►
B
Well,
yes,
I
do
just
earlier
today
we
shipped
the
JSI
PFS
340
release
a
lot
of
good
stuff
in
there,
including
a
new
migration
tool.
I
have
told
people
about
this
many
times.
Please
go
and
read
the
blog
post.
It
is
linked
in
the
notes.
There's
a
little
good
stuff
in
there
and
Alex
released
a
bug
fix
for
0:39
which
stopped
people
from
accessing
their
repo,
so
that
is
now
fixed
as
well.
Yesterday,.
A
Do
they
have
any
one
I
guess
I
can
get
this
update
I
at
the
moment,
the
test
ground
daemon
is
currently
in
progress.
Jesminder
daemon
is
a
certainly
side.
Damon
did
controls
restructure.
The
idea
here
is
that
we
couldn't
intervene
statement
with
test
ground
and
with
like
yes
and
then
github,
hey
I,
want
you
just
add
some
github
pod.
So
hey
what
test
this
thing
per
se
this
test
ground
game
it
would
them
responsible
for
receiving
this
certification
and
spinning
up
which
has
brought.
A
A
A
A
F
Just
just
just
a
super
quick
update,
Oh
updates
from
any
person,
but
with
the
latest
Jess
activist
release
the
Talon
nation.
We
now
have
support
for
by
32
identifiers
on
the
IP
NS
box.
That
includes
both
CID,
be
zero,
converted
to
CID,
be
one
with
like
dark,
be
multicolored
and
the
one
with
the
proper.
A
Okay,
let's
see
I
think
it
go
as
well.
H
Another
round
at
the
sort
of
IPS
value
store
PR,
so
that's
ready
for
review
and
Hugo
and
I
did
some
talking
and
that's
tariqa
about
the
akina's
over
dns
spec
that
to
look
like
I'm.
Currently
assuming
this
is
lower
priority,
but
if
you
want
to
at
least
make
landing
the
spec
a
higher
priority,
that's
cool
too,
but
need
need
some
feedback
there
do
you
know
anything.
I
missed
no.
H
So
working
on
arias
be
ours,
our
database
first
or
it's
Inc
Rinna's
rights,
which
will
in
turn
make
our
stuff
a
lot
faster.
There
are
a
whole
bunch
of
PRS
there.
There
have
been
some
discoveries
along.
The
way
like
some
like
leveldb
was
asynchronous
the
whole
time,
if
you
were
reusing
it,
which
might
explain
some
of
the
performance,
differences
and
just
sort
of
updating
our
tests
and
removing
GX
and
the
various
data
stores
as
I'm
crawling
through
them.
I
just
have
to
do
flat.
E
I've
open
POS
for
paralyzing,
the
import
of
files
into
Joseph's
EFS,
and
yes,
it's
kind
of
like
it's
really
good.
It's
got
about
50%
faster,
so
it
paralyzes
in
two
places
like
the
first.
The
first
one
is
like
the
other.
You
have
a
file
pipeline
and
say
the
file
comes
in
we
chunk
here.
We
write
in
to
the
data
store.
I
will
create
Doug
and
I'm
not
into
the
data
store,
and
then
we
create
the
tree
structure
of
the
directory
that
we're
important
to
previously
that
all
just
kind
of
went
in
one
go.
E
So
now
we
can
paralyze
on
the
file
input,
so
extremophiles
comes
in
and
we
can
fan
out
on
the
files
and
then
the
when
we
start
chunking.
We
can
also
find
out
feature
individual
file
on
the
channels,
and
then
we
found
back
in
where
we
start
writing
the
tree.
The
final
directory
tree,
the
reason
that
we
have
these
two
different
points
of
which
we
found
out
to
say
that
if
we're
importing
one
by
foreign
14
lots
of
small
files
that
typically
are
smaller
than
the
trunk
photos.
E
E
Conversely,
if
you've
got
one
file,
that's
enormous,
then
you
want
to
you
want
to
fan
out
other
chunking
stage
and
not
at
the
file
stage.
So
there's
two
new
parameters
that
are
exposed.
Don't
you
control
that
chunk
invader
and
for
their
in
like
lots
of
small
files,
it's
basically
photos
and
frost.
Now,
if
you're
doing
one
file,
that's
big.
It's
also
fast,
and
the
flipside
of
course,
is
that
you
can
also
make
it
slower
by
having
no
enough
junk.
It's
pretty
cool.
A
I
G
Kick
it
off
if
V
yeah,
so
we've
got
sekai
OS
done
and
integrated
we're
at
the
point
right
now,
where
we're
just
working
on
lip
ptb
core.
So
we've
finished
up
a
lot
of
the
internal
module.
Refactor
circuit,
really
a
DHT
peering
content
routing
are
converted
over
we're
currently
working
on
switching
over
the
the
dialer,
which
is
Cubase
and
to
a
dope,
convey
styler
so
that
we
can
fix
the
per
pair
parallelization
of
dials.
Currently,
that's
all
serial
so
we're
getting
that
out
and
then
flush
fuzz,
working
on
upgrading
the
PDP
daemon
and
interrupt
tests.
G
G
B
I
was
just
laying
the
tracks
in
the
JSF
s.
Httpclient
I've
got
a
second
to
quickly,
explain
it.
It
removes
notes,
dreams
and
pull
strings
in
favor
of
async
iterables,
as
per
hopefully
everywhere
in
the
BTB
and
all
that
ruins
callbacks,
and
it
removes
pier
and
foam
here,
ID
dependencies
in
favor
of
just
sending
back
objects
with
C
IDs,
which
removes
a
whole
wave
of
crypto
libraries
that
are
basically
not
used
in
the
client
and
so
now
now
this
bundle
size
is
92.7
K
down
from
246
K.
B
B
B
Last
week,
I
did
but
Jess
I've
if
I
said
relicense
to
MIT
and
Apache
to.
If
you
haven't
already,
then
please
go
and
visit
the
sign
off
issue
in
between
then
and
now.
Another
pull
request
was
submitted
to
the
HTTP
client.
To
do
the
same
thing.
If
your
name
is
on
that
issue,
then
also,
please
check
it
out
and
ticket
or
copy
and
paste
the
sign
off
message.