►
From YouTube: ⚡️ⒿⓈ Core Dev Team Weekly Sync 🙌🏽 April 29, 2019
A
Okay,
good
afternoon
evening,
good
morning,
whatever
time
it
is
for
you
welcome
to
the
GS
core
dev
team
weekly
sync,
it
is
April
the
29th,
it's
2019,
it's
very
nice
to
have
you.
If
you
are
here,
please
could
you
put
your
name
on
the
attendees
list
on
the
hackpad
I
will
place
the
link
in
the
chat
now
I
may
take
it
today
is
Jacob.
Thank
you
for
volunteering
and.
A
A
Gosh
my
list
was
really
empty
for
some
reason,
so
I
don't
think
I
did
anything
but
I
put
down
something,
so
it
looks
like
I
was
busy
but
anyway.
So,
finally,
finally,
okay
I've
been
ill
for
like
two
and
a
half,
maybe
three
weeks,
so
it's
been
horrible,
but
I
finally
find
me
finalized
the
queue
to
a
chaos,
I've
transferred
them
to
the
spreadsheet.
I
recorded
a
video
today
that
is
kind
of
done
and
dusted
for
this.
This
quarter,
so
we
can
just
get
on
with
it,
which
is
good.
A
Finally,
and
I've
also
been
working
on
a
quartet
called
it
M
Plex,
which
is
a
JavaScript
implementation
of
M
Plex
using
async
iterators.
It's
towards
the
async,
is
rated
async/await
endeavor
and
it's
basically
the
bottom,
the
very
bottom
of
the
stack,
and
hopefully
everything
else
can
kind
of
build
on
top
of
it.
But
what
implicates
stream
multiplexer?
So
you
use
one
stream,
but
you
can
actually
have
lots
of
streams
interacting
over
that.
Just
that
one
stream,
which
is
great
so
initials
tests.
It
looks
to
be
around
20%
faster
than
lipids
p.m.
A
tricks,
the
one
that's
currently
in
production
and
about
10%
faster
than
pol
em
Plex,
which
is
the
one
that
will
soon
be
in
production.
So
that's
good.
It's
also
way
smaller
that
there's
only
about
250
lines
of
code
which
which
it
just
makes
it
a
bit
more
maintainable,
makes
it
a
bit
more
easy
to
understand
and
that's
when
comparisons
around
$700
occurred
in
the
PTM
flex,
which
is
pretty
tricky
to
pick
free.
Once
when
I
first
started
looking
at
it,
so
you
can
see
here
the
results.
A
A
So
what
I've
done
I'm
not
locked
on
anything
I
need
to
do
some
more
testing
for
it.
Mpx.
It's
I've
actually
used
it
in
the
interruptus,
a
built
in
adapter
class
so
that
it
can
look
like
the
old
interface
and
the
the
exchange
files
interrupt
tests
pass
and
they
are
with
comparable
kind
of
performance
to
was
there
before,
which
is
not
really
surprising
because
it's
having
to
go
through
a
bunch
of
extra
hoops.
A
So
that's
good,
yes,
so
testing
for
that
I'm
going
to
send
a
pull
request
for
interface
stream
Luxor,
because
I
made
some
sorry
I
miss
some
interface
changes
to
that
module,
which
will
hopefully
be
acceptable
but
we'll
send
a
PR
and
see
what
you
think
and
if
I
have
time
for
the
remainder
of
this
week,
I
will
be
going
back
and
looking
at
trying
to
enable
the
HD
enjoy
5
FS
for
reals
this
time
and
cool.
So
that's
me
any
quick
questions.
A
Okay,
alright
moving
on
to
so
Dirk
is
not
attending.
He
is
in
the
wrong
timezone
for
this,
but
he
has
been
working
on
adding
the
refs
endpoint
ipfs,
which
is
really
good.
Refs
endpoint
allows
you
to
give
it
a
CID
and
it
will
look
through
the
links
and
give
you
back
all
of
the
links
for
that
particular
CID.
A
A
So
pinning
is
an
interesting,
interesting
tool
because
it
doesn't
nothing's
going
to
get
connected.
So
anyway,
let's
move
on.
It's
also
unable
to
join
the
call
he's
being
working
on
some
gateway
improvements
to
Joe,
so
PFS
streaming
conditional
and
range
request
that
schools
not
blocked,
and
next
he
is
looking
to
expose
I
KN
s,
HTTP
gateway,
identify
what
is
missing
and
sure
that
is
exactly,
but
if
you're
interested
then
go
and
ask
him
good,
that's
low,
let's
move
on
this
Jacob
next.
Would
you
like
to
show
your
update
with
us?
Yes,.
B
So
in
the
Lippe
to
be
switch,
one
of
the
things
we
were
doing
is
adding
observed
addresses
to
our
own
peer
info.
While
this
is
great
for
letting
people
know
what
our
external
addresses
are.
We're
not
doing
any
validation
on
that
we're
just
slapping
every
observed
address
on
there,
which
gives
us
a
very
long
improperly,
probably
very
incorrect,
address
list
that
people
then
have
to
dial
through
so
we've
shut
that
off
for
the
time
being,
until
we
do
a
better
job
of
actually
doing
some
form
of
validation
before
telling
people.
B
B
A
lot
of
this
further
improvements
will
need
to
come
through
network
latency,
but
major
things
that
we
did
was
actually
implement.
Esca
d'emilia
in
terms
of
more
peers,
more
dishes,
showing
paths
in
more
initial
query
peers,
instead
of
just
going
with
3
peers
to
start
it
off
so
that
sped
things
up
a
lot.
Another
thing
was
slow
in
offline
timeouts
right
now.
We
don't
really
have
any
abort
ability
in
the
switch
which
we're
working
on
with
the
async
iterator
stuff,
and
that
creates
a
problem.
B
If
we
get
a
node
that
has
a
bunch
of
addresses
like
JSP
to
be
switch,
libraries
giving
us
a
bunch
of
incorrect
addresses,
it
can
take
us
a
very
long
time
to
dial
them
all.
So
we
can
see
like
two
minutes
or
more
trying
to
dial
a
single
peer,
so
it
ends
up
blocking
up
the
queue.
So
we
added
some
timeouts
in
there
to
avoid
super
long
time.
Outs
on
that.
So
if
you
are
interested
in
seeing
more
about
that,
you
can
check
out
that
PR.
B
That's
just
about
ready
to
go
I'm
going
to
be
doing
some
CPU
testing
against
some
of
the
configuration
options
there.
We
created
a
basic
simulation
in
the
DHT
so
that
we
can
run
future
changes
and
see
what
kind
of
performance
we
have
there.
So
right
now,
it's
just
doing
like
simulation
on
network
latencies
that
are
relatively
accurate
to
what
we're
seeing
now
against
the
live
network,
so
yeah
feel
free
to
poke
around
there.
B
If
you
want
to
I
this
week,
I'm
gonna
be
working
on
pub/sub
issues,
namely
for
pull
in
plex,
but
I
encountered
an
issue
when
I
was
refactoring.
The
interrupt
tests
or
go
publishing
to
Jas
does
not
work
and
that's
against
master.
So
I'm
going
to
be
looking
at
that
this
week,
along
with
the
DHT
CPU
balanced
against
JSI
BFS.
Does
anybody
have
any
questions.
A
C
Ta
is
yes,
sorry.
Some
of
my
window
manager
created
like
1000,
error
messages
somewhere,
okay,
anyway,
hey
so
my
update
is
I've
polished,
the
IPL
D
for
my
cleanup
PR.
So
now
there
should
hopefully
be
really
ready
to
be
merged,
so
another
ton
of
reviews,
but
please
look
good
and
also
more
interfaces
called
synchronous.
So
the
only
a
synchronous
method
now
is
calling
a
cid,
because
you
need
to
calculate
everything
else
is
synchronous,
because
all
our
implementations
only
use
this
synchronous
version
of
it.
C
This
should
hopefully
even
speed
up
the
protocol
buffers
stuff
a
bit
on
the
feelings
of
s
so
we'll
see,
and
it's
almost
done
and
then
next
is
I
open
just
before
the
meeting
a
PR
to
finally
remove
this
rich
description
of
having
packages
in
a
jeer
linting
that
make
it
just
below
version.
One
point:
something
needs
tilde.
Instead
of
a
carrot:
I
guess
it's
super
annoying.
C
If
you
implement
install
and
then
you
need
to
change
the
package
station
manually,
so
this
was
like
a
15
finis
job,
but
it
just
probably
save
everyone
in
this
call
like
an
hour
a
week
or
something-
and
next
is
this-
is
more
like
a
costume
update,
because
I'm
not
really
deep
into
IP
LD
I
work
on
a
thing
called
stack,
it's
something
in
the
Geo
world.
So
if
you're
interested
just
click
on
the
link
and
read
about
it,
so
it's
not
really
that
interesting
for
the
core
team,
but
yeah.
A
Any
questions
for
them.
Ok,
you,
okay,
just
see
your
ad
on
your
appoint
for
the
allowing
carrot
in
pre
1.0
versions.
It
will
save
us
time,
but
we
know
the
rule
so
we're
just
like
dad
forgot
the
rule,
but
it's
mostly
contributor
experience
and
people
who
are
new
to
Jessup
confess
that
this
really
helps
when
you
come
along
and
get
your
PR
failed
for
not
using
a
particular
upgrading
or
installing
a
new
dependency
or
something
and
not
not
using
the
right.
The
right
character
is
is
really
annoying
and
it
means
you
have
to
go.
D
Yeah
say
it
like
only
applies
to
use
the
curry
for
pre.
No
point
one
point:
no
Saudis
told
her
for
pre
nor
point
one
point
nor
right.
So
the
changes
to
at
the
moment
it's
used
the
tilde
for
everything,
pre
one
point:
no,
and
the
changes
to
use
tilde
for
pre.
Nor
point
one
point:
no
alright,
in
short,
like
I,
think
whatever
NPM
does
like
and
not
because
did
yeah,
because
the
developer
experience
is
poor.
There.
C
What
I
did
is
I
just
make
it
change,
basically
that
the
current
behavior
is
the
same
or
like
the
same
like
so
so
it
doesn't
change
anything
on
the
versions
we
wanna
have
I.
Think
making
it
whatever
NPM
does
is
a
separate
discussion.
So
we
see
this
one
is
a
backwards
compatible
change
to
whatever
we
do.
Curly.
A
E
Hi
guys
so
last
week,
I
have
some
holidays,
so
I
didn't
do
much.
I
still
managed
to
work
a
little
bit
on
the
happiness
resolve
command.
Those
peers
are
still
not
merged.
Hopefully,
I'll
get
back
to
them.
This
week
also
works
on
the
on
adding
support
for
the
file
Dom
API
to
JSP
face
it.
She
be
client
and
the
interface.
E
E
I
start
to
start
to
the
bike
shared
on
slack
about
where
to
put
this
shared
logic
and
yep,
we
all
probably
make
an
issue
or
just
shows
the
name
and
make
a
new
ribbon
put
stuff
there.
So
we
don't
spend
too
much
time
talking
about
naming
and
bike
sharing,
I
also
fixed
an
issue.
Actually
Ollie
did
most
of
the
work.
I
only
cleaned
cleaned
it
up
a
little
bit
and
dig
into
it.
So
I
better
understand
the
problem.
Basically
one
of
my
packages
that
transforms
a
pole
stream
to
a
stream
had
about.
E
E
F
That
apps
could
be
a
lot
of
things.
So
it's
just
discussions
at
this
point.
What
we're
going
to
do
and
for
Falcon
try
day
I
did
a
fun
little
project
click
on
and
it
makes
up
funny
combinations
of
words.
It
sort
of
I'm
trying
to
do
sort
of
like
if
you
ever
use
Heroku
and
it
makes
up
random
names
for
your
project
I'm
trying
to
do
it.
Protocol
labs
flavor
of
that
coming
up
after
you
spend
some
more
time
on
pure
base
to
get
that
so
I
can
launch
it.
F
The
pinners
got
some
issues:
I
get
into
production
stable
place.
We
can
live
more
test
lab
or
infest.
Camp
I
want
to
set
up
for
pure
pad
I
want
to
set
up
a
weekly
call
where
people
volunteers
can
just
hang
out
sort
of
our
office
hours
sort
of
thing
and
on
the
project's
working
group
I'm
doing
something
called
integration
mini
projects
and
trying
to
find
ideas
for
little
one-week
projects
it.
We
could
do
ok.
A
D
H
D
A
D
G
D
Know
I
just
been
deleting
the
tests.
I
mean
it's
great,
like
everything's,
so
fast
so
easy
to
do
releases.
Yes,
all
the
functionality
is
intact,
so
I've
been
yeah.
So
I
did
the
PR
for
basically
been
working
on
NFS
the
importer
in
the
exporter,
so
MFA's
has
done.
I've
got
like
enough
of
the
important
exported
working
locally
to
do
when
I
first.
So
then
I
went
back
and
did
the
exporter
properly
now
I'm
doing
the
importer
properly.
So
the
exporter
PR
is
in
the
importer.
D
D
I
also
opened
a
PR
on
Asia
Asia
to
other
linting
rule
to
require
the
wait,
keyword
in
functions
that
have
marked
async
I,
just
wondering
what
people's
feels
are.
Don't
tell
me
now
because
will
bike
shed
for
ages,
but
do
comment
on
the
pier.
The
reason
is
because
it's
super
easy
to
just
throw
async
everywhere
and
when
we
do
that,
we
create
micro
tasks
that
get
executed
out
of
sequence
in
v8
or
at
least
at
the
end
of
the
tick,
which
creates
a
tiny
amount
of
latency.
So
for
really
hot
code
paths.
D
This
can
amount
to
the
via,
like
the
v8
engine,
basically
doing
nothing
for
a
while,
which
is
something
the
near
form
like
discovered
in
our
code
base.
When
they
did
all
the
performance
analysis,
they
would
be
great
to
be
able
to
follow
through
on
their
recommendations
on
that
and
earning
the
linting
rule,
just
kind
of
automatically
stop
us
from
foot
gunning
ourselves
again
anyway.
D
So
there's
a
P
I
open
for
that.
The
links
in
the
notes
go
be
great.
To
have
some
opinions
on
whether
or
not
we
should
merge
that
because
I
know
Volquez
and
opinions
about
like
you
know
you
market
interface,
async,
and
it
says
this
returns
a
promise
which
is
like
stealth
typing,
so
it
should
be
burnt,
but,
like
I,
do
actually
yeah
it's
a
valid
point.
It
makes
the
ergonomics
of
the
API
is
a
bit
more
a
bit
more
nice.
D
So
I'm
gonna
write
so
there's
some
async/await
stuff
is
outstanding
on
that
which
will
unblock
the
repo
async/await
things
so
I'm
gonna
do
a
release
of
all
that.
Probably
tomorrow,
yes,
you
know
the
good
pacing
koi
only
moves
forward,
never
backwards
conquer.
That's
me.
Does
anyone
have
any
questions?
You
go
your
hand
is
in
yeah.
E
So
not
actually
a
question
but
related
to
the
linking
rules
in
Asia
we
already
have
a
back
like
a
backlog
of
some
of
them
like
the
consul,
one,
the
consular
one,
this
one
I'll
try
to
do
to
bundle
them
all
in
one
release
because
needs
me
a
major
because,
when
I
release
that
everything
will
like
fail
in
all
the
ripples,
so
it's
just
to
make
you
aware
all
the
maintainer
x'
that
when
we
do
that,
you
all
need
to
do
some
work
to
make
all
the
ripples
green
again.
The.
A
All
right,
thank
you,
Alex
just
add
on
the
async/await
stuff.
I
can't
remember
if
I
updated
you
all,
but
the
we
are
on
51%
of
the
modules
that
we
said
we
would
convert
to
async/await
for
the
the
weight
endeavor,
so
we
are
getting
there.
Thank
you
for
your
for
your
efforts.
So
far,
it's
been
way
off.
It's
really
good
to
see
the
progress,
so
thank
you.
We
have
like
a
few
minutes
left
for
Christine
updates.
So,
let's
just
quickly
rattle
through
that,
if
we
can
Ollie,
you
are
up.
First,
just.
G
Very
quickly,
it's
already
been
mentioned
on
the
call,
but
there
is
an
office
hours
call
for
one
hour
before
next
week's
ipfs
weekly
call
you
don't
all
have
to
come,
but
I'm
gonna
run
it
for
a
few
weeks
and
use
it
as
an
opportunity
to
answer
questions
from
new
contributors,
people
who
are
wrestling
with
how
best
to
get
started
with
IPS.
So,
even
if
you
don't
turn
up,
I
may
be
tapping
you
on
the
shoulder.
G
A
I
Yeah
just
well,
let
folks
haven't
met
me,
know
I'm
applying
for
the
Jas
core
team
technical
manager,
position,
I,
love
the
way
y'all
work
here
so
I've
been
in
a
lot
of
the
calls
was
my
first
time
in
this
call
that
particular
issue
I
found
looks
like
it
has
a
couple
of
PRS.
That
might
be
like
one
code
review,
which
has
already
been
done:
kind
of
one
refinement
away
from
being
merger.
Bowl
I
was
just
trying
to
look
at
places
where
I
could
lend
a
hand
that
are
you
know.
I
A
H
Protocool,
the
GUI
team
adopted
me
very
exciting,
Thank
You,
Allie
MF
s
tutorial,
has
a
new
sequence
of
lessons
proposed
by
Ollie
and
myself
feel
free
to
click.
If
you'd
like
to
see
it,
if
you
are
building
content
for
ipfs
camp,
which
do
you
think
will
include
a
proto
school
tutorial,
I
would
like
to
know
that
you're
planning
by
doing
that,
so
I
can
start
helping
you
early
instead
of
us
having
to
rush
at
the
end.
H
There's
a
link
here
to
open
an
issue
and
I
did
confirm
with
the
Vita
Molly
that
it's
okay
to
do
that
in
the
public
eye.
So
in
the
proto
school
repo
submitted
conversation
there
I
don't
know
what
I've
mentioned
on
this
column,
what
I
haven't,
but
there's
a
proto
school
weekly
fall
on
Thursday,
so
there's
a
link
there.
If
anyone
who
would
like
to
learn
more
or
join
authors,
chapter
organizers,
all
welcome
yep.
A
Okay,
cool.
Thank
you
very
much
and
we're
done
we're
on
time
and
we're
done,
and
this
is
well
okay.
We
should
really
go
because
it
is
6
o'clock.
Thank
you
very
much.
Everybody
for
coming.
It's
lovely
to
see
your
faces
again
and
I'll,
see
you
next
week
goodbye
good
luck
and
happy
idea.
Fessing
until
next
time,
bye.