►
A
Hi
everyone
welcome
to
the
ipfs
core
implementations
weekly
thing
for
monday,
the
11th
of
january
2021
welcome
back
after
the
break.
This
is
my
first
day
back.
I
am
looking
forward
to
doing
some
awesome
stuff
this
year.
As
I'm
sure
everyone
is
we're
going
to
go
through
our
regular
agenda,
which
is
high
priority
initiatives,
other
initiatives
and
then
the
parking
lot
questions
any
other
business.
That
kind
of
stuff,
so
kicking
us
off
is
upcoming
and
shipped
releases.
B
Something
I
missed
last
week
that
happened
over
the
break.
Go
live,
pdp,
0.13,
released
december
22nd,
or
something
like
that.
That
does
include
some
breaking
changes
technically,
but
they
shouldn't
affect
people
unless
you
are
opening
streams
directly
on
the
connection,
so
just
to
note
there
but
yeah.
I
think
we're
rolling
that
up
or
there's
a
pr
open
to
include
that
in
in.
B
C
A
A
A
That
is
it
for
upcoming
and
ship
releases
and
we
can
move
on
to
the
next
item,
which
is
pinning.
C
Services
on
the
go
side,
I
think
this
is
just
landing
the
mfs
pinning,
which
should
be
happening
this
week.
E
Yeah
and
on
the
gui
side,
I
believe
web
ui
will
wire
up
everything
up
to
the
mfs
and
then
we'll
have
that
commented
out
and
then
wire
it
up
when
it
lands.
I
think
the
configuration
keys
are
stabilized,
so
we
could
assume
they
are
correct
right.
Okay,
so.
C
E
Yeah
so
yeah,
if
it
lands
in
rc
that
then
we'll
write
it
up.
If
not,
we
can
comment
it
out
for
now
and
do
that
later.
F
A
C
Right
yeah,
I
mean
yeah,
ideally
they're
right.
If
we
want
the
web
ui
in
the
final
go
ip
fast
audit
release,
then
we
need
the
http
client
in
beforehand,
so
it
can.
It
can
make
its
way
there
if
anything
changes
like
if
you
know
it
turns
out,
like
we
made
a
mistake.
That's
it's
fine,
because
you're
just
going
to
get
like
an
error
right
of
like
this
thing,
isn't
supported
yet
because
you're
using
an
older
version
of
go
ipfs
and
then
we'll
update
it,
and
then
things
will
work.
A
Okay,
next
up
is
local,
pinning.
C
Yeah,
I
think
things
are,
things
are
pretty
solid
there
and
the
only
thing
currently
is.
We
want
to
do
some
testing
on
evaluating
like
how
long
it
takes
to
do
a
migration
with
lots
of,
if
you
already
have
lots
of
pins
in
your
data
store,
it
shouldn't
be
that
big
a
deal
but
we'd
like
to
be
able
to
set
expectations
with
the
audit
release
as
to
how
long
this
is
going
to
take.
If,
in
fact,
you
have
a
lot
of
pins.
D
That
information
will
be
available.
I
should
have
that
finished
today,
just
going
through
and
producing
it,
something
that
you
can
graph
and
look
at
and
see
what
the
expected
behavior
is
that
way.
If
we
have
anything,
that's
way
outside
of
what
that,
what
that
curve
looks
like,
we
can
see,
there's
something
maybe
on
their
site.
That's
wrong.
D
C
I
mean,
I
think,
once
we
have
stuff
and
and
if
it
looks
if
it
looks
like
what
we're
seeing
internally,
then
then
we're
good
and
we'll
just
like
make
that
we'll
just
add
it
to
the
0.8
release,
notes
github
issue,
all
right,
but
yeah
we'll
just
make
sure
everything
checks
out
internally.
C
C
Hannah
and
alex
are
still
out
so
nothing
there.
Although
yeah,
I
have
something
related
to
data
transfer,
but
we
will
do
that
at
the
end.
A
Neato
js
improves
discoverability
and
connectivity.
G
I
also
created
a
benchmarking
setup,
I'm
just
finishing
some
things
about
it
and
creating
some
docs
so
that
tomorrow
I
can
start
benchmarking,
the
server
itself
and
so
yeah
for
this
week.
I
want
to
finish
the
benchmarking
and
start
working
on
the
examples
in
production
guides,
and
it
would
be
great
if
I
could
get
a
review
for
the
server
this
week
and
that's.
A
It
next
up
is
bi-directional
streaming
and
streaming
areas
in
the
browser,
so
this
got
merged
just
before
christmas
and
it
will
go
out
in
the
next
slide.
Fs
release
along
with
a
blog
post,
explaining
the
new
shiny
and
the
fun
things
people
can
do
with
it.
A
H
C
Yes,
they're
it
badger,
2
well,
the
latest
release
will
then
be
badger.
3
and
the
badger
team
has
said
they
don't
really
have
much
interest
in
doing
anything
other
than
maintaining,
like
anything,
anything
that
is
non-critical
fixes
to
a
version
of
badger
that
is
not
supported
by
d
graph,
so
seems
like
there's.
No
particularly
good
reason.
Why
ipfs,
if
it's
not
using
badger
to
now,
should
continue,
should
use
it
at
all
and
should
just
switch
to
badge
and
just
go
straight
to
badger.
3.
C
we'll
probably
need
to
support
both
for
a
little
while
in
terms
of
like
letting
people
test
out
the
stability
of
badger
3
and
see,
if
they're
happy
with
that,
and
we
also
need
to
set
up
a
system
for
having
multiple
versions
of
data
stores
that
are
the
badger
data
store.
Having
a
single
repository
for
each
version
for
each
major
version
seems
like
a
bad
idea,
which
is
what
we
currently
do.
C
If
you
are
interested
in
that,
I
guess
follow
the
go,
ds,
badger
repo,
where
some
discussions
will
be
happening.
C
C
A
C
Irish
is
back,
he
is
working
on
things.
He's
yeah
he's
talking
with
stephen.
I
think
today
about
some
of
the
multi-stream
issues
and
we're
gonna
try
and
do
some
strategizing
on
what
is
like
the
simplest
way
to
get
at
least
some
of
the
natural
reversal
stuff
done
so
that
we
have
something
soon.
A
Cool
next
up
unit,
suppressively
1.5
and
guide
pfs,
look
like
there's
been
any
update
on
that
ticket.
The
issue
hasn't
really
had
any
comments
on
it
since
the
beginning
of
december.
So.
A
It's
gone
cold
guy
for
sgc
improvements
is
the
next
thing.
D
All
right,
I'm
still
working
toward
seeing
what
we
need
to
do
for
performance
improvements.
As
far
as
dag
awareness
it
looks
like
it
will
actually
work
without
being
gag
aware.
So
that's
a
that's
good
news
and
this
need
to
assess.
Are
there
any
major
performance
impacts
depending
on
how
we
delete
blocks
within
that
dag,
because
that,
if
it
doesn't,
it
still
might
make
sense,
we
need
to
be
aware
just
because
of
performance,
but
it
doesn't
appear
to
actually
break
anything
catastrophically,
which
is
the
good
news.
D
Yes,
last
last
week,
I
finished
all
of
the
modifications
to
the
distributions
site
scripting
and
all
that,
so
it
should
all
be
reviewable
this
week,
there's
a
chunk
of
stuff
to
review
and
it's
not
as
high
priorities
other
things,
so
it
may
not.
We
may
not
get
to
it
right
away,
but
we'll
have
everything
ready
to
go,
certainly
for
the
next
next
release
and
if
we
need
to
backport
it
we
could
for
this
release,
but
at
least
that's
what
you
know.
This
will
be
the
worst
cases.
D
D
Just
just
a
matter
of
review.
A
Great
stuff,
that
is
it
for
the
other
initiatives,
so
the
next
section
is
design
review
proposals
which
is
empty
yeah.
C
Yeah,
I've
got
something
so,
as
I
have
mentioned
previously,
put
a
little
bit
of
thought
into
into
this
issue.
Around
bitswap
has
a
maximum
block
size.
The
maximum
block
size
is
frustrating.
If
you
have
some
already
content
address
data
and
the
blocks
are
bigger
than
one
megabyte
right,
you're
importing
git
blocks
and
some
of
the
git
blocks
are
bigger
than
one
megabyte.
So
now
what
do
you
do?
Can't
use
bitswap?
C
So
what's
the
plan,
so
the
that
proposal
is
one
way
to
go
about
solving
that
there's
a
second
proposal
which
is
related
to
it,
which
says:
hey
well,
if
you
can
already
do
if
you
can
handle
transferring
very
large
blocks,
what
if
you
just
decided
that
all
files
were
just
single,
very
large
blocks.
C
C
These
reviews
need
some
thoughts.
It's
like
a
pretty
different
way
of
thinking
about
things,
and
so
it
may
be
that
this
is
a
bad
idea.
If
it
is
a
bad
idea,
then
please
say
your
comments
so
that
we
don't
do
something
bad.
C
C
Appreciated
also
learned
some
interesting
things
about
how
blake,
3
and
kangaroo
12
could
be
like
super
useful
to
us
in
the
future,
because
they
are
tree
based
hashes
and
we
do
trees
with
hashes,
so
could
be
like
super
useful.
A
Check
out
the
pr
next
up
is
blockers
and
asks
looks
like
lidar's
got
something.
E
Yeah,
just
a
quick
shout
out:
if
you
have
some
spare
time,
install
brave
beta
or
nightly
as
long
it's
it's
1.19,
it
will
have
some
like
built-in
support
for
our
uris,
ipns
and
ipfs
console
slash
so
just
play
with
it.
E
I'm
not
saying
anything
specific
because
it's
useful
to
have
feedback
from
people
who
don't
have
any
expectations.
Just
if
anything,
look
weird
or
does
not
behave,
how
you
would
expect
it
send
feedback
my
way
either
dm
me
or
feel
an
issue
on
in
ipfs
in
web
browsers,
repo
and
I'll
try
to
consolidate
that
feedback.
We'll
try
to.
E
I
don't
expect
big
changes
for
the
initial
release
when
this
initial
support
lands
in
brave
stable,
but
we
will
be
iterating
on
that.
So
for
now
we
want
to
push
this
basic
support
for
resolving
those
uris.
So
that's
the
goal,
so
people
can
resolve
those
uris
either
by
public
gateways
or
local
gateway,
local
ipfs,
node
managed
by
brave,
but
any
feedback
is
welcome.
So
just
let
me
know.
I
And
I
just
wanted
to
note
that
there's
an
ipfs
community
security
meetup
on
wednesday.
This
came
out
of
the
set
of
github
issues
around
themes
for
2021,
one
of
which
being
privacy
and
security
and
a
bunch
of
community
groups
building
on
ipfs
expressed
interest
in
various
forms
of
what
that
might
mean.
I
So
we're
going
to
have
them
all
show
up
and
do
five
minute
things
on
either
what
they've
done
in
regards
to
their
understanding
of
privacy
and
security
on
ipfs
or
what
requirements
they
want
and
then
have
hopefully
a
I
don't
know
I
mean
for
like
half
an
hour.
Discussion
of
you
know
what
ends
up
coming
out
as
themes.
The
most
important
privacy
and
security
work
from
there
to
keep
external
groups
involved.
F
I
have
asked
for
alex
alex
if
you'll
have
a
chance
to
look
at
the
pub
sub
issue.
I'll
put
a
note
and
share
your
opinion
would
be
helpful.
I
propose
specific
changes
to
the
pop
sub
api.
We
have
in
js
ipfs
and
it
would
be
good
to
know
if
you're,
like
them
or
not,.
A
Okay,
I
will
look
at
that
and
respond
thanks
if
that
is
it
for
the
blockers
and
asks
we
move
on
to
questions
who's
got
a.
I
C
I
guess
a
fun
just
bug
I
ran
into
with
the
with
the
crawler
that
I
have
a
apr.
I
have
to
put
up
a
pr
to
fix,
is
multi-hat
multi
adders
are
fun
and
by
fun
I
mean
somewhat
confusing,
because
sometimes
they
end
in
slash
p2p
slash
some
peer
id
and
sometimes
they
don't
and
gilda.
P2P
expects
them
to
not
end
in
that,
because
the
pid
will
come
separately
as
part
of
a
adder
info
object.
C
If
you
aren't
careful
about
this,
you'll
end
up
with
peers
that
are
undialable
or
that
seem
undoubtable,
even
though
you
have
their
addresses
because
those
addresses
don't
seem
valid
so
that
yeah,
you
got
to
be
careful
if
you're
working
with
them
do.
C
Terrifying,
I
suspect
that
there
is
some
there's.
Certainly
a
bug
is
the
bug,
the
fact
that
the
older
version
of
ipfs
was
putting
them.
It
was
putting
the
you
know,
the
slash
p2ps
at
the
end
and
the
dht
somewhere
and
other
ones
were
not
and
we're
all
playing
in
the
same
space
and
that's
causing
problems
sure,
and
so
we
could
have
like
a
more
uniform
dht
and
that
would
fix
things
also.
We
should
probably
be
more
lenient
and
or
more
careful
when,
when
working
with
multi-address.
F
A
If
there's
nothing
else,
then
I
think
we're
done.
Thank
you
for
coming
everyone.
This
has
been
the
ipfs
core
implementations
weekly
sync
from
monday,
the
11th
of
january
2021.
Please
fill
in
your
async
updates,
so
people
know
what's
going
on,
because
it's
really
useful.
I
will
see
you
all
next
week
have
a
have
a
fab
week.
Everyone
bye.