►
From YouTube: From closed to open, a tale about Parse - CodeConf 2016
Description
Present by Florent Vilmart
Shaking the world of many mobile developers and startups, Facebook announced Parse was shutting down in 2017. Alongside that announce, parse-server was released on github. We'll discuss in that session the impact it had on the community of developers as well as the bright future open source givers to the popular backend.
About CodeConf
CodeConf improves the software community by providing a forum for thought-provoking talks and forging social connections. The fourth installment of the CodeConf series took place in Hollywood in 2016. This year's event focused on systems engineering projects, practices, and programs in the open source community.
For more information on this year's CodeConf, go to:
https://codeconf.com/
A
Let's
start
talking
a
bit
about
me:
it
is
a
great
introduction,
so
I'm
software
engineer
and
me
and
I've
been
using
parts
almost
since
its
out
of
beta
and
when
they
went
open-source
I
saw
it
as
an
amazing
opportunity
to
move
forward
with
the
platform
because,
as
a
savage,
it
was
very,
very
constrained
to
the
limits
that
the
parson
wanted
to
put
on
it.
So
path
came
to
fruition
as
a
mobile
backend
as
a
service
as
a
mobile
developer.
You
don't
want
to
care
about.
A
What's
going
on
in
the
backend,
you
don't
want
to
use
your
database,
you
don't
want
to
configure
anything.
You
just
want
it
to
work.
You
want
to
focus
on
the
experience
and
the
code
of
your
mobile
app.
That
was
free,
which
is
very
cool,
and
if
you
gain
popularity,
you
had
all
sorts
of
plan
that
could
scale
up
to
hundreds
of
requests
per
second,
so
that
was
very,
very
appealing
to
do
by
the
developer
crowd.
It
was
schema-less,
so
that
was
very,
very
easy
to
implement
changes
in
your
models.
A
As
you
were
coding,
your
app
oh
I
thought
that
that
you
don't
have
to
deal
with
all
the
my
sequel,
altering
tables
and
all
that
stuff.
You
just
write
it
in
your
code
in
the
app
and
it
just
works
and
the
security
which
is
for
me
the
best
feature
that
was
put
in
what
manage
with
ADL's
and
and
that
made
it
very,
very
easy
to
reason
about
security
and
the
security
on
your
object.
A
At
the
time
there
was
a
lot
of
competitors,
firebase
cloud,
kids,
I'm,
sorry
Bay's
been
acquired
by
Google
and
moving
it
forward.
In
all
cases,
you
don't
want
to
deal
with
the
service
itself.
You
want
to
use
the
SDK
and
the
SDK
is
really
what
makes
the
difference
between
one
and
another
service
the
easier
and
the
best
documentation
you
have
in
the
SDK,
the
more
likely
you
are
to
use
it
so
since
day
one
passes
great
documentation
and
a
is
for
pretty
much
every
platform.
A
So
let's
rewind
a
bit
and
hit
the
timeline,
so
it
started
in
August
2011
acquired
by
Facebook
in
2015.
At
that
time
we
started
getting
a
bit
worried.
It's
a
supernatural.
It
down
and
the
GTO
Kevin
said
no,
it's
a
bit
alright
we're
going
to
move
forward
2013
with
a
very,
very
different
time.
If
you
think
about
it,
Facebook
Adam
made
a
big
push
on
mobile.
A
They
were
not
generating
significant
revenue
from
advertisement
in
mobile,
so
maybe
that
makes
sense
at
that
time
to
keep
path
up
and
acquire
mobile
developers
and
build
a
great
ecosystem
of
mobile
apps
in
2015.
They
started
open
sourcing.
Everything
I
mean
everything
for
The
Client
SDK.
So
you
got
all
the
SDKs
that
are
open
source
in
2015
and
then
early
this
year
we've
got
the
notice
in
January.
A
Ok,
we're
going
to
shut
down,
but
unlike
many
Silicon
Valley
acquisition,
where
the
stuff
is
just
shuts
down
and
the
users
has
to
find
another
solution,
they
provided
the
solution
which
was
open,
sourcing
the
platform,
and
then
we
expect
the
final
clap
next
year
in
January.
So
if
anyone
is
running
past
just
get
ready
at
that
time,
because
it's
not
going
to
work
any
more
so
pubs
got
a
great
history
with
open
source.
These
are
all
the
SDKs
that
are
available
and
they
are
all
open
source.
You
can
contribute
to
it.
A
A
That's
that's
one
of
the
great
benefits
that
we
have
right
now
with
with
the
platform,
and
so
they
added
those
two
new
major
component
to
their
open
source
contributions,
which
is
the
server
and
the
dashboard
I
contributed
more
to
the
server
because
I'm
not
really
a
front-end
JavaScript
guy.
But
if
any
one
of
you
is
in
love
with
react.js,
you
should
check
out
the
dashboard.
This
is
one
of
the
best
architectured
front-end
application
that
I've
ever
seen.
There
is
a
ton
of
components.
If
you
want
to
learn
more
about,
we
exist.
A
So
pass
was
a
back
end
as
a
service
that,
at
more
than
a
million
users,
more
than
750
apps
100,000
weekly
running
apps
when
it
went
open-source
the
target
is
just
one
user
and
we
are
all
almost
so
for
engineers
here
and
we
know
we
don't
design
software
the
same
way
if
you
want
to
address
a
million
users
to
one
user,
so
the
platform
has
to
be
outdone.
The
product
is
completely
different,
is
under
the
same
requirements
because
the
product
is
different.
A
We
went
from
a
massively
scalable
system
for
millions
of
users
to
a
platform
that
only
addresses
one
user,
the
technology
exchange.
You
don't
make
the
same
technical
choices.
If
you're
building
an
app
for
million
users,
then
you
are
for
one
and
then
because
the
technology
has
to
change
the
audience
and
shifted
to
this
is
not
the
same
audience.
This
is
not
a
service
anymore.
This
is
a
software
that
you
run,
and
this
is
what's
interesting
so
release
in
what
January
Openshaw
server.
We
started
having
a
lot
of
complaints.
A
Oh
man,
why
is
it
not
working
release
the
original
source
code
and
after
a
while
I
guess
the
people
at
Facebook
enforce
that
to
be
tired
of
hearing
that
complaint,
that
the
code
was
not
the
original
password?
And
that
was
the
answer.
You
don't
be
upset
we're
the
same
way.
So
this
is
why
it
went
through
a
full
rewrite
of
the
code
base
comes
with
bugs.
We
usually
don't
rewrite
code,
but
do
you
want
to
install
that
just
for
you
know,
making
a
little
experiment
in
the
mobile
application?
A
So
under
the
belly
is
being
billed
as
an
extra
J
as
up
it.
Friends
are
no
js'.
Obviously
that
makes
it
very,
very
easy
to
integrate
with
your
current
web
server
sake.
We
also
built
a
command-line
interface.
I
spend
a
lot
of
time
on
it,
because
I
don't
want
to
code
to
manage
my
pass
server.
I
just
want
to
configure
it
and
make
it
work.
The
same
way.
I
was
clicking
buttons
on
past
comm.
I
just
want
to
finish
server
and
have
the
CLI
do
everything
for
me.
A
So
the
CLI
is
as
functional
as
it
can
be.
You
can
configure
pretty
much
everything
you
want
to
configuration
size
or
environment
variables,
it's
built
on
top
of
es6,
so
with
battle
and
all
that
stuff,
because
es6
is
not
available
yet
really
it's
heavily
tested
with
the
coverage
of
over
93%
modern
thousand
tests
and
the
test
matrix
is
over
eight
environments,
so
two
different
versions,
three
versions
of
MongoDB
and
PostgreSQL.
So
it's
really
rock-solid
and
another
constraint.
We
wanted
it
to
be
easy
to
deploy
and
easy
to
run.
A
A
Open
source,
so
whether
it
is
I
did
not
expect
that
traction
10,000
stars
in
just
five
months,
so
that
two
thousand
stars
a
month
you
know
the
co-creation.
It
is
really
really
impressive
to
see
that
many
people
wanting
expressing
an
interest
in
the
project.
We
must
500
full
requests.
This
is
massive.
Yes,
there
was
a
lot
of
verbs
and
there
was
a
need
for
the
spore
request
to
be
merged,
but
this
comes
from
more
than
80
people,
so
the
community
is
growing
and
growing
at
that
refund.
A
Technically
you
started
with
that.
Just
that
force
you
don't
need
to
know
anything
else.
So
as
a
user,
that's
that's
just
great
one
single
point
of
contact
if
it
doesn't
worth
of
just
buzz.
That
is
done
now.
I
mean
that's,
that's
problematic.
Yeah!
Remember
those
people
are
mobile
developers.
What
do
they
know?
How
do
you
pick
your
back-end?
How
do
you
pick
your
infrastructure?
Should
you
go
over
AWS?
Should
you
use
Google
cloud
IBM
just
really
do
bluemix?
Why
not
should
you
do
Erica?
A
What
are
the
limits
of
all
those
platforms
of
all
those
constraints
and
it's
a
bit
complicated
for
us
to
educate
the
current
users,
because
we
are
not
here
to
educate
those
people
we
are
not
here
to
have
them
make
those
choices.
We
are
here
to
provide
an
open
source
project
that
works,
and
then
we
don't
go
into
the
database,
etc,
etcetera.
So
this
is.
This
is
really
complicated
for
everyone.
A
Originally
those
front-end
developers
mobile
web,
they
have
no
backing
skills
and
then,
if
you
just
want
to
take
an
hour
and
skim
through
the
issues,
you'll
see
that
how
how
do
I
set
an
environment
variable.
You
don't
like
those
basic
questions
that
most
of
us
can
answer
are
some
questions
that
are
asked
every
day
over
and
over,
because
the
people
have
no
backend
chaos,
they
never
touch
them
for
most
of
them,
an
infrastructure.
They
don't
know
what
AWS
is
doing.
They
don't
know.
What
is
the
difference
between
a
container
and
energy
initiative
instance?
A
How
does
it
work?
You
know
and
that's
problematic,
because
they
want
they
want
everything
to
work
and,
let's
not
get
into
managing
a
database
like
MongoDB,
looks
easy
from
the
outside,
but
it's
not
that
easy.
Should
I
use
a
replica
set,
how
much
replicas,
how
do
I
make
sure
I
got
availability?
What
do
I
do
when
there
is
a
failover?
How
do
I
manage
all
that?
This
is.
This
is
something
very,
very
complex.
That's
been
added
on
top
of
something
very
simple,
so
the
product
is
different.
It's
not
a
service.
A
So
our
existing
users-
and
we
tend
to
address
you-
know
all
this
crowd
and
all
those
questions
or
existing
users.
What
they
care
about
is
read
the
continuity
of
service.
When
you
get
code
in
your
mobile
app
and
you
build-
and
you
spend
you
know
a
year
to
your
three
years,
growing
your
community
growing
your
product,
you
don't
want
to
rewrite
everything
from
one
day
to
the
other.
If
you
want
to
rewrite
it,
you
want
to
do
it.
You
know
small
step
at
a
time,
so
they
want
continuity's
of
service.
A
They
want
disability,
they
want
to
rely
on
the
open
source
code
as
they've
been
relying
on
the
service
and
that's
really
important
too,
and
for
some
of
them
they
want
to
migrate,
and
we
can't
do
anything
about
that.
The
even
if
the
power
server
project
was
perfectly
rolling
and
super
stable.
Some
people
just
don't
want
to
deal
with
the
back
end,
but
they're
gonna
move
over
to
firebase,
which
made
a
great
job
as
trying
to
ping
poke
people
around.
But
that's
that's
just
fine.
You
know
we
can't
address
everyone.
A
A
Looking
for
a
back-end
I
would
reconsider
pause
first,
because
it's
open
source,
it's
highly
open
it
opinionated.
Yes,
but
this
is
just
object-oriented
database
management,
you
create
your
objects,
but
your
properties
and
boom.
It
gets
safe.
So
that's
pretty
yep,
honest
opinion
and
simple
opinion.
It's
not
like
firebase
when
you
know
you've
got
a
whole
new
concept
of
database
to
wrap
your
head
around
and
the
edgy
case.
The
client
SDKs
our
battle-tested
I've,
been
out
for
more
than
five
years
and
they
are
working.
A
So
you
can
rely
on
the
code
that
is
in
your
client
to
be
stable
and
efficient,
and
the
new
kind
of
users
are
the
software
as
a
service
companies
that
are
starting
to
pop
up
and
what
they
want
is
make
money.
Then
that's,
that's
understandable.
They
want
to
help
people
go
from
path
to
another,
to
their
service,
without
disrupting
the
continuity
and
without
thinking
of
the
backend.
So
we
see
a
few
of
those.
A
A
From
the
ground
up,
we
designed
it
the
people
at
Facebook
designed
it
to
be
extensible.
You
can.
If
you
go
to
your
back-end,
which
is
Google
cloud,
then
you
want
to
use
the
Google
cloud
file
service
to
store
your
files.
You
can
do
it
if
you
want
to
build
a
new
one,
you
can
plug
and
play
different
components
of
the
server,
and
that
is
that
is
a
good
pop,
as
you're
contributed.
Google
contributed
like
AWS
contributed.
A
So
we
see
a
lot
of
interest
from
the
different
cloud
providers
to
contribute
those
extensions
onto
path
server
without
modifying
the
core,
the
core
of
the
system,
so
you
can
easily
swap
parts
of
the
implementation,
something
really
fun.
That
happened
is
the
live
queries,
so
basically
that
a
real-time
deal
see
now
to
listen
to
changes
from
the
database
into
the
client.
If
you
want
to
build
a
more
reactive
application,
instead
of
falling
for
changes,
you
can
listen
to
changes
and
they
are
fired
up
in
your
app.
A
So
this
is
a
feature
that's
been
inside
of
a
since
the
beginning,
and
now
it's
on
Passover
and
it's
working
great
people
are
loving
it
next.
One.
A
native
social
authentication
is
part,
only
add
Twitter
and
Facebook
for
login,
which
is
cool,
but
not
enough.
So
one
of
the
first
things
I
contributed
is
a
really
flexible
way
to
manage
authentication,
so
you
can
bring
in
any
old
token-based
authentication
to
pass
server
very
easily,
with
just
a
few
lines
of
code
and
you're
good
to
go.
I
think
we've
got
eight
github,
Instagram,
LinkedIn,
etc,
etc.
A
Next
big
thing
is
progress.
We
are
working
actively
on
it.
We
are
very,
very
close
to
have
it
working
I'm
really
excited
about
that,
bringing
back
the
schema
constrained
database
by
Postgres
allowing
transactions.
If
you
have
a
player,
if
you
are
building
a
system
that
requires
those
features-
and
this
is
something
that
would
have
never
happened.
If
Foss
was
you
know,
keeping
working
as
it
is
right
now,
so
that
is
great,
we
are
I
would
say,
amount
away
from
shaping
it.
A
It's
already
being
tested
right
now
we
are
very,
very
close
and
then
obviously
cockroach
GB
I
look
forward
to
that
because
it
works
with
the
Postgres
language
and
then
last
feature
that
we
won't
work
on
really
hard.
With
the
event
of
you
know:
micro
services
and
all
that
stuff.
We
want
cloud
code
to
go
to
the
next
level
and
we
want
you
to
run
a
very,
very
small
snippet
of
code
to
modify
the
behavior
of
your
installation
very
easily.