►
From YouTube: Mapping a pandemic and Deno - Jozi.js October
Description
How to map a pandemic
Richard Young (@richardyoung00)
Being the most populated province and a major economic hub, Gauteng has seen roughly a third of all confirmed COVID-19 cases in South Africa. To manage the spread of cases the Gauteng Provincial Government partnered with IBM Research, Wits University and the GCRO to build a web based analytics platform to help answer questions such as where the current hot spots of infection are, where the virus is predicted to spread, and which communities are the most at risk.
The platform has been used internally by the Provincial Government for a while, but was recently released to the public at https://gpcoronavirus.co.za. In this talk I will take you through the technologies we used to build the platform, some of the challenges we had to overcome, and how to get started with mapping out geo-spacial datasets in the browser.
Deno: Is it just Node, but alphabetically?
Gergana Young (@gerybbg)
Node.js was the first, and for a long time the only, runtime that allowed us to run JavaScript outside the browser. Its popularity is massive, and there are tons of systems running it in production. However, even its creator, Ryan Dahl, has been heard talking about the bad parts. This is why he started a new project, Deno.
Deno is a new JavaScript (and TypeScript) runtime for the server built in Rust. It is still very new, with version 1.0 only having been released in May 2020. It is meant as an alternative to Node, and was built to avoid a lot of Node's pitfalls. Although having some disadvantages of its own, the creators are working on ironing out the kinks and making it suitable to run in production.
Join me for this, code heavy, talk where I will start off by taking you through the advantages and disadvantages of Deno. I will then show you how to get set up, and we will go through an example of creating a RestAPI in Deno.
❓Q&A: slido.com - #11285
📢Submit a talk: https://goo.gl/forms/66MrplT1FogDX88Q2
💛Code of conduct: https://docs.google.com/document/d/1j-OfcTwYejEj-WiQq9F9AcVv2L-Ki81RuhFTjquDJyk/edit?usp=sharing
Meetup: https://www.meetup.com/Jozi-JS/events/273637015/
A
We
hope
you're
all
doing
fine,
but
it's
been
it's
been
heavy
in
johannesburg,
so
it's
been
really
really
hot
over
the
last
week,
or
so
I
hope,
you're
we're
all
coping
and
at
home.
Without
your
air
conditioning
with
whatever
poultry
fan
you
can
manage
and,
yes,
we
are
still
remote.
We
are
still
home,
we're
hoping
that
this
year
will
end
uneventfully
and
then
next
year
we'll
be
able
to
return
to
in-person
meetups
and
are
really
really
looking
forward
to
it.
A
A
A
We're
also
always
looking
for
talks.
So
if
you
are
interested
in
speaking
potentially,
please
feel
free
to
submit
the
talk.
It's
the
qr
code
on
your
left,
and
I
appreciate
the
irony
of
putting
qr
codes
on
something
where
we
could
just
put
hyperlinks,
and
these
are
also
in
the
descriptions
below
so
feel
free
to
click.
Those,
and
I
also
appreciate
the
irony
of
putting
a
qr
code
for
watching
all
of
the
past
talks.
A
The
very
many
of
them,
all
of
which
have
been
uploaded
to
this
very
youtube,
channel
that
you're
on
right
now,
so
you
know
where
they
are.
You
can
click
around
and
please
let
us
know
what
you
think:
there's
some
incredible
wisdom
from
local
speakers,
we're
going
to
have
two
talks
tonight
and
for
each
talk
we're
going
to
be
using
slido
with
the
following
details
for
any
questions
and
answers
that
we
have.
So
please
put
your
questions
in
slido
and
then
I
will
read
them
out
to
the
speakers
and
get
them
answered
yeah.
A
B
So,
first,
a
little
bit
about
me,
here's
a
picture
of
what
I
looked
like
before.
I
started
working
from
home
five
days
a
week,
I'm
a
software
engineer
at
ibm
research
based
in
johannesburg
for
more
information
about
the
ibm
research
lab
in
johannesburg.
You
can
check
out
that
url
that's
on
screen
and
you
can
also
find
me
on
twitter
and
github
at
those
details
that
are
there.
B
B
So
when
covet
19
infections
first
started
picking
up,
we
started
working
quite
closely
together
with
the
khartang
provincial
government.
As
well
as
vitz
university
and
the
gcro
to
help
visualize
and
make
sense
of
some
of
the
data
that
they
had,
the
data
that
they
had
was
the
locations
of
positive
cases
and
some
basic
info,
such
as
age
and
gender,
demographic
and
risk
factor,
data
about
areas
in
khateng
put
together
by
the
gcro,
which
is
the
khateng
city
region
observatory
and
data,
from
prediction
models
about
how
cases
would
spread
throughout
khartang,
put
together
by
this
university.
B
They
wanted
to
be
able
to
answer
questions
like
which
suburbs
had
the
most
number
of
new
cases
in
the
previous
few
days,
which
of
these
hot
spots
are
most
likely
to
get
out
of
control,
which
areas
are
most
likely
to
become
hotspots
in
the
future,
and
they
also
wanted
to
know
things
like
the
age
distribution
of
positive
cases
in
a
hotspot,
because
a
breakout
of
cases
in
an
old
age
home
is
going
to
be
a
lot
more
dangerous
than
a
breakout
of
cases.
In
a
preschool.
B
B
So
the
initial
version
of
the
website
wasn't
open
to
the
public.
It
was.
There
was
a
login
that
you
had
used
to
get
to
it,
but
we
did
recently
get
approval
to
make
the
sites
available
to
the
public.
So
if
you
want
to
go
and
visit
it,
it's
at
gpcoronovirus.co.c
I'm
going
to
take
you
through
a
quick
tour
of
the
site
so
that
you
have
context
of
some
of
the
things
that
I'm
going
to
talk
about
later.
B
B
At
the
bottom,
we
have
a
timeline.
If
we
select
a
particular
area,
we
can
see
the
individual
timeline
just
for
that
area,
so
I've
selected
marshall
time
here,
which
is
enjoyable
cbd
and
we
can
see
a
timeline
of
all
of
the
cases
from
march
to
present,
as
well
as
the
changes
per
day.
We
can
also
aggregate
that
by
week,
which
just
gives
you
a
slightly
smoother
view,
so
you
don't
get
all
of
the
the
spikes.
B
B
B
Another
thing
that
we
can
do
is
we
can
filter
to
a
specific
time
period.
So
at
the
bottom
here
we've
got
a
timeline.
We've
currently
got
the
whole
time
period
selected,
but
we
can
filter
to
just
cases
that
occurred
in
october
and
now
we
can
quite
easily
see
which
areas
are
the
more
recent
hotspots
where
there's
lots
of
infections
occurring.
B
Something
else
that's
useful
about
this
is
it
allows
us
to
drill
down
to
specific
periods?
So,
for
example,
if
I
choose
march,
we
can
quite
easily
see
that
the
main
hotspot
of
cases
in
march
is
centered
around
santan,
which
is
most
likely
due
to
people
coming
back
from
overseas
trips
that
were
infected
and
busy
infecting
their
co-workers
and
family.
B
Let
me
just
reset
everything
right,
so
we
also
implemented
a
heat
map
functionality
which
you
can
see
up
here.
We
can
turn
that
on
and
off,
and
the
reason
we
did
that
is
because
the
wards
in
south
africa
are
not
all
equal
sizes.
You
can
see.
Some
of
them
are
quite
big,
and
some
of
them
are
quite
small
and
the
problem
that
that
creates
is,
if
you've
got
an
area,
that's
got
lots
of
small
wards.
It
becomes
difficult
to
see
if
that
area
is
potentially
becoming
a
hot
spot.
B
A
good
example
of
this
is,
if
I
zoom
in
a
little
bit,
you
can
see
this
area
around
sueto
over
here
looks
like
it's.
It's
largely
been
untouched
by
the
the
coronavirus,
but
it's
actually
because
there's
lots
of
small
wards
that
the
cases
are
distributed
between.
If
we
turn
on
the
heat
map,
it
does
spatial
aggregation
and
we
can
see
that
actually,
there
was
quite
a
few
cases
in
that.
B
Area
we
also
have
the
ability
to
see
the
the
top
areas
within
south
africa.
So,
for
example,
we
can
see
a
ranking
of
all
of
the
provinces,
which
ones
have
the
most
cases,
but
we
can
also
see
which
of
the
worst
districts
sub-districts
and
wards
within
the
whole
country,
and
we
can
see
information
about
each
of
those
and,
as
we
click
on
one
of
these,
we
can
see
all
of
the
other
charts
updates.
B
At
the
same
time,
we
have
a
couple
of
other
metrics
that
we
just
that
we're
displaying
so
we've
got
deaths,
recoveries
tests
hospitalization
the
number
of
active
cases.
We
also
have
cases
with
a
hundred
thousand,
which
is
a
really
good
metric.
If
you
want
to
try
and
factor
in
the
the
population
of
a
particular
area,
not
all
metrics
are
available
at
all
levels.
B
So
the
next
data
set
that
we
integrated
was
metrics
and
indices
compiled
by
the
gcro,
which
we
called
risk
factors,
and
they
they
mapped
out
all
of
the
areas
in
khateng,
based
on
various
factors
that
are
all
listed
here.
You
can
see
that
we've
got
these
two
indices
and
then
all
of
the
things
that
make
up
each
of
these
indices
and
what
they're
looking
for
is
areas
that
are
likely
to
be
susceptible
to
covert
19.
B
So
we
set
this
up
in
such
a
way
that
it's
quite
easy
for
us
to
choose
a
particular
area
that
we're
interested
in
and
swap
between
the
cases
and
the
risk
factors,
and
it
becomes
really
easy
for
us
to
start
identifying
areas
that
might
be
high
risk
as
soon
as
they
start
getting
an
increase
in
cases.
We
know
to
start
investigating
those
areas.
B
They
ran
a
number
of
different
models,
and
these
models
have
actually
changed
quite
a
few
times.
So
the
the
most
recent
data
that
we
got
just
predicts
the
level
level
one,
but
previously
they
had
models
that
had
predictions
for
all
of
the
the
different
levels
from
one
to
five
and
that
gave
policymakers
a
good
idea
of
which
areas
are
most
likely
to
see
an
increase
in
cases
and
also
how
many
cases
are
we
expected
to
have
so.
B
This
timeline
at
the
bottom
here
now
shows
us
what
these
predictions
are
for
a
future
date,
and
we
also
have
the
ability
to
choose
a
particular
date
and
see
how
that
affects
the
map
up
at
the
top
here.
So,
as
we
select
a
different
date,
we
can
see
that
the
map
updates
there's
a
couple
of
other
things
that
they
have
as
part
of
their
prediction
models,
such
as
fatalities,
recoveries,
total
cases.
B
They
also
have
quite
detailed
predictions
around
hospitalization,
and
this
was
very
useful
for
determining
how
many
hospital
beds
does
the
the
kaiten
government
need
to
provide
they've.
Also
split
it
into
public
and
private,
as
well
as
if
it's
icu
high
care,
isolation
award,
general
ward,
that
kind
of
thing
so
you.
B
B
Unfortunately,
a
lot
of
our
officially
released
data
in
south
africa
looks
a
little
bit
like
this.
It's
an
image
that
gets
posted
on
social
media
or
maybe
a
table
that
gets
embedded
into
a
pdf
report,
and
this
makes
it
really
difficult
to
collect
and
process
this
data
because
it's
not
machine
readable.
B
Thankfully,
there's
a
research
group
at
the
university
of
pretoria
that
realized
the
importance
of
having
machine,
readable
data
and
the
data
science
for
social
impact
research
group
created
a
github
repo
with
all
of
the
publicly
released
data
around
covet.
19
and
they've
got
a
whole
team
of
volunteers
that
help
keep
this
data
up
to
date
on
a
daily
basis,
so
we
ended
up
with
quite
a
lot
of
data.
In
fact,
we
we
have
currently,
I
think,
more
than
one
and
a
half
million
unique
data
points
that
we're
displaying.
B
We
run
batch
jobs
every
four
hours
on
the
server
that
fetch
and
process
new
data
from
all
of
our
various
sources,
and
we
pre-process
a
lot
of
this
data
into
static
files.
We
can
then
serve
this
content
to
the
front
end
and
cache
as
much
as
possible,
and
the
last
thing
that
we
do
is
we
try
to
optimize
our
file
formats
as
much
as
possible
to
to
make
the
files
easy
to
download.
B
An
example
of
some
of
our
data,
so
here
we
can
see
two
different
forms
of
the
same
data.
Javascript
really
lacks
working
with
json
data,
but
json
data
kind
of
has
a
problem,
and
you
can
immediately
see
by
looking
at
this
that
there's
quite
a
lot
of
duplication
in
the
json
data.
We're
writing
out
the
object
keys
for
every
item,
which
we
don't
really
need
to
do
in
comparison.
The
csv
we've
just
got
all
of
the
keys
at
the
top,
and
then
we
can
just
separate
all
of
our
data
by
by
commas.
B
So
we
served
all
of
our
data
as
csv,
instead
of
json
to
reduce
the
file
size
and
then
we
convert
it
to
javascript
objects
in
the
browser.
Once
it's
download
downloaded
there's
a
couple
of
other
formats
that
may
reduce
the
file
size
even
more
things
like
bson,
but
csv
is
still
human
readable,
which
we
thought
was
quite
an
important
factor
in
this.
B
So
how
much
of
a
difference
did
it
actually
make?
Well,
I
took
an
example
of
our
data
snapshot
from
earlier
this
week.
That's
the
the
cases,
deaths,
recoveries
tests
and
hospitalization
data,
and
there
was
77
700
items
and
we
can
see
that
in
json
format
it
was
10.1
megabytes,
unformatted
and
in
csv
format
that
same
data
was
2.8
megabytes.
So
that's
quite
a
big
difference,
but
because
we're
serving
our
data
with
gzip,
we
actually
need
to
compare
the
gza
file
size.
B
One
way
around
this
is
to
do
data
aggregation
here.
I
have
an
example
of
our
case
demographics
data
where
we've
got
the
date.
The
case
was
reported
on
the
ward,
that
it
was
reported
in
gender
age
group
and
then
what
we
did
is
we
grouped
the
the
cases
that
were
in
the
same
ward,
and
we
also
created
age
buckets
for
the
purposes
of
this
dashboard.
B
B
So
how
much
does
this
actually
save
us?
Well,
we
can
see
that
the
unaggregated
data
was
5.7
megabytes
and
when
we
aggregated
by
ward
and
age
group
it
came
down
to
4.5,
megabytes
didn't
seem
like
that
much
of
a
difference,
but
when
we
gzipped
it,
it
actually
got
much
better
performance
and
almost
half
the
size
from
811
kilobytes
to
600
461
kilobytes,
let's
dive
into
the
front
end
a
little
bit
and
how
we
actually
built
that
we
built
it
using
pure
modern
javascript.
B
We
didn't
use
any
frameworks,
and-
and
that
means
that
we
didn't
need
to
look
for
any
react
or
angular
specific
charting
libraries,
which
can
always
be
a
little
bit
tricky
to
to
work
with
optimization,
was
also
really
easy.
We
didn't
have
to
dig
into
the
framework
internals
and
we
had
full
control
over
how
and
when
rendering
gets
done,
which
is
important
when
rendering
your
page
generates
an
svg
chart.
That
has
a
hundred
thousand
elements
in
it.
B
B
We
use
the
charting
library
called
dcjs,
which
is
really
good
for
creating
connected
charts,
so
all
of
the
charts
that
interact
with
each
other,
that's
dcjs,
which
is
built
on
top
of
d3,
which
is
kind
of
the
de
facto
standard
for
for
writing.
Charts
these
days
and
dcjs
also
uses
something
called
cross
filter
for
filtering
and
processing.
B
Dataview
allows
us
to
have
one
copy
of
the
data
set
in
memory
with
a
number
of
different
filters
and
views
that
depend
on
those
filters
and
when
one
of
the
filters
is
changed,
the
views
that
depend
on
that
filter.
Update
this
library
kind
of
really
tests
the
data
processing
capabilities
of
javascript,
but
I
think
that's
a
whole
other
talk
on
itself,
so
I
won't
dive
into
that
too
much.
B
B
B
It's
got
a
type
of
feature
collection
and
it's
also
got
a
list
of
features
and
each
of
those
features
contain
some
properties
associated
with
them.
So
here
I've
set
the
area
to
harten
I've
set
a
date.
I've
set
the
total
number
of
cases
on
that
date
and
I've,
given
us
an
intensity
property
which
will
be
a
number
between
zero
and
one
and
that'll,
be
how
much
we
want
to
highlight
that
area.
B
It
also
has
some
kind
of
geometry
and
that's
usually
going
to
be
a
polygon
or
a
point
or
a
line
in
this
case.
We're
mapping
out
an
area,
so
polygon
makes
sense,
and
then
we
have
a
list
of
coordinates,
and
these
are
like
your
standard
coordinates
that
you
use
on
a
gps,
and
these
make
up
all
of
the
different
points
of
the
boundary
of
this
area.
B
B
B
B
B
B
B
B
Now.
If
we
want
to
update
our
data
dynamically,
we
can
update
the
properties
in
our
geo.
Json
source
object,
we'll
find
kha
teng
by
searching
through
the
the
features
in
our
geo
json,
we'll
change
the
date
and
set
the
the
new
number
of
cases
on
it
oops,
and
then
we
can
set
the
intensity
value
as
a
fraction
of
the
maximum
number
of
cases
that
way,
the
areas
with
more.
C
B
C
B
B
One
option
is
you:
can
try
simplifying
your
geojson
files?
You
can
use
a
website
called
mapshaper.org
to
simplify
the
boundaries.
What
it
does
is
it
reduces
a
number
of
points
that
make
up
each
area
and
it
tries
to
choose
them
smartly,
so
that
the
shapes
don't
change
too
much,
and
this
is
a
an
animation
showing
how
it
changes.
Based
on
how
much
you
want
to
simplify
map.
B
Shaper
actually
does
all
of
this
processing
in
the
in
the
browser
on
the
client
side,
which
is
actually
quite
impressive,
so
map
shaper
in
itself
is
quite
an
impressive
tool,
but
anyway
we
can
only
simplify
the
data
so
much
before
you
lose
too
much
detail
and
some
of
the
wards
in
south
africa
are
really
small.
So
we
had
to
look
for
another
solution.
B
B
If
you
look
in
the
network,
tab
of
your
browser,
you'll
see
it
loading
all
of
the
tiles
as
you
scroll
around,
and
each
tile
is
encoded
into
the
url,
its
location.
So
we've
got
the
x
y
and
z
coordinates
of
the
tile,
as
well
as
the
zoom
level
that
we're
in
and
what
this
does
is.
It
allows
us
to
not
have
to
load
the
entire
data
set
all
at
once.
We
can
just
load
the
the
data
set,
that's
visible
at
the
moment.
B
B
B
Now,
if
we
have
a
vector
tile
set
that's
stored
in
mapbox
rather
than
our
own
geojson,
how
do
we
actually
update
and
style
it?
Well,
we
have
the
ability
to
query
the
currently
rendered
features
that
are
visible
on
the
map.
So
what
we
can
do
is
we
can
say,
map
dot,
query
rendered
features
and
tell
it
the
layers
that
we're
interested
in
and
we'll
get
back
a
list
of
features
that
have
currently
been
rendered.
They
come
back
in
something
that
looks
similar
to
geojson,
but
it
obviously
won't
be
the
full
set.
A
B
B
You
can
check
out
geojson.io,
which
has
got
that
really
cool
tool
for
building
your
own
geojson
files
and
mapping
them
out
or
geojson.org
to
check
out
the
spec
you're,
also,
probably
going
to
need
some
sort
of
a
ui
library
to
render
a
map
you
can
use
the
mapbox
gljs,
which
is
the
obviously
the
mapbox
one.
They
don't
force
you
to
use
their
back-end
services,
which
require
you
to
register
an
account
and
you
can
just
use
their
front-end
library
and
and
use
things
from
openstreetmaps.
B
You
can
use
leaflet.js,
which
isn't
affiliated
with
any
sort
of
hosting
service.
It
has
slightly
fewer
features
than
the
mapbox
and
I
think
mapbox
was
originally
built
on
top
of
leaflet.js,
and
I
believe
you
can
also
do
similar
things
with
the
google
maps
developer
api.
But
I'm
not
too
sure
I
haven't
dived
into
the
features
of
that
for
data
processing.
A
Wow
richard
that
was
phenomenal,
like
I
think
I'm
gonna
have
to
go
through
the
whole
thing
again
in
half
speed
and
pause
and
take
notes.
That's
it's
a
brilliant
job
and
an
amazing
cause.
You
must
be
really
really
proud.
There
are
two
questions,
so
the
first
question
said
you
were
pulling
data
from
multiple
sources.
So
did
you
include
some
sort
of
caching
for
the
data
sources.
B
Yeah,
so
so,
there's
quite
a
few
different
layers
of
caching.
So
so,
firstly,
we're
only
pulling
the
data
every
four
hours
and
then
we're
storing
it
all
on
on
our
server
and
then
we're
caching,
it
at
various
levels
on
both
the
back
end
and
the
front
end.
So
there's
quite
a
lot
of
caching,
which
we
did
just
to
to
make
our
labs
easier
and
not
have
to
scale
up
our
application
too
much.
So
if
there
aren't
the
questions
about
caching,
then
the
answer
is
yes.
Lots
of
caching.
B
So
not
not
too
many.
We
try
to
make
most
of
its
static
files.
So
a
lot
of
the
the
processing
happens
in
a
batch
job
where
it
doesn't
really
matter
how
long
that
takes
as
long
as
it
you
know
takes
less
than
four
hours
or
whatever
that
we've
got
between
batch
jobs
running,
but
most
of
it
is
served
statically
actually
just
with
with
nginx
and
there's
very
few
parts
of
it
that
actually
hit
our
application
server.
So
we
don't
have
to
have
too
many
instances
running.
B
So
the
the
I
guess,
the
way
this
project
came
about
was
all
quite
rushed,
so
we
didn't
have
too
much
time
to
do
that,
much
planning
up
front
and
it's
still
an
ongoing
project.
So
we
may
still
do
this
in
the
future.
But
one
of
the
things
that
we
can't
do
at
the
moment
is
when
we
query
data
from
the
server.
We
query,
all
of
it
at
the
same
time,
for
the
entire
time
period.
A
Okay,
cool,
that's
a
good
answer!
Thank
you
very
much
richard.
It
was
an
awesome
talk.
Definitely
looking
forward
to
publishing
that
one
more
great.
So
next
up
we
have
somebody
who
the
jersey,
js,
faithful,
are
very
accustomed
to
seeing
and
she's
going
to
take
it
away
with
a
little
something
about
dino.
C
Awesome
hide
the
zoom
controls.
I
recently
learned
how
to
hide:
zoom
controls,
it's
the
best
feature
ever
cool
all
right,
hi.
Everybody
just
want
to
get
my
notes
up
over
here.
Thank
you
all
for
being
here
coming
joining
us
once
again
for
another
josie
js,
so
yeah.
As
mike
said,
I'm
going
to
be
talking
to
you
about
dino.
C
Before
we
get
started.
Let
me
introduce
myself
for
those
of
you
who
don't
know
me.
My
name
is
gergana,
but
I
usually
go
by
jerry
because
that's
a
little
bit
easier
to
pronounce
and
you
can
find
me
on
twitter
or
on
my
website,
for
if
you
want
to
have
a
look
at
some
of
the
other
weird
projects
that
I
work
on
in
my
spare
time.
C
I
currently
work
for
a
company
called
cora.
I
is
my
pension,
I'm
a
co-organizer
of
the
jersey,
js
meetup
group.
If
you're
here,
then
you
know
that
already
probably
and
I'm
also
a
microsoft
mvp,
I
was
given
the
mvp
award
because
of
my
contributions
to
the
south
african
dev
community,
in
particular,
to
iot,
and
because
I
have
this
weird
thing
of
taking
javascript
and
iot
and
putting
them
together.
I
know
two
very
unreliable
technologies
working
together,
but
it's
my
thing.
C
Unfortunately,
this
time
this
talk
will
not
contain
any
things,
but
there's
still
javascript,
so
it
should
still
be
good.
I
want
to
start
off
with
a
little
bit
of
the
history
of
javascript
and
a
lot
of
the
facts.
I'm
going
to
mention
come
from
a
talk
called
the
birth
and
death
of
javascript
by
gary
bernhardt
and
whether
you
love
or
hate
javascript.
You
should
definitely
watch
that
talk.
It's
a
very
good
talk,
but
I'm
going
to
take
you
back
to
1994.
C
C
C
They
wanted
more
from
the
web
so,
together
with
brendan
ich,
they
created
javascript
in
1995,
netscape
navigated,
to
the
second
version
of
that
gui
browser
was
released
and
javascript
was
included
in
it
story
goes
that
the
name
was
officially
livescript,
but
a
few
months
later
was
renamed
to
javascript,
because
at
the
time
java
was
the
new
popular
language.
Everyone
was
writing
in
it,
so
they
wanted
to
jump
onto
that
bandwagon
and
use
the
name.
C
It
has
been
the
most
popular
and
the
one
that
everybody
has
heard
of
together
with
npm
nodes
package
manager
and
event-driven.
I
o
and
the
v8
engine
node.js
made
server-side
development
accessible
to
all
javascript
developers
and
to
this
day,
node
is
used
by
many
companies.
Big
and
small.
Some
of
the
giants
still
use
it
and
despite
many
doubts
in
the
language,
thanks
to
node,
javascript
became
the
most
used
programming
language
in
the
world,
but
then
in
2018
ryan
doll
did
a
talk
about
the
10
things
I
hate.
I
regret
about
node.js
in
it.
C
C
C
So
let's
talk
about
some
of
the
good
things
of
dino.
Now,
as
I
said,
dino
was
made
based
on
the
bad
parts
of
node.
So
a
lot
of
the
advantages
I'm
going
to
talk
about
I'm
going
to
compare
that
to
node,
firstly,
and
probably
most
importantly,
dino
is
secure
by
default
when
it
comes
to
node.
When
you
run
a
node
program
on
your
computer,
it
has
access
to
absolutely
everything
your
file
system,
the
internet,
everything
secure
by
default,
means
that
dnode
does
not
have
this
access.
C
Unless
you
explicitly
grant
it
and
you'll
see
an
example
of
that
when
we
get
to
the
demo
a
bit
later,
it
supports
typescript
by
default.
Now
I
know
typescript
is
a
very
debated
topic
and
some
people
love
it.
Some
people
hate
it.
Some
people
that
see
that
see
it
as
the
only
way
to
write
javascript.
Some
people
don't
see
the
point
of
it.
That's
not
the
point
of
this,
though.
C
The
fact
that
it's
there
as
a
default
gives
people
an
option,
whereas
in
node
we
have
to
install
extra
third-party
libraries
like
ts
node
in
order
to
run
typescript
it
ships
as
a
single
executable.
Now
this
is
really
nice,
because,
if
you
ever
need
to
update
you
know
you
just
write,
you
know
you
just
run
dino
update,
not
like
node,
where
you
have
to
install
the
node
version
manager,
so
you
can
manage
which
version
of
node
you're
actually
running.
C
C
It
also
follows
the
web
browser
standards.
Now
this
is
one
of
the
things
that
node
did
not
do
very
well,
and
most
of
the
reason
for
that
is
because
a
lot
of
node
was
created
before
a
lot
of
these
things
were
web
browser
standards,
but
now
that
there
are
standards,
a
server
side,
language
following
the
same
standards
makes
it
a
lot
easier
to
go
between
the
two,
and
it
has
a
bunch
of
built-in
tooling,
which
is
quite
cool.
Things
like
testing
and
formatting
and
linting
are
all
built
in
in
into
dino.
C
So
you
don't
need
to
install
third-party
libraries
based
on
these
advantages,
and
this
is
not
a
small,
a
small
amount.
Even
though
there's
more,
you
would
think
that
dino
is
the
best
thing
ever
and
we
should
all
switch
to
dno
right
away,
but
it
would
be
unfair
if
I
didn't
tell
you
about
the
bad
parts
as
well.
C
C
C
C
C
C
Is
we
can
use
es
module
imports?
So
we
can
import
hello
from
hello
needs
to
be
in
integrated
commerce
and
importantly,
the
difference
from
node.
Is
we
have
to
provide
the
extension
in
node
when
you
use
the
import
or
require
syntax?
You
don't
have
to
provide
an
extension
in
dino,
you
do
so
we
run
and
we
have
the
same
result
and
that's
cool.
I
mean
console
logging.
C
C
And
we
just
have
a
url
for
random
dog
pictures,
but
now
this
returns
a
promise.
So
do
we
have
to
go
response.then
or
no,
because
dino
also
supports
top
level
weights,
and
this
is
something
that
I'm
really
excited
about.
Top
level
awaits
are
the
best,
and
then
we
have
to
get
the
json
from
the
response.
C
C
What
else
can
you
do
with
the
note?
Well,
you
can
use
browser
standards
as
of
this
version,
so
version
1.4,
which
was
released
very
recently.
You
can
even
do
things
like
web
sockets
like
natively,
how
you
would
do
them
in
a
browser.
You
would
do
it
the
same
in
dino.
You
can
also
create
an
http
server.
It's
admitted
most
of
our
node
programs
are
actually
just
http
servers,
so
we
can
import
the
native
server
from
the
dino
standard
library
and
create
a
new
server.
C
C
The
dot
message
now
we
can
start
this
up,
let's
take
out
this
and
give
it
access
to
the
internet,
because
we
now
have
a
server
that
will
need
to
be
able
to
access
from
localhost
and
we
have
a
server
started.
So
if
we
go
to
postman
and
send
a
get
request
to
localhost
8080,
we
have
our
random
dog
picture.
C
Cool
this
is
the
default
way
with
the
standard
library
to
create
a
dino
server,
but
I
want
to
show
you
another
way
of
creating
it
now
most
of
you
have
probably
used
express
or
cola
or
something
to
that
effect
in
in
node.
I'm
going
to
use
an
external
library
called
oak
for
those
of
you
who
have
seen
a
cover.
C
You'll
find
this
very
familiar
oak
was
built
to
be
the
dino
equivalent
of
koa,
but
that
cover
is
for
node
we're
importing
it,
and
you
can
see
an
x
over
there
that
x
over
there
specifies
that
the
thing
we're
importing
is
a
third-party
library.
It's
not
one
of
the
standard
libraries
that
dino
has
created.
Someone
else
has
created
it
as
an
open
source
library.
C
C
C
C
Then
last
thing
we
need
to
do
is
start
this
thing
up
to
do
that
we
specify
that
our
application
is
using
the
roots
and
allowing
these
methods
and
we
listen
on
the
port
we
created.
Now.
If
we
stop
this
start
it
again,
we
can
go
to
postman
again
and
we
have
our
hello
josie.js,
that's
cool.
So
now,
let's
build
something
useful:
let's
build
a
to-do
application,
I'm
going
to
create
a
new
folder
over
here
and
call
it
controllers.
C
C
And
import
get
to
do's,
don't
forget
the
extension,
stop
that
restart
and
if
we
send
this
we're
getting
our
data.
So
now,
let's
get
some
actual
data.
Now,
when
I
was
planning
this
demo,
I
wanted
to
connect
it
up
to
to
show
you
how
all
of
that
works,
and
this
is
the
problem
with
working
with
something
that
is
so
new.
C
So,
as
I
said
earlier,
dino
was
recently
updated
to
version
1.4,
but
the
package
I
found
hadn't
been
updated
to
work
with
1.4,
which
meant
that
I
couldn't
use
unless
I
downgraded
my
dino
version.
So
that's
why
I
decided
to
rather
just
use
a
json
file
which
has
a
few
to-do's
in
it
and
show
you
how
it
would
work
with
files.
C
C
C
C
C
C
C
C
Roots
there
we
go
cool,
stop
start,
there's
actually
a
watch
flag,
which
was
added
recently
as
well,
but
it's
still
unstable.
So
it
doesn't
actually
work
that
well,
okay,
let's
try
that
post
again,
create
it.
So
now,
if
we
go
back
to
the
get
we
should
have
four
to
do's
with
mark
is
complete
is
one
of
our
to
do
in
the
interest
of
saving
some
time.
C
C
C
C
C
One
other
thing
I
want
to
talk
about
a
little
bit
is
modules,
so
you
saw
me
using
these
imports
like
the
dino
dinoland
oak
module
and
it
uses
es6
modules,
which
is
which
is
nice
and
it's
imported
directly
from
a
url
or
you
can
download
it
locally
and
import
it
from
a
file
path,
and
all
of
these
things
are
immutable
and
they
get
cached
locally.
So
you
can
do
a
reload.
So
if
we
did
a
reload
over
here
reload
it
will
redownload
all
of
these
external
dependencies,
so
it
doesn't
download
them
every
time.
C
The
one
problem
that
I've
seen
with
it
is
it's
very
it's
kind
of
difficult
to
manage.
So,
like
I'm
importing
just
oak,
I'm
not
saying
which
version
I
can
specify
exact
version,
but
then,
if
I
need
to
import
this
in
another
file
as
well,
then
I
need
to
re-import
it.
You
know,
oh,
a
suggestion
that
they
have
a
recommended
way
of
doing
dealing
with
this
on.
The
dino
website
is
that
you
have
a
dependencies
file
where
you
import
all
of
your
dependencies
and
export
them
from
there.
C
The
other
problem
with
this
is
that
it
also
becomes
very
difficult
to
check
for
vulnerabilities
in
these
packages
because
they
could
be
imported
in
so
many
different
places
and
they're
not
like
ever
downloaded
locally
or
whatever,
it's
very
difficult
to
check
for
the
vulnerabilities
in
packages.
So
that's
one
of
the
things.
That's
currently
a
problem
as
well
with
you
know.
C
Something.
That's
kind
of
cool,
though,
is
how
you
can
do
testing
in
dina.
So
there
is
a
dino
test
function.
Currently,
it
only
supports
assertions
and
cover
and
code
coverage,
but
you
don't
have
to
install
any
external
things
in
order
to
do
testing,
which
is
pretty
cool,
because
I've
always
found
it
difficult
to
choose
a
library
to
do
testing
within
node.
I
don't
know
if
anyone
else
has
that
kind
of
problem.
C
Then
some
of
the
other
tools
that
you
get
for
dino,
which
are
default
installing
so
installing
this
and
distributing
executable
code.
So
you
can
package
up
your
package
and
give
it
to
someone
else
to
install
this
format.
So
built-in
code,
formatter,
dino
fmt,
will
format
your
code
for
you
bundling
so
take
your
entire
program
and
bundle
it
into
a
single
javascript
file.
It's
nice
that
it
puts
it
into
just
one
file.
C
C
If
we
do
something
like
this,
you
know
info
on
main
dot.
Js
we
get
all
of
the
dependencies
and
their
dependencies,
so
you
can
see
and
you
can
see
how
big
they
are
as
well.
You
can
see
this
that
this
one
has
quite
a
lot
and
then
there's
also
lint,
which
is
a
code
linter.
It
uses
the
standard,
eslint
rules
as
far
as
you
know,
yeah,
and
follows
those,
and
you
just
run
the
knowledge
and
you
get
that.
C
The
answer
to
that
question
is
no,
at
least
in
my
opinion,
I
think
dino
is
a
is
going
to
be
a
good
alternative
to
node,
it's
not
there.
Yet
I
don't
think
it's
production
ready
yet
because
node,
if
we
think
about
it,
node
has
been
around
for
what
like
11
years
now.
You
know
it's
been
around
for
six
months,
so
it
it
needs
some
time,
but
it
does
build
on
the
good
parts
of
node.
It's
it's
fun
to
work
with
it's
it's
quite
nice.
C
The
fact
that
it
follows
so
many
of
the
browser
standards
is
really
is
really
nice
because
it
you
can
do
the
things
that
you
already
know
how
to
do
in
the
browser
inside.
You
know
and
a
lot
of
those
things,
so
I
think
one
day
it
will
get
there,
but
for
now
I
think
it's
just
fun
to
learn
and
keep
on
top
of
just
so.
You
know
what's
going
on
with
it,
that's
it.
I
spoke
very
fast.
Thank
you.
A
C
C
C
C
Cool
yeah,
and
so
now
so
now
it
will
wait
and
if
I
make
a
change
here,
I
don't
know
what
kind
of
change
to
make
you
see
the
watcher
file
change
detected,
restarting
but
yeah,
it's
it's
it's
a
little
bit
iffy.
Sometimes
when
you
hit
an
error,
the
it
doesn't
actually
restart.
I
guess
because
it's
still
new.
C
So
yeah
there
was
the
dependencies
file
thing
I
spoke
about
it.
Doesn't
it
doesn't
have
a
package.json?
That
was
one
of
the
things
that
ryan
doll
mentioned
in
his
talk
about
not
liking
about
dino.
So
so
you
would
create
a
dependencies.ts
and
import
all
of
your
modules
in
there
and
export
them
from
there,
so
that
any
of
your
other
files
would
use
it
from
that
file.
C
So
you
don't
import
different
ones
in
different
places,
and
it's
actually.
What
I'm
doing
here
is.
Another
thing
is
not
recommended.
You
have
to
pin
to
a
version.
You
shouldn't
just
rely
on
the
latest
version.
You
should
be
pinning
to
a
specific
version
because,
especially
with
the
third-party
libraries,
like
I
said,
with
the
one
they
do
they
break
when
there's
updates.
A
Okay,
cool,
I
think
that's
fantastic,
very,
very
concisely
done
so.
Thank
you
for
the
great
talk
sharing
both
of
the
youngs.
Thank
you.
We
had
a
matched
pair
this
evening,
which
is
great
again
if
I
can
remind
everyone
if
anyone
is
keen
to
speak,
we're
full
up
for
the
rest
of
the
year,
which
well
it's
basically
just
november,
but
we're
definitely
keen
to
accept
new
people
from
next
year.
We're
new
speaker
friendly.
So
we'll
help
you.
If
you
need
it
and
yeah,
you
must
come
and
speak
and
we
will
see
you
in
november.