►
Description
How we have integrated IPFS, Tauri and Mintter Hypertext Documents to enable open collaboration. We will explain our requirements for an Open and local-first Hypertext System and how IPFS and Tauri have help us to build the system.
A
A
A
It's
everything
is
open
source
GitHub.
You
can
go
to
Minter
team
Minter
repository
GitHub.
It's
I
mean
you
can
download
the
application
if
you
want,
but
I
only
recommend
it
for
very,
very
patient
people.
Still
it's
a
bit
unstable
but
happy
to
reach
us
and
to
tell
us
feedback
and
any
bugs.
You
find
looking
forward
to
connect
with
ipfs.
A
So
we
are
team
F5
with
working
for
the
last
almost
three
years
in
ipfs,
and
it's
very
exciting
to
be
here
because
it's
the
first
time
that
we
are
really
interacting
with
the
community.
So
it's
like
it's
been
like.
We've
been
in
the
cave
for
three
years,
and
now
we
are
rich
in
the
community,
so
we
are
very
excited
to
be
here
really
looking
forward
to
collaborate
and
connect
and
talk
to
all
of
you.
A
A
Science
is
broken,
journalism
is
broken,
even
universities
are
close
to
being
broke,
and
this
matters
so
much
this
matters
so
much
so
we
believe
that
we
need
to
focus
on
fixing
it,
and
we
think
that
we
need
a
new
generation
of
digital
tools,
so
we
need
to
run
away
from
this
Doom
scrolling
emotional
emotions
control
that
and
when
it
a
new
generation
of
detailed
tools.
That
brings
not
only
dialogue
we
already
have
but
dialogue.
That
creates
knowledge.
That
brings
knowledge
creation.
A
So
this
started
this
started
three
years
ago
and
we
started
searching
for
answers.
How
can
we
fix
it
so
that
Nelson
Doug
Engelbert,
like
we
have
read
it
all
and
seen
all
the
videos,
and
and
it's
incredible
that
they
fix
everything
like
50
years
ago
and
now
we
are
finally
building
what
they
envisioned,
which
is
a
lot
related
with
the
how
humans
work,
that
we
associate
ideas
in
our
minds
and
how
to
make
that
happen
in
the
digital
metaverse.
A
So
what
we
have
built
on
top
of
all
these
old
ideas
trying
to
solve
the
Urgent.
Our
urgent
matters
is
a
decentralized
knowledge
collaboration,
application
powered
by
a
sinkable,
Knowledge
Graph
that
we
believe
that
can
fix
public
deliberation,
a
public
deliberation
that
creates
knowledge,
let's
see
so
basically,
first
is
decentralized.
We
need
decentralization
oops.
A
We
need
a
decentralization
and
that's
something
that
decentralized
identities
bring
us
because
allows
permissionless
participation
in
computer
networks,
where
people
can
share,
communicate
collaborate
without
central
servers
or
without
Gatekeepers
or
like
the
old
CIS
operators
bbss.
Where
there's
someone
watching
us
and
it's
so
creepy
and
it's
a
time
for
building
like
networking
applications
and
that's
what
decentralized
identities
were
well.
A
A
So
now,
let's
call
our
collaboration
requires
permanent
links,
so
in
the
knowledge
creation
process
like
if
you
are
writing
an
article
and
you're
going
to
refer
another
article,
you
cannot
have
this
like.
You
won't
do
it,
because
the
risk
of
having
a
404
is
so
big
that
you
won't
use
the
web
to
refer
someone
else's
article
in
your
own
article,
because
the
risk
of
breaking
links
is
so
high.
A
So
thanks
to
apfs
now
we
have
permanent
links
and
also
we
have
our
Mentor
documents
syncable.
So
if
you
refer
to
another
document,
you
can
bring
that
document
to
your
computer,
so
you
can
keep
it
forever,
so
your
you
will
make
sure
that
that
link
is
not
going
to
get
broken
and
also
somehow
one
of
the
coolest
things
we
have
built
is
our
own.
A
Now,
let's
think
about
knowledge.
Graph
decentralized,
single
Knowledge
Graph
is
a
dag
of
changes
from
multiple
devices.
We
have
Multi-Device
so
well
similar
how
git
works,
but
we
use
a
crdt
for
automatic
conflict
resolution
so
so
that
allows
to
in
your
local
machine.
You
have
a
set
of
meter
documents
that
just
sync
with
the
people.
A
A
So
conclusion
we
have
built
an
open
access
to
knowledge
assistant
with
permissionless
participation,
permanent
links
to
refer
to
reference
documents.
Versioning,
that
enables
knowledge
collaboration
without
a
central
server
and
the
freedom
to
associate
reuse
and
campaign
and
compose
and
I
think
that
that's
the
way
we
can
solve
public
deliberation
that
creates
knowledge.
A
We
have
been
very
inspired
by
local
first
principles,
that
I
recommend
I,
recommend,
ink
and
switch
paper
on
on
the
principles
of
local.
First
and
like
with
this,
we
have
discovered
like
using
this
layer
of
peer-to-peer
storage
that,
like
the
architecture
of
the
applications,
changed
a
bit
from
the
normal
web
Technologies
architectures,
because
the
like
the
the
data
layer
is
completely
separated
from
the
from
the
user
experience.
A
So
you
have
zero
latency
interacts
interacting
with
data
and
that,
like
upload
our
mind
with
whole
new
possibilities
in
the
development
of
hypertext,
so
instant
navigation,
which
you
know
those
five
seconds
you
have
when
every
time
you
have
two
seconds
when
you
reload
another
website.
So
here
you
don't
have
it
it's
completely
instant.
A
A
So
at
the
very
beginning
we
tried
to
build
winter
on
top
of
the
web
browser,
but
it
it
we
felt
so
constrained
with
the
with
a
web
browser
to
to
express
all
these
hypertext
functionalities.
We
wanted
to
express,
and
also
we
need,
we
needed
like
a
direct
access
to
the
file
system,
as
we
have
this
decentralized
knowledge
graph
integrated
with
ipfs
and
sqlite.
A
So
basically,
we
use
tari,
which
is
a
very
light
library
that
instead
of
shipping
the
whole
web
browser
like
electron.js,
does
it
just
takes
the
native
it
calls
to
the
native
webview
of
the
of
the
party
system
so
includes
so.
Every
single
operating
system,
like
Mac
OS
Windows
Linux,
includes
a
library
with
a
webview
that
you
can
instantiate.
A
So
what
it
provides
us
is
a
a
very
small
size,
bundle
and
very
secure
so
Tory.
They
focused
from
first
principles
to
make
a
very
secure
software
and
it
feels
good.
We
have
contributed
a
lot
to
Tory
like
the
multiple
windows
plugin,
the
Sentry
plugin,
the
CI
pipelines,
plugin,
a
native
to
Bar
plugin
story.
Api
plugin
we've
been
contributing
a
lot
to
the
Atari
project.
A
I,
don't
know
how
much
time
I
have
but
well
this
is
architecture,
so
we
have
the
edema.
We
have
an
ipfs,
we
have
a
well.
The
Atari
calls
the
IP,
the
the
demon,
the
the
demon,
connects
with
a
web
use
with
a
erpc
API.
So
there's
no
direct
calls
to
ffs,
but
it's
the
the
demo
who
calls
the
ipfs
connectivity
and
well
we
just
right.
Now
we
just
react
and
the
different
components.
I
don't
know.
If
someone
has
questions
please
and
we
we
liked
I
got
well
I.
A
So
we
went
to
multiple
windows
approach
instead
of
canvas
or
something
or
those
kind
of
approaches
which
still
allow
us
the
parallel
pages,
so
the
instead
of
having
tabs
the
idea
of
having
parallel
Pages
means
that
you
are.
You
have
the
the
information
in
front
of
you,
so
it
allows
to
make
the
connections
to
read
the
connection,
to
see
the
visible
connections
between
texts
and
also
to
create
to
to
have
a
more
easier
citation
process.
So
this
is
Minter.
So
these
are
all
the
windows
Minter.
A
A
You
can
reuse
date,
content
keep
in
the
authorship,
we
call
them
transclusions,
that's
that
Nelson
set
and
it's
working
using
ipfs.
All
every
document
is
a
CID
copy
link
boom
so
easy.
This
is
our
way
to
find.
So
here
we
are
using
the
DHT
to
our
searching
the
network
boom
Define,
the
network,
the
document.