►
From YouTube: 📦Package Managers WG Weekly Sync June 11, 2019
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
B
E
A
D
F
Hi
I'm
Benna,
can
you
hear
me
yeah?
Okay,
I
am
working
or
names
of
one
running
repositories
for
experimental
ones,
and
they
have
cumulative
size
of
4,
terabytes
now
and
I'm
gaining
some
real-world
experience
with
what
the
problems
are
and
some
resource
leaks
and
missing
propagation
of
announcements.
Other
of
these
things
and
IPS
is
not
working
as
I
would
like
it.
So
it's
things,
I
see
but
long
term
I
think
six
integration.
D
A
D
D
B
Have
one
open
item
for
the
three
of
you
and
that
is
before
I
go
and
print
out
10,000
sheets
of
paper,
reflecting
what
I
posted
in
about
if
I
I
can
proceed
with
going
and
freaking
out
individual
ones
of
those
just
for
a
disaster-free
thing
for
our
for
our
eighty-five
session,
but
we'll
just
need
to
all
get
together
in
person
and
pull
out
the
ones
that
we
actually
want
to
pick
apart.
I
think
that's
probably
the
best
thing
seeing
is
how
I'm
getting
on
saying
tomorrow
said:
I'll
make
sense.
Everybody
do.
D
D
B
B
B
Rayland,
this
whole
thing:
sorry
about
that,
if
we
could
cut
them,
if
we
could
cut
the
combined
list
of
both
problems
and
benefits
down
to
about
20,
that's
still
probably
a
lot
functionally
dot
vote
on,
but
I
think
we'll
end
up
with
you
know.
If
we
start
with
20,
we
can
knock
it
down
in
person
and
see
how
we
go.
I
just
need
to
go
to
the
service
bureau
this
afternoon
to
print
out
that
and
some
other
camp
stuff.
So
if
y'all
could
look
through
it
like
the
next
hour,
that
would
be
superb.
C
C
C
But
it's
kind
of
cool
to
be
able
to
turn
your
Wi-Fi
off
and
say,
publish
and
then
install
in
and
it
works
across.
The
only
thing
doesn't
work
right
now
is
get
dependencies
which
is
annoying
because
there
are
J's
ipfs
libraries
that
have
kit
dependencies.
I've
got
a
plan
to
make
it
mostly
work,
but
it
won't
work
100
and
offline,
and
the
other
plan
is
to
tell
the
GSI
PF
has
people
to
remove
all.
C
Way
but
I
won't
do
a
demo
of
anything
now
I'll
just
try
and
get
the
lightning
talk,
video
ready
and
but
yeah.
B
Sorry
forgot
to
ask
one
other
question
related
to
our
deep
dive.
Andrew
and
Alex.
Are
you
two
still
okay
on
collaborating
on
a
here's
where
we
are
right
now
level
setting
presentation
at
the
beginning
of
that
deep
dive
which
I
say
either
you
could
do
it
via
the
magic
in
video.
If
you
wanted
to
capture
yourself,
Andrew
Alex,
if
you
wouldn't
mind
giving
the
introduction
and
then
you
know,
staying
in
the
deep
dive
as
much
as
you
see
fit,
that's
the
one
piece.
B
B
D
A
No
boats
anything
just
been
yeah
working
on
you
know
getting
the
MPM
rescue
mera
running
on
the
Raspberry
Pi
so
that
we
can
use
ipfs
NPM
camp
is
broadcasting
its
availability
by
mdns.
Now
I've
got
a
like
a
pre-release
of
NPM
IPF
I
still
look
for
that
mdns
name
on
the
local
network,
and
it's
going
to
be
well
good.
You
were
gonna
mother.
You
were
going
to
give
me
a
list
of
things
to
also
make
available
a
variety
of
s
on
the
network.
D
D
Have
not
that
list
from
from
the
electives,
yet
thinking
that
you
know,
maybe
they
have
pre
work
that
they
want
pre-installed,
hopefully
that
stuff
I
can
send
everyone
an
email
and
check
to
see
if
that
those
things
have
solidified,
but
I'm.
Also,
you
know,
as
a
one
of
the
cat
herders
for
content
being
created
for
camp
they're,
literally
gonna,
be
making
me
install
stuff
when
we
practice
running
through
their
workshops
and
so
from
there.
I
will
definitely
get
a
list
of
those
things,
but
I'll,
try
and
see.
F
So
currently,
I
fetch
the
latest
rolling
release
updates
of
openSUSE.
That
code
opens
a
tumblr
mate
and
that's
like
true
to
fire
for
our
gigabytes
per
day,
but
I
always
do
a
full
snapshot,
so
I
get
new
CID
and
then
I
add
the
CID
to
the
oil
tree.
There's
some
object.
Ipfs
object
up
pitch
pitch
some
pitch
thing.
So
I
have
this
coded
and
adds
a
pitch.
Then
I
have
a
new
CID
with
all
the
old
stator
in
the
history
tree
and
the
current
one
replaced
with
the
current
data
and
that's
very
useful.
F
So
if
all
the
old
data
in
there,
but
also
the
current
one
in
the
constant
place
and
then
I
publish,
updates,
but
I
tried
to
publish
that
over
IP
NS
and
every
time
I
resolve
it
over
IP
NS.
It
takes
60
seconds
and
then
it's
cached
for
a
minute
and
then
it
takes
again
60
seconds.
It's
not
something
you
can
use.
Rather
so
I
went
with
some
do
DNS.
We
just
put
an
ID
FSC
ID
in
there
every
day,
that's
working
reasonably
well,
but
then
I
notice
it
it's
ipfs
process
is
growing
and
growing.
F
F
D
A
The
on
the
62nd
APNs,
resolute
name,
resolution
thing:
that's
because
the
demon
like
it
waits
to
get
a
number
of
confirmations
it.
It
makes
a
bunch
of
requests
for
the
latest
hash
of
a
given
type,
ENS
name
and
compares
them,
and
if
you
know
to
make
sure
that
it's
got
the
latest
one
and
every
now
and
again
it'll
hit
make
it
requests
to
appear,
that's
behind
an
app
and
it
will
timeout
and
that's
why
I
always
takes
that
60
seconds.
A
A
I,
don't
use
random
I
think
it.
There
is
a
no
rhythm
that
he
uses.
I'm
not
hung
up
central
more
days,
but
if
I
see
it
makes
a
bunch
of
connections
to
verify
that
it's
got
the
latest
version
of
the
IP
NS
name.
You
have
the
CID
that
it
was
going
to
and
if
one
of
those
is
behind
in
that
which
should
love
them
are
x
out,
which
is
why
it
takes.
A
D
Something
that
I've
heard
other
people
doing
I'm
not
sure
how
much
we
recommend
this.
But
if
you
reduce
the
number
here's
you
need
to
get
a
confirmation
from
from
20,
which
is
relatively
high
to
something
more
like
eat
or
or
something
like
that.
You're
just
you're
doing
fewer
round
trips
you're
more
likely
to
get
responses
from
peers
without
having
to
like
wait
for
your
timeout.
D
D
I
think
we'd
be
really
interested
to
know
like
kind
of
like
what
about
you
excited
about
putting
openSUSE
on
ipfs
and
kind
of
yours,
your
top
level
goals
like
what.
What
is
most
appealing
about
that
to
you,
because
we're
we're
trying
to
make
sure
that
you
know
people
have
that
opportunity
and
then
it's
a
great
experience
for
everyone.
So
the
more
background
info
you
can
give
us
on
that
I'm,
where
we
can
hopefully
help.
F
So
I
like
cystic,
wants
the
integrated
deduplication
because
there's
the
snapshot
and
every
snapshot
has
a
lot
of
the
same
data
as
a
previous
snapshot.
So
I
think
one
snapshot
is
98
gigabyte
total,
but
there's
only
two
to
four
gigabyte
new
data
in
there.
So
I
can
keep
a
history
without
adding
90
gigabyte
each
day,
but
only
I
think
about
each
day
and
then
I'm
doing
it
for
the
producer.
F
With
this
thing,
so
whenever
I
want
to
clip,
use
a
binary
and
it's
the
would
buy
any
set,
they
use
like
old
compiler,
and
so
it
as
I
chains
and
normally
is
not
around
anymore.
And
then
you
can't
refuse
artifacts
anymore,
but
know
it
has
its
history
and
I
can
pick
the
items
they
have
a
small
database,
5a
I,
say:
okay,
if
it's
using
this
versions
and
it's
this
IP
FSC
ID
and
then
it
can
just
fix
the
cid
and
turning
it
into
my
dependencies
and
use
it
during
that.
F
That's
really
nice
nice
feet
and
at
the
same
time,
I
could
use
it
to
update
my
laptop
at
home
because
I
fetch
it
once
and
then
it
can
update
the
same.
That
as
the
next
laptop
next
to
it,
without
fetching
all
the
data
from
the
internet
again,
so
this
automatic
mirroring.
That's
really
a
nice
feature.
That's
what
I
liked
this
bit
remember
patron,
does
really
not
suited
to
this
use
case.
F
So
that's
the
main
use
cases
on
my
side.
Maybe
a
used
surface
quit,
but
the
script
has
these
problems
that
we
have
a
download
read
you
liked
on
opposite
servants
if
needed,
electro,
sometimes
points
this
way
and
sometimes
points
another
way,
and
it's
good
doesn't
notice
it.
It's
actually
the
same
fire
on
this
different
mirrors,
so
it
keeps
downloading
duplicates
and
then
I
started
pointing
to
one
specific
mirror
and
only
use
that
one.
F
But
then
sometimes
f1
was
outdated
or
inconsistent
copy,
so
some
fights
where
from
the
new
your
lip
went
somewhere
wouldn't
and
it
was
missing
missing
things
up
and
yeah.
So
it
was
not
not
perfect
and
sometimes
even
fights
got
replaced.
So
normally
a
lot
of
soft
assumes
that
I,
it's
not
replaced.
So
if
you
do
a
new
ability
get
a
new
build
number,
but
sometimes
every
renames
and
sirenians
course
builder
must
to
go
down
again
and
yeah
it's
the
same
file.
F
F
F
Okay,
I
hope,
I'm
wasn't
so
the
course
I
saw
some
182
the
plate,
no
gigabyte
give
it.
One
has
80
gigabytes
passing
by
my
server
in
a
few
days
since
it's
not
so
myself,
okay,
but
if,
if
the
gateways
passes
through
so
if
people
really
go
through
the
gateways
which
P
and
the
Gateway
switches
from
Isis-
and
it's
put
some
load
on
the
gateways,
not
sure
if
the
public's
that's
for
these
first
gateways.
D
Yeah,
so
if
you're
getting
about
a
you
said
a
gigabyte
passing
through
your
node
in
a
day,
I
think
the
gateway
in
general
is
serving
something
thing
between
like
ten
and
twenty
five
gigabytes
or
terabytes
per
day.
So
there's
a
lot
there's
a
lot
of
other
traffic.
So
don't
worry
your
that
is.
That
would
not
be
overloading
at
all.
Okay.
F
D
Of
the
things
you
live
like
you
know,
this
is
what
I'm
doing
so.
I
want
to
get
out
is
like
one
right
on
on
our
expectation
of
like
some
cool
stuff,
and
it's
super
exciting
to
us
that,
like
you're
picking
it
up
and
running
with
it.
So
thank
you
so
much
also
for
coming
and
chatting
with
us
about
it,
where
you're
learning
stuff,
by
talking
to
you
and
excited
to
get
help
and.
B
F
Okay,
any
questions,
yes,
four
terabytes,
but
it's
actually
only
a
few
hundred
gigabytes
of
data
and
because
of
CD
duplication.
That's
currently
stored
on
my
server
and
oh,
the
next
thing
we
haven't
found
any
definite
things
yet
is
once
it
grows
too
much
for
this
one
server.
So
can
I
put
it
on
some
offline
machine,
so
it
puts
a
history
on
some
offline
machine
or
ISM,
not
not
so
very
connected
machine,
because
mostly
the
current
data
will
be
used.
F
F
D
Any
give
you
ideas
cluster,
and
so
you
add
a
new
machine.
Then
it
will
start
doing
the
balance
across
multiple
machines
and
I.
Do
this
cluster
kind
of
lets?
You
have
a
close
that
are
pinning
content
across
different
places,
and
so,
if
you
start
pinning
things
to
cluster
it'll,
be
like
oh
well,
it
already
has
the
stuff
pinned
on
this
machine,
but
maybe
I
want
you
know:
X
x,
factor
of
replication,
so
all
like
as
well,
and
when
you
get
new
stuff
like
this
one
has
space.
D
D
Don't
know
if
it
necessarily
helps
like
migrating
to
cluster
means
that
you
you're
gonna,
have
to
like
do
that.
Configuration
versus
and
you're
mentioning
that
you're
kind
of
already
in
a
place
where
you're
manually
pinning
all
of
these.
These
historical
snapshots
right.
F
F
Because
sometimes
trained
to
convict
and
advices,
you
need
to
restart
at
a
a
man,
but
there's
no
restock
command
and
no
conflict
no
command.
That's
also
something
I
was
wondering,
because
many
many
classic
UNIX
demons
have
sent
a
secret
and
if
he
sent
us
a
Cooper
Theresa
could
begin
and
use
a
nucleus
instead
of
the
old
ones.
That.