►
From YouTube: Panel: Rust in reality - EuroRust 2022
Description
Rebecca Rumbul (Curator)
Executive Director at Rust Foundation
Rebecca Rumbul will talk with representatives of companies that bet on Rust about their motivations to choose Rust, as well as challenges they are facing and ways in which they benefit from their decisions.
Panelists:
- Chris Konstad, Production Engineer from Meta
- Armin Ronacher, Principal Architect at Sentry
- Luca Palmieri, Senior Software Engineer at Amazon Web Services (AWS)
- Florian Gilcher, Managing Director at Ferrous Systems
A
Thank
you,
hi
everyone,
I'm
Beck,
Rumble,
I'm,
executive
director
and
CEO
of
the
Russ
Foundation
and
yeah.
We
are
hopefully
going
to
have
just
a
lovely
chat
about
rust
in
reality,
using
it
in
business
and
where
it's
going
in
the
future.
Can
I
just
ask
my
panelists
to
introduce
themselves
quickly.
B
C
D
C
A
Awesome,
thank
you
all
for
joining
me
today.
So
I
think
Florian.
If
we
can
kick
off
with
you,
you
know,
rust
is,
is
an
amazing
language
and
you
have
made
you
have
made
a
business
out
of
it.
Can
you
tell
us
a
bit
about
that.
D
Sure
so
I
think
one
of
my
favorite
papers
about
programming
languages
is,
if
you
want
to
make
a
business
out
of
it.
No
one's
adapts
programming
languages
for
features
it's
the
hard
truth,
but
because
they
have
a
need
and
the
one
thing
that
I
found
that
I
see
is
the
reason
as
a
success
and
the
business
success
investors,
because
it
comes
at
a
point
where
a
lot
of
companies
are
rethinking
the
tooling
stack,
everything
becomes
faster.
Everything
becomes
connected.
The
area,
we're
anchoring
and
currently
in
automotive
companies
need
to
think
about.
D
We
can't
do
the
five-year
development
cycle
anymore,
because
suddenly
our
cars
are
on
connected
to
the
internet.
We
need
to
crash
patch,
like
all
of
these
kind
of
things,
and
rust
makes
an
offer
for
a
both
convenient
to
use
programming
languages
with
stability
guarantees
that
he
would
usually
find
in
the
functional
space
and
that's
the
direction
that
the
view
is
going.
We
need
new
tools
and
then
tools
are
selected
rather
than
constantly
wanting
to
replace
tools
and
just
do
better
on
fun
things.
D
A
You
kind
of
you
were
ahead
of
that
very
steep
learning
curve
we
all
hear
about,
but
some
businesses
you
know,
are
having
to
make
the
decision.
Are
we
going
to
invest
in
rest?
Are
we
going
to
use
it?
How
are
we
going
to
do
that?
Luca?
You
gave
a
really
great
presentation.
A
couple
of
weeks
ago
that
I
was
I
was
part
of
talking
about
how
the
organization
you
were
working
for
at
the
time
kind
of
started
off
that
process.
A
How
did
that
go?
Yeah.
E
Sure
so,
for
contacts,
my
previous
company
was
100
developers,
150
developers,
originally
a
c-sharp
shop,
so
a
lot
of
investment
into
the
c-sharp
ecosystem
and
we
started
adopting
grass
a
couple
of
years
back
so
we're
looking
at
2020
end
of
2019
and
at
the
end
of
2022
that
was
ramped
up
to
40,
50,
rust
developers.
E
So
almost
half
of
the
development
Force,
oh
Jesus,
the
choice
for
us
as
Florian
was
saying
like
you,
come
at
the
language
with
a
need
and
some
of
those
needs
are
Technical,
and
some
of
those
needs
are
more
social.
If
you
want,
we
came
across
from
a
technical
perspective
because
we
needed
better
domain
modeling,
so
we
needed
algebraic
types
we
need
to
handle
null
ability
we
needed
to
handle
the
possibility
of
failures
in
a
way
that
was
more
manageable
than
exceptions.
E
E
It
was
not
easy
to
get
started
like
in
the
beginning
of
2020.
Our
use
case
was
began
development,
so
web
I
think
a
way
they've
been
stabilized
in
November
2019,
so
it
was
very
fresh.
The
ecosystem
was
going
through
multiple.
You
know
waves
of
churn
on
various
kind,
but
it
was
enough
to
prove
it
on
a
business
case.
In
our
case,
a
banking
ledger.
So
a
fairly
you
know
safety
critical,
but
not
in
the
definition
certificate
you
would
use,
but
safety
in
the
sense
of
you
know.
E
If
we
make
some
mistake,
this
can
cost
us
several
millions
of
pounds,
depending
on
the
common
mistake
that
we
make
and
so
correctness
there
was
valued
enough
for
us
to
make
a
bet
to
try
something
new
and
then
we
found
out
it
actually
wasn't
that
much
more
difficult
to
write
than
what
we
were
using
before
and
so
from
there
kind
of
spread
through
simply
developers
in
the
company,
knowing
that
it
was
a
golden
pathway
which
was
supported
and
choosing
to
use
it
and
kind
of
iterating
over
that.
So
that
was
a
little
bit
the
process.
A
Thanks:
Luca
and
Armin
how?
How
was
the
process
at
century.io.
C
Yeah
I
think
for
us
it
was
mostly
so
when
we
started
we
were
a
python
shop,
so
we
did
most
of
the.
C
In
fact,
the
entire
implementation
of
all
of
Sentry
was
all
in
Python
and
I
was
playing
around
with
rust
already
for
a
while,
just
as
a
as
an
interesting
hobby
project,
because
I
find
it
interesting
and
at
one
point
we
actually
figured
out
that
one
of
the
sort
of
hobby
implementations
of
the
source
map
processing
that
I
did
in
in
Rust
was
actually
outperforming
the
production
python
one
and
because
rust
was
already
at
the
time
very
trivial,
to
use
as
an
extension
to
python.
C
So
you
could
call
into
rust
code
very
trivially
from
Python,
and
we
sort
of
had
an
obvious
way
to
put
it
in
and
the
results
were
very
promising
from
a
performance
perspective
and
then,
at
the
same
time
we
also
had
a
small
little
utility
a
command
line
tool
and
if
you've
ever
used
python,
the
distribution
of
command
line
utilities
to
customers
with
python
is
very
tricky,
and
so
rust
was
It
was
supposed
to
do
the
same
thing
as
the
server
is
already
doing.
C
So
we
we
had
some
use
to
reuse
the
code
between
the
command
and
client
and
the
server,
and
so
we
sort
of
started
slipping
into
this
and
over
the
next
couple
of
years.
C
I
think
we
had
just
more
and
more
Rascals
could
appear,
naturally,
in
the
code
base
and
I
think
we
have
probably
like
almost
seven
years
of
rust
at
Century
in
one
form
one
another
and
because
we
already
had
started
building
some
of
our
processing
pipeline
in
in
Rust
over
the
years
just
started
accumulating
more
and
more,
and
the
initial
benefits
was
almost
entirely
performance.
C
But
one
of
the
core,
critical
components
of
Sentry
is
that
we
process
native
stack
traces.
So
when
you
have
a
crash
coming
from
an
iOS
device
or
an
ndk
device,
or
it's
just
plain,
C
plus
you
need
to
process
the
debug
information
files
and
a
lot
of
the
historic
Tools
in
this
space
are
actually
intended
to
be
run
on
trusted
data.
So
you
would
have
your
own
debugger
running
locally
and
you
kind
of
trust,
your
own
compiler,
to
not
create
nonsense.
C
But
if
you
process
debug
symbols
from
customers,
you
kind
of
are
always
exposed
to
this
inherent
risk
that
this
is
doing
something
it's
not
supposed
to
do,
and
the
rust
ecosystem
started.
Building
more
and
more
of
these
tools
around
to
actually
process
debug
information
files
and-
and
we
were
able
to
contribute
back
into
this-
and
we
are
very
happy
now
that
I
think
as
I'm
pretty
sure.
As
of
today,
with
the
exception
of
a
tiny,
tiny
component,
all
of
the
simple
and
and
debug
processing
is
actually
fully
running
on
Rust.
C
So
we
replaced
the
entirety
of
the
C
plus
plus
code
base
that
we
used
for
this
before
and
I
think
it
worked
out
really
well
in
terms
of
how
it
actually
functions
in
a
company.
It's
very
fun
to
work
on
I
would
say-
and
it's
also
easy
for
us
comparatively
to
make
people
excited
to
work
on
this
code
base.
That
wasn't
the
case
with
C
plus
plus
before
foreign.
A
Matter
is
a
behemoth.
It's
a
huge
huge
huge
company
was
the
Journey
of
getting
arrested
and
starting
to
use
it
similar,
or
was
that
a
different
proposition?
It.
B
Was
it
was
a
a
long-running
process
that
took
many
many
years
so
first,
the
source
code
management
team,
looked
at
adopting
rust
in
about
2016.
engineers
at
meta,
have
a
lot
of
flexibility
in
choosing
the
tools
that
they
use
to
get
the
job
done,
including
choosing
the
language.
That's
used.
That
being
said,
there's
a
few
officially
blessed
languages,
and
if
you
stick
down
that
route,
it
is
easier
to
get
work
done
and
it
is
easier
to
like
you,
don't
get
pushback
from
people
of
saying
like.
B
Why
are
you
using
you
know
this
random
language
that
just
came
out
but
anyways,
so
we're
looking
at
doing
some
work
in
Mercurial,
there's
a
bunch
of
python
engineers
and
moving
from
python
to
something
more
performant,
and
so
the
default
option,
for
that
would
be
C,
plus
plus
and
one
of
the
engineers
at
the
time
didn't
like
the
idea
of
trying
to
onboard
bunch
of
python
first
Engineers
into
the
very
complex
C
plus
language
and
everything
everything
in
meta
is
like
networked
and
concurrent
and
async
and
all
that
stuff.
B
So
it's
it's
a
tall
order
to
go
from
nothing
to
that
in
C,
plus
plus,
and
this
is
before
Co
routines
right,
and
so
they
started
looking
around
at
rust
and
started
importing
rust
and
that
worked
out
really
well
so
later
on,
like
other
teams,
kind
of
started
piggybacking
off
their
work,
so
they
had
to
like
integrate
the
rust
compiler
with
our
build
tooling
and
all
that
kind
of
stuff.
B
So
once
that
work
was
done,
other
teams
kind
of
started
hopping
on
and
it
was
kind
of
like
watching
a
forest
fire.
You
get
all
these
little
Sparks,
these
little
fires
all
over
the
place
and
then
eventually,
like
you
know,
you
can't
it
it's
hard
to
find
a
team.
That's
not
interested
or
or
already
starting
to
use
rust.
It's
absolutely
everywhere.
B
B
So
when
you,
when
you
join
the
company,
you
go
through
this,
like
technical,
all
these
technical
classes
and
then
you
go
sit
around
with
different
teams
and
try
to
find
which
team
you
want
to
join
and
that's
how
you
decide
which
team
you
go
on
and
while
you're
sitting
there
doing
your
classes,
you
can
get
tasks
from
real
teams
and
do
real
production
work
that
are
like
bite-sized
little
bits
and
that's
one
of
the
ways
that
we
kind
of
help
onboard
people
onto
rust
is
we
have
a
lot
of
these
little
rest,
specific
boot
camp
tasks
that
people
can
try
out
and
honestly
like
for
us,
for
my
team,
like
using
rust,
has
been
a
a
benefit
in
hiring
people
onto
the
team.
B
You
know
new
people
come
into
the
company
and
come
out
of
boot
camp
and
they
say
I
want
a
team
that
does
rest,
and
so
that's
been.
That's
been
super
helpful
too.
There's
a
bunch
of
latent
demand
out
there.
A
Is
yeah,
that's
really
exciting,
so
a
few
of
you
have
actually
spoken
about
hiring.
You
know.
We've
we've
seen
recently
the
this
year
rust
was
like
the
Most
Wanted
language,
not
just
the
most
loved
in
the
snack
overflow
survey.
A
This
is
kind
of
speaking
to
a
demand
for,
for
us
developers.
Is
that
something
that
you're
finding
easy
to
to
recruit
or
is
it
difficult?
Are
there
gaps
in
the
market?
I
don't
know.
Do
you
want
to
go
first,
Chris,
okay,.
B
B
Some
come
with
experience,
some
don't
so
it's
been
like
easier
for
teams
that
have
this
rust
exposure
to
like
hire
out
of
boot
camp.
So
for
us
it's
it's
from
my
experience.
It's
been
a
pure
win.
B
You
know
we'll
have
new
people
come
into
the
come
into
the
company
and
be
like
I
want
rust.
Specifically,
don't
don't
show
me
anything
that
doesn't
do
that,
so
that's
a
competitive
Advantage
because
it
can
be
kind
of
competitive
trying
to
get.
You
know
the
top
Engineers
out
of
boot
camp
from
other
teams
that
also
want
them.
D
If
I
can
tune
the
question
a
little,
we
have
no
problem
hiring
rust
Engineers
with
our
profile,
but
I.
Think
there's
there's
an
interesting
conversation
going
on
about
this,
because
we're
both
getting
questions
from
clients
and
potential
clients
about.
Can
we
find
Rust
engineers
and
I
think
this
still
a
huge
pool
of
people
that
would
actually
like
to
switch
a
job
to
one
workout.
They
can
do
more
rust.
That's
the
one
side
and
the
other
thing
is
from
when
I
talk
to
programmers.
D
They
say
like
there's,
not
many
job
postings
out
there
for
us
and
I.
Think
one
of
one
of
the
things
we're
seeing
there.
We
are
currently
retraining
cohorts
of
teams
into
rust
and
I.
Think
that's
a
if
you,
if
you
look
at
it
closely
a
very
reasonable
process.
Companies
don't
adopt
new
Languages
by
hiring
new
developers
with
new
languages
on
a
new
team.
That's
at
least
one
new
too
much
so
the
usual
thing
they
do.
Is
they
retrain
an
internal
team
that
they
trust
to
execute
something?
Try
a
small
project.
D
Try
two
small
projects
grow
the
project
and
then
at
some
point,
they're
starting
to
hire
new
rust
people
into
the
team,
because
the
most
important
bit
at
the
beginning
is
that
you
have
developers
that
have
the
context
that
know
the
requirements
that
know
how
they
have
the
company
works
and
that
trust
the
management
to
also
say
hey.
By
the
way
this
project
is
going
well
or
it
is
going
not
as
well
as
you'd
expect
it
for
also
for
a
new
developer.
D
That
probably
is
a
bad
position
to
be
in
saying
hey
by
the
way
you
picked
this
language
and
I
now
have
to
tell
you
that
thing
isn't
going
as
well.
So
there's
all
these
kinds
of
Dynamics
going
on,
and
it
has
been
our
experience
in
the
last
one
to
two
years
that
we're
working
with
more
large
Enterprises
retraining,
whole
teams
in
the
hundreds
into
rust,.
E
I
mean
it's
absolutely
true:
we
did
the
same
thing
like
you,
didn't
throw
a
new
project
in
a
new
language
to
new
people.
Like
you
usually
start
with
some
internal
people,
you
get
them
to
do
it
also
because
it's
a
matter
of
trust
right
like
if
you
move
existing
team
members
into
using
a
new
language
and
the
new
language
doesn't
work
out,
they're
still
part
of
your
company.
E
If
you
are
a
bunch
of
people,
and
you
say
we're
gonna
do
Rust
what
happens
if
the
project
doesn't
work
out,
and
so
you
don't
want
to
be
in
that
situation,
nor
you
as
a
company
nor
them
they
join.
You
know
believing
that
we're
going
to
use
a
certain
technology
to
find
out
two
months
afterwards
that
actually
like
was
very
fun,
but
it
doesn't
fit
our
requirements.
C
So
for
us,
it's
I
think
so.
First
of
all,
I
we
have
I
I
generally
feel
like
more
people
that
want
to
do
Rust,
and
we
actually
have
rust
code
at
Sentry
because
we
have
a
lot
of
JavaScript.
Still,
we
have
a
lot
of
python.
We
have
gazillions
of
sdks
and
all
kinds
of
languages,
so
I
feel
like
we
are
now
at
this
point
where
there
is
actually
more
sort
of
inbound
demand
for
rust
position
than
we
actually
have
available,
and
we
also
went
on
the
same
path.
C
It's
originally
when
we
started
the
whole
thing
that
we're
not
really
rust
programmers.
So
every
single
person
that
joined
up
to
a
certain
point,
probably
learned
for
us
at
the
company,
so
there's
internal
retraining
for
sure
is,
is
how
we
went
down
this
and
yeah
generally.
It
feels
like
this,
this
a
growing
demand
of
people
to
actually
find
the
place
where
they
can
do
this
yeah
I
think
it
worked
well
for
us
it
was.
C
A
So
for
individuals
who
are
looking
to
to
get
a
job
in
Ruskin,
one
of
your
companies
may
be
how
what
what
do
you?
What's
your
advice
for
those
people?
If
you
know
it's,
it's
more
a
case
of
at
the
moment,
you're
training
people
internally
to
move
into
things.
Is
it
should
they,
you
know,
do
a
do
personal
projects,
there's
things
that
they
can
do
to
bulk
out
their
CVS
to
be
attractive.
E
That's
right
to
answer
that
I
think
it
really
depends.
Obviously
different
companies
of
different
hiding
cultures,
like
some
people
want
to
experience
with
the
language.
Other
companies
don't
really
care
much.
My
personal
suggestion
is
once
again
not
to
go
for
too
many
news
like
if
you
are
applying
for
a
position
and
that
position
is
in
a
certain
domain,
it
requires
rust
for
you
to
be.
You
know
a
candidate
that
is
relevant
to
that
position
either.
E
You
need
to
know
the
language
you
need
to
know
the
domain,
if
you
don't
know,
either
becomes
very
difficult
to
step
into
that
as
the
candidate
position,
especially
for
companies
that
are
adopting
grass
at
least
that
was
our
process.
In
my
previous
place,
we
try
to
balance
out.
You
know,
domain
expert
and
language
experts.
A
C
In
so
I,
ironically,
get
a
lot
of
questions
about
how
to
get
rust
into
a
company
already,
so
one
of
the
I
think
sort
of
to
go
with
this.
A
little
bit
is
like
you,
probably
already
have
a
problem
where
Russ
can
fit
in
somewhat
so
I
come
up
from
the
python
community
performance
issues
is
a
classical
reason
for
people
to
look
for
Solutions
and
the
rust
ecosystem
is
actually
pretty
strong
in
certain
areas.
C
For
us,
obviously,
debug
information
file
processing
was
a
huge
Advantage,
but
there's
there's
a
lot
of
utility
in
the
Creator
ecosystem,
and
if
you
really
are
in
this
mood
that
you
want
to
try
to
see,
if
you
can
introduce
rust
in
a
company,
maybe
you
actually
have
a
performance
issue
where
you
can
relatively
trivially
work
with
the
existing
ecosystem
to
see.
If,
if
you
can
demonstrate
that
this
is
something
that
actually
can
benefit,
it's
incredibly
hard
to
build
a
new
service
in
a
new
language
with
no
prior
experience.
D
Yeah,
so
on
the
question:
if
you're
individually,
searching
for
a
job,
I
can
highly
recommend
two
things:
HeadHunters
on
bad
and
they're
specialized
was
to
Head
Hunters
they.
If
you
want
to
switch
to
rust
career
contact,
one
of
them.
The
second
is
there's
a
danger
of
comparison.
Probably
everyone
here
on
stage
is
a
wrong
comparison
to
compare
yourself
to
for
your
skill
in
Rust.
D
Rust
is
still
a
young
young
language,
and
that
brings
the
advantage
of
It's
relatively
easy
to
become
an
expert
in
any
one
of
those
things
that
Armin
mentioned
Implement
a
protocol
or
something
really
well
or
something
like
that.
It
is
currently
doable
with
a
reasonable
amount
of
time
to
become
an
expert
in
a
domain
where,
in
JavaScript,
Java
or
any
one
of
the
traditional,
very
broad
languages
isn't
possible
and
to
point
the
light
at
you
and
I.
Think
that
that's
one
of
the
reasons
why
don't
don't
compare
to
anyone
else?
D
Look
for
the
niche
where
there's
nothing
yet
and
no
one's
doing
it
in
Rust,
and
it
can
be
a
very
easy
one
to
two
blog
posts.
It's
still
the
point
where,
with
one
to
two
blog
posts
you
can
set
in
a
niche
just
to
make
it
clear.
First
system
needed
three
blog
posts
to
be
the
experts
in
safety,
critical.
A
B
I
mean
I've
met
as
far
as
getting
a
rust
job.
In
particular,
you
just
kind
of
get
into
the
get
into
you
know
through
the
interview
process,
so
any
programming
language
in
the
world
will
get
you
through
that
and
then
landing
on
a
team
that
has
rust
like
you're
gonna,
probably
be.
You
know,
jockeying
for
position
against
people
that
maybe
have
rust
experience,
and
then
you
know
what,
if
you
don't
like
rust,
you
should
probably
try
it
out.
B
First,
just
to
make
sure
you
actually
do
like
it,
because
you
know
there's
a
bit
of
a
learning
curve
and
it's
totally
worth
it.
It's
100
worth
of
learning
curve.
I
would
like
do
it
again
in
a
heartbeat,
but
it
was
you
know
it's
something
that
you
know
might
take
a
time
or
two
to
get
past
and
then
once
you're
past
it
it's
it's
great,
but
yeah,
but
you
know
you
might
want
to
do
that
before
trying
to
land
on
the
team
that
you
want
to
land
on.
B
So
it's
probably
good
to
have
some.
You
know
experience
ahead
of
time.
A
So
yeah,
that's
that's!
That's
really
encouraging
in
terms
of
you
know
that
you're
going
to
run
out
of
internal
Engineers
to
retrain
and
rest,
though
at
some
point
and
the
learning
curve
is
something
that
I
think
everyone
here
has
heard
over
and
over
and
over
again
is
there
a
point
at
which
we
should
be?
You
know
more
actively
trying
to
to
educate
people
in
rest
that
they
don't
have
to
go
through
three
or
four
languages.
First,
to
get
to
rest,
like.
D
So,
first
of
all,
I've
been
training
with
us
since
2015,
and
the
first
thing
is,
the
learning
curve
has
flattened
both
in
in
education
experience
in
code
out
there
there's
much
more
code,
examples
that
are
really
good
and
rather
easy
to
use.
D
It's
the
interesting
thing
about
programming
languages
for
the
first
couple
of
years.
The
program
language
Community
itself
needs
to
figure
out
how
it
works
and
how
it's
easy
and
that's
been
the
phase
from
2015
to
2018.
Something
like
that.
D
The
other
thing
is:
there's
no
pre-requirement
to
learn
any
languages
before
rust.
It
is
actually
a
problem
that
previous
knowledge
in
programming
languages
interferes
with
learning
new
languages.
D
That's
the
thing
that
trainers
do
teach
you
the
subject
without
interference
with
other
subjects,
and
that's
so
there's.
No.
You
need
to
be
this
tall
to
learn
rust.
That
needs
to
be
understood
both
from
companies
and
and
also
by
individuals.
The
hardest
groups
to
train
are
people
who
believe
that
they've
done.
They
are
done
with
systems
programming
and
they
know
it
all,
because
you
need
to
tell
them
that.
Yes,
they
know
it
all,
but
they
can
do
it
on
another
language.
D
As
well
and
the
last
bit
is,
it
is
often
overstated
how
much
the
programming
language
matters
in
our
trade
there's
so
much
more
knowledge,
how
the
hardware
Works,
how
protocols
in
the
space
Works
how
the
environment,
Works
codes,
names,
protocols,
all
of
these
kind
of
things
and
Russ
doesn't
do
away
with
all
of
that.
C
C
I
have
done
python
courses
in
the
past,
where
I
was
trying
to
teach
it
to
people
and
you.
You
basically
spent
the
first
50
of
your
like
getting
started
because,
like
someone
has
a
Windows
receiver,
they
have
a
different
version
of
python
that
doesn't
do
anything.
Then
there's
Anaconda,
there's
like
they
somehow
broke
their
local
paper
or
whatever
and
I've
never
had
this
problem.
That
I
showed
someone
rust
and
it
wouldn't
work
immediately.
C
But
the
frustration
comes
if
you're
already
used
to
something
and
then
it
doesn't
work
anymore
like
particularly
if
you
come
from
a
language
like
C
or
C,
plus
plus,
where
you
grow
up
with
this
idea
that
you
can
do
anything
for
as
long
as
you
can
touch
the
memory.
It
just
doesn't
work
like
that
in
Rust.
It
will
prevent
you
and
if
you're,
if
you're,
really
trying
to
sort
of
get
this
pattern
that
you're
already
used
to
working
in
Rust.
You
just
keep
running
into
walls.
C
A
couple
of
years
ago,
I
wrote
the
blog
post
about
like
the
best
way
to
program.
Rust
is
to
learn
which
things
you
can't
do,
and
this
is
still
I
still
get
email
responses
to
that
particular
blog
post,
because
for
a
lot
of
people,
this
sort
of
is
their
experience.
If
they
come
from
not
languages
like
they're
trying
to
do
this
and
then
only
after
they
learn
somewhere
in
the
internet,
that
you
can
never
do
that
they
stop
trying.
E
I
think
another
data
point
there
is
that
there
are
learning
curves
I.
Think
for
us
like
it's
a
lot
of
people.
Get
scared
by
the
surface
of
the
language
like
past
is
a
big
language.
There's
a
lot
of
things
in
there,
but
in
reality,
depending
on
what
you're
trying
to
do,
you
only
need
a
small
subset
of
those
things
like
there
are
certain
parts
of
the
language
which
have
never
touched
because,
for
example,
I
never
did
see
ffi
like
I
didn't
need
it.
I've
been
doing
grass
professionally
for
three
years,
I.
E
Consider
myself
decent
and
I
leave
happily
like
a
lot
of
people,
go
to
Russ,
saying
I
need
to
learn
all
those
things
and
there's
a
lot
to
learn,
but
it's
actually
not
necessary
and
I.
Think
we
are
seeing
the
different
subset
of
the
ecosystems
as
Florian
was
saying,
are
evolving.
You
know
their
own
learning,
curves
they're,
evolving,
their
own
resources
and
their
own
ramps
to
doing
web
to
doing
games
to
doing
embedded
whatever.
E
That
is,
and
it's
those
resources
you
know,
accumulate
and
those
ecosystems
stabilize
good
examples
are
gonna,
go
out
and
it's
gonna
get
increasingly
easier
and
easier
and
easier,
and
you
know
Frameworks
are
going
to
get
more
complete
and
so
on
and
so
forth,
which
goes
back
to
what
Florian
was
saying.
Like
there's
a
lot
of
space
to
do
things.
One
way
is
looking
at
things
that
are
underserved.
Another
way
is
looking
at
projects
from
you
know
five
years
ago
that
are
unmentained
and
just
refreshing
them.
B
Yeah
kind
of
hopping
onto
that
I've
been
writing
rust
code
professionally,
since
2019
and
I
can
count
on
one
hand,
the
number
of
times
I've
had
to
use
like
explicit
lifetimes,
like
clone,
carries
you
really
far,
and
you
probably
don't
need
to
worry
about
it
and
like
I
love
that
when
I'm
writing
rust,
that
clone
sticks
out.
B
So
that
way,
you
don't
accidentally
copy
too
many
times,
but
it
kills
me
when
I
watch
people
trying
to
learn
rust
for
the
first
time
and
they're
getting
caught
up
on
lifetimes
and
trying
to
avoid
that
one
clone
that
happens
in
Maine
once
and
you're
like
it
doesn't
matter
just
clone
it,
it's
fine
it'll
be
okay,
so
you
know,
like
you
know,
don't
don't
sweat
that
stuff
right,
the
documentation
is
amazing.
The
book
is
great.
B
The
amount
of
blog
posts
that
have
already
been
written
and
will
be
written
in
the
future
is
is
amazing,
and
you
know
the
compiler
errors.
I
mean
I
come
from
C,
plus
plus
yeah
I
mean
anyone
from
C
plus
will
tell
you,
you
get
one
character
wrong,
and
your
compiler
error
is
just
this
inscrutable,
40
page
output
that
you
can't
even
find
the
line
that
caused
the
error,
even
if
you
wanted
to
like
it's
a
skill
to
develop
how
to
like
debug
those
issues
and
like
the
rust
compiler
error.
B
E
A
Could
just
go
home?
There's
got
to
be
challenges,
though.
What?
What
do
you
see
is
kind
of
the
biggest
challenges
in
terms
of
you
know,
increasing
use,
advice,
breast
adoption,
I,
don't
know
who
wants
their
kids?
You
want
to
jump
in
first
yeah.
E
Sure,
okay,
let's
go
in
I,
have
a
list
so
as
foreign
was
saying,
the
learning
curve
flattened,
but
in
reality,
when
you
wander
in
some
Corners,
then
it
gets.
You
know
very
curly,
so
it
can
get
recursive
asynchronous
codes
can
get
quite
nasty
when
it
comes,
for
example,
to
compiler
errors,
and
you
know
getting
to
do
certain
things.
It's
very
easy.
E
You
know
to
start
from
something
that
looks
innocent
and
you
kind
of
step
in
the
wrong
way
and
then
you're
off
a
cliff
and
you're
like
I'm
gonna
go
rescue
me.
So
that's
that
that's
challenges
in
the
ecosystem.
E
Obviously,
when
you
write
software
commercially
like
the
ideal
way,
is
you
can
add
it
once
you
can
get
updates
forever
and
it's
gonna
run
without
any
problems.
You
know,
that's
the
dream
in
reality,
being
a
young
ecosystem
does
a
lot
of
churn
in
certain
parts
of
the
ecosystem
and
what
that
means
is
that
you
require
more
maintenance
that
perhaps
you
would
require
if
you
were
using.
You
know
aspnet
core,
which
modern
life
doesn't
change
much
in
the
last
few
years.
E
Other
challenges,
I
think,
will
just
be
getting
more
companies
to
use
it,
and
that
requires
getting
more
experienced
for
us
developers
on
the
market,
but
requires
also
a
general.
You
know
getting
Consulting
companies
which
are
able
to
do,
for
example,
self
augmentation.
That
was
a
thing
when
we
were
using
it
a
true
layer
we
were
like.
You
know.
We
need
a
few
extra
tabs
because
we
need
to
get
this
project
done
on
these
dates
and
really
there
wasn't
much
on
the
market.
E
If
you
wanted
to
get
someone
to
come
at
your
company
and
help
you
build
that
project
unless
you
wanted
to
train
them,
but
at
the
point
where
you're
bringing
them-
and
you
know
that
kind
of
doesn't
work
out.
So
those
are
some
of
the
challenges
and
obviously
scaling
the
project.
I
think
is
going
to
be
difficult
when
the
community
size
increases
and
the
number
of
contributors
increases
and
managing
you
know,
keeping
the
language
learnable
while
trying
to
satisfy
more
use
cases.
Probably
there
are
better
people
than
me
to
talk
about
the
project.
D
Yeah
so,
from
my
perspective,
what's
currently
needed,
one
of
the
situations
I'm
very
often
in
is
obviously
I
talk
to
clients
to
develop
a
set
of
clients
and
about
a
lot
of
things
what
the
intentions
are,
but
the
problem
is
that
message
doesn't
just
stop
there:
everyone's
under
NDA
and
sometimes
even
at
this
conference.
You're
in
conversations
with
three
people
are
at
a
table
and
are
talking
vaguely
about
areas
and
places
where
people
could
use
maybe
rust
on.
Maybe
that
platform
sometimes
like
absurd
theater
and
the
solution
for
that
is
simple.
D
If
you're
using
rust
talk
about
it,
if
you
have
a
manager
that
does
not
know
whether
someone
else
is
using
rest,
try
to
find
someone
who's
willing
to
talk
about
it,
because
all
of
these
companies
and
organizations
are
in
a
situation
where
they're
looking
left
and
right
and
they're
thinking.
There
could
be
something
in
the
room
but
they're
not
quite
sure
yet,
and
that's
a
huge
problem
that
we
currently
have
in
our
industry.
D
Rust
is
also
kind
of
like
a
silent
language,
I'm
country,
from
the
Ruby
space
and
at
Ruby
there's
the
joke
that
even
companies
that
don't
say
they
use
Ruby
somewhere,
have
a
huge
rookie
stack
somewhere
in
the
Ops
Department,
something
like
chef
or
puppet,
but
they're,
not
using
Ruby
and
we're
in
a
similar
situation.
You
have
a
lot
of
companies
that
are
effectively
using
rust
somewhere.
You
won't
find
it
on
their
page
or
something
else.
So
it's
important
that
we
have
spaces
this
conference,
maybe
more
specialized
conferences.
D
Events
meetups,
where
people
can
actively
say
hey
by
the
way
I'm
from
company
ABC
we're
having
a
rest
Department
there,
because
that
builds
Trust
in
the
gold
standard
would
be
blog
about
it,
write
about
it
with
whomever
and
have
that
conversation
more
out
in
the
open,
because
I
can
say
a
lot
on
this
stage
on
where
rust
is
being
used.
You
can
trust
me
or
not.
That's
to
be
honest,
not
a
position.
D
C
Some
I
think
coming
from
a
little
bit
of
a
different
angle,
which
is
that
I
don't
actually
care
if
people
use
it
commercially,
but
I
do
know
that
people
do
do
and
one
of
the
I
see
two
challenges
at
the
moment.
A
little
bit
in
this
one
is
that
we
have
a
very
visible
public
ecosystem
of
Open
Source
software,
and
that
is
running
One
Direction,
and
then
you
have
companies
that
often
are
constrained
in
some
form
or
another,
and
they
are
not
as
visible
as
the
open
source
projects
are
and
I
think.
C
This
is
a
little
bit
of
a
problem
at
the
moment
because,
for
instance,
we
are
right
now
sort
of
I
think
in
the
sixth
month
of
secret
sort
of
reshifting
of
the
rust
ecosystem,
where
a
lot
of
projects
are
trying
to
move
up,
the
minimum,
supported
rest
version
and
that's
very
fine
in
the
open
source
ecosystem,
but
there's
a
very
well
hidden
commercial
base
that
can't
move
the
compiler
as
often
as
the
rest
of
the
ecosystem
does
and
I
think.
This
is
a
little
bit
of
a
problem
that
we
don't
really
talk
about
much.
C
That
sort
of
there's
no
rule
there
was
sort
of
a
hidden
rule.
The
last
couple
of
years
that
you,
a
lot
of
projects,
are
very
conservative.
All
this
time,
everybody's
moving
up
I
feel
this
is
something
that
probably
needs
to
be
discussed
in
one
form
or
another,
and
the
other
thing
that
I
think
is
still
a
problem
for
commercial
use
of
rust
is
actually
compile.
C
Time
still
are
not
ideal,
or
at
the
very
least,
the
it's
still
super
easy
to
build
a
massive
crate
with
a
lot
of
dependencies
and
everything
else,
and
quite
a
few
people
go
down
this
path
of
building
this
massive
crate
and
then
have
the
problem
splitting
up
later.
The
support
for
split
crates
is
still
not
ideal,
so
I
think
as
projects
are
growing
and
getting
larger
and
larger
and
rust.
This
is
probably
going
to
help
us
a
little
bit.
B
I
mean
you
know:
definitely
all
those
challenges,
like
the
churn
and
the
environments
compile
times
for
sure
are
a
huge
problem
and
also
like
being
able
to
integrate
with
other
languages.
B
Is
super
important
right
like
if,
if
you're
like
a
larger
company-
and
you
want
to
adopt
rust,
you're,
not
adopting
it
on
these
isolated,
Greenfield
projects
or
you'd
like
to,
but
you
don't
always
have
a
bunch
of
those
and
you'll
have
to
integrate
with
existing
C
plus
libraries,
you'll
have
to
become
a
python
library,
or
you
know,
integrate
with
these
other
languages,
and
so
having
you
know,
solid
support
for
that
is,
is
pretty
important.
You
know,
I
mean
it's.
B
It's
really
easy
to
write
like
a
python,
a
rust
library
that
python
can
call
into,
for
instance,
and
that's
been
absolutely
huge
for
adoption
within
meta.
Very.
E
Okay,
that's
a
tough
question
like
General.
My
suggestion
would
be
to
start
small
and
start
off
the
critical
path
like
play
with
the
language.
Do
some
clis
do
some
build
tooling?
Do
something
that
you
know
it's
not
gonna
break
your
business.
If
you
go
sour
and
then,
if
you
like
it
build
on
top
from
there,
if
you
don't
like
it,
just
stop
and
try
again
in
a
couple
of
years
as
simple
as
that.
D
So,
from
my
perspective,
it's
it's
incredibly
important
to
be
honest
about.
Do
we
actually
have
a
problem
that
rust
solves
you've
seen
that
fail,
but
it's
also
there's
a
lot
of
things
that
was
solved
at
businesses
and
if
you
have
the
hunch
that
just
could
be
a
solution,
the
moment
to
look
into
it
and
to
start
with
it
is
now.
C
Yeah
I
think
I
tend
to
agree
with
this.
One
other
thing
that
if
you
have
a
company
that
wants
to
play
around
with
rust,
that
really
doesn't
have
a
use
case
is
it's.
It's
actually
quite
useful
to
have
this
sort
of
code
Champs
or
like
blessed
events
where
people
can
toy
around
those
different
things,
and
sometimes
out
of
that
comes
something
interesting,
and
particularly,
if
you
look
at
the
Advent
of
code,
the
number
of
people
that
actually
choose
to
do
it
in
Rust
is
pretty
impressive
and
I've.
B
It
can
also
be
helpful
to
find
other
people
in
your
company
that
are
interested
in
Rust,
there's,
probably
more
of
them
than
you
know,
and
you're,
probably
all
interested
in
working
on
the
same
stuff
together.
A
Amazing
Luca
Florian
Armin
Chris.
Thank
you
so
much
for
coming
on
the
panel
I'm
sure
everyone
else
really
appreciated
it
so
yeah.
Thank
you
very
much.
Thank
you.