►
Description
Read the meeting notes at https://github.com/ipfs/team-mgmt/issues/992#issuecomment-688410970
For more information on IPFS
- visit the project website: https://ipfs.io
- or follow IPFS on Twitter: https://twitter.com/IPFS
Sign up to get IPFS news, including releases, ecosystem updates, and community announcements in your inbox, each Tuesday: http://eepurl.com/gL2Pi5
A
Hi
everyone
welcome
to
the
ipfs
core
implementations
weekly
sync
for
monday,
the
7th
of
september
2020..
I
am
making
brain.
I
will
be
your
host.
We
are
going
to
go
through
our
high
priority
initiatives,
other
initiatives
and
then
block
titans
parking
lot.
Q,
a
that
kind
of
thing.
It's
going
to
be
awesome,
we're
going
to
find
out
what
is
happening
in
the
world
of
ipfs
on
this
a
wonderful
monday.
A
So
the
first
thing
to
talk
about
are
upcoming
and
ship
releases.
A
I
can
talk
about
jsrpfs
quickly.
We
have
shipped
jsrpfs
0.50,
which
is
incredible.
It
has
a
pinning
revamp,
which
is
gonna,
deliver
some
pretty
significant
speed
improvements
when
you're
pinning
things
which
you
are.
When
you
add
things
so,
if
you're
working
on
large
data
sets,
you
will
very
much
want
to
upgrade
to
this.
It's
going
to
make
your
life
a
lot
easier.
A
We
also
have
started
accepting
un8
arrays
as
inputs
for
things
which
is
really
cool,
so
we've
always
kind
of
accepted
them
in
the
shape
of
buffers,
but
internally
they've
been
converted
which
is
slow
and
annoying
and
like
the
best
way
to
do,
it
is
basically
to
pass
buffers
as
they
are,
which
makes
no
sense
in
the
browser,
because
buffers
aren't
a
thing
so
now
everywhere
we
accepted
buffers.
We
now
accept
you
in
eight
arrays.
A
The
difference
is
that
now
we
only
give
you
back
a
unit
array
for
things,
so
this
is
going
to
impact
you,
because
if
you
previously
coded
to
the
buffer
interface,
you'll
be
used
to
doing
things
like
tostring
and
passing
like
a
base64
in
there.
A
That's
not
going
to
work
anymore
because
we're
using
just
the
una
array
interface.
So
you
need
to
code
to
that
in
the
future,
and
your
code
will
run
in
node
and
the
browser
and
dino
and
wherever
javascript
takes
us
next
on
this
crazy
journey
that
we're
all
on
please
code
to
you
and
eight
arrays.
Also
ships
with
the
p2p
0.29,
which
gives
us
signed
peer
records,
gossip
sub,
1.1
and
ed2
5.
5
1
9
keys
properly,
which
is
great
so
please
do
check
it
out.
A
B
Do
that
yeah,
so
we've
got
a
few
bug,
fixes
that
we're
slating
we're
gonna
do
our
second
rc
for
the
0.7
release
this
week.
I
believe
that
should
be
the
last
rc
and
then
we
will
throw
0.7
over
the
fence
next
week
for
everyone
to
consume,
in
addition
to
the
psychiatric
deprecation,
there's
lots
of
extra
good
stuff
in
there.
So
I
think
we
have
a
meet-up
later
this
week
to
talk
about
it,
so
be
there.
C
Yep,
so
we
got
two
discussions
on
the
cosmetic
api
spec
changes.
One
is
about
clarifying
the
way
we
expect
error
responses
to
look
like
and
the
second
one
is
being
very
specific
how
a
name
filter
works,
namely
that
it's
like
partial
much
and
it's
case
insensitive.
I
think
that's
the
latest
consensus
and
we
are
having
a
meeting
tomorrow
to
go
over
integration
between
go
ipfs,
pinning
api
integration,
pinning
clients
and
I
believe,
the
way
web
ui
will
consume
that,
as
that
will
be
the
first
user
of
that
api.
A
B
Yeah,
so
for
the
client,
I
think
dean
ended
up
generating
using
something
like
the
this
swagger
because
they
have
the
open
api
generators.
So
he
just
generated
the
client
and
then
created
an
adapter
on
top
of
that
to
abstract
out
that.
So
there
are
some
pretty
pretty
decent
generators
for
for
js
as
well,
and
I
think
there's
like
a
nice
pretty
decent
like
prism
test
client
that
mocks
it
out,
but
yeah
would
be
good
to
do.
A
Approach
makes
a
lot
of
sense,
cool,
moving
on
ed
two
five,
five,
one,
nine
keys.
B
Yeah
no
update
on
those
it's
still
on
track
to
land
in
go
ip
episode.
Seven
next
week
same
thing
with
sec,
io
removal,
nothing
new
there
it's
happening
it's
coming
so
we'll
also
need
to
like.
I
need
I'll
open
up
a
pr
this
week
to
actually
remove
that
from
from
js.
So
we
can
launch
that
in
the
next
release.
A
Awesome
next
thing
on
the
list
is
the
rust
ipfs
initiative
week,
37
in
big
brother
house.
D
Yeah
well,
first
off
there
was
a
really
great
community
contribution
that
landed.
A
lot
of
you
probably
know
rudiger
from
actics
contributed
the
dns
lookup.
So
we
can
resolve
our
domains
now
from
within
rust,
fps,
which
is
really
cool.
D
But
apart
from
that,
we
just
finished
a
pretty
big
cycle
of
work
which
lands
us
pinning
dht
dns
and
a
very
rudimentary
file
system
store
for
blocks
based
on
multi
hashes,
which
is
really
cool,
so
we're
catching
up
and
the
performance
of
the
rust
ipfs
is
admittedly
slightly
slower
than
but
comparable
to
the
go.
Implementation
of
adding
large
data
sets,
and
things
like
that.
So
that's
where
we're
at.
E
B
D
No,
we
I
mean,
like
a
lot
of
the
ipfs
stuff
we
just
implement,
whatever
implementation
of
p2p
is
on
the
ecosystem,
so
we've
just
kind
of
wired
up
the
rustler,
p2p,
cademlia
and
and
all
that
stuff.
So
one
thing
that's
name
is
autonet,
so
dht
publishing
is
like
we're.
Not
we're
not
really
sure
if
we're
actually
publishing
anything,
because
we're
not
doing
any
nap
punching,
but
there's
a
pr
in
place
for
that
in
the
p2p.
So
we're
just
kind
of
waiting
around
for
that.
E
A
Not
gonna
be
around
next
item
on
the
list
is
js
improves,
discoverability
and
connectivity.
F
Yeah
so
for
the
first
part,
the
author
relay,
I
got
back
to
to
it
last
week
and
I
basically
finished
the
review
that
jacob
provided
me
a
while
ago,
and
I
added
some
other
stuff
that
I
intended
to
add.
But
for
the
first
cycle
of
feedback,
I
didn't
finish
it.
So
it's
now
the
first
iteration
it's
ready
for
review
jacob
will
hopefully
review
it
this
week
and
yeah.
F
So
the
next
iteration
on
the
author
relay
will
basically
include
what
is
missing,
which
is
updating
the
network
when
appear
multi-other
changes
so
that
we
propagate
that
information
to
all
the
other
peers
and
also
to
find
remote
realizing
the
network.
F
While,
basically
now
you
will
connect
to
appear
and
if
the
peer
as
the
relay
protocol
and
also
supports
up,
you
will
basically
add
it
as
autorelay
according
to
the
limits
that
you
can
configure
before,
but
yeah.
But
the
goal
is
to
also
try
to
find
realize
if
you
are
not
connected
to
them.
Regarding
the
connection
manager.
F
Overall,
I
spent
the
the
end
of
last
week
and
today's
morning,
basically
finishing
a
epic
issue
which
contains
basically
six
different
milestones
of
all
the
work
that
I
think
we
need
to
do
regarding
the
connection
manager.
F
Overall,
the
big
goals
here
are
basically
to
align
with
the
rendezvous
and
the
author
relay
work,
and
we
need
to
the
connection
manager
to
be
more
intelligent
so
that
we,
for
example,
don't
close
a
connection
that
we
have
with
the
auto
relay
node
that
we
are
buying
to
and
all
other
kind
of,
similar
flows,
and
also
this
is
really
important
for
the
dht.
That
should
be
our
work
after
this
and
so
yeah.
F
I
think
the
connection
manager
will
be
a
big
thing
for
the
next
couple
of
releases,
so
the
the
first
three
milestones
that
I
created
are
the
ones
that
I
think
we
will
need
to
tackle
first,
hopefully,
for
the
next
release,
which
basically
are
to
improve
the
watermark
observations.
We
currently
have
basically
a
naive
implementation,
where,
if
we
are
getting
to
the
maximum
threshold,
we
just
disconnect
from
a
peer
without
any
like
kind
of
knowledge.
F
For
example,
we
should
check
if
the
peer
that
we
are
going
to
disconnect
as
running
protocols
that
we
want-
or
it
basically
appear,
that
we
are
connected
to
result,
network
and
other
kind
of
characteristics
that
you
should.
We
should
check,
and
I
have
kind
of
described
them
in
the
milestones.
F
If
you
want
to
check
it,
and
then
we
also
need
to
guarantee
that
the
connections
that
we
want
to
are
kept
alive,
which
is
not
a
thing
right
now,
and
also
another
thing
that
go
already,
as
that
we
should
have,
is
basically
to
protect
connections
which
is
kind
of
what
I
talked
about
before
that
we
need
to
protect.
The
connections
on
the
peers
that
we
really
rely
on
other
example
would
be
if
we
want
to
be
attached
with
a
circuit
relay
node.
F
A
That's
awesome
that
is
it
for
the
high
priority
initiatives
so
moving
on
to
the
other
initiatives,
anything
on
the
sub
domain
gateway.
C
It's
in
support
for
ipns
names,
with
those
shorter
keys
scheduled
to
land
with
0.7,
and
it's
mostly
we
are
waiting
for
that
for
a
blog
posts
and
because,
like
if
we
post
a
blog
post,
then
we
expect
partners
like
cloudflare
to
also
support
that,
and
it
takes
time
to
trickle
down
all
those
updates.
So
it
will
be
a
few
weeks
before
that
lands.
But
that's
the
only
remaining
part.
I
think.
A
Cool
there's
been
no
update
on
unix
of
westerly
1.5
incarcerated.
As
far
as
I
know,
moving
on
to
the
pinning
system,
revamped
that
is
shipped
totally
shipped
all
aboard
in
a
shipped
which
is
awesome,
no
erratically
so
I'd.
Imagine,
there's
no
update
on
the
dac
service
or
improving
the
web
ui
file.
C
App,
I
can
give
a
quick
update
on
web
ui
file
ads
the
need
for
buffering
entire
thing
before
you
import
it
over
http
api
that
patch
by
iraq
landed
and
it's
scheduled
to
be
released
with
web
ui
211.
Hopefully,
this
week,
additional
improvements
around
progress
reporting
will
land
in
the
next.
C
A
Cool
and
that's
it
for
the
other
initiatives.
Does
anybody
have
anything
they
want
to
propose
for
a
design
review.
A
So
we
do
like
user
agent
sniffing
on
all
requests
that
come
in
via
the
http
api,
and
I
want
to
switch
the
electron
renderer
to
use
native
fetch
instead
of
the
port
of
the
node
http
module.
A
So
it
needs
a
an
update
to
the
go
ipfs
check
on
the
user
agent,
which
I've,
pr'd
and
nadine
has
just
merged,
and
hopefully
that's
going
to
get
out
in
the
next
rc.
So
it
will
be
in
0.7,
which
will
be
radical
because
it
will
neatly
halve
the
time
it
takes
to
run
electron
tests
in
ci
because
they
are
really
really
slow.
A
A
Cool
any
questions.
A
Anything
for
the
parking
all
right-
that's
amazing!
This
has
been
the
arguments.
Confrontations
weekly
sync
for
monday
september.
The
7th
2020
please
fill
in
your
async
updates
review
and
have
a
nice
time
on
the
internet.
Stay
safe,
don't
touch
your
face
just
because
we
bought
a
chronovirus
doesn't
mean
it's
over
see
you
all
in
the
future.