►
Description
Weekly updates: https://gitlab.com/gitlab-org/incubation-engineering/jamstack/meta/-/issues/5
Experience Issue: https://gitlab.com/gitlab-org/incubation-engineering/jamstack/meta/-/issues/13
A
Hi
and
welcome
to
update
number
five
for
the
jumpstack
incubation
engineering
group.
I
have
a
few
different
things
that
I
want
to
talk
about
today.
One
is
the
progress
update.
I
don't
have
a
lot
to
say
on
this.
I've
spent
most
of
the
time
last
week
with
getting
the
code
ready
for
review
and
have
that
run.
In
the
background
from
now
on.
Meanwhile,
I've
started
my
preliminary
work
on
the
next
task,
which
is
the
cdn
delivery
of
pages.
A
A
So,
let's
assume
I
have
a
very
little
application
that
compiles
to
static
files.
This
is
a
nox.js
application
which
is
built
on
view,
but
it
could
be
next.js
or
anything
else
would
be
very
similar.
So
let's
say
we,
we
have
this
small
little
application.
So
this
is
what
it
looks
like
locally,
very
small
compiles
to
a
static
site,
but
very
simple.
So
and
I
have
a
repository
in
gitlab
where
this,
where
this
project
is
stored,
it
doesn't
have
a
gitlab
ci
yaml
file.
A
A
We
could
add
a
new
domain
and
redirect
in
the
domain
to
that,
but
still
the
page
would
be
delivered
by
gitlab
pages
servers
which
aren't
cached
or
anything
and
they're
mostly
located
in
the
us.
So
we
get
a
lot
of
load
times
for
non-us
clients,
so
the
idea
is
to
use
fastly
in
front
of
of
of
this
page
instead.
So
how
are
we
doing
that?
This
is
fastly.
A
A
A
Okay,
I'm
gonna
add
that,
so
this
is
the
domain
that
my
website
is
going
to
be
reachable
under
for
the
world
having
been
cached
now
on
the
back
end,
this
section
basically
tells
fastly
whenever
you
get
a
request
to
this
original
domain,
get
the
files
for
that
from
from
this
host,
and
here
is
where
it
gets
a
little
bit
confusing
for
the
user.
So
I
was
misled
to
think
that
I
needed
to
create
a
domain
for
this
particular
application
first,
but
it
turns
out.
I
don't.
A
This
is
my
organization
on
gitlab,
dot,
git
and
the
gitlab
pages
domain.
Now,
where
I
was
confused,
is
this
is
a
host
for
every
single
project
used
by
every
single
project
inside
the
gitlab
org
organization.
So
how
is
gitlab
pages
able
to
discern
that
this
is
actually
going
for
my
project?
I
mean
in
this
url
it
does
it
by
identifying
this
whole
path.
This
is
a
path
that
points
to
my
project
and
in
here
I
can't
give
the
whole
path
it.
It
won't.
A
A
A
That
I've
used
for
fastly
and
just
add
it
here,
and
I
don't
even
need
to
go
through
the
all
the
validate
steps,
and
this
is
maybe
something
that
we
need
to
to
improve
because
now
this
domain,
some
jam
stack,
is
connected
with
this
with
his
pages
project,
and
if
I
go
here
and
activate
this
service
in
fastly,
I
can
now
access
this
url.
It
will
probably
take
a
little
while
again.
A
A
You
see
that
the
us
has
good
response
times,
but
everywhere
else
is
rather
slow.
If
I
load
that
page
through
fastly
the
first
time,
the
initial
call
is
even
slower
than
calling
it
with
pages.
The
reason
is
because
fastly
needs
to
connect
to
pages
first
to
build
their
cache,
but
once
it's
cached
here
subsequent
load,
it's
much
much
much
faster
and
that
is
across
the
world,
because
fastly,
obviously
living
on
the
edge
a
proper
edge.
A
For
example,
just
having
this
use
case,
we
could
instead
of
having
fastly
connect
to
the
pager
server.
We
could
deploy
the
com
compiled
applications
to
the
edge
and
we
get
rid
of
the
caching
steps,
so
the
initial
load
would
be
much
faster
as
well
or
looking
even
further
in
the
future
is
we
could
create
services
or
allow
services
that
build
on
demand?