►
From YouTube: ⒿⓈ⚡️ js-ipfs Core Dev Team Weekly 2018-03-26
Description
A
Boom
welcome
everyone
to
the
Jessup
EFS
quickly
they
have
stand
up.
Today
is
March
26,
2008
een
and,
as
usual,
we
will
try
to
make
a
very
short
call
to
share
our
updates.
What
we
have
done
last
week,
what
we
got
plucked
and
what
we
are
planning
for
next
week
and
also
answer
any
questions
that
people
might
have
as
usual.
A
B
A
You
yeah
all
right
so
I'm
MVDs
and
my
bet
is
that
I've
advanced
on
the
leaf
peer-to-peer,
config
refactor
there's
a
lot
of
work
there.
A
lot
of
work.
That's
will
influence
the
overall.
We
factor
I'm,
just
doing
it
backwards,
so
there's
a
dumpster
as
usual,
review,
PRS
and
merges
and
releases
all
things
that
people
ask
me
and
yeah,
like
I,
didn't
manage
to
do
everything
I
had
planned
last
week,
just
because
of
end
of
the
quarter,
administrative
tasks
that
I
have
to
do
as
well.
A
A
I
think
it
will
be
very
beneficial,
especially
because
some
of
us
already
are
doing
that
and
I
want
to
make
explicit
so
that
the
contributors
and
the
maintainer
is
know
that,
like
they
always
assign
to
what
I
want
to
finish,
the
liquid
beer,
config
pull
request
and
I
want
to
work
with.
All
of
you
on
ok,
ours
for
quarter
two
and
my
proposal.
A
There
is
like
to
really
like
take
inspiration
from
all
the
web
browsers
in
desert,
like
we
start
with
a
poor
request
on
just
a
TFS
back
to
a
document,
and
we
can
use
that
document
to
share
feedback
to
each
other
and
then
once
we
feel
that
document
is
finished,
then
I
will
pour
it
over
to
the
spreadsheet.
Google
Spreadsheets
is
not
really
great
for
collaborative
editing,
because
the
comments
feature
doesn't
really
work.
Well.
B
B
It
was
combined
with
a
code
sprint
yeah
I
talk
a
lot
of
people
about
ipfs
and
also
noise,
the
other
project
I'm
doing
and
therefore
much
more
than
anything,
because
I
haven't
written
any
real
work,
and
but
next
is
working
on
the
Oakley
hours.
Obviously,
and
I
still
changes
for
the
tag
API
in
the
queue
for
the
CLI,
MDM
and
HTTP
API.
So
I
want
to
finish
those
and
I've
seen
that
they
are
quite
a
few
flaky
tests
on
Jas
ideas.
B
A
Last
suggestion
is
given
that
you're
going
to
tinker
with
tests
right
now
there
has
been
a
proposal
to
create
a
demon
pool
or
a
note
pool
one
of
the
things
that
makes
the
tests
take
a
lot
of
time
is
really
spawning
your
demons,
and
so
if
we
create
something
that
perhaps
at
the
FSD
CTL
and
just
like
spawns
like
ten
demons
or
whatever
the
number
of
demons
is
that
is
required
for
the
tests,
then
every
time
that
occurs
form
is
called
then
then,
like
you,
don't
need
to
wait
for
the
creation
of
it.
Dimitri.
C
A
Like
there
is
two
things
attacking
I'm
playing
the
RSA
keys
and
like
spawning
the
process,
creating
RSA
keys
is
definitely
probably
going
to
cut
off
all
the
time.
That's
already
spent
they're
pulling
the
demons.
I
was
not
proposing
to
reuse
the
demons,
but
just
like
to
create
a
demon
pool
that
is
already
available
so
that,
instead
of
like
you
waiting
ten
seconds
for
a
demon
to
be
ready,
thank
you
ma'am,
but
this
is
like
a
suggestion
for
your
Booker,
even
if
you're,
okay
with
that,
won't
make
your
work
easier
by
making
this
faster.
A
B
A
D
D
Well,
this
was
my
onboarding
week,
my
first
week
being
the
code.
It
was
a
nice
challenge,
I
must
say
so.
I've
been
a
small
pea,
are
fixing
winter
error
that
already
got
merged.
Then
the
vid
mentioned
me
in
an
issue,
so
I
started
working
on
that
I
published
a
module,
a
small
module
that
compares
see
most
he
said
of
comparing
places
instances
we're
using
like
a
structure
not
name
to
to
check.
If
an
instance
is
part
of
a
quest
but
I
think
we
suppose
be
a
little
patient.
D
That
suggests
that
we
used
symbols
yeah,
so
why
there
was
an
echo,
so
I
publish
the
module
doing
that
I
started
working
on
it,
applying
it
to
some
other
APIs
but
available.
The
transpiler
was
having
some
problems,
so
I
had
to
rework
that
and
I'm
doing
that
at
the
moment,
and
yet
next
week,
I
plan
to
to
make
some
PR
students
and
modules
I
think
that's
it
I
sent
you
an
email
with
it
with
something
like
the
hwachun
working.
If
you
want.
D
A
F
A
A
That's
why
I
was
reported
or
where
the
main
issue
were
like.
We
initially
thought
about
the
problem
to
track
all
the
work
that
needs
to
happen.
All
the
other
ebooks
like
just
just
doing
a
list
or
like
okay
I
need
to
change
this
repo
nice
trip,
one
release
and
etc,
like
just
like
all
those
tasks
helps
everyone
else
understand
what
is
the
status
of
that
desk
because
well,
there
is
no
way
to
to
really
have
an
issue
across
multiple
repos.
You
really
have
to
link
all
the
things
a
little
bit
manually,
yeah.
D
A
A
The
contract
that
we
have
with
each
other
is
when
you
open
a
PR
that
requires,
like
other
peers,
from
other
modules.
You
just
say:
okay,
these
PR,
in
order
for
it
to
work,
depends
on
these
PRS
for
all
the
other
modules
and
the
first
thing
that
is
testing
I'm
reviewing
that's
the
of
like.
If
groaning
NPM,
install
NPM
going
all
the
way
and
testing
it.
Okay,
again,
it's
kind
of
like
what
we
live
with
NPM,
but
like
it's
the
thing
that
worked
well
pretty
much
for
us
all.
A
A
Let's
catch
like
later
or
maybe
tomorrow,
given
up
now,
you
have
spend
your
first
week.
So
then
we
can
start
like
identifying
major
verticals
for
you
to
like
start
focusing
on
instead
of
I,
just
picking
random
issues
from
the
pool
sounds
good
I'll
send
you
my
link
so
that
you
can
forget
I'm
cool
next
up,
Aisling
brain
also
known
as
Alex
pot
sites.
All.
E
Right
so
this
week
it
is
some
onboarding
and
had
the
initial
attempts
taking
over
chunks
of
streams
merged
into
the
UNIX
FS
engine.
Repo
I
started
looking
at
kind
of
experimenting
with
with
changing
the
max
chunk
size
for
files
that
are
stored
in
IPA
first
to
see
if
we
could
get
kind
of
better
seeking
performance
by
loading,
smaller
chunks
of
the
files
and
one
go
and
I
discovered
that
the
change
that
I'd
made
doesn't
work
when
you
have
lots
and
lots
of
chunks,
because
the
tree
structure
gets
deeper
and
the
approaches
and
I
even
didn't.
E
A
Sounds
good
sounds
good.
That's
great!
So
can
you
two
mildly.
Some
things
the
PR
that
you
need
me
to
review
and
merge
so
they're,
my
favorite
hazard,
yeah
sure.
Well
thing
you
already
see
same
thing
for
UI
likes
like
we
should
catch
up
tomorrow,
just
like
now
that
you
have
done
the
first
week
so
that
we
can
find
some
verticals
for
you
to
like
focus
more
a
mid
long
term
and
not
just
pick
random
issues
from
the
lost
world
sounds
good.
Yeah,
okay,
Rob.
G
A
E
E
A
So
now
I
understand
to
qualify.
I
guess
essentially
like
even
go
what
you
support
different
chunk
choices
like
we
just
like
default
to
256
K,
but
we've
been
doing
that
forever
and
what
Alex
was
experimenting
with.
Is
that
like
when
you
have
this
graph,
and
you
want
to
do
seek
like
the
the
wings
on
the
graph
to
tell
you
certain
sizes
about
what
is
below
that
branch
and
so
with
smaller
chunks?
A
You
can
quickly
quickly
go
to
the
to
the
branch
that,
as
the
the
offset
that
you're
looking
for,
but
it
is
I
guess
I
can
experiment,
it's
not
like
changing
the
default
value.
The
default
will
still
be
256,
K
or
just
like,
maybe
like
in
the
tutorial.
Explain.
If
you
really
want
to
do
six,
maybe
you
want
to
import
a
files
in
a
different
way.
G
I
mean
sort
of
so
brief
frames.
This
is
like
an
experiment
in
what
ways
using
IPS
works
well
or
not,
I've,
really
a
change
to
genocide,
beef
s
per
se,
so
any
advice
that
comes
out
of
this
should
be
exactly
the
same
for
somebody
who's
gonna
be
of
s.
So
that's
I,
guess
somewhat
important
to
understand
that
is.
A
C
Hey
so
last
week,
I
worked
on
I
proposed
a
ctrl
in
Romans,
mostly
detecting
running
notes,
which
I
think
ideas
required
for
because
that's
the
ability
to
use
to
ingest
environment
variables
to
set
the
demon
and
default
into
a
flag
to
default
to
the
demon
addresses
I
also
worked
on.
I
also
started
looking
into
the
I've
been
making
some
changes
to
live
it
to
be
multiplex
related
to
an
issue
that
was
discovered
and
then
I
realized
that
there's
an
issue
to
rework
it
with
full
streams.
So
I've
been
looking
into
that
as
well.
A
Sounds
good
yeah
like
that
these
will
be
super
important
yesterday.
That's
way
faster
because
that
we
have
a
conversion
between
those
jet
streams
and
pools,
streams
there
and
also
reduce
the
panel
size
of
the
peer-to-peer,
because
then
we
don't
need
to
import
this.
The
same.
So
that's
our
mission
to
work.
Thank
you
for
picking
a
recipe
Rob.
G
A
H
Cool
so
sir
last
week
took
a
look
at
the
IP
six
ipv4
like
validations
and
Multi
adder,
and
just
try
to
make
sense
of
what
was
going
on
and
compared
that
to
the
go
library.
So,
given
that
doing
the
ipv6
validation,
this
is
actually
just
a
lot
of
code.
To
write
and
like
verify
is
correct.
It
seemed
like
using
the
IP
address.
Library
would
be
a
good
option.
I
think
it's
maybe
used
already
within
lit
p2p
switch
so
I
just
wrapped
that
with
the
buffer
and
submitted
PR.
If
that
makes
sense.
H
Alternatively,
there's
a
long
reg
X
that
we
could
use
to
apply
against
the
ipv6
and
then
the
other
thing
was
I
started
working
on
the
JSI
PLD
like
put
to
support
like
a
custom,
hash
I'll
go
on
the
CID
and
so
I
just
filed
a
PR
for
that
as
well
or
the
start
of
that.
So
it
involves
going
through
all
of
the
utilities
and
changing
the
like
shape
of
the
request.
A
B
Yeah,
just
for
saying
I'm,
the
the
isn't
we
are
changing
the
IP
of
this
back
a
bit.
I
will
put
it
to
your
issue
as
well,
but
the
API
for
the
serializing,
the
CID,
will
change.
So
perhaps
it
automatically
resolved
encoding
problem
so
patiently.
Well,
you
pick,
though,
I
get
the
takeaway,
probably
better,
picking
out
others.
You,
okay,.
A
H
A
Yep,
when
we
know
these
things,
it's
always
good
to
have
with
on
the
issue.
Okay,
I
guess
issue
is
we're
going
to
get
solved
by
this
other
and
ever
or
it's
blocked
by
its
own
endeavor.
So
that's
people,
that's
don't
have
access
to
these
calls
or
go
participate
in
these
calls
also
have
that
information
cool
alright.
A
I
Sir,
so
I
added
a
spec
to
the
interface
repo
for
adding
CID
instances
to
the
files
cat
API,
that's
been
merged
into
excuse
me
into
JSI.
Pfs
also
did
in
update
some
small
updates
for
the
multi-plex
dependency.
So
that's
now
using
em
Plex
instead
of
the
deprecated
moles
our
Plex
and
then
one
of
the
globs
in
the
testing
suite
was
incorrect.
So
that's
been
corrected.
I
This
week,
I'll
probably
meet
with
David
to
again
look
at
vertical
slice
for
the
application
and
right
now
I'm
just
continuing
to
deep
dive
into
the
ecosystem
to
get
up
to
speed.
A
Sounds
good,
okay,
I'll
share
my
calendar
with
everyone,
let's
well,
it's
like
a
child
as
soon
as
possible,
so
that
everyone
can
like
be
productive
for
the
rest
of
life,
yeah,
Kyle
and
new
go.
Do
you
have
anything
to
share?
Ask
effect,
I!
Think
it's
the
first
time
you're
caught
I,
say
hi
and
I
just
guys,
yeah.
J
F
I,
let's
see
I'm
just
reversing
in
but
I'm,
so
it's
when
I
mentioned
quickly
that
this
week,
I'm
sort
of
finalizing
all
the
stuff
for
how
we're
gonna
communicate
with
web
browsers
on
getting
interfaces
that
we
can
use
to
improve
GSI
PFS
in
the
future.
Things
like
you
know,
being
able
to
get
access
like
UDP,
sockets
and
stuff,
like
that,
like
pretty
pretty
long
term
like
ambitious
stuff
that'll
make
it
so
that
you
know
web
RTC.
Isn't
the
600
connection?
F
A
Alright,
thank
you
that
exciting
cool
okay.
So
we
have
like
one
minute
left
and
I
guess
there
is
like
one
thing
that
I
want
to
cover,
which
is
the
waffle
works,
so
we're
like
three
waffle
words
he's
like
I
can
bend
style
an
awful
more.
It
was
like
backlog,
ready
in
progress
and
done.
There
is
a
video
on
the
first
issue
or
the
first
time
about
the
the
thread
for
the
weekly.
Is
that
X
point
that
we
use
it?
A
There's
a
document
MGMT
on
giuseppe
doesn't
explain
how
to
use
it
and
what
the
thing
that
I'm
like?
Is
it
to
me?
The
thing
that
I'm?
What
to
get
to
by
explaining
is
is
that
I
feel
that
we
are
not
using
the
waffle
boards
as
they
were,
designed
to
be
used,
as
in
like
it
is
really
hard
to
look
at
the
motherboard
and
understand
like
what
people
are
working
on
because
like
when
you
touch
pull
requests
like
it
tells
you
that
it
becomes
in
progress
and
that's
great
because
it's
automatic.
A
A
Why
do
we
want
this
like
we
want
this
to
it
understand
if
we
are
actually
tackling
the
top
priority
issues
like
right
now,
when
I
do
the
filter
by
p0
and
p1
and
so
on.
It
is
clear
that
not
everyone
is
tackling
the
most
important
issues,
at
least
from
the
the
eristic
that
we
use
to
declare
what
is
an
important
mission
or
not.
A
A
A
J
A
We
can
have
a
separate
level
right
now.
All
we
have
been
making
a
distinction.
Is
that
like
when
someone
is
assigned
to
an
issue,
it
means
like
that
person
is
owning
the
issue,
and
so,
if
something
on
sound
ready
that,
like
is
assigned
to
me
you,
you
should
not
try
to
pick
it
up
like
it's
on
my
plate,
and
so
there
is
a
way
to
filter
that,
like
by
empty
issues,
so
that
you
can
find
something
new
to
work
on.
A
G
A
I'm
good,
like
you,
you
alright
having
a
level
will
be
useful,
but
but
that's
pretty
much
it
that's
it.
Essentially.
What
means
like
when
someone
opens
the
waffle
board,
it
is
in
progress
like
they
should
they
have
a
clear
picture
like
what
each
person
is
working
on
and
they
should
see
the
movement
of
the
issues
going
to
the
down
column
and.
A
Typically,
people
should,
like
always
pick
the
most
urgent
first.
Of
course,
there's
like
some
that
are
like.
There
is
some
questions
here
because,
like
there
might
be
some
urgent
issue
and
then
because
of
that,
you
might
solve
another
not
as
urgent
and
now
you
just
pile
them
up
and
I
keep
them
as
one
package,
and
so
you
end
up
doing
less
urgent
things
and
then,
depending
on
like
the
time
pressure
or
if
we
are
preparing
ourselves
for
a
demo
or
or
if
we
have
a
feature
that
we
really
really
support.
A
Some
use
case,
like
the
the
way
that
we
perceive
that
extra
added
work
might
change,
and
so
sometimes
like
it's
really
important,
just
like
focus
on
the
critical
ones.
The
pick
need
to
be
shipped
as
soon
as
possible.
Other
times
it's
okay
to
just
level
up
the
entire
project
as
we
go,
but
then
again
always
like,
as
a
golden
rule,
just
like
try
to
take
back
all
the
peer
ones
and
p0s
first
and
get
that
waffle.
Word
updated.
A
A
It
so
the
LPRs
4GS
ipfs,
specifically
I,
will
open
a
PR
to
the
Jessup
FS
repo,
with
dot.
Md
and
I
will
also
link
to
the
IP
passing
web
browsers
working
group
markdown
file,
because
what
they
did
was
just
greater
like
my
on
file
with
objectives,
key
results,
priorities
and
idea.
The
discussion
on
a
PR
which
makes
it
really
nice
to
the
discussion
and
then
once
they
were
confident
that
the
list
was
right.
A
Then
they
added
its
to
the
spreadsheet
and
so
yeah
I'll
woven
a
PR
and
I
will
tag
everyone
and
and
we'll
just
take
it
from
there
sweet
give
me
tree.
You
got
in
your
hand,
I
know
you're,
just
waiting
cool
all
right.
Thank
you.
So
much
everyone
for
your
time
and
have
a
great
week
see
you
on
the
interwebs.