
I'll
be
trying
to
do
this
week.
D
C
D
Just
try
to
get
it
right
on
multiple
machines,
but
there's
a
lot
of
code
review,
reusing
rolls
latest
stuff
and
he
has
to
take
a
look
at
it.
I
think
like
cuz
I,
don't
even
know
how
I'm,
using
that
you
guys
and
thinks
the
way
he
intended
them
to
be.
Yes,
so,
but
it's
pretty
cool
stuff
with
the
system
he
created
so
looking
forward
to
iterate
on
it.
E
I've
been
doing
some
changes,
this
soul
started
up
like
a
petting
for
like
a
much
more
opposed
to
steal,
async
version
and
also
the
little
setup
that
we
do
when
we
configure
CTL
then
pass
that
CTL
instance
to
the
interface
forth.
We
do
this
either
in
jessup,
EFS
or
HTTP
query
and
we'll
probably
also
will
do
it
in
interrupt
and
as
I
started.
Looking
to
porting.
E
All
of
this
to
facing
CTL
I
found
a
bunch
of
problems
stuff
that
we
can
confuse
sure
which
version
was
running
even
if
we
pass
by
th
to
be
client
from
Jessica
Fest.
So
he
tells
you,
okay,
I,
want
you
to
run
this
fortune,
not
your
own
version
and
stuff
wasn't
working.
All
we
couldn't
be
sure
of
which
version
of
it
should
be
client
in
Ibrahim,
JSF
EFS
was
being
run.
It
is
specially
in
the
browser
and
also
a
bunch
of
the
best,
the
interface
Cortez.
B
F
A
A
G
G
C
G
Your
own
test
point
also
related,
like
we
will
be
putting
more
effort
alright,
now
that
we
have
gone
through
this
iteration
and
understand
better
I'll
test
downwards
and
like
issues
more
will
be
able
to
make
more
progress,
I'm
like
specifying
a
test
plans,
and
so
it
might
happen
soon
that
you
might
hear
from
me
like
thinking
about
like
hey,
like
he's
there
at
that
planet,
where
you
want
to
see
written.
Do
you
have
time
to
contribute
to
test
run
by
writing
a
test
plan?
Can
we
make
this
useful
for
you
today?
G
Right
and
so
still
like
some
some
things
to
tune
up
like
so
as
you
just
like
heard
from
Jim
like
I,
just
got
the
tube
content.
Your
set
up
to
work.
We
need
to
check
in
with
our
role
like
as
mean
like
very
busy,
with
I
call
the
events,
and
so,
let's
make
sure
that,
like
we
didn't
take
the
project
liking
that
completely
like
off-track,
but
I
I
believe
like
we
are
in
the
right
track.
You
just
want
to
make
sure,
but
but
yeah
like
if
you
do
at
the
time
the
interest
agree.
G
Our
city,
like
I,
heard
from
Burke
on
the
previous
meeting
that
he
wants
to
test.
Swab
like
you,
are
in
work
because
that's
exactly
the
best
plan
that
Jim
is
writing
at
the
moment.
So
so
Jamie's
really
writing
the
best
time
to
check
out
like
what
Big
Shot
is
doing
and
yeah
like
another
thing
that
you
might
find
and
I
calculate
Jim
here
it
was
his
idea
because,
like
des
Khurana
as
I
go
SDK,
but
that
you
just
like
import
the
test
and
therefore
you
can
like
just
plug
into
the
whole
text
constructor.
G
G
So
if
you're
interested
on
like
that
Union
web
assembly,
and
if
you
want
to
give
it
a
try
to
compile
the
go
SDK
eating
to
web
assembly
and
see
if
you
can
get
Ike
a
test
plan
with
Jessica
vests
and
that's
like
a
very
fun
thing
to
do
and
yeah
like
as
long
I,
also
pointed
out
like
we
have
this
like
Trello
board,
a
Kanban
like
the
link
is
from
the
Remy.
Let
me
know
if
you
don't
have
access
to
it.
I
can
give
you
access.
A
G
G
That
said,
like
I,
don't
feel
it
such
as
like
a
a
top
priority
for
this
group,
because,
like
so
many
things
into
yeah
like
if
you
do
have
the
time
to
spend,
even
if
it's
just
like
10
minutes
per
helper
problem,
there
is
seven
total
to
thank
you
a
little
bit
over
an
hour
just
like
blessing
true
see
if
you
learn
something
new
that
might
be
the
case
or
I
see.
If
there's
like
something
that,
like
you,
it's
missing
there,
that
will
you
be
little
basically
useful.
A
G
A
H
Okay,
so
basically
subdomain
gateways
are
blocked
by
the
lack
of
support
for
PIDs
in
base
52,
and
for
that
we
need
to
implement
pure
this
as
CA
DS.
That's
working
I
linked
the
tracking
issue.
Basically
I
did
not
get
to
this,
but
I
want
to
at
least
find
some
time
this
week
and
implement
support
plus
test
in
a
way
that
does
not
flip
the
switch
of
the
default
representation
but
like
if
someone
passes
the
peer
ID
I
see
id
v1.
It
should
just
work
because,
basically,
like
Delhi
p2p
internally
workers
only
about
the
multi
cache.
A
Listen.
Thank
you
later,
quick
questions
for
Ledo,
no
ok.
Moving
on
then
to
distributed.
Signaling
is
still
on
hold
pending
async/await
refactor
still
the
case.
Yes,
yes,
okay
moving
on
then
to
IP,
NS
and
Aiden
is
out
for
holiday
okay.
So
we
will
skip
over
that
one
as
well.
Unless
people
want
to
say
anything,
Hugo.
E
I
E
C
Do
we
have
a
spec
for
I
penis
over
dere
house
I,
know
that
there's
a
this
was
implemented
in
JSI
P
of
s
and
and
the
aim
for
q3
was
to
have
that.
But
what's
the
latest
status
on
that
I
know
that
Dominic
had
built
something
in
go
but
that
we
needed
the
spec
in
order
to
make
sure
that
we're
building
the
same
thing
across
both
languages
and
then
that
kind
of
got
blocked.
Yeah.
F
C
F
C
A
A
Do.
D,
codec
and
I
forgot
that
we
I
think
there
was
a
decision
a
few
weeks
or
months
back
to
just
in
to
print
out
I've
killed.
You
draw
blocks
in
that
case
because
we
just
don't
have
the
information
anymore,
and
so
just
so
you
know
that
was
the
solution
there,
but
we
can
go
ahead
and
change
that,
and
that
means
that
we
have
no
longer
have
to
do
the
dual
CID
version.
Look
up
every
time.
A
We
want
to
get
something
out
of
our
repo,
we
don't
check
version
zero,
Co
DS,
and
if
it's
not,
there
check
version
one
co
DS,
and
if
it's
not
there,
it's
definitely
not
there,
and
if
it
was
there
on
on
either
version,
then
it
was
there.
So
that's
that's
the
performance
went
and-
and
it
gets
us
one
step
closer
to
CID
base
32
by
default.
So
that's
happening
any
questions
on
that.
A
A
A
Cool
okay,
so
any
questions
for
duck
before
him
move
on
his
work
should
make
bit
swap
infinitely
better,
rounded
down
somewhat,
maybe
twice
it's
good
times:
good
I!
Don't
anyway,
okay!
Next
up,
okay,
it's
this
one
again,
the
async/await,
refactor
Libby
to
be
Vasco
and
Jacob.
Would
you
like
to
click
up
there.
K
Yes,
so
we've
got
the
dialer
upgrader,
the
initial
part
of
that
has
been
merged
and
hooked
up
to
the
PDP,
so
you
can
dial
and
stuff
now
using
all
the
new
stuff,
which
is
great
my
area.
We
also
added
an
interface
for
crypto
this
week.
This
weekend
we
did
not
have
an
interface
for
it.
So
now
we
do
so
SEC
IO
can
migrate
to
that
and
we'll
also
be
doing
a
plaintext
to
based
all
of
that.
We've
also
migrated
all
of
the
interface
repos
into
a
new
jeaious
interface
repo.
K
So
all
the
async
refactor
can
just
pull
from
there
and
will
deprecated
all
of
those
those
old
repos.
We
just
have
one
place
for
the
interfaces,
and
then
we've
got
a
slew
of
pub/sub
PRS
out
and
ready
for
my
eyes
to
look
at
so
I'll,
be
reviewing
those
this
weekend,
giving
much
little
feedback
and
then
Bush
was
working
on
interpreting
that
into
the
p2p.
So
hopefully
we
should
have
pub/sub
done
this
week
and
then
we're
also
working
on
migrating.
The
pure
store
over
I
think
that's
mostly
it.
Oh
we're
pausing
on
WebSocket
star
for
right.
K
Now,
we'll
probably
pick
that
up
back
at
the
end
towards
the
end
of
the
refactor,
but
we
base
in
sweer
subset
sun
setting
the
protocol.
We
may
work
on
a
different
approach
to
get
us.
The
connectivity
like
circuit
relay
and
then
another
pure
discovery
system
that
we
lose
when
we
lose
WebSockets
are
but
we're
postponing.
That
now
is
the
lower
priority.
A
Nice
I
was
saying
in
an
earlier
meeting
that
we've
async/await
it
feels
like
the
beginning
of
the
end
of
things,
have
bubbled
up
significantly
Lippe
to
be
I.
Think
in
the
according
to
the
issue,
there's
something
like
8,
8
repos
that
are
in
progress
for
async
refactor.
There
know
how
complete
that
is,
but
there's
nothing
there's
nothing
more
to
star
on
they
are.
A
It's
been
a
long
slog,
so
thank
you.
Everyone
who
has
contributed
because
couldn't
have
done
it
without
you
there's
something
like
67
replays,
alright,
ok,
let's
move
on
so
oh
yeah,
so
the
news
I
had
was
that
I
did
make
a
start
on
this
for
and
for
jess
IVFs,
and
I
did
some
work
refactoring,
the
kind
of
boot
logic.
So
when
you
first
created
a
node
and
so
like,
we
have
distinct
kind
of
states
for
initialized
and
started
and
stopped,
and
what
it
does
it
kind
of
restricts
the
AP.
A
A
It
also
adds
ipfs,
add
in
the
new
kind
of
new
API
style
and
as
in
using
async,
actually
using
async
iterators
as
part
the
API.
So
if
you're
interested
in
that
go
and
check
out
the
br
and
yeah,
there
are
a
bunch
of
other
new
things.
So
Pedro,
you
said
you
put
down
your
engine,
yes,
you've
been
working
on
it
and
drop,
and
interface
cortés
and
I
have
a
mental
note
to
get
to
that
and
review
them.
A
A
A
I
Be
brief,
so
mainly
I've
just
been
iterating
on
feedback
with
the
poll
requests
related
to
some
of
the
amount
stuff,
mainly
there's
some
ambiguities
like
we
have
some
flexibility
as
to
how
we
can
do
some
things
in
regards
to
our
plugin
the
api
and
like
the
actual
protocol
that
we're
using
for
the
file
operations
and
stuff.
So
things
work
right
now,
but
they
could
work
another
way
that
may
be
better.
So
some
of
that
stuff
just
has
to
be
discussed
and
and
like
written
down.
I
So
our
users
know
what
to
expect
and
we
can
actually
write
tests
around
it
and
stuff
like
that.
So
that's
been
on
my
mind
and
then
I've
also
been
doing
a
lot
of
upstream
stuff
in
regards
to
like
portability,
adding
a
lot
of
platform
support
an
architecture
support
for
some
of
the
things
that
we
depend
on
and
writing
and
debugging
tests,
so
that
I
can
go
from
saying
it
should
work
to.
It
does
work
and
that's
wrong.