►
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
B
A
B
B
The
48
of
you
who
are
watching
now
they're
just
trickling
in
there's
no
handsome
in
this
week,
so
we
don't
get
his
amazing
power
to
reach
out
into
the
ether
webs
and
how
people
come
to
the
show.
It's
just
us,
but
that's
all
right
and
we
had
a
very
successful
show.
Last
week
the
blazer
stuff
went
off
was.
A
B
B
Blazer
is
so
hot
right
now
I
think
we
did
a
pretty
good
job
of
promoting
it.
We
had
a
lot
of
people
went
like
240,
odd
people
watching
live
at
one
point,
and
we've
had
4,000
views
of
that
episode
in
a
week,
which
is
you
know,
back
to
sort
of
mid
last
year,
numbers
which
is
fantastic
back
to
the
handsome
in-stream
handsome
and
channel
numbers.
So
we're.
B
I
will
say
what
we're
going
to
be
doing
for
the
next
few
weeks,
which
I
think
I
mentioned
two
weeks
ago.
Is
that
the
two
one
release
we've
been
talking
about?
A
lot
is
imminence
coming
along
I'll,
give
a
status
update
in
a
little
bit,
and
so
we're
going
to
be
using
the
show
to
try
and
highlight
sort
of
you
know
feature
areas
of
to
one.
So
today,
I'm
gonna
try
and
give
a
very
quick
look
at
the
some
of
the
little
improvements
we've
made
in
razor
pages
into
one.
B
So
I'll
do
that
today,
there's
not
many
of
them,
so
I
won't
take
long.
Next
week,
we've
got
Xavier
on
to
look
at
the
ability
to
put
razor
in
a
class
library
and
sort
of
the
changes
to
the
razor
compilation
system
that
we've
done
in
2.1,
so
that'll
be
super
cool
and
then
I
think,
two
weeks
after
all
week
after
that,
I'm
supposed
to
be
going
over
I
think
the
the
HTTP
by
default
and
the
HCBS
configuration
improvements
that
we've
made
and
the
cookie
consent
and
stuff
like
that
in
the
templates.
B
Then
the
week
after
that,
I
think
we
have
Ryan
Novak
back
on
from
the
MVC
team
and
he
was
going
to
go
through
the
API
controller
attribute
and
the
improvements
that
that
brings
including
an
API
result
of
T
the
conventions
around
Auto
validation
of
model,
state
bedroom
errors
for
DD
serialization
errors
on
input.
Things
like
that
stuff,
we've
done
for
web
api's,
basically
and
then
I
think.
It's
me
again
talking
about
something
and
then
after
that,
like
all
the
way
through
into
March,
something
is
Sarab.
B
Who
will
be
talking
about
the
improvements
to
the
HTML
core
module
that
is
hosting
a
Shamir
core
in
IAS,
so
look
at
the
in
process
hosting
model
and
how
people
can
try
that
out
in
preview
one
because
it
will
not
be
on
by
default
in
preview.
One
you're
gonna
have
to
sort
of
do
some
work
to
use
it,
but
we
would
like
people
to
try
and
let
us
know
before
we
ship
the
preview
to
so
anyway.
The
next
we
have
one
lined.
A
B
A
B
A
A
So
you
mentioned
blazer,
here's
the
deep
in
dec,
in-depth
technical
introduction,
Steve
Sanderson
wrote
up,
and
here
he
kind
of
explained
all
this
stuff,
here's
how
it
actually
works.
Here's
how
it
loads
stuff,
you
might
not!
You
know,
realize,
like
I
naively
at
the
very
beginning
of
this
first
time,
I
poked
at
it
was
thinking
everything
was
getting
turned
into
a
web
assembly.
No,
that's
not
true.
The
web
assembly
is
loading
and
running
your
your.
You
know
your
msi
LOL.
A
Yeah
so
I
mean
just
all.
The
kind
of
this
is
everything
you
need
to
know
the
deep
kind
of
technical,
here's,
how
it
loads
and
stuff
he
also
kind
of
towards
yam,
goes
into
well
things
like
JavaScript,
typescript
Interop,
which
is
very
cool,
and
then
talking
about
some
optimization.
You
know
here's
things
that
can
happen
in
the
future
as
far
as
optimizing.
So
just
you
know,
this
is
like
you
mentioned.
It's
very
cool.
A
First
of
all
that
he
he
kind
of
does
a
full
walkthrough,
here's
how
I'm
creating
the
motor
context
and
model
and
stuff
he
talks
about
how
he's
mapping
the
you
know,
things
like
on
get
on
post
to
credit
operation,
x'
that
make
sense
for
ajax.
So
you
know
creating
things
like
on
get
select
all
on
gets
like
by
ID
and
then
he's
using.
B
The
named
handler
feature,
you
know
exactly
what
the
what
it
was
intended:
the
ability
to
be
able
to
have
multiple
handlers
that
you
know
you
can
different
differentiate
per
command.
There's
a
couple
of
things.
I
would
you
know
I
would
tweak
about
this
if
I
was
to
give
feedback.
But
overall
it's
it's
it's.
You
know
it's
a
good
explanation
and
a
good
example
of
how
you
could
use
the
razor
pages
primitives,
for
you
know
to
send
them.
You
can
do
sort
of
other
stuff.
B
With
you
know,
we
don't
have
a
first-class
JSON
response
type
because
we're
like
you
know
it's
not
really
idiomatic
to
return
Jason
from
your
page
model,
but
you
can
because
any
action
result
will
get
executed.
So
you
can
absolutely
have
an
AJAX
endpoint,
essentially
that
you
call
from
JavaScript
and
return
Jason
it'll
just
work
and,
as
you
can
see
here,
it
does
and
you
can
give
all
the
code
in
sort
of
one
place.
So
it's
kind
of
cool
one.
A
B
Reason
that
I
think
the
only
reason
is
because
he's
genericized
kind
of
the
handler
he
hard-coded
the
handler
name
suffix
into
the
client
side,
so
he's
passing
through
the
named
handler
as
a
Christian,
parameter
and
I'm,
assuming
that
he
didn't
want
to
have
to
differentiate
those
operations
which
required
a
named
Handler
and
those
that
didn't
because
on
delete
would
just
execute
if
a
delete
verb.
You
saw
right.
B
To
insert
handler
equals
blah
at
the
top,
their
handler
equals
insert
hand
goes
blah.
Now
it
turns
out.
That's
actually
not
super
generic
he's
done
it
in
each
button
handler,
so
he
could
have
done
it
that
way,
but
then
it
wouldn't
be
yeah
potentially
consistent
and
I
mean
either
it's
fine
I
think
that's
all
just
like
whatever
the
person
decides.
None
of
that
is
something
you
can't
change
or
or
have
an
effect
on.
So
that's
just
opinion
at
that
point
very.
B
The
other
thing
I
would
change.
The
thing
I
was
talking
about
was
he's
passing
the
handler
through
as
a
query
string
argument
and
the
client
side
is
hard
coding
that
using
the
routing
and
the
named
named
rat
parameters,
you
can
have
that
you
know
go
through
as
a
segment
part
of
the
URL
in
as
a
path
rather
than
it
has
the
as
a
hand
as
a
crucian
variable.
B
Both
will
like
if
you
pass
handler
through
the
query
string-
and
you
marked
it
in
the
segment
in
your
route
as
optional,
it
should
still
dispatch,
but
you
can
set
it
up
such
that
it
would
all
be
sort
of
it
would
just
honor
what
you
had
configured
the
page
to
do,
and
rather
than
hard
coding
the
urls
in
Java
scope.
What
I,
typically
like
to
do
is
omit
them
into
the
HTML
in
data
data
variables.
You
know
using
the
URL
helpers.
B
So
again
they
always
honor
what
your
router
is
set
up
to
do,
and
then
you
just
you
harvest
those
from
the
JavaScript
by
grabbing
them
out
of
the
data
variables
and
then
use
them
that
way.
So
you
don't
have
to
do
with
this
hard-coded
in
your
all
stuff
and
that
you
know
there's
nothing
wrong
with
this
approach.
Cool.
A
A
Ok,
so
he
went
through
and
use
that
and
then
showed
how,
by
just
configuring
a
lot
of
it
is
just
saying
you
know
turn
it
on
there
are
some
worries
went
through
and
set
specific.
You
know
specific
options
for
that
then
went
through
and
then
you
know,
everything's
green,
so
I
mean
there's,
there's
definitely
more
detail
here
and
he
does
link
out
to
specifically
why
you
would
want
to
that's
one
thing
that
I
always
like
with
his
post
is.
A
He
gives
a
lot
of
kind
of
footnotes
at
the
end
so
towards
the
end
of
this
he's
got
plenty
of
links.
There
was
also
some
good
discussion
on
Twitter,
I,
think,
buried,
Oren's
and
other
folks
commented
back
and
said:
hey
I,
don't
use
this
or
you
know
I
recommend
that
so
he
went
through
and
made
several
updates
afterwards
and.
B
B
Moving
to
marketing
pages
that
aren't
HTTP
is
non
secure
if
you're
doing
data
entry
and
then
in
2
1,
as
I've
mentioned,
where
we're
going
to
support
HCBS
by
default
on
new
projects,
and
we
have
some
new
middleware
and
some
new
experiences
that
make
that
better
over
time.
I
think
you'll
see
us
do
more
for
that
as
well.
Some
of
these
things
are
a
little
hard
to
do
by
default
because
they
require
you
to
think
carefully
about
things
like
content,
security
policy
and
some
of
the
other
ones
they're
require.
B
You
know,
have
an
effect
on
the
features
of
your
application.
If
you
don't,
if
you're,
not
aware
of
them,
and
so
just
doing
them
by
default
out
of
the
box
with
a
default
policy,
can
be
a
little
intrusive.
But
we
continue
to
look
at
ways
that
we
can
just
improve
and
ratchet
up
the
default
security
status
of
all
new,
a
suite
core
apps.
The
first
one
is
a
spirit.
Core
is
HBS
with
HSTs
into
one
and
we'll
continue
to
tune
to
look
at
what
things
that
we
can
do.
Moving
forward
cool.
A
Ok,
so
here's
one
from
Jeremy
Campbell
and
he's
writing
here
specifically
about
blue
milk,
so
he's
always
got
interesting
names
for
things.
A
blue
milk
is
an
OSS,
does
a
few
things.
It
is
an
inversion
of
control
tool.
It's
also
a
you
follows.
A
lot
of
conventions
is
pretty
similar
to
structure
maps.
A
Well,
there's
a
lot
here.
This
is
pretty
interesting,
so
here
he's
got
some
kind
of
standard
default
conventions
and
then,
based
on
that
he'll
generate
code
for
you
and
work
with
IOC,
so
I,
you
know
honestly,
I've
read
through
this
I
would
need
to
play
with
it
a
bit
more
to
kind
of
understand
in
more
detail,
but
I
do
like
how
Jeremy's
working
to
build
both
like
opinionated
convention
based
things,
but
he
worked
specifically
so
that
this
integrates
well
with
the
asp
net
core
di
system
right.
A
So
this
this
is
built
for
kind
of
a
drop-in.
You
know
here's
how
you
can
use
it
with
asp
net
core,
pretty
cool
atom
store.
This
is
pretty
straightforward
here.
He
has
built
out
something
using
web
hooks
and
here
he's
built
at
a
sample
using
Azure
alert
web
hooks.
So
he
shows
how
to
configure
a
shower.
A
B
B
A
B
A
And
you
know,
a
lot
of
this
is
like
how
do
you
feel
about
it?
What
do
you
like
you
know,
but
I
I
definitely
see
the
I
do
tend
to
try
and
get
rid
of
strings
where
I
can
without
going
overboard,
because
if
a
string
misspelling
is
something
you
won't
catch
until
you
know
you
hit
a
unit
test
or
whatever
right.
So
the
idea
here
is
replacing
HTML,
dot,
partial
partial
and
then
in
quotes
dinner
form
with
something
like
this.
So.
B
A
So
what
this
is
doing,
this
is
our
for
MVC.
This
is
a
Rozlyn
powered
code
generator
and
the
idea
is
that
this
you
have
a
command
that
you
run
it.
So
this
is
a
PowerShell
command
and
it
generates
code
and
then
so
it's
not
using
tt4
templates,
it's
using
a
Rozlyn
powered
generator
and
then
it'll
give
you
kind
of
hard-coded
things
that
you
can
then
strongly-typed
refer
to
things
as
you'll
see
here.
This
is
a
small
scroll
bar
here.
He's
got
tons
of
you
know,
there's
yeah
documentation
here
so
I.
B
Wonder
how
it
this
will
work
alongside
the
new
razor
build
stuff.
That's
going
to
happen
into
one
where
we're
going
to
default
to
building
razor
at
build
time.
It'll
still
be
a
sort
of
a
companion
assembly.
It
won't
be
in
your
application
assembly,
so
you
still
can't
reference
it
used
by
type
but
yeah.
This
is
an
interesting,
interesting
approach.
Yeah.
A
Cool
all
right,
this
is
coming
up.
This
Friday
Jeff
Fritz
had
a
thing
as
he
started.
His
livestream
he's
been
doing
live
streaming
several
days
a
week.
I
think
he
just
did
one.
This
morning
today
he
was
looking
at
the
difference
between
partial
views
and
view
components,
but
he's
been
doing
a
lot
of
just
different
life
coding
is
Pinet,
showing
people
how
to
you
know,
do
stuff
with
get
and
github
all
kinds
of
stuff.
So
he
kind
of
issued
a
challenge.
He
said
hey.
If
we
hit
500
followers,
we're
gonna,
give
a
free.
A
You
know
big
SP
net
core
workshop
yeah,
so
after
hitting
that
here's
here's
the
thing
for
that.
So
that
is
this
Friday.
He
talked
me
into
co-presenting
some
sessions
with
him.
We've
got
Shane
boy,
we've
got
Julie
Lehrman,
he
said
possibly
some
others
I,
don't
know.
If
he's
you
know
tried
to
tried
to
talk,
you
know
folks
on
the
team
into
it,
but
definitely
he's
you
know.
There's
this
will
be
a
lot
of
fun.
So
there's
more
details
coming
come.
B
A
A
So
looking
forward
to
that
and,
of
course,
it'll
be
recorded
because
it's
going
out
on
the
livestream
and
then
it'll
be
up
on
this
stretch
and
mixer
stuff
later
all
right,
Andrew
Locke
continuing
now
I
will
totally
confess
some
of
this
seems
to
be
going
over
my
head
today,
but
what
he
did
he
previously
had
looked
at.
Optimizing
docker
builds
using
cake
and
he
was
he
was
talking
about
optimizing
it
so
that
he
was
pushing
out
to
docker
hub
and
and
doing
his
little
kind
of
repeated
work
as
possible
mm-hmm
here.
A
What
he's
doing
is
in
order
to
optimize
that
he
wants
to
push
out
the
CS
proj
files
only,
but
not
the
rest
of
the
code
and
part
of
the
reason
for
that
is
so
that
when
he
makes
a
change
to
a
single
file
like
a
CS
file,
it's
not
going
to
have
to
rebuild
that
docker
layer
that
it's
able
to
get
as
much
docker
reuse
as
possible
mm-hmm
and
what
he
found
as
he
was
trying
to
do.
This
is
he
first
tried
the
naive
solution
and
just
tried
to
do
a
simple
copy
copy.
A
A
Why
not
is
is
tearing
other
things,
so
he
said
you
know
why?
Don't
we?
Why
don't
we
tar
everything,
including
the
the
solution
and
then
new,
get
config
files?
So
that's
what
he's
doing
there
and
then
he
shows
here's.
My
you
know
completed
build
script
and
docker
file,
so
I
yeah
I
thought
that
was
kind
of
interesting
and
just
a
way
as
you
go
through
and
optimize
your
build.
B
A
And
I'm
just
reminded
going
way
back
when
we
first
started
doing
these
stand-ups
I,
we
said
like
hey
this
stuff's
open-source
get
involved
and
one
of
the
people
that
heard
that
was
Ben
Adams
and
that's
actually
how
he
got
inspired
was
off
this
stand-up.
So
here
you
know
the
next
Ben
Adams,
please
step
up
for
grabs,
and
this
is
the
the
whole
idea
with
up
for
grabs
tags.
Is
these
are
supposed
to
be
issues
where
they're
easy
to
kind
of
get
it
get
started?
It's
not
it's.
B
It's
more
than
that
too.
It's
also
things
that
we
have.
We
don't
consider.
I
can
use
my
words
carefully
or
I
won't,
and
then
people
can
just
infer
what
I
mean
like
they're,
obviously
not
important
enough
for
us
to
just
do
like
we
are,
but
we
have
other
work
that
we
are
prioritizing
above
these
items
for
the
engineers
here.
Otherwise
we
would,
just
you
know,
fix
these
bugs.
B
It'll
make
the
few
people
that
it's
annoying
happy,
but
there's
just
not
affecting
enough
people
that
warrants
us
going
off
and
spending
time
on
it.
If
that
makes
sense,
and
if
someone
does
submit
a
PR
for
it
and
fix
it,
that's
very
low
chance
of
it
having
a
side
effect
that
would
then
adversely
impact
a
whole
bunch
more
people,
because
you
know
sometimes
we
have
bugs
that
affect
a
small
amount
of
people.
It
affects
them
severely.
B
A
A
You
know
like
I've
I've,
had
simple
things
in
an
open-source
repo
where
I'm
like
yeah
put
it,
make
it
up
for
grabs
and
somebody
digs
into
it,
and
it's
like
you
might
not
have
realized.
But
there's
all
these
implementations
and
I've
run
into
this
too,
where
it's
like.
Oh
I'll,
just
fix
this
thing
in
this
open
source
library
and
like
six
hours
later,
I'm
like
regretting
I
ever
heard.
A
All
right,
one
last
one
that
I
thought
was
interesting
and
I
forget
I
forget
actually
this
person's
name.
Let
me
see
if
we
can
find
it
quickly.
I,
don't
know
this
gosh
anyhow
I
saw
this.
I
saw
this
linked
out.
This
was
on
yeah,
all
right,
so
you're
by
RT
for
now,
but
this
was
on
the
front
page
or
of
hacker
news
this
past
week,
some
interesting
stuff
in
here
this
is
a
kind
of
opinionated
list
of
here's,
some
things
that
this
person
does
for
new
asp
net
core
projects
as
I
went
through
I
thought.
A
You
know
yeah.
Some
of
this
is
stuff
I
turn
on
some
of
it
was
stuff.
I
disagreed
with
and
I.
Think
that's
a
good
way
to
look
at
this
is
this
is
one
person's
opinion
of
things
that
they
like
to
do?
You
know
kind
of
by
default.
So
some
things
like
you
know,
validating
model
state
automatically,
which.
A
Know
some
some
different
things
about.
There
are
some
new
get
packages,
for
instance
that
they
pull
in
specifically,
this
is
their
own
sorting,
filtering
pagination,
you
get
packaged
so
kind
of
cool
and
then
also
their
own
authentication.
Some
of
these
things
are
interesting
where
it's
like,
while
I
rewrote
my
own
everything
you
know,
but
I
felt
like
this
was
kind
of
a
good
mix
of
some
of
those
things
that
they
liked
doing,
and
then
some
other
specific,
like
I,
just
like
to
turn
this
in
bud
on
default.
So
funny.
B
I
mean
I
liked
it
there's
a
lot
of
very
real
stuff
and
it's
funny.
These
are
always
the
post.
I
get
a
lot
of
comments,
because
people
look
at
this
and
go
I.
Don't
do
it
that
way,
I!
Do
it
this
way.
Why
would
you
ever
do
that?
We
don't
do
that
in
our
apps,
but,
like
a
lot
of
this
stuff
is
literally
yeah
when
we
built
this
app
like
there
was
one
section
there
about
changing
the
user
during
development
and
I.
Remember
working
in
all
types
of
apps,
so
we'd
have
these
companion.
B
You
know
these
companion
tools
that
you
would
write
specifically
just
to
help
you
do
stuff
when
you're
coding
right,
you
know
you
would
be
able
to
like
set
the
entire
site
to
run
as
this
one
user
or
rewind.
This
shopping
cart
three
steps
from
wherever
it
currently
is
or,
like
you
know,
all
these
crazy
utilities
that
you
drive
as
a
div
just
to
help
you
get
your
stuff
done
quicker,
and
some
of
these
are
just
like.
B
Maybe
you
think
think
of
that
so
yeah
like
for
the
type
of
apps
that
I
usually
write,
these
things
make
a
lot
of
sense
and
the
other
thing
I
also
used
to
say
a
lot
to
people
was
you
know
it's
offering.
The
software
craftsmanship
is
something
that
people
talk
about
a
lot
and
you
know
there's
been
various
up-and-down
movements
around
that
and
but
I
was
always
a
bit
advocate
of
you
know.
If
you
can
give
up
a
few
things
that
are
like
you
know,
I
really
like
things
to
work.
B
If
I
could
just
change
it
to
do
this,
I
would
be
I
to
sleep
so
much
better
or
you
could
just
not
die
on
that
hill
and
just
do
it
the
way
the
framework
wants
you
to
do
it
and
you
never
have
to
worry
about
it
again.
Just
like
the
fact
that
the
framework
wants
you
to
live
this
way
and
if
you
do
you're
gonna
save
yourself
a
crapload
of
time,
that's
not
always
true,
obviously,
but
like
especially
in
web
forms.
B
Like
it,
it
was
more
that
because
they
started
back
then,
when
there
wasn't
those
things,
it
was
very
difficult
for
them
to
to
move
to
them
because
they
just
never
did
that
work.
I
think
the
same
is
true
of
any
framework
right.
Some
people
choose
to
not
like
things
and
they
implement
their
own.
That's
fine!
That's
totally
fine,
but
I.
Do
encourage
people
to
take
at
least
have
a
look
have
a
look
at
the
way
it
wants
you
to
do
it
think
very
hard
about
whether
it's
really
worth
you
not
doing
it.
A
That's
that
I
I've
come
to
deeply
appreciate.
Some
developers
are
like
that
because
they
end
up
doing
amazing
things
like
sometimes
somebody
an
example
there
we
talked
about
earlier
Jeremy,
Miller,
I,
love
that
he
continues
to
kind
of
push
in
different
directions
and
work
towards
conventions
he
likes
and
build
out
things
that
tons
of
people
have
used.
But
there
is,
for
you
know,
Joe
developer,
Joe
development
shop,
there's
a
cost
to
that
right.
B
A
A
B
Right
I'm,
sorry,
I,
said
I
would
give
people
a
2-1
update
so
yeah.
We
when
I
wasn't
on
the
show
last
week,
I
think
they
would
have
talked
about.
We
put
out
the
two
on
road
map
blogposts
last
week.
Did
1ef
did
one
and
he's
been
at
core.
Did
one
and
we
put
a
fair
bit
of
effort
into
trying
to
cover
most
of
the
improvements
that
are
planned
for
or
actually
have
already
been
done
for
the
upcoming
2
1
preview?
With
the
you
know,
because
a
couple
of
code,
samples
and
things
like
that?
B
So
if
you
haven't
read
that
post,
please
go
and
do
so
now
all
those
posts,
I
should
say
some
highlights.
You
know
global
tools
for
dotnet
core
there's
the
command
line.
It's
gonna
be
cool,
so
you
can
now
very
similar
to
NPM
s
global
tools
feature
you
can
write
a
dotnet
console
app.
You
can
pack
it
up
as
a
tool
the
NuGet
package
and
then
anyone
else
with
no
core
install
can
install
that
on
their
machine
from
the
command
line,
and
it
goes
under
the
path
for
that
user
and
then
they
can
invoke
that
tool.
B
Any
time
I
will
be
using
that
for
some
of
the
tools
that
we
ship
so
like
dotnet
watch,
which
is
the
command
line,
a
project
watching
tool
that
you
use
when
you're
doing
sort
of
iterative
development
and
the
command
line
or
with
vs
code,
is
now
going
to
be
a
global
tool.
In
2.1,
we
have
a
new
global
tool
to
do
with
managing
developer
certificates
for
the
new
HTTP
experience,
which
we'll
look
at
in
a
couple
weeks.
I
think
it's
another
one
that
we
just
watched
over.
B
Probably
a
couple
I
think
there's
like
a
seek
the
the
tool
that
we
use
to
set
up
the
sequel
tables
for
our
distributed
caching
provider
for
sequel,
as
now
a
global
tool
it's
after
etc.
Then,
of
course,
we
want
folks
to
build
their
own
tools,
but
today
I
was
gonna.
Just
do
a
very
quick
little
bit
of
demos
on
the
razor
pages
improvements
there
aren't
many
of
them.
B
Let
me
get
my
screen
sharing
going,
but
I
wanted
to
quickly
focus
on
that
today
and,
as
we
said
up
front,
people
will
be
coming
into
demo
other
stuff.
So
this
is
the
blog
post.
I
was
talking
in
today's
been
a
core
blog
post,
and
this
is
the
section
on
razor
pages
improvement.
So
I
was
just
gonna
quickly
sort
of
demonstrate
those
so
over
here,
I
have
a
2-1
build
installed.
B
I
was
actually
hoping
to
sneakily
share
with
the
hundred
or
so
people
watching
how
to
get
this
build
because
there
are
some
folks
using
to
one
now,
it's
quite
difficult
doing
it
the
traditional
way.
We
have
a
new
build
process
that
all
the
teams
are
working
on.
That
is
gonna
help,
hopefully
solve
a
lot
of
our
sort
of
reliability
and
coherence,
issues
with
getting
and
reliably
getting
a
build
sort
of
everyday
and
we've
been
using
that
internally.
B
But
we
haven't
yet
Apple
sort
of
updated
the
public
links
that
people
are
used
to
using
to
get
Knightley's
for
the.net
core
CLI,
so
I
mean
there
are
places
you
can
get
them,
but
the
URLs
are
Horrible's
like
this
is
the
URL
for
the
latest,
one
that
I
just
installed
right.
So
this
is
product
build
201,
800
209,
which
you
can
probably
guess,
means
it's
February.
The
9th
third
build
that
day
and
then
some
other
number
on
the
end
of
it,
and
so
you
know
this.
B
This
link
here
is
literally
to
the
windows
installer,
and
you
can
you
know
you
can
see
that
the
URL
here
it
is
public
like
if
you
can
find
it,
you
can
install
it.
It's
just
that
this
thing
is
not
in
a
branch
like
it's
literally
in
a
random
commit
hash.
So
you
know
good
luck
finding
it,
but
this
will
be
updated
shortly,
such
that
it's
it'll
be
easier
for
people
to
consume
these
bills
and
the
great
thing
about
these
bills
like
I
say
this
is
the
are
coherent.
B
You
don't
have
to
worry
about
getting
a
nice
phonetic
or
set
of
packages
from
here,
and
it
said
a
templates
from
there
and
a
dotnet
called
runtime
from
there
and
then
a
matching
CLI
in
SDK
from
over
there,
like
you,
literally
install
this
thing
on
your
phone
for
your
platform
of
choice
and
everything
inside
that
was
built
together
right.
So
the
ace
minute
core
templates
use
the
right
runtime,
which
use
the
right
a
spin,
your
core
runtime,
which
depend
on
the
right
versions
of
the
packages
it
has.
B
The
fallback
folder
slash
new,
get
cache
correctly,
so
you
don't
even
need
to
feed
if
you
just
want
to
use
the
templates
and
all
that
stuff
just
works,
so
this
is
going
to
make
it
much
easier,
and
you
know
this
will
build
everyday.
The
whole
idea
is
that
this
will
be
available,
so
you
could
always
pick
up
the
latest
build
so
I
have
that
installed.
I've
use
quite
a
few
of
these
bills
now,
and
this
is
what
we're
doing
your
validation
or
a
verification
on
the
hope
quickly.
I'll
interject
is
that
we
will.
B
We
are
locking
down
this
release
right
now,
so
our
hope
is
that
we're
going
to
be
able
to
get
this
out
to
folks
in
the
next
couple
of
weeks
we're
pushing
really
hard.
We
still
got
a
couple
of
issues
that
we're
working
through,
but
you
know
that's
our
goal.
That's
I
just
so.
People
have
an
understanding
of
where
we
are
right.
Now
we
are
trying
to
get
this
out
in
the
next
couple
of
weeks,
so
hold
on
you'll.
Be
able
to
use
these
bits
on
official
channels
very
soon.
B
So
I've
done
I've
created
a
new
to
one
app
here.
If
I
look
in
the
CS
prize,
I
can
prove
that
so
you
can
see
here,
I
am
using.
Oh,
I
feel
I
targeted
full
framework,
which
I
didn't
mean
to
do.
Let
me
change
that.
I
was
testing
some
full
framework
scenarios
yesterday,
and
so
it
must
have
stuck
there
see.
I
did
a
trace
full
framework.
So
let's
do
that
again,
one
dotnet
core,
because
I
actually
wanted
to
see
it
using
have
a
look.
B
You
can
see
pretty
well
from
what
I
can
tell
so
we
now
support
pages
in
areas,
so
I
could
go
ahead
and
create
an
areas
folder
in
here
and
then
watch
the
CS
product.
Eight,
that's
not
particularly
fun,
so
we
can
close
this
name,
I,
guess
and
we'll
close
the
getting
started
page
and
then
inside
here
I
can
add
a
pages
folder.
Okay,
so
remember
areas
are
kind
of
like
sorry,
I
should
add
a
named
folders
or
a
named
area.
B
B
So
I'll
need
to
obviously
fix
my
model
because
I've
copied
code,
that's
in
the
wrong
spot,
so
we'll
say
this
is
now
in
dot
areas,
dots
demo
pages.
This
thing
still
won't
find
it
because
I
haven't
got
a
view.
Imports
that's
setting
the
namespace
of
this
file,
so
I
could
do
a
couple
of
things.
I
could
just
fully
qualify
this
index
model
in
here
or
I
could
add
a
view.
B
Imports
like
the
one
we
have
in
the
root
that
you
can
see,
sets
the
namespace
to
the
folder
I'll,
just
fully
qualify
it
for
the
sake
of
making
this
simple,
so
I'll
just
copy
that
main
space.
Put
that
on
the
front
of
that.
So
now
you
can
see
it
finds
that
index
model
alright.
So
in
theory
now
and
I'll
get
rid
of
all
this
noise
so
that
we
can
I
should
be
able
to
run
this
and
navigate
to
this
page
in
that
area.
So
this
will
behave
exactly
the
way.
B
I
hope
you
would
expect
it
to
so.
Once
the
app
is
running
it's
going
to
launch
to
the
home
page,
you
can
see
at
htps
by
default,
which
we'll
talk
about
in
a
couple
weeks.
I
should
be
able
to
now
go
to
the
demo
area
and
go
and
on
theory
yeah
there.
You
go
so
like
over
the
demo
area
and
it
goes
yep
the
index
for
that
works,
and
this
is
from
my
demo
area.
So
now
we
get
pages
in
areas
just
as
working
the
way
as
I
said,
I
hope
you
would
expect
it
with.
B
Else
do
we
support.
There
was
my
list
of
things
to
cheat
support
for
pages
shared
now.
This
is
quite
subtle.
It's
actually
a
little
bit
more.
That's
going
on
with
support
the
pages
shared
than
first
meets
the
eye,
so
everyone
will
know
that
you
know
we've
had
MVC
has
this
convention
of
in
the
views
folder,
you
can
have
a
shared
folder
and
then,
when
you
ask
MVC
to
find
a
view
or
a
partial,
by
name
it
will
fall
back
to
the
shared
location.
As
the
last
place,
it
looks.
B
Okay,
so
it'll
look
in
the
very
specific
places
yet
first,
based
on
the
context
in
which
you
said,
please
might
find
me
the
the
fou
view
and
if
it
doesn't
find
them
there,
it'll
find
them
in
the
shed.
Now
we
didn't
have
that
same
support
for
pages,
but
you
know
very
quickly.
After
we
shipped
pages
people
said
I
want
pages,
slash
shared,
so
we
now
support
that,
and
you
can
see
in
here.
B
We've
got
a
shared
folder
in
a
template
and
it
has
the
cookie
consent
partial
in
it,
which
means
that
if
you
have
an
MVC
view,
for
example,
that
is
looking
for
a
partial
of
this
name.
It'll
first
look
inside
the
view
hierarchy,
including
the
shared
folder,
but
then,
if
it
doesn't
find
that
it'll
go
and
look
inside
the
pages
hierarchy
and
look
inside
its
shared
folder,
so
now
you
can.
It
just
increases
the
ability
to
share
assets
between
sort
of
traditional
views
and
razor
pages,
because
you
now
have
that
shake
convention
there
now.
B
The
other
thing
it
does,
though,
which
is
kind
of
cool.
If
I
can
add
a
full-blown
page
in
here,
so
I
can
add
a
razor
page,
I
can
say,
share
the
page
and
I'll
go
ahead
and
just
add
a
heading
again
and
just
cheat
and
say
this
is
a
shared
page
now,
and
so
this
isn't
in
the
normal
location,
it's
in
the
shared
folder
and,
as
I
said,
usually
you
treat
the
shared
folder
as
a
fallback.
So
what
do
we
do
with
an
asset?
That's
actually
routable.
Remember
pages
are
routable
they
get
served.
B
So
how
does
this
work?
Well,
this
just
acts
as
if
it
wasn't
in
the
shared
folder,
okay,
so
now
I
could
should
be
able
to
go
to
my
app
and
I
should
be
able
to
actually
just
browse
a
shared
page.
Perhaps
rebuilding
and
other
work
it
explodes,
because
I
did
something
wrong.
What
did
I
do
wrong?
One
of
the
devs,
hopefully,
is
watching
a
little
burst
in
here
and
tell
me
what
I
did
wrong
will
send
me
a
text
because
we
fought
to
rebuild,
because
that
always
fixes
everything.
But
I
did
this
demo
in
London.
B
A
B
Shouldn't
be
because
areas
are
like
a
part
of
the
or
another
thing
as
well
like
they
have
their
own
part
of
the
of
the
root
right,
so
they
have
a
thing
called
demo.
It
shouldn't
matter
at
all.
It
should
have
no
impact
on
any
other
page
that
you
have
so
I'm,
not
really
sure
what's
going
on
there,
but
one
thing
that
you
can
do
with
areas.
B
So
the
extension
of
this
demo,
which
won't
work
now
because
this
doesn't
work,
is
that
the
fallback
logic
for
finding
pages
supports
overriding,
and
so
you
can
do
things
like
have
a
page
in
an
area
called
demo.
And
then,
if
that
was
brought
in
by
a
librarian,
Harvey
is
going
to
show
this
stuff
next
week.
You
can
override
that
in
your
app,
based
purely
on
the
fact
that
the
part
of
the
the
paths
to
that
view
match
okay,
and
then
we
have
a
very
simple
algorithm.
B
That
says
that
when
we
try
and
find
a
view,
if
we
find
two
assets
with
the
same
name
at
that
path,
then
we
will
figure
out
which
one
wins
based
on,
basically,
which
one
is
closer
to
your
app.
So
what
it
effectively
lets
you
do
is
have
pages
or
partials
of
views
in
a
library,
we're
in
a
class
are
being
a
new
get
package
or
now.
B
In
your
app
by
just
popping
them
in
your
app
in
the
pages
folder
or
the
area
folder,
if
the
library
you
brought
in
had
in
an
area
with
the
same
name
and
then
it'll
win
because
it'll
go
oh
I
have
to,
but
you
have
one
in
your
app
so
that
one
wins
right
and
so
doing
that
you
get
this
really
nice,
simple
composition,
model
for
providing
UI
and
then
over
running
it.
But
how
beers
going
to
do
a
bunch
of
damage
on
that
next
week?
B
So
I
won't
steal
that
thunder,
because
his
demos
will
be
much
better
rehearsed
and
they
won't
fail,
like
my
interests
in
I'm,
so
the
loved
couple
other
things
on
pages
that
I
want
to
show.
You
can
now
bind
all
properties
on
the
page.
So
when
you
do
a
page
that
takes
parameters
so,
instead
of
doing
it
on
get
say,
we
wanted
to
add
an
on
post
here
right
and
then
we
had
a
property
to
support
binding
like
a
customer
or
something
so
as
if
hey
I
can
have
a
public
customer
model
of
some.
B
B
What
I
will
you
have
whatever
you
want
in
here,
though,
you
might
have
your
ID
property,
and
you
have
your
straining
first
name
and
in
your
method
you
would
you
would
you
know
you're
basically
binding
this
thing,
and
then
you
are
copying
from
whatever
was
bound.
You
know,
check
your
model
validation,
then
copy
EEF
model,
whatever
it
is,
I'll
use,
Auto,
mapper
or
something
along
those
lines.
It's
a
yes,
that's
gonna
hate
me
for
doing
that,
so
what
you
can
do
now
previously
the
way
this
would
work
is
you
would
mark
this
as
buying
property?
B
Okay.
So
that
means
when,
when
an
on
get
request
is
made,
please
attempt
to
do
a
complex
model,
binding
operation
based
on
this
type
under
this
property.
Now,
in
some
pages,
you
might
have
more
than
one
of
these
because
you're
doing
named
handlers
the
thing
that
we
briefly
talked
about
before
so
as
well
as
on
post
you
might
have
like
on
post,
foo
or
fee
and
on
post.
B
You
know
bar
to
handle
different
buttons
on
your
pages,
for
example,
or
something
like
that
and
you
might
want
to
bind
different
properties
per
method
or
you
just
might
have
multiple
properties
that
you
want
to
bind
for
whatever
reason.
So
now,
the
very
simple
thing
is:
we
support.
You
know
putting
this
thing
up
here.
So
if
you
want
to
just
bind
everything
on
the
on
the
on
the
page
model,
you
can
do
that
now.
We
initially
had
this
in
the
prototype.
A
few
years
ago,
we've
decided
to
remove
that.
B
We
said
we'd
wait
for
feedback,
we've
got
the
feedback.
People
wanted
us,
so
we've
added
that
support
now,
so
you
can
just
decorate
the
entire
page
as
saying
yep.
I
always
want
to
bind
everything
on
here.
You
know,
don't
make
me,
do
it
property,
a
property
and
then
the
last
thing
that
we've
got
very
strong
feedback
on?
Was
people
wanted
to
be
able
to
have
code
that
ran
before
every
Handler
or
after
every
handler
right?
B
And
so
in
this
example,
here
I've
got
a
you
know
a
whole
bunch
of
handlers,
but
these
there's
no
place
to
put
common
code
other
than
the
constructor
all
right.
So
now
page
models
all
implement
a
page
filter,
so
you
can
just
override
like
on
page
handler
executing
and
you
can
run
some
code
before
any
of
these
things
run
okay
and
then,
similarly,
we
could
override
on
page
handler
executed
and
you
can
run
code
after
and
like
normal,
like
any
other
filter
right,
you
get
a
specific
context
and
you
can
shortcut
things.
B
You
can
change
the
result.
New
inspect
change
mutate
do
whatever
it
is
that
you
want
to
do.
You
could
obviously
put
this
code
in
an
actual
filter
class,
and
then
you
decorate
it
as
an
attribute
on
this
page,
but
if
you
just
want
to
run
code
in
the
page
model
just
for
this
page
before
or
after
you
can
now
do
that
very
easily,
because
all
page
models
implement
our
page.
So.
A
B
Well
also,
like
the
constructor,
you
have
absolutely
no
context
about.
What's
going
to,
X
are
right,
see
this
context.
Tells
you
well
I'm
going
to
execute
this
action.
You
know,
or
this
handler
like
these,
are
the
arguments
are
going
to
be
passed
to
the
executing
handler.
This
is
the
method
I'm
going
to
execute.
You
can
change
it
right
like
this
handler
method.
B
Descriptor
is
mutable,
so
you
can
go
in
here
and
say
well,
it
said
it
was
going
to
execute
the
pon
post,
but
for
whatever
reason,
I
wanted
to
execute
this
other
one
right
and
you
can
do
all
this
stuff
in
a
page
filter
today,
but
now
you
can
do
it
directly
inside
the
page.
If
that
suits
your
scenario.
Ok,
so
again
that
was
feedback.
We
go
back
to
oh
and
that's
something
you'll
know
and
be
able
to
do
into
one.
B
This
feature
already
exists
in
controllers,
like
the
base
controller
class,
implements
a
filter
so
that
you
can
just
execute
some
code
before
a
heart
or
after
any
action
method
that
runs
on
there.
So
it's
basically
supporting
the
same
thing
on
here
now.
The
other
feature
that
is
now
so
is
now
meant
to
be
into
one
that
but
I
just
found
out
today
that
we
watched
it
and
accidentally
it
won't
be
in
preview.
It
will
be
in
preview,
too,
is
the
feedback
that
people
want
to
be
able
to
decorate
these
properties
simple
properties
here.
B
So
this
is
a
bad
example,
because
that's
a
complex
property,
but
if
I
had
something
like
this
all
right,
this
is
quite
common,
so
you
did
like
an
on
get
research
page
and
you
accept
a
filter
or
a
query.
You
may
want
to
say
this
is
required.
We
don't
have
any
support
for
doing
that
today,
because
this
was
analogous
to
an
action
method,
parameter
on
an
NBC
controller,
and
you
can't
mark
parameter
as
required.
Using
the
validation
features
right.
You
can.
B
One
so
now
you
could
actually
do
this
and
I
could
bring
in
the
appropriate
namespace
for
the
day
orientation
and
that
will
now
actually
validate
that
when
you
run
the
model
validation
for
this
page
and
say
well,
if
there's
no
valid
value
for
filter,
then
this
will
throw
knowing
the
little
model
state
dictionary
will
contain
an
error.
Unfortunately,
the
code
to
make
this
work
in
page
model
didn't
land
in
preview.
B
1,
it's
being
done
right
now,
so
it'll
be
in
the
builds
right
after
preview
1,
because
we're
shutting
down
for
preview
1
as
I
said,
but
it
will,
it
has
landed
in
controllers.
So
if
you
want
to
use
validation,
attributes
on
actually
methods
into
one
preview,
one
you
will
be
able
to
do
that
and
you
will
be
able
to
and,
as
I
said
it's
coming
to
pay
trolls
so
yeah
they
were.
The
improvements.
I
wanted
to
show
sorry
about
the
failure.
I'm
gonna
have
to
go
off
and
figure
out
why
that
didn't
work.
B
B
So,
as
I
said,
we'd
only
bind
properties,
even
when
you
mark
them
with
the
attribute.
They
only
get
bound
on
non
get
requests,
but
there
is
a
parameter
to
the
attribute
that
you
can
set,
which
will
make
it
bind
for
all
request
types.
So
again
that
was
just
the
default
behavior
that
we
chose
it's
a
convention.
You
can
change
it
in
the
configuration
of
razor
pages
and
your
startup
code.
You
can
change
that
convention,
but
that's
the
convention.
We
chose
out-of-the-box.
B
I
create
changing
and
that's
all
the
current
parameters
on
the
query
changing
one
in
a
reasonable
way.
Such
a
common
sorting
scenario
think
I,
miss
okay,
I
think
what
he
means
is
that
there
are
scenarios
where
you
do
a
get
or
a
post
to
a
page
with
safe
for
pushing
parameters,
and
you
want
to
render
a
result
but
change
just
one
parameter:
bright,
toggle
or
sending
the
deciding
yeah
like,
for
example,
render
a
link.
I
should
say
that
uses
the
current
thing,
but
changes
a
single
parameter.
B
Yes,
that
that
scenario
is
a
little
cumbersome
today,
you
basically
have
to
reconstruct
that
thing.
That's
a
really
good
suggestion!
Nick!
You
should
log
that
issue
on
the
MVC
repo
and
we
might
even
be
a
few.
If
it's
well
defined
and
fairly
simple
to
do.
We
might
even
be
able
to
put
that
in
preview
too,
so
I'm
not
promising
there,
but
that's
a
good
one.
Iii
know
I
have
struggled
with
that.
One
myself,
he
also
asked.
Is
there
any
work
happening
on
the
SDK
pile
up?
B
The
previews
are
great,
but
the
number
of
SDKs
that
get
installed
really
starts
to
rack
up.
Even
on
just
the
vs
previews
much
less
days,
so
I
think
what
he
is
referring
to
is
that
if
I
go
to
to
like
my
absent
features
section
if
you're
installing
daily
builds
of
SDKs
their
installs
right,
and
so
you
end
up
seeing
them
all
in
your
list
of
stuff-
that's
installed
on
your.
You
know
your
on
your
machine.
B
So
if
I
share
my
screen
again,
we
can
see
what
this
means
so
I
have
up
my
apps
and
features
from
the
Windows
10
settings
app
and
I've
filtered
on
their
core.
So
I
have
you
know,
don't
they
call
one
one?
Five,
the
runtime,
which
comes
with
Visual
Studio
for
the
the
1x
apps
I,
have
these
SDKs
202
and
203,
which
are
again
versions
that
come
with
Visual
Studio.
B
They
were
the
releases
up
and
until
November
last
year,
so
up
until
15,
5vs
I
believe
then
they
switch
to
the
versioning
of
two
one
for
the
SDK,
which
was
very
confusing,
but
we
are
retconning
that
over
time.
So
these
two
one,
two
two
one,
three
two
one
four
these
were
and
you
can
see
they
were
installed
last
year
and
earlier
this
year
these
were
releases
that
happen
as
part
of
our
servicing
for
dotnet
court,
and
then
these
releases
here
to
1/100
to
1/100
and
I.
B
Think
if
you
click
on
them,
you
get
the
full
number
so
273
to
6
versus
735,
for
these
are
the
new
versioning
scheme
that
we're
using
for
servicing
of
Oh
Donna
core,
oh
and
one
300
is
the
version
of
the
SDK
that
contains
dotnet
core
2.1.
Now
it's
all
a
bit
weird
now,
because,
unfortunately
we
did
this
versioning
thing
as
I
say
where
we
went
to
one
for
the
SDK
before
the
runtime.
B
That's
not
happening
moving
forward,
so
when
two
comes
out
the
SDK
will
be
and
the
runtime
will
be,
and
then
the
patch
number,
the
third
number
of
the
SDK
will
has
its
own
versioning
scheme.
That
encapsulates
both
feature
changes
and
patches,
and
so
it
basically
goes
up
by
a
hundred
when
there
is
level
change
or
the
you
know
to
1/100
is
a
patch
for
the
to
Oh
sort
of
but
product.
It's
the
it's.
The
OH
SDK
effectively
and
300
is
the
SDK
for
dotnet
core
2.1
okay.
B
So
we
do
a
patch
for
that
it'll
be
301
and
302.
But
if
we
have
to
add
a
new,
you
know
support
to
a
new
version
of
C
sharp
or
a
new
version
of
new,
get
that
we
go
up
to
1
400.
But
when,
if
dotnet
core
comes
out,
the
whole
SDK
will
move
to
2.2
dot.
You
know
whatever,
so
the
2
1
release
is
always
going
to
be
a
little
weird
because
it
straddles
both
the
o
runtime
and
the
1
runtime.
B
So
you
do
need
a
bit
of
a
decoder
ring,
but
moving
forward
after
that
it
will
get
a
lot
clearer.
So
to
the
next
point,
yeah
I
have
a
lot
of
build
so
I've
installed
what
1,
2
3
4
of
those
new
build
types.
I
talked
about
on
this
machine,
and
so
I
have
you
know
eight,
oh
one,
six,
eight!
Oh
two,
six,
zero,
eight
one,
one
eight
and
eight!
Oh
one,
nine!
All
on
this
machine!
Now,
that's
because
the
the
SDK
installation
behavior
is
side
by
side.
B
That's
that
was
the
choice
that
was
made
as
one
of
the
sort
of
tenant
features
of
dotnet
core.
Is
that
installs
a
side
by
side
so
that
there's
less
there's
always
a
chance
for
you
to
after
you
do
an
install?
If
you
do
find
something
wrong,
you
can
modify
a
project
to
pin
yourself
to
an
earlier
version
of
the
SDK.
You
know
we
all
know
the
runtime
so
side
by
side
right,
there's
a
folder.
It
has
the
runtime
in
it.
B
It
has
a
version
name,
so
you
could
they're
always
side
by
side
and
the
SDK
installs
include
the
runtime,
but
the
SDK
like
the
command-line
tools.
Donate
blood
themselves
are
also
side
by
side
side
by
side
by
default,
but
we
do
roll
you
forward
to
the
newest
one
by
default.
That's
why
you
end
up
with
them
all
in
your
add/remove
programs
list.
B
If
you
pave
your
machine
and
install
again
from
scratch,
you'll
generally
just
install
the
latest
one
and
then
you'll
just
have
that
one
installed
again
right
and
if
you
don't
want
the
previous
ones,
you
can
always
go
and
safely
uninstall,
the
previous
one
after
you've
installed
the
new
ones
and
the
installers
should
be
offered.
You
know
correctly
fingers
crossed
so
that
if
you
want
to
install
a
previous
version,
it
doesn't
you
know
stuff
up
the
new
one
that
you
just
installed,
because
it
does
the
right
ref,
counting
and
all
that
type
of
thing.
So
Nick.
B
If
that's,
if
that's
a
super
concern
for
you,
you
should
be
out
of
one
install
it's
a
support
scenario
of
what
earlier
versions
and
assuming
that
you
don't
have
a
global
Jason
that
pins
a
particular
project.
To
that
specific
version
that
you
want
installed
there
shouldn't
be
any
ramifications
of
that.
But
I
do
appreciate
that
folks,
who
just
want
to
be
on
the
latest.
It
does
result
in
those
things
it's
an
extra
step
required.
You
have
to
uninstall
the
previous
one.
A
B
Lots
of
talk,
no
firm
plans,
there's
a
lot
of
talk
about.
We
talked
about
the
Indy
sono
scenario
we
just
showed
Wright
was
using
in
like
OS
level
installers
to
install
this
stuff,
but
we
know
some
folks
like
to
use
like
on
Linux.
You'll,
often
use
that
guess,
like
their
package
management
system
to
get
it
or
you
might
just
want
tar
files
or
zip
files,
then
you
want
some
command-line.
You
can
use
to
easily
acquire
runtime
updates
or
SDK
updates
or
switch
to
a
different
version.
B
Yes,
it's
being
talked
about
dean
vm
functionally
was
pretty
good.
The
way
it
was
implemented
was
was
was
a
little
less
suboptimal.
I
think
you
know
changed
your
path
and
had
all
these
options
for
global,
changing
versus
current
session
versus
user
and
stuff.
The
biggest
issue
is
when
you
cross
the
streams.
B
Well,
do
you
expect
that
to
elevate
like
do
you
expect
it
to
to
detect
that
you're
running
with
a
version
of.net
that
was
originally
installed
in
Program
Files
on
Windows,
for
example,
and
that
if
you
try
and
install
a
new
runtime
that
it
should
say
well,
you
can't
cuz
you're
not
running
as
and
a
minx
it
what's
to
put
it
next
to
it.
Well
that
won't
show
up
in
ever
new
programs,
because
it
wasn't
built
using
MSI
right,
or
would
it
be
okay
to
have
that
command,
install
runtime
to
a
different
place?
B
But
then,
when
you
say
like
list
me
the
runtimes,
it
has
to
know
about
all
the
places
and
it
might
find
runtimes
right
like.
Oh,
you
have.
You
got
in
your
user
hive
here,
plus
you've
got
the
global
ones,
you've
installed,
and
maybe
you
have
some
run
terms
just
in
the
current
repo,
because
the
Dante
CLI
does
include
some
features
for
like
discovering
things
just
in
the
current
folder
in
the
dot
with
the
dotnet
folder,
it
gets
really
complicated.
B
B
We
think
we're
going
to
jump
from
being
nowhere
close
to
being
a
contender
to
you
know
top
5,
maybe
even
better.
That's
our
hope.
So
that's
looking
really
good,
and
then
we've
worked
closely
with
the
sequel
client
team
to
help
identify
some
some
bottlenecks
that
they've
had
in
their
Linux,
client
and
I
got
told
today
that
a
fix
has
gone
in
which
is
completely
removed.
That
bottleneck.
So
we
expect
to
see
some
good
improvements
for
folks,
using
sequel,
client
on
Linux
as
well,
and
then
from
there
we're
going
to
continue.
B
This
is
just
focused
on
fortunes
so
far,
which
is
the
read-only
benchmark
because
we
have
to
start
somewhere
and
then
from
that
we'll
probably
move
into
the
update
benchmark
to
figure
out.
You
know
what
are
the:
what
are
the
set
of
issues
we
have
to
tackle
in
terms
of
performance
when
doing
updates
over
Adia
don't
met
on
post,
Grayson
sequel,
but
we've
made
really
really
good
progress
very.
A
B
B
B
B
Is
they
have
to
be
installed
globally,
which
means
that
when
we
wanted
to
make
a
chain
there
Singleton's,
so
women
want
to
make
a
change
to
the
aged
Minette
core
module
when
we
want
to
make
a
change
to
the
AC
unit
core
module,
the
problem
was
that
it
was
very
risky
right
if
we
make
a
change
there
to
fix
every
eight
minute
core
app
running
on
that
machine.
Okay.
B
So
in
two
one,
what
we've
done
is
we've
split
that
so
now
what
we
have
the
ACE
minute
core
modules
effectively
become
a
shim
that
can
includes.
You
know
it's
basically
freezing
in
time
the
current
status
of
the
asp
net
core
module,
but
also
supports
finding
the
other
half
of
itself
inside
the
deployed
application.
So
you
can
effectively
bin
deploy
the
actual
functional
part
of
the
ACE
minute
core
model
with
your
application,
like
literally
the
native
code.
So
that
means
that
we
can
now
make
changes
to
the
a
spirit
core
module
native
part.
B
The
bit
that
integrates
deeply
with
is
our
version
of
without
danger
of
affecting
all
the
other
apps.
Turning
on
the
machine,
it's
just
that
it's
more
of
the
same
right.
Getting
side-by-side
support
allows
us
to
make
to
move
fast
and
break
things
without
breaking
everything
which
is
super
important,
so
that
might
be
what
they're
referring
to
I'm.
Looking
at
the
EC,
looking
at
the
last
commit
in
the
ace
minute,
core
module,
I'm
looking
at
it
and
it
says,
create
issue
template,
maybe
start
by
saying
this
repos
I.
B
Yet
so
this
repair
is
up,
sleep
move
it
to
so,
it
looks
like
they
may
have
just
moved
the
code
so
rather
because
of
the
changes
I
just
talked
about,
we
basically
consider
the
ACE
minute
core
module
and
the
is
integration
all
part
of
the
same
product.
So
it
seems
like
they've,
made
the
choice
just
to
move
it
all
into
one
repo
and,
for
example,
it
means
they
build
together.
B
They
always
build
together
as
part
of
the
same
unit
now,
because
there
is
tight
integration
between
the
two,
so
I
think
it's
just
a
case
of
code
moving
around
there's
nothing
to
be
alarmed
about
we're
just
moving
these
things
to
into
one
repo,
so
that
they're
always
together.
We
can
report
issues,
we
can
code
on
them
together
and
they
can
get
built
together.