►
From YouTube: 🖧 IPLD Every-two-weeks Sync 🙌🏽 2019-07-08
Description
A meeting every two weeks to sync up on all IPLD related topics. It's open for everyone and recorded. https://github.com/ipld/team-mgmt
A
A
Guess
we
haven't
really
done
one
of
these,
though,
since
Barcelona
started,
so
a
lot
of
stuff
shifted
around
in
Barcelona,
not
necessarily
because
of
any
activity
that
was
there,
but
mainly
because,
like
me
and
Eric
had
some
time
to
sync
up
and
I
had
some
time
to
think
up
with
other
people.
A
That's
the
sort
of
big
change-
oh
yeah
and
Hannah's
coming
over
to
do
the
selected
stuff
as
well
as
graph
sync.
The
protocol
work,
which
frees
up
air
to
spend
more
time
on
schemas
and
the
priority
for
cement
in
general,
is
just
like
bumped
way
way
up.
So,
if
schemas
encode
Jen
in
a
big
red
ball
kind
of
bumped
up
in
Prior,
that's
my
update.
A
B
I,
didn't
actually
think
this
was
very
approachable,
and
apparently
it
was
because
in
the
next
couple
of
hours,
that
person
came
back
to
me
with
a
couple
of
questions
by
like.
Is
this
a
palace
electric?
Does
this?
Do
it
I
think
it
does,
and
in
the
case
of
little
squares
of
this
whole
zone
yeah?
Actually
so
that
was
really
cool
and
I.
Think
a
selector
specs
are
kinda
bottom
track
as
a
result
of
these
forms
of
feedback
and
I.
B
Think
some
of
the
other
folks,
like
the
foul
coin,
guys
seem
to
be
looking
at
doing
a
bunch
of
coins
back
work
that
they're
on
in
the
next
couple
of
weeks
for
my
ability
schemes.
So
quite
a
lot
of
attention
be
that
so
that's
kind
of
exciting
and
he's
gonna
put
a
lot
of
pressure
on
us
to
like
deliver
the
thing
and
do
succeed
at
user
goals
in
the
next
I.
Don't
know
time
T,
so
yeah
I'm
really
excited
about
that
and
also
it
does
seem
to
feed
into.
C
C
Was
some
there
were
some
comments
last
week
from
Jeremy
that
that
need
some
experimenting
so
and
then
is
also
that
Oh
that
conversation
about
context
and
parameters.
C
It's
just
that
I
get
so
frustrated
with
that
conversation,
because
it's
I
know
that
I
think
I
understand
the
desire
behind
it,
but
it's
like
we
don't
have
tools
to
do
this
stuff
right
now.
I
just
want
to
do.
I
want
to
do
stuff
with
these
data
structures,
and
you
know
I
can't
do
stuff
with
them
with
this
idealized
version
of
of
how
they
should
all
fit
together
when
we
don't
have
the
bits
to
do
that.
C
C
A
No
yeah
I
think
two
different
things
are
getting
confused
there,
so
you
have
if
one
thing,
which
is
like
the
definition
of
the
data
structure,
which
is
effectively
like
the
same
thing
for
every
one
of
these
hashmaps,
regardless
of
what
those
parameters
are
right.
So
this
is
just
like
hey.
This
is
the
name
of
this
thing,
and
you
know
at
some
point
in
the
future.
Here's
like
the
web
assembly
function
for
now.
It's
just
literally
like
this
is
a
special
thing.
This
is,
as
far
as
like
the
parameters
for
the
actual
algorithms.
C
B
A
Even
if
we
just
like
add
a
note
that,
like
hey
the
sections
under
discussion
or
whatever
I'm,
adding
a
bunch
of
this
to
the
cosmic
spec
to
do
just
like
we're
still
actively
discussing
blah
blah
just
that
we
can
get
something
like
in
because
ideally
I'd
like
like
to
just
get
it
merged
so
that
we
can
have
these
discussions
as
separate
pr's.
Rather
than
like
this
one,
big
PR.
That's
getting
kind
of
hairy
too,
to
really
get
through.
B
A
Yeah
so,
like
you
know,
if
the
implementation
that
you
you've
worked
on,
has
the
parameters
and
come
to
the
roadblock
I
would
just
put
that
in
the
spec.
Put
a
note
that,
like
hey
we're
discussing
this
still
and
then
merge
it,
and
then,
if
Jeremy
specifically
wants
a
way
to
not
encode
the
parameters
and
select
infer
some
kind
of
default.
Behavior
then
like
he
can
start
a
PR
for
that,
or
somebody
can
start
up
here,
for
that
we
can
discuss
with.
That
way,
would
be
like
my
ideal
solution
to
this
yeah.
C
Okay,
I
mean
okay,
I
mean
if
someone
has
a
language
for
that,
then
that
would
be
good
because
other
what
otherwise
it's
like
we're
discussing
this,
but
I,
don't
even
I,
don't
even
fully
understand
why
we're
discussing
it.
It
just
doesn't
come
with
that,
but
anyway,
I
saw
I
also
started
to
I'm
interested
in
using
these
things.
To
do
some
practical
stuff
with
and
I
was
looking
at.
C
The
let
last
word
on
there
actually
in
yesterday,
as
well
as
looking
at
the
question
of
temp
repair,
is
that
when
you're,
using
these
things
and
doing
hearing
mutations-
and
you
know
you're,
creating
lots
of
intermediate
blocks
that
you're
gonna
throw
away
and
potentially
having
very
large
data
structures,
so
you
can
have
potentially
a
lot
of
these
things.
I
was
looking
at
the
question
of
block
storage
when
you're
doing
these
mutations,
so
that
you
can
both
throw
away
and
collect
later
state
efficiently.
C
What
does
that
look
like?
So
it's
just
you
know
just
playing
with
that
little
piece
of
the
puzzle
to
see
if
there
was
a
an
obvious
obvious
efficient
way
to
do
that,
because
I
want
to
I'd
like
to
start
using
them
for
some
ideas
but
yeah,
that's
a
rabbit
hole
I
want
to
get
it
too
too
far
down,
because
there's
there's
easy,
there's
there's
easy
versions
of
that
that
we
can
use,
but
anyway
and
then
the
other
thing
was
two
on
my
plate
was
coming
up
to
spend
some
more
quality
time
with
go.
A
D
D
A
A
B
D
B
A
B
A
B
A
Right
right,
yeah,
that
would
work
too.
That's
that's
kind
of
scary
yeah,
usually.
A
D
The
other
thing
that
I
have
been
working
on
is
in
the
JSON
schema
work
for
the
did,
methods
and
I
think
what
might
be
really
helpful
is
a
very
good
an
issue
filed
this.
We
if
we
can
get
IP
LD
submitted
as
an
I
as
a
valid
scheme,
name
for
a
fully
formatted
URI,
and
then
you
can
do
some
pretty
cool
stuff
with
schemas
so
actually
like
you
can
have
self
describing
schemas
with
that
without
a
authority.
D
B
B
A
Yeah,
it
seems
like
there's
there's
to
sort
of
separate
contacts
right
like
there's.
There's
the
hey
I
already
know
what
this
thing
is
from
somewhere
else.
Right,
like
like
the
fat
point
of
discussion,
gets
to
that.
But
then,
even
within
the
discussion,
when
you
hit
like
intermediary
nodes,
you
have
a
context
that
you
carry
over
from
prior
nodes
and.
D
Yeah,
so
this
example
is
just
a
the
json
schema
for
the
verifiable
credentials
stuff
and
in
the
ID
is
actually
is
it's
probably
formatted,
IP
LD
syntax,
then.
Actually
it
is
you,
even
though
it's
the
this.
It
doesn't
exist
in
this
when
you
dereference
it
and
actually
get
this
from
IP
LD.
It's
at
least
resolvable.
D
No
Jason
application
JSON.
So
sorry,
that's
just
yes
in
mime
type.
No,
so
this
is
actually
the
identity.
The
ID
of
the
the
field
doesn't
necessarily
mean
that
it's
resolvable,
so
this
is
a
I
can
give
it
a
random
number,
a
UUID,
but
basically
since
I
mean
this
is
resolvable
to
a
dag
Tibor
sterilize
to
json.
It
itself
describe
its.
It
says
this:
this
is
the
idea
of
this
thing,
and
so,
when
you're
doing
definitions
and
so
down
below
here,
there's
I'm
referencing
here
yeah.
D
A
Just
I'd,
like
the
only
thing
that
I
wonder
about,
is
that,
unlike
other
scheme
names,
there's
no
transport
inferred
by
IPL
B,
so
like
it's
an
identifier,
but
it
doesn't
necessarily
tell
you
how
to
get
it,
whereas
most
schemes
that
I
can
think
of
what
would
be
implying
how
you
go
and
get
the
data
like
then
yeah.
The
scheme
actually
gives
you
some
kind
of
context
on
how
to
get
it
mm-hmm.
D
But
like
in
jason
schema,
the
idea
is
not
you're
not
supposed
to
necessarily
be
able
to
resolve
it.
So
a
lot
of
people
actually
do
resolve
it
and
if
you
put
Jason
schemas
even
in
like
this
vs
Microsoft
code,
visual
code,
those
do
try
to
actually
go
out
and
go
to
http
resource
and
actually
download
them.
But
they're
not
technically
supposed
to.
D
Yeah,
it's
supposed
to
be
that
you
have
all
your
schemas
and
you
can
reference
your
different
schemas,
a
lot
of
them.
Actually,
if
you
have
them
in
the
same
folder
they're
supposed
to
resolve.
You
know
it
with
relative
naming
but
I.
Think
using
this
IP
LD
solution
might
be
more
elegant
and
simple
and
self,
describing
in
standalone
without
having
a
Thorat,
ativ
source
and.
A
How
often
so
I
know
that
some
Seema
systems
are
actually
going
and
resolving
all
those
are
old.
Like
you
said
my
muscles
any
of
the
Jason
LD
stuff,
but
back
when
I
actually
tried
to
do
some
RTF
stuff,
it
got
to
the
point
where
you
would
just
use
them
as
string
identifiers,
you
wouldn't
actually
go
and
try
to
resolve
any
of
the
URLs
once
you've
done
four
different
schemas.
Is
that
still
the
case
in
Jason
Aldean,
most
in
time
that
people
don't
actually
go
out
and
get
the
discrimination
so.