►
From YouTube: Podcast 001 - Matthew Hodgson
Description
Welcome to our first episode of Jitsi and Beyond! On this very first episode we have a chat with Matthew from the Matrix project. Let's talk about life, protocols, decentralization, encryption, etc!
Cheers!
Emil and Saúl.
Links:
https://matrix.org/
https://anchor.fm/jitsi-and-beyond/episodes/001---Matthew-Hodgson-e1cq7ik
A
Hey
everyone
greetings
from
sue
winnemill,
we're
here
today
with
matthew
hodgson
from
matrix,
the
founder
of
the
of
the
matrix
project
and
we're
very
excited
to
have
him
for
our.
You
know
this
new
experiment
that
we're
doing
with
podcasting
to
the
gt
crowds
we're
very
honored
to
have
matthew
as
our
as
our
first
guest
as
a
first
guest.
Here
we
are
hoping
to
have
some
pretty
nice
discussions
today,
we'll
see
where
the
conversation
takes.
Us
matthew
welcome,
very
happy
to
have
you
here.
Yeah.
A
B
A
B
B
Yes,
so
I
ran
the
voip
function
for
this
little
company
in
west
london
called
mx
telecom,
and
for
many
years
we
were
building.
First
of
all,
the
sets
off
switch,
I'm
using
the
reciprocate
c,
plus
plus
stack,
which
was
a
horizontally
scalable
carrier
grade
soft
switch
and
it
had
gateways
through
to
h323
and
pstn
e1
cards.
What
else
do
we
have
h324
ram,
so
circuit
switch
video
calling
for
interactive
two-way
video
on
3g
umts
phones
back
before
the
crazy
smartphones
we
have
today.
B
Me,
I
guess
this
was
straight
out
of
university,
so
in
2003
had
been
like
23
or
something
right,
so
we
were
doing
that
for
about
10
years
and
it
was
a
lot
of
fun
because
we
learned
all
about
zip.
We
also
built
our
own
media
stacks.
Our
own
rtp
srtp
iced
unturned,
basically
ended
up
with
something
that
felt
a
lot
like
webrtc
before
webrtc
existed
and
then,
when
the
smartphones
came
along,
we
ported
our
server-side
stack
and
ran
it
client-side.
B
So
this
probably
is
something
very
familiar
in
terms
of
the
stack
that,
of
course,
on
the
jetsea
side,
fujitsu
mate,
the
communicator
app
was
doing
and
we
ended
up
with
a
very,
very
similar
product,
admittedly
or
c
plus
plus
rather
than
java,
but
we
ended
up
running
it.
Both
on
ios,
using
like
native
z,
plus
burst,
link
from
the
objective
c.
A
B
B
A
B
B
As
you
say,
we
got
lucky
and
andox
realized
that
if
they
mixed
together
my
void
team
with
the
mobile
development
team
that
they
had
also
acquired
a
month
later,
a
company
called
stromezo
in
france,
which
was
a
spin-off
from
france
telecom
based
up
in
ren
in
brittany
that
if
they
combine
those
guys
with
my
guys,
then
you
could
start
building
smartphone
apps
with
voip
stack
in
it,
as
well
as
messaging.
An
address
book
and
merged
address
book.
All
of
these
kind
of
features
and
start
competing
with
skype
or
whatsapp,
or
that
sort.
A
Of
it
was
all
the
rage
back
then
right
competing
with
skype.
Everyone
was
that
was
kind
of
the
gold
standard
of,
or
rather
I
think
what
what
people
perceived
is.
Oh
yeah,
I
guess
that's
how
telecommunication
has
to
look
now.
Look
at
how
successful
skype
is.
Let's
try
and
do
the
same
thing
yep.
This
was
2011.
B
A
And
you
could
argue
actually
that
whatsapp
was
pretty
similar
to
skype
in
its
and
in
its
way
of
perceiving
the
world,
and
essentially
you
have
the
same.
Siwo
approach.
Look
we're
one
service
provider,
we
give
you
clients
and
services
and
everything
and
we
try
to
make
it
as
easy
as
possible
for
you
to
to
start
chatting
right.
Would
you
agree
with
that?
Yeah.
B
C
B
C
And
extensions
were
numbers
and
what
skype
brought
was
the
same
communications
pattern,
except
that
it
was
body
list
driven
and
then
it
was
what
driven
sorry
body
list
driven.
So
you
had
your
body
list
and
then
these
were
people
and
it
was
names
and
then
it
was
also.
I
think
it
was
using
presence
in
the
way
that
it
was
kind
of
originally
envisioned
by
the
people
like
the
xmpp
crowd
and
the
sip
crowd.
So
it's
like.
A
Yeah,
that's
not
really
the
whatsapp
right
that
was
just
a
matter
of.
We
actually
became
available
all
the
time.
It's
not
just
that
your
presence
stayed
changed.
You
you,
your
personality
changed
because
now
we
walk
around
with
these
connectors
to
the
internet
all
the
time
and
we
are
available.
It
is.
C
It
is
true,
I
mean
the
one
thing,
for
example,
one
of
the
things
that
whatsapp
didn't
have
day
one,
but
they
added
later
was
the
double
tick.
I
believe
so
the
read
receipts
and
I
think
that
further
cemented
these
these,
like
eagerness
to
to
say
when
is
my
reply
coming,
I
just
texted
you,
I
see
two
ticks.
It
means
your
phone
has
and.
C
So
it's
I
mean,
I
think
those
things
are
innovations
in
their
own
right,
because
they
do
match
a
little
bit
of
how
we
communicate
today.
But
the
whatsapp
model
came
in
in
my
mind
in
a
very
similar
way
to
how
skype
well
to
what
skype
already
had
settled,
which
is
like
we're
not
going
to
not
use
phone
numbers.
B
There's
a
slight
twist
like
I
would
say
that
fiber,
for
instance,
when
it
came
on,
was
the
first
one
I
think,
which
actually
plundered
your
contact
list
in
order
to
identify
people
based
on
their
e164
numbers
rather
than
using
a
random
display,
name
identifier
like
skyped
it,
and
that
was
a
minor
twist,
but
it
was
quite
revolutionary
and
then
once
I
did
the
same
thing
and
just
totally
change
that
vibe
and
you're
right,
the
presence
is
dead.
You
know
we.
B
Basically
we
hardly
use
it
on
matrix
at
all,
it's
implemented,
but
nobody,
it's
not
turned
on
by
default,
because
people
use
read
receipts
instead
and
frankly,
it's
a
better
experience.
Because
not
only
do
you
know
whether
the
person's
online,
you
also
know
whether
they're
reading
your
message
or
not.
A
That's
what
we're
here
for
so
so
certainly
amdux
takes
an
interest
in
hey.
Maybe
we
can
we
can
build.
How
would
we
code
that
you
know
a
new
generation
communication
app
and,
and
you
start
working
on
it-
I
I
remember
your
first
presentations
from
matrix.
You
were
very
much
still
with
amdocs.
It
wasn't
amdux
perceived
as
an
amdux
project.
What
happened
then,
so.
A
B
B
B
I
think
there's
a
lot
more
overlap
than
anybody
would
like
to
admit
it's
almost
like
we're
adjacent
countries
or
something,
but
it's
almost
like
england
is
a
whole
bunch
of
french
people
from
a
thousand
years
ago.
But
anyway,
what
happened?
Is
we
spent
the
two
years
before
matrix
selling
this
clone,
basically
of
what's
apple
skype
to
telcos?
It
was
really
interesting
because
we've
got
to
learn
how
to
run
a
business,
but
also
deploying
a
carrier
grade
scale
was
fun.
We
finally
got
to
take
our
soft
switch.
B
The
app
in
brazil
was
called
blah,
which
apparently
is
a
very
cool
word
in
brazilian
portuguese,
which
I
don't
really
believe,
but
either
way
it
was
blah.com
sadly
defunct
now,
but
that's
where
we
really
created
the
matrix
team,
but
not
the
product.
And
after
two
years
of
doing
that
we
had
a
bit
of
a
crisis
because
it
was
working.
You
know
we
were
profitable.
The
deals
were
like
tens
of
millions
of
dollars.
It
was
a
good
gig.
B
A
B
A
B
Honestly,
first
move
for
advantage:
global
network
affects
possibly
that
it's
not
branded
as
a
telco
and
that
it's
an
indie
app
and
also
they
were
moving
much
faster
than
us,
because
we
were
certainly
somewhat
slowed
down
by
the
process
of
going
viral
telco.
I
mean
we
liked
in
brazil
they
were
great
yeah
yeah,
but
still
there
are
okay.
B
B
Just
yeah
not
me
personally,
but
it
was
using
again
the
same
void
stack
like
a
webrtc
stack
that
we
had
for
that
yeah
it
worked,
and
that
was
like
four
years
before
skype
did
video
conferencing
on
mobile
and
I
think
we're
one
of
the
first
in
the
world
to
do
it
as
a
service
at
all.
But
it
was
it
just
didn't.
Have
the
network
effects
the
whatsapp
had
already.
A
B
Well,
it's
more
having
shipped
the
initial
app
with
the
features
that
we
had.
There
obviously
needed
to
be
constant
evolution
and
iteration
on
it
to
keep
it
up
to
speed,
whereas
a
typical
telco
customer
at
the
time
and
possibly
even
now,
views
an
app
as
a
thing
that
you
buy.
It's
like
buying
a
soft,
it's
yeah
like
buying
a
soft
switch
or
buying
a
network
switch
or
a
base
station.
B
You
know
you
go
and
if
you're
a
telco,
you
can
buy
your
4g
network
from
ericsson
and
you
bought
it
and
it's
done
and
in
10
years
you
buy
a
5g
network
and
they
were
thinking
of
it
in
the
same
way,
whereas
obviously
whatsapp
was
just
constantly
doing
middle-income
incremental
improvements,
nothing
massive,
but
I
don't
know
it
might
have
been
captioned
images.
I
think
came
out
about
that
time
or
the
ability
to
bundle
images
into
a
set.
And
obviously
we
could
have
done
that
too.
B
But
the
the
velocity
and
the
agility,
when
selling
to
a
telco
is
very
different
to
if
you're,
just
doing
it
as
a
startup
like
whatsapp,
we're
doing
it
at
the
time
all
right,
and
so
we
reckon
that
we
needed
to
change
something
massively.
If
we
were
ever
going
to
displace
facebook
and
whatsapp
and
that's
where
matrix
came
from.
It
was
basically
the
experience
of
10
years
of
doing
sip
and
xmpp,
and
all
these
other
protocols
and
saying
if
none.
B
B
A
A
Else
that
was
interesting.
I
I'll
probably
just
add
that
to
the
question,
because
your
next
thing,
as
you
are
reimagining
things
from
the
ground
up
and
you
immediately
go
to
what
matters
for
developers,
which
is
interesting,
that
you
would
think
that.
Why
won't
matter
why?
Why
do
you
believe
that
the
the
developers
here
are
are
those
who
have
to
be
satisfied,
you're,
obviously,
thinking
about
an
ecosystem
you're,
not
thinking
any
more
about
the
sidewalled
approach
and
I'm
curious
what
took
you
there
as
well?
So
I
guess
it
is
the
same
question.
B
Yeah
yeah,
I
guess
this
is
the
fundamental
thing.
Well,
I
built
a
protocol
and
an
open
network
rather
than
yet
another
silo,
and
the
realization
was
that
even
at
that
point,
whatsapp
was
getting
towards
a
billion
users
and
they
already
have
massive
critical
mass.
So
how
are
you
going
to
go
up
against
an
incumbent
like
that?
The
only.
B
A
B
I
guess
the
bet
that
we
have
been
playing
on
is
that
in
the
long
term,
you
can
build
a
more
exciting
and
more
successful
and
more
creative
and
transformative
thing.
If
you
do
it
in
an
open
ecosystem,
rather
than
as
a
silo
as
a
silo,
you
can
get
rich
quick.
You
can
capture
100
million
users
and
serve
yourself
not
once
but
twice
to
ebay
and
microsoft.
B
If
your
skype,
however,
it
screws
everybody
else
over,
it
deprives
the
rest
of
the
species,
whether
they
are
end
users
or
developers
from
there
being
an
open
environment
where
they
can
be
creative
and
built
on
an
open
platform.
So
none
of
this
would
be
happening
if
the
internet
didn't
exist
as
an
open
platform
that
is
unencumbered.
B
So
matrix
is
very
much
a
ideological
attempt
to
fix
that
balance
before
it's
too
late
and
finally,
add
the
missing
open
communication
there
for
the
open
web,
so
that
there
is
just
a
standard
way
to
store
and
publish
and
subscribe
to
real-time
data.
And
it's
important
to
understand.
It's
not
message
passing
it's
not
my
xmpp.
It's
not
my
smtp.
A
So
you're
you
you're.
I
I'd
like
to
understand
your
opinion
on
this
because
you're
talking
about
this
as
you
you,
you
perceive
the
current
situation
in
the
world
to
be
a
dystopian
one,
a
broken
one
yeah
and,
and
you
believe
that
you're
enabling
a
better
one
is.
Is
that
fair
or
is
it?
Do
you
believe
that,
at
the
end
of
the
day,
everyone
should
should
end
up
using
something
matrix-based?
Do
you
think?
A
B
Want
them
to
have
the
option
to
use
it,
and
also
we
want
to
be
able
to
bridge
through
to
people
who
are
otherwise
trapped,
so
you
could
natively
be
using
slack
or
discord
or
teams
or
whatever,
but
the
fact
that
there
is
a
open
fabric
that
you
can
link
the
conversation
to
and
participate
with.
People
within
that
wider
network
is
good
enough
for
us.
So
it's
very
much
not
trying
to
just
bulldoze
away
all
of
these
silos
and
have
everybody
sitting
on
top
of
matrix.
B
B
Then
it's
going
to
be
a
safer
solution
than
having
76
000
random,
assist
admins
running
their
own
servers,
which
is
what
happens
in
matrix,
and
some
of
them
are
not
professional
assistants.
I'm
sure
many
matrix
servers
on
the
public
internet
today
are
wide-open
security
disasters
and
therefore
ends
up
being
less
secure
than
if
you
put
all
of
your
eggs
in
one
basket.
B
So
it's
a
trade-off
between
privacy,
hopefully
theoretically,
and
freedom
on
the
matrix
side
of
things.
So
that
would
be
one
example
of
the
benefit
of
a
silo
approach.
I
still
think
it's
a
dystopia,
because
if
I
work
to
some
national
nation-state
security
agency,
I
wanted
to
recap:
I
know
which
secure
messenger.
I
would
put
all
of
my
effort
trying
to
compromise.
C
But
don't
you
think
that
also
kind
of
in
the
same
way
as
we
gravitated
away
from
presence,
we
have
gravitated
towards
these
siloed
things
like
everything,
did
the
one
thing
that
remains
federated
is
email,
and
even
so
people
that
host
their
email
today
in
2021,
you
need
to
be
really
careful.
If
and
and
do
your
job
very
diligently,
if
you
want
to
receive
email
from,
you
know
gmail
and
the
likes.
C
So
you
did
mention
that
that
these
two
different
approaches
like
having
these
interconnected
islands
has
has
certain
challenges
like
federation
in
and
of
itself.
Has
this
it's
a
lot
harder
to
ensure
security,
safety
and
twin?
How
do
you
deal
with
rogue
agents?
So
what
is
like?
Do
you
think
this
will
be
where
things
will
end,
or
do
you
envision
a
day
where?
Well,
maybe
you
end
up
having
element
be
like
signal,
a
big
central
thing
hosted
by
you
like?
Do
you
have
numbers
on
matrix
federation
that
show
you
how
healthy
the
ecosystem
is.
B
B
It's
so
much
easier
to
build
a
silo
than
a
decentralized
system,
and
I
know
firsthand
because
for
10
years
we
built
proprietary
silos
and
then
for
the
last
seven
years,
we've
been
building
precisely
the
same
thing,
except
now:
it's
open
source
and
it's
decentralized
in
its
standard
space
and
it's
the
same
literally
the
same
team
of
people
who
have
done
both-
and
I
would
say
it's
about
six
to
ten
times
more
effort
to
make
it
work
in
a
byzantine
fault,
tolerant,
decentralized,
open,
federated
manner
than
if
you're
just
doing
it.
It's
a
centralized
product.
B
So
why
would
you
do
that?
If
you,
if
I
walk
out
of
the
office
and
go
on
a
chiswick
high
road
and
ask
some
random
guy
on
the
street?
How
much
is
it
important
to
you
that
your
messenger
is
decentralized
and
built
on
an
open
standard
and
is
end-to-end
encrypted?
The
99.9
chance
is
that
they
don't
care
at
all.
So
why
would
they
even
use
this
rather
than
whatsapp
or
the
alternative,
and
the
fact
is
that
it
has
to
be
better.
B
It
has
to
be
at
least
as
good
in
terms
of
features,
but
if
you
have
put
in
that
effort
to
go
and
create
a
new
protocol,
and
then
you
stack
in
a
new
industry
and
a
new
ecosystem
of
people
participating
within
it,
then-
and
it
is
as
good
as
the
centralized
alternative-
then
it's
more
should
I
use
an
encrypted
product
or
an
unencrypted
product.
They
do
the
same
thing.
It
costs
the
same.
To
me.
C
B
A
Oh
man,
I'll
take
you
up
on
that
one,
because
I
I
think
people
are.
You
know
it's
way
too
easy
to
bash
on
on
on
the
facebooks
and
the
youtubes
for
for
their
algorithms
and-
and
I
think
people
have
been
taking
that
facility
more
than
they
should,
because
you
know
you
and
I
and
so
were
around
in
the
days
where
people
were
on
irc,
and
we
remember
that
the
addicted
types
existed
back
then
as
well.
I
I
know
people
who
were
just
going
through
the
night
having
pointless
conversations
and
fights
and.
A
Unencrypted
so
so
I
would.
I
would
actually
argue
that
that
entire
addiction
is
much
more
a
function
of
our
desire
of
the
fact
that
now
communication
is
so
easy
and
that
we
get
to
cherry-pick
what
communication
we
have
and
what
information
we
receive,
and
I
would
I
would.
I
would
posit
that
the
algorithms
here
are
probably
helping,
but,
but
only
only
so
much
so.
I.
B
Totally
agree
that
humans
are
addictive
animals
and
speaking
as
they're,
not
very
reformed,
irc
addicts
nowadays
completely
carrying
matrix
addicts,
people
can
get
themselves
addicted
to
all
sorts
of
things
without
any
help.
However,
it's
particularly
toxic
feedback
loop
you
get
when
the
content
that
you're
being
produced
is
deliberately
targeted
in
order
to
make
you
more
shocked
and
horrified
and
addicted
and
disgusted,
and
generally
it
doesn't
optimize
for
quality
by
any
metric
whatsoever.
It
is
literally
just
trying
to
feel
that
feedback
loop.
So
I
see.
A
B
It
collapses
now.
This
was
the
real
real
eye-opener
that
we
had
in
about
2016
on
matrix
that
we
got
the
decentralized
comms
working
relatively
well,
it
pulled
hot.
No,
this
works.
I
can
see
this
replacing,
I
can
see
it
replacing
xmpp.
Potentially.
This
is
great,
oh
wait!
How
in
a
second,
anybody
can
join
this
network,
it's
end-to-end
encrypted
and
anybody
can
run
a
server
and
you
can
use
it
for
anything.
B
You
like
what
could
possibly
go
wrong
and
obviously
you
end
up
with
an
awful
lot
of
people
using
it
for
an
awful
lot
of
different
things,
and
some
of
them
are
fine
and
some
of
them
are,
depending
on
your
particular
viewpoint,
potentially
really
not
fun
at
all.
So
you
have
to
provide
the
tools.
You
have
to
empower
the
users
to
be
able
to
filter
out
the
stuff
that
they
don't
want
to
see.
So
basically
they
need
an
algorithm
to
surface
content.
B
You
can't
really
have
a
centralized
reputation
system
that
goes
and
feeds
these
sort
of
that
goes,
and
does
the
filtering
for
you.
Instead,
you
have
to
embrace
the
moral
relativism
here.
It's
a
subjective
thing.
One
guy's
terrorist
is
going
to
be
another
guy's
freedom
phone
to
one
guy's.
Adult
inventory
is
another
guy's
illegal
pawn.
A
I
I
actually
think
that
that
is
exactly
what
creates
the
addictivity
in
the
echo
chambers,
letting
the
users
choose
for
themselves,
so,
and,
and
and
and
essentially
that's
what
the
algorithms
at
facebook
and
youtube
do
as
well.
They
try
to
encourage
you,
but
ultimately
they
don't
force
you
into
anything.
A
So
they
had
to
have
a
little
bit
of
everyone,
a
little
bit
of
everything
for
everyone
and
you
end
up
kind
of
getting
alternative
viewpoints
in
spite
of
yourself.
Just
because
you
know,
there's
a
there's
a
show
about
the
cows
and
you
hear
about
the
problems
with
the
cows
and
now
you
know
it
at
the
back
of
your
mind
that
people
worry
about
the
cows
and,
if
you
think
of
it,
that's
kind
of
the
main
value
of
of
of
sunday
church
gatherings
as
well.
A
Right
that
you
you
get
to
rub
shoulders
with
your
with
your
peers,
the
ones
that
you
wouldn't
necessarily
talk
to.
But
now
you
hear
that
you
know
this
family's
kid
is
sick
and
the
house
for
these
other
people
burned
and,
and
and
that's
because
they
were-
you-
know
this
other
guy
with
the
tractor
burnt
it
down
accidentally.
So
now
you
know
the
tractors
can
be
a
problem
for
some
people.
So
it's
I
I.
A
I
actually
think
that
when
you
it's
not
that
I'm
disagreeing
with
you
because
there's
a
part
of
what
you
said
that
I
very
much
agree
with
while
or
maybe
all
of
it.
You
know
it's
at
the
end
of
the
day.
Maybe
that's
a
little
bit
too
lazy
for
us
to
rely
on
public
broadcasters
to
to
to
service
our
information
diet
because,
obviously
now
they're
in
control
and
and
can
spin
things
in
a
way
which
they
always
have
been
to
a
limit
right,
because
they
couldn't
afford
to
do
it.
B
Is
the
key
balancing
act
right?
I
think
it's
a
really
good
analogy
that
if
you're
an
irc
network
20
years
ago,
you
type
slash
list
you
get
like
100
000
channels
about
every
random
thing
in
the
world,
and
you
have
to
kind
of
vaguely
pick
your
way
through
it.
You
don't
have
any
good
tools
to
pick
your
way
for
it.
You
might
eventually
find
the
correct
hash
linux
or
whatever
it
is
that
you
were
looking
for.
B
B
A
A
You
had
two
channels
and-
and
it
didn't
have
that
much
of
a
choice
and-
and
I
agree
that
we've
been
traveling
toward
the
world
of
today-
you
know
the
minute
that
you
had
cable
and
then
two
operators
became
20
or
200.
We
started
traveling
toward
the
world
today,
which
is
where
you
have
you
know,
maybe
two
billion
information
for
possible
information
sources,
but
in
the.
B
Newspaper
world
you've,
almost
oh,
unless
you're
in
a
dictatorship
where
the
government
is
deliberately
suppressing
the
media
sort
of
range
there's
always
been
at
least
two
contradictory
viewpoints
from
like
the
1600s,
and
before
that,
it's
almost
who
do
you
go
to
in
the
village?
To
get
a
point
of
view?
B
Do
you
talk
to
that
elder
or
that
elder
and
they're
going
to
tell
you
the
left
wing
or
the
right-wing
kind
of
viewpoint
or
whatever,
whereas
on
a
facebook
world
that
you
just
don't
have
that
there
is
one
algorithm
that
has
been
created
for
you,
based
on
the
things
which
you
seem
to
be
interested
in
and
you
just
can't
change
the
viewpoint.
There
is.
A
I
would
I
you
know
I
would
I
would
again
I'm
very
I'm
not
convinced
at
all
that
the
algorithm
is,
what
really
determines
you,
but
essentially
what
we're
arguing
about
is.
Do
people
get
their
information
bites
because
primarily
they're
served
to
them
by
algorithms,
or
is
it
primarily
because
they
seek
out?
You
know,
I
I
want
to
see
what
alexandria
cassio
court
said
next
or
or
what
ben
shapiro
said
next
or
and
they
just
go
and
and
seek
out
the
same
sources
again
and
again.
I'm
sure
that
the
algorithms
play
a
role.
B
It's
a
second
order
effect.
I
agree
on
top
of
the
fundamental
one
of
what
people
were
looking
for
in
the
first
place,
but
I
guess
we
see
a
responsibility
with
matrix
to
try
to
let
people
literally
create
their
own
algorithms
and
also
visualize
what
their
algorithm
is
doing
and
saying:
hey
we're
showing
you
this
aoc
post,
because
you
keep
stalking
her
twitter
feed
or
where,
by
the
way,
we're
showing
you
this
trump
thing
and
99
of
the
world
might
disagree
with
what
this
says,
but
we're
showing
it
to
you
anyway.
B
By
the
way,
do
you
want
to
even
visualize
your
algorithm
and
we'll
be
playing
around
with
webgl
and
doing
kind
of
map-based
visualizations
of
the
communities
which
exist
in
the
public
matrix
network?
Obviously
you're
not
looking
at
any
private
metadata,
but
just
looking
at
the
public
room
as
out
there,
and
you
can
literally
do
cluster
analysis
to
draw
big
blue
continents
and
big
red
continents,
and
you
can
put
a
pin
in
the
map
and
say
right
now.
Your
viewpoint
is
here.
By
the
way,
do
you
want
to
see
what
happens
to
the
room
list?
B
If
you
go
and
move
it
over
here
and
no,
I
could,
if
I
screen
show
right
now,
we'll
be
able
to
see.
Yes,
let
me
just
go
and
quickly
pull
up
a
thing
called
sierra
leone,
which
hopefully
is
online.
It's
a
clone
of
twitter
that
we
built
on
top
of
matrix
I'll
just
have
to
find
it
and
what
it
has
in
it
is
the
ability
to
literally
curate
your
own
algorithms
as
an
experiment,
so
that
you
basically
subscribe
to
reputation,
feeds
which
are
rooms
and
matrix
ipads
topics
and
matrix,
and
they
have
lying
this
user.
B
This
room,
the
server
this
community
is
plus
100
minus
100,
and
it's
entirely
up
to
you
which
ones
to
subscribe
to.
You
can
then
blend
them
together
to
different
percentages,
and
then
you
can
republish
that
and
a
bit
like
pandora
playlists,
allow
you
to
go
and
subscribe
to
different
music
tastes
and
then
kind
of
recommend.
A
C
B
It
was
very
specifically
the
room
list
on
matrix.org
was
the
first
starting
point
here,
because
at
the
moment
it
feels
like
an
irc
channel
list.
It's
got,
I
think,
60
000
rooms
in
it
and
it
is
everything
from
hentai
to
maths,
to
games
to
all
sorts
of
different
things
and
it's
just
totally
unusable
and
we
wanted
to
allow
people
to
kind
of
curate
us
and
say
hi,
I'm
interested
in
the
linux
stuff,
and
can
you
get
rid
of
the
nsfw
stuff
or
whatever
it
might
happen
to
be,
but
it
becomes
even
more.
B
So
let
me
just
show
my
screen
and
show
my
whole
screen
for
expedient
if
in
a
kind
of
twittery
style,
micro
blogging
context.
So
hopefully
you
can
see
here
my
sierra
leone,
client
and
here
I've
got
and
I've
subscribed
to
hash
evil
on
dendrite.matrix.org,
and
this
was
a
room
if
I
say
that
I
like
nsfw,
I
like
this,
so
it
turns
out
that
some
of
the
some
of
these
things
I
can
adjust
the
levels
on
to
literally
gray
them
out.
B
So
let
me
go
to
say
this
post
here
and
it's
a
threaded
set
of
conversations.
It
happens
that
it's
done
via
matrix,
but
it's
got
the
same
kind
of
semantics
as
a
hack
and
you
style
thing
or
a
twitter
style
thing.
And
if
I
go
into
here
and
I
say:
hey,
I
don't
want
to
see
evil
stuff.
I
don't
know
if
I
have
any
evil
users
here.
Let
me
find
a
better
test
thing
quickly.
B
B
It's
not
it's
the
the
dancing
man
or
whatever
it's
great.
We
also
have
bob
in
here
somewhere,
bob
wow
bob
sorry
get
confused
by
the
captions
being
bored.
Do
we
actually
have
bob
in
here?
No,
it's
just
the
the
dancing
man,
but
either
way
you
get
the
idea
that
you
can
subscribe
to
different
views
and
then
you
can
go
and
say
I
dislike
evil.
A
You
know,
there's
there's
two
aspects
of
this:
one
of
them
excites
me
and
the
other
one
doesn't
excite
me
at
all.
The
one
that
doesn't
excite
me
is
that
it
means
more
work
because
I
now
have
to
go
and-
and
you
know,
understand
these
concepts
and
and
and
tweak
them
and
keep
them
in
mind
and
maybe
readjust
them,
and
I
don't
know
how
much
I
I
would
like
to
do
that
now.
The
part
that
excites
me
is
about.
A
It
is
not
so
much
about
what's
happening
and
what
has
to
happen
in
the
work
that
I
have
to
do
it's
it's
about
the
work
that
someone
doesn't
have
to
do.
What
I
mean
by
this
is
you
know
I
in
the
in
the
early
2000s
I
I'd
say
I
was
you
know
a
die
hard
fan
of
federation
as
peter
saint
andre
would
say
you
know
federate
or
die.
A
It
really
wasn't
obvious
to
me
that
federation
was
bringing
anything
of
substantial
value
and
it
was
obvious
to
me
that
it
is
making
things
harder
in
terms
of
innovation,
because
you
have
to
keep
moving
parts
up
at
the
same
level
and
then
act
accordingly
when
they're
not
at
the
same
level
and
the
the
complexity
was
hugely
increased.
So
I
I
guess
at
one
point
I
was
just
I
don't
know
why
why
you
would
bother.
A
I
don't
know
what
the
problem
is
with
having
10
apps
on
your
phone
and
just
using
10
different
networks,
I'm
not
gonna
visit.
There
was
a
problem
and,
interestingly
enough,
what
brought
federation
in
as
as
a
necessary
thing
was.
You
know
it's
kind
of
what
you
were
saying
about
compromising
moxie
in
in
signal.
A
Interestingly,
though,
it's
not
the
secret
services
that
compromise
that
single
party-
you
know
it's,
not
the
nsa,
it's
not
the
cia.
What
I
think
we've
seen
way
more,
often
and
way
more
obviously,
is
we
the
mob
compromising
them?
So
if
you
look
at,
if
you
look
at
facebook
and
and
twitter
now,
you're,
essentially
they
are
moxie
in
their
own
silos.
They
run
they
which
makes
them
look.
A
If
I
have
a
problem
with
what's
going
on
that
platform,
I
know
who
I
have
to
go
and
shout
it
and
and
these
people
are
going
to
cave
in
and
if
the
mob
is
big
enough.
So
you
end
up
having
twitter
and
facebook
essentially
having
to
pick
sides
in
in
ideological
debates,
philosophical
debates
really
where
they
really
shouldn't
have
to
pick
sites,
and
once
they
pick,
they
end
up
curating
views
that
are,
you
know,
morally,
not
outrageous,
but
simply
incompatible
with
one
way
of
viewing
the
world
and-
and
it's
inevitable.
A
We
know
we
just
took
taking
them
down
from
from
two
app
stores
and
and
which,
which
is
outrageous,
but
at
least
I
can
understand,
and
then
amazon
piles
in
which
felt
almost
like
it's
like.
Oh,
we
want
to
be
part
of
the
cool
kids
too.
We
want
to
show
everyone.
No
one
knew
that
power
was
using
amazon,
but
we're
going
to
tell
everyone
that
that
we're
also
cool,
because
we
can
also
ban
them
look
and-
and
it
was,
it
was
so
outrageous
that
that
these
things
were
happening.
A
You
know
tar
and
feathering.
Essentially,
so
so
that's
what
excites
me
about
about
the
idea
of
federation
about
the
idea
of
leaving
those
responsibilities
to
people.
I
think
they
would
hate
it,
but
at
least
it
wouldn't
be
a
single
party
responsible
for
it
anymore.
I
mean
this
takes
me
to
to
to
to
my
next
line
of
questioning
and
which
was
going
to
be
how
much
more
time
do
you
have
do
you
have
another
15
minutes
or
so
or
oh,
I've
got
a
bit
of
a
stop
on.
A
Right
all
right,
so
you
guys
just
went
over
and
had
a
very
successful
funding
round.
Congratulations,
I'm
really
happy
for
you
all.
I
remember
I'm
actually
very
admirative.
You
had
a
very
hard
very
hard
beginning
after
amdocs
money
was
scarce
and
you
persevered
you
tried
patreon,
you
tried,
you
tried
contracts,
you
tried
a
bunch
of
different
things
and
you
finally
found
some
peace
and
calm,
and
you
know
the
ability
to
project
yourself
in
the
future
and
and
is
it?
A
That
means
business
that
has
money
and,
and
that
can
pay,
for
you
know
for
various
services
that
you're
offering
and
and
in
in
it's
now
easier
to
understand
the
you
know
why
why
why
element
will
be
a
successful
business
enterprise?
Would
you
would
you
agree
that
that's
kind
of
a
big
part
of
the
reason?
Oh.
B
A
B
Our
severance
money
together-
and
it
said
that
we
worked
together
for
three
months
on
matrix
to
see
if
we
could
make
a
go
of
it
and
luckily
were
able
to
prove
that
it
had
legs
enough
to
raise
investment
and
then,
since
then,
it's
been
pretty
good
and
the
way
in
which
we
proved
that
there
was
appetite
for.
This
was
first
from
public
sector
deployments,
the
entirety
of
the
french
state
coming
on
board.
Why.
B
Interested
in
you
because
they
want
to
control
their
own
communications,
they
don't
want
to
be
sitting
unencrypted
in
america.
If
you
are
a
country
and
you
want
to
communicate,
then
you
want
to
have
it
under
your
own
roof
and
you
certainly
don't
want
it
unencrypted,
given
that
it
could
be
used
for
really
sensitive
things,
and
that
has
been
our
initial
niche,
but
it's
then
expanded
all
over
the
shop.
B
Obviously,
in
the
open
source
and
the
kind
of
nerdy
domain
that's
been
very
popular,
but
also
big
private
sector
companies,
particularly
those
who
interface
with
the
public
sector,
find
themselves
wanting
to
be
on
the
same
communications
network
that
the
public
sector
guys
are
on.
So
you
literally
get
a
network
halo
effect
a
bit
like
the
early
days
of
email
and
the
internet
that
sure
you
could
be
on
compuserve
or
sprint
net
or
a
t
and
d
net,
or
you
can
use
this
slightly.
B
C
B
B
It
would
not
federate.
Basically,
I
mean
I
think
rocket
chat
has
got
basic
private
federation
going
on
these
days,
but
it's
not
like
you're
part
of
a
much
larger
network
and
it
doesn't
have
the
partition
resilience
that
you
get
in
matrix
that
even
if
there
is
a
net
split
briefly,
it
goes
and
merges
up
seamlessly
afterwards
and
that's
the
real
kicker
in
france.
B
We've
got
65
deployments
and
it's
roughly
three
per
ministry
and
you've
got
the
public
facing
one
and
then
you've
got
various
sort
of
air
gaps,
internal
ones,
and
they
all
can
talk
to
one
another.
So
it
gives
the
it
departments
complete
self
sovereignty
as
to
whether
they
run
it
on
premise
in
the
private
cloud
public
cloud,
wherever
to
run
it
on
windows
or
run
it
on
linux.
It
doesn't
matter
because
in
the
end,
it
can
all
go
and
participate
in
that
wider
network.
B
When
you
have
other
countries,
also
coming
on
board
like
in
germany,
there's
the
entire
health
service
now
gematic
mandating
matrix
is
the
way
for
the
health
industry
to
intercommunicate
or
the
bundeswehr
the
armed
forces
again
using
it
throughout
the
whole
thing
as
the
one
way
to
communicate.
Then
it's
it's
a
no-brainer
that
the
french
minister.
Instead,
these
army
will
want
to
go
and
talk
to
the
bundeswehr
and
have
the
same
protocol
for
into
communicating.
B
Then
we
provide
bridges
to
xmpp
for
talking
through
on
other
coalition,
comms
and
nato
stuff,
where
there's
lots
of
x
and
pp,
and
I
think
the
only
alternative
really
would
be
an
xmpp
based
approach.
But
for
whatever
reason,
there
doesn't
seem
to
be
a
competitor
to
slack
called
teams
or
discord
which
is
built
on
xmpp,
and
if
there
was
then
perhaps
you
wouldn't
have
gone
off
and
written
one
on
top
of
matrix.
A
B
And
so
we're
just
trying
to
provide
tools
to
let
people
basically
make
up
their
own
minds
matches
they
do
in
the
real
world.
There's
nobody
in
the
real
world
telling
you
forcing
you
to
read
the
guardian
rather
than
the
sun
or
whatever,
and
likewise
this
is
giving
the
same
kind
of
flexibility
for
the
users
to
decide
which
team
they
want
to
go
on
and
hope
that
the
good
ones
for
your
definition
of
good
win.
A
So
matt
this
was
matthew.
This
was
awesome
and
I
I
actually
think
that
I
we
could
spend
another
hour
and
we're
going
to
try
and
set
that
up
to
see
if
you
have
availability
talking
about
I'm
very
interested
in
diving
into
the
government
use
case,
and-
and
you
know
what
exactly
works
for
them,
what
doesn't
would
so,
let's,
I
hope
it's
all
right
with
you
that
we
try
and
schedule.
B
A
Exactly
so
so,
let's
try
and
do
that
and
pause
here
for
the
time
being,
it
was
really
tons
of
fun.
Hearing
your
perspective
on
all
these
on
all
these
things
and
I'm
very
very
happy
that
we
get
to
start
this
with
you.
Thank
you
for
agreeing
to
spend
time
with
us,
and
so
did
you
want
anything?
Do
you
want
to
add
anything
and
at
closing.
C
I'm
very
much
looking
forward
to
another
session
when
we
dive
into
some
other
topics
as
well.
We,
we
I'll
admit
we
did
have
a
a
little
bit
of
planning.
We
we
put
some
topics
in
the
list
and
I
think
we
barely
scratched
the
surface,
but
that's
a
good
thing,
exactly
yeah,
it's
it's
it's
a
great
sign
and
and
for
example,
you
did
briefly
touch
on
on
the
peer-to-peer
aspect.
I
think
that
that's
another
really
interesting
topic
where
we
can
we
can
dive
into
because
it
has
evolved
a
lot
across
time
and
yeah.
B
Breakthrough
a
couple
of
months
ago,
with
a
whole
new
routing,
algorithm
called
snack,
which
I
would
love
to
talk
about,
but
yeah
thank
you
for
the
opportunity
to
chat
and
to
be
part
of
the
first
broadcast.
I'm
I'm
honoured
and
it'll
be
really
fun
to
follow
it
up
at
some
point
in
the
future
sounds
good.
Enjoy
your.