►
From YouTube: ASP.NET Community Standup - Oct 10, 2017 - OSS Real Talk
Description
This week's community links: https://www.one-tab.com/page/s_1FhFUzRzuwkuJSoUPMLA
B
B
I
can
set
up
yeah.
The
microphone
is
the
same
microphone,
but
it's
now
on
a
stand
with
a
pop
filter.
So
now,
when
I
type
in
theory,
it
shouldn't
be
ringing
for
everybody
is
as
I
typed
now,
because
it's
in
a
spider
mount
or
a
shock
mount
whatever.
You
call
those
things
yeah
and
I.
Don't
know
what
the
pop
filter
does.
I'm,
not
an
audio
person.
I
mean
I,
appreciate,
good
audio,
but
I,
don't
know
what
the
pop
filter
does.
You
can
tell
me
the.
A
B
Well,
going
high-tech
now
so,
hopefully
I
sound,
even
better
than
you
and
Scott
now
Scott
will
be
joining
us
he's
having
some
machine
issues
as
per
usual
I'm.
Assuming
he's
got
some
random
daily
build
of
14
different
things
on
his
machine,
which
is
why
he's
audio
and
current
doesn't
work.
We
even
find
out
yep,
but
we
are
here.
Yes.
A
A
Cool
I
I'm
juggling
think
so
we
have
been
looking
at
switching
over
to
using
like
to
do
it
we're
interested
if
people
are
seeing
this
in
HD
or
will
see
the
recording
and
HD,
especially
because
so
I've
changed
the
latency
mode
and
then
we're
looking
at
setting
it
up
with
OBS,
but
there's
some
complications
there
and
how
we
do
it.
So
yeah.
B
Once
we
get
a
sort
of
self
production,
it's
a
whole
massive
jump.
We'd
have
to
choose
our
video
calling
tech
we'd
have
to
obviously
wouldn't
do
the
the
focus.
Switching
when
people
talk
we'd
have
to
direct
that
manually
we'd
be
doing
the
encoding
ourselves
like
it.
Would
you
know
it's
all
doable,
it's
just
a
lot
of
work.
Yeah
right
now
we
just
wrap.
We
just
trying
to
call
yeah.
B
A
B
B
A
Okay,
so
I'm
a
hair
went
through
and
he
he
won
the
ride
about.
Well,
he
was
inspired
by
David
Fowler's
tweet
about,
like
you,
gotta
use
doctor,
so
he
went
through
and
I
pointed
out
before
the
the
recent
video
by
Mark
Rendell,
and
so
basically
he
kind
of
live
blogged
his
going
through
this
kind
of
stuff.
So
he
just
kind
of
said:
here's
what
I
did-
and
you
know,
here's
how
it
worked
for
me
and
all
that
kind
of
stuff.
So
you
know
the
cool
thing
with
this
I
mean
really.
A
B
Never
I
think
I've
installed
it
once
at
all
and
I
think
I've
literally
never
run
it
and
I
feel
bad
about
it.
To
tell
you
the
truth,
like
I
really
need
to
do
it,
but
I
just
haven't
had
that
tasks
that
need
hasn't
arisen
yet
that
I've
forced
myself
to
sit
down
and
go
I'm
gonna
make
a
docker
thing.
Yeah
build
my
app
in
docker.
I
just
haven't
done
it
yet.
The.
A
B
It's
far
more
impactful
when
you
running
like
what
the
Fowler's
been
looking
at
it
in
the
scope
of
we're
doing
some
more
tech
and
power
work
again
and
he's
wanting
to
replicate
other
frameworks,
results
and
so
he's
using
docker
to
get
all
the
prerequisites
set
up,
because
you
know
some
of
them
are
Java
apps.
Some
of
them
are
C++.
Some
of
them
are
this
and
that
he
doesn't
have
to
install
all
the
various
frameworks
and
dev
tools
and
things
to
get
that
happening.
B
He
just
wants
to
start
a
docker
container
for
a
given
benchmark
and
have
it
do
all
the
building
and
stuff
inside
of
it
and
then
just
go
away.
Yeah!
That's
where
you
really
see
the
benefit
of
things.
I
build
servers,
and
you
know
when
you
have
all
lot
of
prereqs.
That
requires
setup,
and
now
you
do
have
to
do
it's
just
all
exactly
image
thing.
Well,.
A
B
A
So
this
is,
this
is
a
big
ones,
Dominic
writing
about
the
identity,
server
for
v2
release.
So
obviously
the
big
big
feature
here
is
asp
net
core
to
support.
Also,
some
other
things
like
back
channel,
logout,
spec
federation
scenarios
and
terminal
cleanup
and
then
support
for
asp
net
core
config
system
also
also
really
nice.
That's.
A
Identity
server,
for
this
is
the
big
like
major
yeah
I,
guess
almost
like
the
product
name,
just
on
chat
or
something
but
and
also
cool.
Here
he's
got
89
contributors
to
this
or
you
know
they
did
so
nice
all
right,
so
Chris
announcing
here
the
ws
Federation
200
preview,
one
out-of-band
release.
Yes,
so
I
don't
know.
If
you
want
to
say
anything
about
this
honestly.
B
It
had
a
bunch
of
downstream
or
upstream
I,
don't
know
which
way
that
is
dependencies,
that
hadn't
been
ported
to
done
in
standard,
slash,
tonic
or
yet
that
work
is
now
all
complete,
and
so
we
were
able
to
retrospectively
fix
that
by
shipping
a
to
o
package
atom
and
obviously
after
202
ad
WS
Federation
support.
So
this
is
the
first
preview.
B
A
A
A
B
Picked
some
issues
that
came
up
as
a
result
of
the
first
alpha
that
people
reported
we've
made
some
API
changes
and
yeah.
This
is
why
this
is
an
alpha,
but
this
is
nowhere
near
done
yet
right
we're
trying
to
put
out
a
minimum
sort
of
set
of
scenarios
so
that
we
can
get
feedback
on
some
of
the
big
changes
and
the
new
improvements,
and
so
we're
thankful
for
people
who
try
out
these
very
early
releases.
B
They
really
really
helped
shape
the
product,
and
you
know
I've
been
harping
on
about
it
recently,
but
we
have
struggled
to
get
a
lot
of
preview
coverage
recently,
especially
broad,
like
we
tend
to
get
the
same
people
who
use
our
previous,
which
is
good.
We
get
a
good
feedback
from
them,
but
I'd
love
to
see
us
get
a
little
more
broad
sort
of
usage
of
our
preview.
So
we
can
get
more.
B
Broad
feedback
is
often
the
case
that
we
don't
get
key
feedback
items
until
well
after
we've
shipped
in
ITM,
and
then
it
becomes
very
hard
for
us
to
change
things
in
a
real
way.
So
yeah
I
certainly
encourage
people
to
go
and
try
this
out.
As
I
said
bunch
of
bug
fixes
as
the
typescript
JavaScript
client
a
c-sharp
client.
The
binary
hub
protocol
is
in
there
using
message
pack
and
we
fixed
some
leaks
and
stuff.
So
hopefully
it
should
be
a
little,
but
it's
not
ready
for
production.
So
don't
ask
it's
an
alpha.
B
A
B
B
You
know
X
or
if
you
have
some
time
available,
and
you
have
a
signal,
our
app
and
you
want
to
try
and
reproduce
some
of
that
on
a
snit
call,
and
the
other
should
be
enough
to
do
that
now
and
give
us
some
feedback
based
on
the
changes
and
stuff
that
we've
made
or
things
that
you
think
we've
missed
or
things
that
you
really
wanted
us
to
fix,
but
we
haven't
done
yet,
etc,
etc.
So
cool.
A
Okay,
so
featuring
another
of
Jax,
you
know
I,
I,
always
kind
of
like
I
go
back
and
forth
about
featuring
a
series.
Cuz
I
don't
want
to
keep
giving
the
same
person
too
much
airtime,
but
this
is
a
really
really
good
post,
so
he's
posted.
What
he's
been
doing
is
digging
into
a
repo
and
going
beyond
they.
Just
kind
of
like
here
are
the
features
or
here's
how
this
one
thing
works
and
really
kind
of
digging
more
in
depth.
A
So
here
he's
looking
at
action
results
which
are
very
important
for
for
NBC
and
how
NBC
functions.
So
he
goes
through
some
things
like
what
is
a
view
result
and
things
like
that.
What
I
think
is
really
kind
of
great
here
is
that
he's
got
some
some
nice
diagrams
here
and
so
he's
showing
you
know
how
how
a
view
result
is
executed
and
explaining
that
out
explaining
the
executor.
This
is
I.
A
Remember
the
first
time
kind
of
tracing
through
this
when
I
was
really
trying
to
learn
MVC
like
writing
a
book
and
really
really
wanted
to
deeply
understand
it,
there's
more
to
it
than
you'd.
Think
you
know
as
far
as
how
these
things
and
it
does
allow
it
to
be
very
pluggable,
but
it
also
there's
a
good
amount
in
there.
So
this
is,
this
is
very
you
know
view
localization
how
you
can
customize
it
all
that
kind
of
stuff,
yeah.
B
And
I
hope
I
mean
there's.
There
is
complexity.
Obviously
we
hope
that
the
complexity
is
laid
in
such
a
way
that
you
don't
really
pay
for
it
until
you
need
it.
So
if
you're
not
returning
views,
then
you
the
whole
branch
to
do
with
view
execution.
There
goes
away
if
you're
just
doing
I
have
a
controller
with
some
methods.
You
know,
there's
a
filter
pipeline
before
that.
If
the
pipeline
is
empty,
then
there's
not
much
going
on
there.
Obviously
there's
model
binding.
B
If
you
choose
to
use
it,
I,
don't
know
how
you
wouldn't
but
nomination
those
type
of
things,
and
they
are
all
like.
You
said,
sort
of
subsystems
that
you,
the
more
you
dig
in
the
more
you'll
learn
about
them.
One
of
the
efforts
we're
looking
at
moving
for
today's
minute
core
is
trying
to
understand
what
parts
of
like
this
system
of
MVC
really
should
be
lower
down
in
the
stack.
When
we
build
a
snake
hole,
we
didn't
really
take
anything
that
was
in
NBC
and
Baker.
We
never.
We
never
did
it.
B
We
didn't
kind
of
say:
okay,
that's
the
new
unit
of
composition
for
an
a
spinet
core
app,
and
that
needs
to
be
in
the
core.
Look.
We
just
did
that
for
middleware.
That
was
really
it
on
the
eye,
but
things
like
routing
are
still
separate.
You
know
routing
is
only
given
when
you
add
MVC,
it
adds
its
own
router.
B
You
know,
gathering
metadata
and
storing
metadata
friend
points
in
the
app
all
those
type
of
things,
and
then
we
wonder
why,
over
time
as
I
said,
look
at
what
an
MVC
and
decide
are
there
more
things
that
need
to
be
that
a
common
concerns
that
really
should
be
brought
down
the
whole
concept
of
an
action
result.
If
you
build
raw
middleware,
hello
Scott,
this
is
mobile.
Scott
I
am.
C
C
C
B
A
C
C
C
B
C
B
It
looks
at
the
request
on
the
way
in
it
does
something
the
request
keeps
going
and
then
maybe
looks
at
the
way
out
and
modifies
it
and
whatever
or
it's
terminal,
it's
an
endpoint
in
the
application.
If
it's
an
endpoint
and
you
want
to
return
stuff
you're
on
your
own,
like
all
we
give,
you
is
the
response
dream,
there's
no
composition
model
for
returning,
HTML
or
Jason
or
any
of
the
other
stuff
in
the
middle
web
pipeline.
All
that
stuff
is
obviously
in
MVC,
which
is
sort
of
much
higher
up
the
stack,
which
is
unfortunate.
B
You
know,
there's
there's
a
lot
of
class
of
middleware,
whether
they
be
like
our
own
diagnostic
pages,
for
example,
that
show
the
developer
Diagnostics
page
or
the
error
page
things
like
swashbuckle
that
have
a
pre-canned
UI,
those
things
it's
it's
probably
too
much
to
have
to
have
all
of
MVC
or
to
have
to
force
the
app
owner
to
plug
those
into
their
instance
of
MVC.
You
probably
just
want
a
very
simple
way
to
return
a
result,
so
here's
a
result
execute
it,
and
then
you
can
decide
what
that
is.
A
B
A
A
A
I
am
post,
it's
pretty
cool,
though
it's
just
like
hey.
Look
at
all
this
great
stuff.
Here's
what's
built
in
you
know:
here's
you
know:
cats,
trolls,
cool
writers,
cool,
Dockers,
cool,
just
all
this
stuff
so
and
down
at
the
bottom.
He
actually
does
a
disclaimer
I,
don't
work
for
Microsoft
I,
just
love
it.
So
it's
kind
of
cool
to
see
this
overall
feedback.
Oh
there's
that
and
then
Hashem
talking
about
using
P
o
file
localization
in
asp
net
core
to
using
orchard
core,
and
this
is
something
I've
talked
to
Sebastian
recently.
A
B
B
If
you
just
want
to
do
the
PIO
file
localization,
you
can
use
orchards,
pfr
localizer
just
plugs
straight
into
any
HP
net
core
app
as
if
it
was
like
one
that
you
know
the
SP
a
core
team
make
I
mean
Seb
works
in
the
area
net
caught
him.
He
does
this
stuff
for
us
nobody's
doing
it
out
of
the
auto
project,
but
that's
similar
some
other
things.
They
have
an
asset,
a
client
asset
manager,
API
with
a
set
of
tag,
helpers
and
server-side.
B
A
Asset
manager,
I
did
a
code
conversations
interview
with
them
and
so
yeah,
it's
like
you're,
saying
it's
a
really
full-featured
asset
pipeline
thing
and
it
handles
some
pretty
cool
stuff
like
it.
It
may,
if
you
reference
things
multiple
times
that
make
sure
that
they're
not
referenced.
But
you
know
what
I
mean
that
they're
not
included
multiple
times
it
does
a
lot
of
kind
of
intelligent
stuff,
so
yeah
cool.
So
here's
here's
just
the
code
sample.
This
is
basically
all
there
is
to
it
to
to
use
a
KO
file
which
is
I,
guess.
C
C
B
You
look
better
than
Tibet
just
stand
there
and
don't
move
you
look
great
there
we
go
there,
we
go
so
what
kicked
it
off?
Is
you
did
a
blog
post,
Scott
I,
don't
know
four
weeks
ago
and
it
was
highlighting
some
open
source
framework.
Let
me
see
if
I
can
find
it,
and
you
know,
as
you
know,
you
usually
do
you
do
some
great.
You
highlight
great
stuff,
you
see
in
the
community.
You
talk
about
stuff
that
we
do.
You
can
kind
of
do
a
whole
bunch
of
different
stuff
right.
B
C
B
C
B
And
so
you
write
your
blog
post,
you
hollowed
it,
and
one
of
the
first
comments
was
a
couple
of
the
comments.
If
I
recall
were
people
who
were
quite
clearly
disappointed
that
we
hadn't
bothered
to
do
that
work
and
put
it
in
the
box
and
I
think
it
was
along
the
lines
of
I.
Don't
understand
why
Microsoft
doesn't
include
this
type
of
functionality
and
build
this
themselves
and
that
that
struck
a
nerve
with
me.
B
I
have
to
say
you
know,
because
we
I
think
that
represents
the
other
side
of
the
spectrum,
to
what
many
a
lot
of
the
other
feedback
that
we've
gotten
from
the
open
source
community
in
the
past,
which
is
you
know,
we
say
we
want
to
go
off
and
build
X
feature
X,
and
then
we
get
us
or
why?
Don't
you
just
use
this
thing
that
already
exists,
sure
which
is
the
act
opposite
right?
It's
like
we
want
to
go
and
provide
into
an
experience,
or
we
want
to.
B
C
B
B
Know-
and
we
have
examples
where
we've
done-
that,
whether
it's
something
like
Jason
net,
where
we
decided
to
take
the
clear
incumbent
and
ship
that
and
III
found
that
that
that
that
paradox,
though
those
two
viewpoints,
those
two
ends
of
the
spectrum
striking
and
it's
something
that
I
think
I'm,
not
sure
people
think
about
how
we
have
to
walk
that
sort
of
that
line
between
those
two
opposing
viewpoints
and
I.
Think
they
both
have
merit.
I.
B
Think
I
think
that
both
of
those
come
from
a
genuine
place
and
it's
something
I
just
wanted
to
say
that
we
constantly
sort
of
we
we
talk
about.
We
constantly
battles,
a
strong
word,
lowercase
B,
but
battle
with
whenever
we
want
to
take
on
you
know,
building
any
type
of
new
feature
or
looking
at
an
end-to-end
experience
that
we
want
to
build.
B
Sometimes
it
can
be
difficult
to
build
end-to-end
experiences
that
meet
the
vision
that
we
have
using
four
different
libraries
built
by
four
different
authors
for
four
different
purposes,
with
different
visions,
and
sometimes
it's
not
sometimes
we
can
but
take
something
very
generic
and
utilitarian,
and
we
can
plug
it
in
at
a
certain
layer
and
it's
good
for
everyone,
but
there
are
side
effects.
Every
time
we
take
dependencies
on
third-party
libraries,
there
was
a
discussion
we
had
recently
on
on
Twitter.
Oh.
A
B
About
about
Jason
Dunn
net
in
particular,
right,
which
was
interesting
one.
He
was
complaining
that
the
the
side
effect
of
us
taking
dependency
on
something
like
Jason
net
is
that
if
Jason
net
releases,
a
new
version
that
introduces
some
type
of
breaking
change-
and
it
can
be
very
subtle
right-
doesn't
have
to
be
a
large
one.
It
practically.
It
makes
it
practically
impossible
to
use
the
new
version
of
that
library
in
an
a
it's
been
at
core
app.
B
If
we've
taken
a
dependency
on
the
current
behavior
okay,
because
if
you
update
I
mean
you
get,
will
let
you
update?
That's
perfectly
fine,
but
once
you
update
your
a
genetical,
app
may
now
be
broken
depending
on
what
that
third-party
library
author
decided
to
do
in
that
version.
Okay-
and
you
know
that-
can
devolve
into
arguments
about
december
and
how
it
applies-
inverted
net
and
you
get
which
frankly,
I
think
we
should
do
a
better
job
of
of
coaching
and
leading
in
the
ecosystem
on
it's,
not
simple.
B
So
yeah
I
just
wanted
to
I.
Guess
said
when
that
when
that
comment
came
up,
I
found
it
I
found
it
really
striking
and
that
I
hadn't
heard
that
articulated
in
a
while
and
it
was.
It
was
the
opposite
end
of
the
spectrum
to
a
lot
of
the
feedback
we
have
received
recently
and
I.
Just
wanted
people
know
we
do.
We
do
here
both
sides
and
we
do
constantly
think
about
what
the
right
choice
is
for
the
the
majority
of
customers
and
we,
when
we
have
to
make
a
default
choice
which
leans
one
way.
B
A
C
I
was
talking
with
her
about
how
it
was
going
and
how
her
work
was
and
she's
like
she
says:
I
thought,
I,
understood
and
how
complicated
things
were
working
for
a
company
that
ship
software
to
a
billion
pcs,
and
then
you
start
looking
at
security
checks
and
polly
checks
and,
and
you
know,
localization
and
testing
and
and
while
it's
easy
to
say
hey
that
was
that
was
a
bad
bill.
That
was
an
easy
bug
and
you
should
see
that
or
why
didn't
you
just
use
food
net?
That's
a
thing.
C
B
B
Doing
it
in
such
a
way
that
people
can
still
understand
it
like
we
still
I
mean
we
have
versions
soup,
as
it
is
right
now,
we've
got
net
core
and
we're
working
very
hard
to
really
make
sure
we're
cognizant
of
every
version
that
shows
up,
and
it's
there
for
a
reason,
and
can
people
understand
it
and
when
do
they
change
and
is
understandable
that
you
know
this
one
changes
now
and
that
one
changes
then,
but
it's
difficult
ever
ever.
Every
way
you
decide
to
go
has
has
it
has
a
side
effect,
has
many
side
effects.
B
There
are
ripple
effects
that
I
think
are
difficult
to
see
and
I
mean
when
I
started,
I
always
said,
I
didn't
I,
didn't,
know
them
and
to
be
fair,
I
know
a
lot
of
them
now
and
I
still
about
them
internally.
If
I
come
across
a
bug
like,
oh
my
god,
how
did
that
team
not
find
that
and
of
course
we're
just
as
guilty
of
that
everybody,
like
me,
I
understand
what
it's
like
to
be
in
an
app
developers:
shoes
when
you
hit
that
type
of
issue.
It's
like.
How
did
they
not
see
this?
B
It's
complicated,
it's
just
really
really
complicated
stuff
and
we
do
constantly
have
to
walk
the
line
about.
We
are
literally
trading
off
between
maybe
14
different
variants
on
any
given
decision,
and
we
it's
not
as
simple
as
coming
out
with
one
guiding
sort
of
flowchart
decision
tree
policy
that
we
use
that
we
it's
quite
nuanced-
and
you
know
Scott
knows
this
and
John
knows
this
and
Scott's
been
in
some
of
our
triage
meetings
before
and
seen
the
sausage
being
made.
B
A
Thing
that
I
see
now
is
apologies
for
cutting
in
I
hope
it's
along
with
the
flow
here,
but
one
thing
that
I've
seen
change
a
lot
over
the
time
I've
been
at
Microsoft.
Is
that
how
open
you're
able
to
be
now
with
your
decision-making
process?
So
you
know,
because
the
code
is
developed
on
github,
because
you're
making
announcements
of
intention
and
your
shipping,
you
know
what
I
mean
and
changes
are
being
made
a
while
before
the
new
get
package
which
is
a
while
before
the
meta
pack
is
a
big
ship
date.
Does.
B
That
make
it
easier
or
harder
or
jump,
I
think
it's
it's
I
mean
again
I
think
it
depends
on
your
perspective,
I
know
for
a
fact
that
there
are
people
who
think
we
are
anywhere
near
open
enough
and
I
sometimes
think
that
as
well.
There
is
a
cost
to
being
completely
open
and
we're
not
completely
open.
B
We
know
we
openly
admit
that,
like
we've
been
going
through
sort
of
the
I
call
that
the
business
phase
of
our
of
our
product
planning
right
now,
where
we're
looking
at
what
our
directors
want
our
business
goals
to
be
for
the
next.
You
know
in
months
or
years
for
the
next
set
of
versions,
we're
working
on
and
then
they
hand
us
a
list
of
requirements
that
aren't.
You
must
have
this
API
in
this
feature.
B
We
do
previews
and
all
that
stuff
is
very
iterative.
But
I
know
for
a
fact
that
you
know
that
I'm
sure
plenty
of
people
would
like
us
to
it
would
have
a
would
like
to
have
a
lot
more
input.
Somehow
will
democratize
somehow
the
the
actual
features
that
get
built
right,
I
know
other
projects
you
know
do
have
things
like
that,
but
they
don't.
They
certainly
don't
all
work
that
way
and
I
know
some
people
they're
better
or
worse
than
bad
as
a
definition
of
open
source.
But
that's
certainly
not.
B
B
We
thought
we
reach
out
to
customers
and
actually
go
and
do
phone
calls
with
them
and
sit
with
them
and
generally
to
discuss
a
particular
problem
area
that
we
think
is
interesting
that
we
think
might
be
solvable
by
some
set
of
features,
and
so
we
try
and
find
summers
that
we
think
are
experiencing
those
issues
and
we
have
them
tell
us
what
their
issues
are
and
then
we
try
and
find
the
common
thing.
You
know
all
the
sort
of
standard
sort
of
public
planning
stuff
which
I
imagine
a
lot
of
our
customers
themselves
do.
B
B
Both
sides
of
the
argument
and
all
I
can
say
is
that
we
continue
to
try
and
refine
that
we're
not
always
going
to
get
it
right
for
everybody
and
I
apologize
in
advance,
yeah,
sometimes
I
reach
out
to
Scott
and
I,
say
Scott.
Can
you
talk
me
off
the
ledge
and
help
me
figure
out
how
to
this
particular
change?.
C
A
A
C
C
C
B
No
I
think
I
think
forums
like
this.
This
is
one
of
the
very
reasons
we
created.
This
show
right.
If
you
think
right
back
to
over
two
years
ago
or
three
years
ago,
when
we
started
a
spinet,
cordeen
ax,
whatever
was
at
the
time,
and
we
were
seeing
a
lot
of
FUD
in
the
community
about
people
would
see
code
and
had
heard
some.
B
Some
third
party
person
pontificate
they'll
have
an
opinion
about
something
about
what
we
were
doing,
not
hearing
it
from
us
and
then
something
would
start
it
was
like
you
know:
we've
got
to
be
able
to
talk
frankly
in
an
open
forum
and
be
able
to
have
people
talk
to
us
yeah
weekly,
so
this
is
kind
of
why
we
started
this
show
and
I.
Think
we
haven't
talked
about
some
of
this
stuff
and
I
dislike
is
that
this
comment
is
kind
of
bloated
up
and
I.
B
Think
it's
worthwhile
sure,
like
you
said
touching
point
on
this
every
now
and
then
and
acknowledging
that
yeah
we
yeah
we
can
we're
guilty
as
anyone
of
perhaps
over
trivializing
certain
people's
problems.
Sometimes
all
I
ask
is
that
I
think
I
think
it
goes
both
ways
I
think
sometimes
people
trivialize
what
our
problems
are
as
well,
and
they,
as
you
said
before
it
feels
very
everyone,
has
a
narrow
perspective
when
it
hits
them,
and
it
can
be
very
difficult
to
see
the
wider
picture.
B
And
a
lot
of
it
is
working
through
language
as
well,
in
the
sense
that,
when
someone
uses
phrasing
like
this
is
completely
broken,
we
need
to
understand
what
that
means
to
that
person,
because
that
can
mean
that
the
performance
is
unacceptable
or
the
level
of
complexity
is
unacceptable.
I
can't
figure
out
how
to
use
it.
B
It
could
be
a
particular
type
of
error,
they're
getting
under
certain
conditions-
and
you
know
a
lot
of
it-
is
just
that
type
of
diagnosis
and
we've
had
a
lot
of
that
recently,
trying
to
ensure
that
we
we
understand
what
people
mean
when
they
say
this
is
not
working
for
me.
It's
fundamentally
broken.
It
can
sometimes
take
a
fair
bit
of
digging
to
really
get
to
the
crux
of
the
issue.
Quite.
A
A
question
for
you
on
that
I've
seen
some
projects,
you
know
on
github,
use
github
issue
templates
and
pull
request
templates
to
kind
of
here's
a
form
when
I
create
a
new
issue.
It
says:
what
version
are
you
running?
It
does
you
know
and
some
even
use
the
kind
of
checklist
syntax?
Is
it
a
hard
crash?
Is
it
just
frustrating?
Is
it
do
we
use
us
at
all
I,
don't
think.
B
B
You
issue
they
kind
of
like
spits
a
whole
bunch
of
stuff
into
the
issue
and
then
you're
supposed
to
fill
it
out,
but
a
lot
of
people
just
control
they
delete
and
then
type
their
stuff.
That's
right,
like
I
mean
it's
a
suggestion.
Yeah
you
fill
those
things,
it's
more
likely
that
we
won't
have
to
be
follow-up.
Questions
it'll,
be
easy
to
triage
and
those
type
of
things,
but
sometimes
it's
pretty
clear-cut.
You
can
just
paste
a
stack
trace
and
say
this
is
what
I
called.
B
Obviously
this
shouldn't
happen,
but
you
know
just
things
like
a
developer
is
your
product
and
of
a
similar
size
and
you
were
kind
of,
but
you
wanted
someone
to
report
the
issue
of
you
just
share
as
much
information
as
you
possibly
can
and
it'll
help
us.
Obviously,
that's
only
way.
I
got
that
off
my
chest.
We're
all
I,
really
I
I
want
people
to
keep
giving
us
that
feedback
like
I
want
to
understand
all
ends
like
both
ends
of
the
spectrum
and
everything
in
between.
B
So
it
was
really,
as
I
said,
interesting
to
see
that
comment
on
Scott's
blog
when
I've
heard
so
much
of
the
other
side.
I
hadn't
really
heard
that
side
for
quite
some
time
and
there's
something
that
we've
been
talking
about.
We've
used
that
I
don't
want
to
say
defense,
but
we've
used
that
reasoning
for
some
decisions
we've
made
in
the
past
about
when
we
want
something.
That's
in
box
there
were
reasons
we
can't
always
take
dependencies
on
third
parties,
etc,
etc,
etc,
and
it
was
interesting
to
get
an
outside
opinion
along
the
lines
of.
B
Why
doesn't
Microsoft
just
build
this
thing?
I,
don't
understand
whether
the
community
had
to
do
it.
I
think,
obviously,
there's
a
happy
equilibrium
somewhere,
where
we
provide
enough
inbox
value
that
you
know
you
get
a
really
good
feature
set
out
of
the
box.
Some
of
that
might
come
from
some
open
source,
libraries
that
are
on
the
periphery,
but
we
provide
by
default,
which
you
can
swap
out
with
a
different
one.
B
If
you
want,
then
other
things
a
little
more
neechal
a
little
more
specialist
and
the
community
does
a
really
good
job
of
running
with
those
and
providing
good
solutions
for
them,
and
then
there's
always
that
you
know
how
do
we
as
big
corporate
Microsoft?
Do
it
the
best
job
we
can
of
promoting?
You
know
those
type
of
things.
That's
something
I
know
Scott.
B
You
know
we've
been
trying
to
tackle
that
beast
for
a
long
time
and
you
know
there's
a
lot
of
eyes
that
go
into
things
like
what
templates
show
up
in
visual
studio
by
default,
which
one
is
ticked
by
default.
A
lot
of
opinions
are
filtered
and
sifted
through
at
Microsoft
before
that
decision
is
made
and
so
yeah.
It's
not
always
easy
just
to
make
everyone
happy.
Obviously,
so
there's
a
bunch
of
questions
in
here.
Do
we
want?
Oh
there's.
B
Couple
things
we're
working
on
recently:
I
just
wanted
to
touch
on
we've
been
doing
those
some
looks
at
tech
and
power
again,
David
fell,
I
created
a
repo.
If
you
want
to
look
at
his
github
called
platform,
tech
empower,
I,
think
so.
Tekken
Power
has
a
couple
different
excuse
me:
ratings
like
levels
for
their
benchmarks.
B
They
have
a
platform
which
is
the
lowest
level,
basically
yeah,
no
one,
no
one
codes
directly
against
the
platform,
because
you
won't
be
productive
right.
It's
like
it's
a
roar,
HTTP
server
and
there's
no
dispatching
logic
and
there's
there's
friendly
API
is
for
returning
strings.
You
like
you're,
doing
bytes
and,
like
you,
like
you
view
under
Jason,
you
pulling
your
own
json
library
and
stuff
like
that
right.
So
then
there's
the
micro
above
that
which
is
like
oh
okay,
it's
a
micro
framework.
You
have
dispatching.
B
You
might
have
some
helper
methods
for
dealing
with
text
and
file,
uploads
and
stuff
like
that,
and
then
there's
what
they
call
full,
which
is,
you
know,
would
be
our
MVC
framework
as
a
full
framework.
Our
middleware
layer
is
what
we
call
a
micro
framework
and
we
don't
have
a
platform
entry.
We
don't
have
a
platform
entry
for
dotnet
at
all,
and
so
what
we've
been
looking
at?
What
would
a
platform
entry
for
attack
and
power?
B
Look
like
specifically
for
the
plaintext
and
Jason
scenarios
right
now,
because
most
of
the
one
well
Millie
all
of
the
top
ten
our
platform
entries
they're,
not
micro,
framework
or
full
framework.
You
know
one
of
the
things
we
pride
ourselves
on
is
a
spirit
core
in
tech
and
power
for
plain
text
he's
actually
one
of
the
top
micro
frameworks
and
one
of
the
top
full
frameworks,
it's
in
the
top
three,
but
everything
else
above
us
other
than
those
ones
rated
as
platforms.
Okay.
B
So
there
are
a
lot
lower
level,
so
it'll
be
interesting
for
us
to
have
a
platform
submission
just
see
how
fast
we
can
really
push
dotnet
as
an
engine
with
you
know,
an
HTTP
server
and
took
the
bits
that
make
up
kestrel
with
a
very,
very
thin
layer.
On
top
of
that,
you
know
for
the
purposes
of
the
benchmark,
so
we've
been
doing
some
investigations
there
and
the
people
who
are
really
interested
in
that
type
of
stuff.
B
One
of
those
things
there's
also
that
we
get
to
see
where
we
lose
performance
so,
as
we
add
back
features
from
the
actual
framework
on
top
of
this
platform
level
thing
excuse
me
how
much
perfect
we
use
you
know
what
does
it
really
cost
to
have
a
nice
incoming
header
dictionary
where
everything's
been
turned
into
a
string
for
you
already,
rather
than
dealing
with
the
raw?
You
know,
memory
mapped
a
bunch
of
chunk
of
memory
that
came
off
the
network.
What
does
it
really
cost
you
to
have
nice
methods
for
writing
out?
B
You
know
certain
file
types
or
nice
methods
for
doing
dispatching,
like
routing
type
stuff
other
than
you
manually,
looking
at
the
HTTP
request
line
and
then
figuring
out
what
the
path
is
like
every
one
of
those
niceties
has
a
cost,
and
by
doing
this
exercise
we
can
actually
start
to
ascertain
what
those
various
layers
cost
us
in
terms
of
a
very
raw
benchmark
like
that.
So
it's
actually
a
really
interesting
exercise
for
people
should
they
say
wish
to
see
another
thing
that
was
in
a
couple
other
things
excuse
me:
I've
had
a
cold.
B
B
So
that's
coming
up
and
we'd
love
to
get
people
to
watch
that
and
give
us
feedback
on
that,
so
that
we
can
get
involved
in
the
2.1
stuff,
nice
and
early.
The
other
thing,
as
part
of
that
before
that
we're
discussing
some
potential
improvements
around
how
a
spinet
core
apps
run
on
top
of
a
platform
right
now.
So
in
oh,
we
introduced
your
sink
called
the
runtime
store.
If
you
recall
it's,
you
know
it's
all
the
stuff
is
in
there
it's
engined
already.
B
So
when
you
publish
your
app,
we
trim
it
out,
so
you
don't
have
to
take
it
with
you,
because
it's
assumed
you
already
have
that
stuff
in
the
runtime
store
and
it
works
pretty
well,
it
does
turn
out.
It
has
some
issues
now
that
it's
out
in
the
wild
and
we're
thinking,
and
we
have
to
think
about
servicing
it
deploying
it
to
multiple
environments
there.
There
we
have
run
into
some
design
issues
with
the
runtime
store
that
we're
working
through
some
potential
changes
to
fix
those
right
now
the
plan,
hopefully,
would
be
very
transparent.
B
It
would
be
basically,
you
wouldn't
notice
any
change.
You
would
install
some
200
X
version
of
the
platform.
You
would
open
your
project
and
change
the
a
spinet
core
package
from
200
to
a
X,
25
or
whatever
it's
gonna
be,
and
that's
it
you
would
rebuild
and
run,
and
everything
would
just
work
and
be
running
on
top
of
whatever
this
new
improved
way
of
achieving
the
same
things
are
I
just
want
to
give
people
a
heads
up
today.
B
Now
that
we're
seeing
it
out
in
the
wild-
and
it
should
be
good
for
people
I-
think
the
end
result
is
going
to
be
better
in
a
lot
of
ways,
but
the
we
have
a
very
strict
goal
of
making
sure
it's
as
little
disruption
if
any
to
people
who
are
already
running
on
to
our.
Obviously,
we
want
that
to
be
as
seamless
as
possible
so
right
now.
We
think
the
plan
will
make
it
so
that
it's
literally
no
trouble
at
all.
You
just
installed
a
new
version.
B
B
B
Think,
if
I
throw
all
the
way
back
up,
Nick
asked
a
question
in
context
of
when
I
was
talking
about.
The
idea
of
dispatching,
which
is
like
global
routing,
is
like
it
was
global
routing.
Why
didn't
you
just
call
it
global
routing,
I,
don't
know
we
couldn't
call
it
routing,
because
we
already
have
a
thing
called
routing,
so
we
wanted
to
give
it
a
name
that
denoted
where
it
really
fits
in
the
pipeline.
B
I
think
the
current
plan
is
that
in
this
thing,
basically
sits
at
the
end
of
the
middleware
pipeline
and
has
the
job
of
dispatching
if
something
hasn't
done
it
already,
based
on
the
metadata
that
came
with
the
request,
a
lot
type
of
thing,
so
it
is
quite
literally
the
dispatcher
but
yeah
its
naming.
It
is
what
it
is.
B
Jonathan
Chanin
pointed
me
to
Botwin,
which
is
one
of
his
projects,
which
seems
to
be
very
much
along
the
lines
of
something
stuff
I
was
talking
about,
which
was
what
would
it
mean
to
take
some
of
these
core
MBC
principles
or
primitives?
Like
return?
An
action
result
and
bake
them
a
lot
lower
down.
He
has
a
thing
called
Botwin.
That
is
seemingly
an
attempt
to
do
that.
B
There
has
some
basic
routing,
though
dispatching
type
stuff
and
the
ability
to
return
results
which
is
kind
of
interesting,
so
if
you're
interesting,
and
that
he
calls
it
the
a
library
that
allows
Nancy,
S
crowding
for
use
in
a
spinet
core.
So
if
you're
interested
in
just
the
routing
part
of
Nancy
in
a
spinet
core,
then
look
at
bahtman
that
looks
like
it's
worthwhile.
B
Luke
Latham
has
asked
question:
dropping
Bower
and
bundler
minna
fire
for
library
manager
pac-man.
How
will
pac-man
work
across
tooling
outside
of
the
yes
doesn't
need
project
file
entries
to
work
outside
of
yes,
so
I
think
I
might
have
mentioned
this.
Nice
stand
up,
I,
don't
know
current
plan
is
we're
going
to
drop
our
net
other
templates
in
the
next
patch
because
bear
is
effectively
dead
and
then
we're
going
to
replace
that
sometime
after
that,
it
will
not
be
at
the
same
time
we
will
be
dropped
with
library,
manager
or
pacman.
That's
been
called.
B
This
is
the
prototype
that
Mads
has
had
flooding
around
for
a
couple
years
and
similarly
the
bundler
minna
fire.
We
have
a
new
thing
that
meds
has
been
working
on,
which
is
a
common
that
supports
both
run
time
and
build
time
bundling
and
minification
and
supports
dispatching
into
node
so
that
you
can
even
do
you
know
your
bundle
config.
B
You
know
at
runtime
call
out
to
a
node
based
NPM
task
right,
very
zen,
that's
using
the
same
work
that
we
do
in
our
spire
services
to
run
web
hat
on
the
server,
and
so
you
get
the
benefit
of
development
time
everything
being
super
dynamic
and
fast.
You
just
change
your
file
and
things
get
recompiling
the
fly
and
then,
when
you
publish
it,
all
gets
pre
done
for
you
and
the
output
goes
along
with
your
application.
So
that
is
the
direction.
The
timing
is
still
being
figured
out.
Altom
utley!
B
I
want
to
get
this
out
as
quick
as
possible.
I
want
us
to
move
to
one
the
pac-man,
which
is
the
not
a
package
manager
package
manager
and
it
will
work
outside
of
es.
So
it's
actually
going
to
be
integrated
into
msbuild,
just
like
the
rest
of
the
project
system.
So
we're
even
toying
with
the
idea
of
having
a
new
item
type
in
the
project
file
to
denote
that
you
have
basically
a
client
package
reference.
B
If
you
want
jQuery
or
your
angular
whatever
might
be,
then
you
would
just
denote
that
in
the
CS
projects,
like
you
do
package
ref
today,
you'd
have
a
different
one,
and
then
you
can.
The
thing
is
pluggable
itself
so
where
it
actually
goes
off
and
gets,
the
libraries
from
is
is
an
independent
concern
and
Marvis
we'd
have
two
faults
for
that.
So
yep,
that's
that's!
Coming
along
what
else?
What
else?
What
else?
B
Could
you
explain
how
the
automatic
so
from
Frederick
had
Azure
automatic
light
up
of
app
insights
and
agile
works,
even
though
you
do
not
say
anything
about
app
insights
in
your
code?
Explain
the
magic
injection.
So
in
200
we
added
a
feature
called
hosting
startup.
Excuse
me,
and
by
way
of
environment
variables,
you
can
set
an
environment
variable
with
a
special
name,
SP
Nick
or
hosting
startup
assemblies.
I
think
it
is
which
is
a
semicolon.
B
B
Because
done-day
core
is
a
lot
more
strict
about
learning
assemblies,
then
dotnet
framework
it
doesn't
have
a
fusion
doesn't
have
the
assembly
loading
logic
that
doesn't
framework.
Does
all
of
the
assemblies
that
your
application
can.
Reference
are
actually
baked
into
that
depth
stock
JSON
file.
When
you
build
the
application,
if
it's
not
in
there,
then
you
only
get
very
rudimentary
assembly
learning
support.
So
the
additional
depth
support
is
the
ability
to
have
a
precooked
sort
of
third
depths
file.
You
have
the
frameworks
file.
B
You
have
the
apps
file,
then
you
can
have
this
additional
file
that
you
can
kind
of
paste
on
top
of
the
app
at
runtime
via
an
environment
variable,
and
then
the
host
will
allow
those
assemblies
to
be
loaded.
If
someone
says
please
go
and
load
them.
So
it's
the
combination
of
those
two
features,
one
from
the
runtime
and
one
from
a
smear
core
that
lets
us
basically
find
an
assembly
of
a
given
name,
find
the
types
inside
of
it.
B
They
implement
a
certain
interface
and
then
call
a
method
on
them
at
the
beginning
of
the
application
starting
up
when
the
web
host
is
being
built,
and
then
that's
where
application
insights
in
you
know
in
in
Visual
Studio
and
in
Azure
can
inject
itself
in
and
you
know,
get
its
environment.
You
get
see.
An
instrumentation
key
from
the
environment
and
all
the
stuff
to
the
DI
container
adds
the
tackle
the
component,
etc,
etc.
So
yeah,
it's
all
first-class
features
that
you
can
use
yourself.
B
If
you
want
to
to
do
scrolling
down
trying
to
find
more
questions,
it's
actually
very
helpful
when
people
just
mentioned
my
name,
because
in
a
highlight
in
orange,
I
can
design
ice.
So
the
people
who
ask
me
questions
directly
are
getting
preferential
treatment.
So
Maria
asked
what
level
of
detail
should
we
be
giving
to
provide
you
with
meaningful
issues?
Oh,
is
that
what
you
were
it's
the
funny
is
that
what
you
were
talking
about
before?
Well,.
B
Yeah
yeah
so
I
mean
I
mean
it
depends
on
the
issue.
Basically,
we
need
to
have.
We
need
enough
detail
such
that
we
can
reproduce
the
issue.
If
we
can't
reproduce
it,
we
can't
fix
it.
That's
what
really
comes
down
to
so
I
mean
the
usual
things.
What
platform
where
you
one?
What
version
are
you
using
if
you
can
share
your
project
file
and
your
startup
CES?
That
is
immensely
helpful.
A
lot
of
the
time
that'll
make
sure
that
we
that
we
there's
no
stale
code.
There's
no
misuse
of
API
is
going
on.
B
You
don't
have
mismatching
versions
or
anything
like
that
that
we
can
diagnose
very
quickly.
It
also
gives
us
an
overall
sense
of
what
your
application
is.
If
you
can,
the
best
thing
is
to
try
and
reproduce
the
issue
in
a
much
smaller
representative
application,
just
like
distill
the
issue
down
to
its
core
and
then
share
that
project
with
us
is
really
obviously
quotes
it's
more
work,
and
sometimes
you
can't
do
that,
because
the
issue
is
so
complex
that
it
requires
a
whole
bunch
of
moving
parts,
but
that's
really
helpful
when
you
can't
do
that
and.
A
Sometimes
it's
funny,
but
sometimes
the
act
of
doing
that
shows
you.
Maybe
if
there's
an
issue
with
your
code,
you
know
like
I've,
seen
that
myself,
where
I'm
like
this
is
a
problem.
Let
me
write
a
sample
and
then
you
get
it
down
and
you're
like.
Oh,
it
doesn't
reproduce.
Oh
my
database
was
offline.
There's
something
stupid
right
right,
yeah.
B
B
A
B
To
put
them
in
just
to
address
that
right
now
we
ship,
angular
and
Redux,
react
in
Redux
in
the
Box,
because
they
are
overwhelmingly
the
most
used
of
those
templates
based
on
before
we
had
them
in
the
box
right
like
by
a
many
of
an
order,
the
truth,
I
believe
and
so
putting
aralia
and
view
in
there.
We
felt
would
just
create
confusion
for
people
who
were
looking.
You
know
to
start
out,
and
it
just
wasn't
seeing
the
usage
people
can
install
the
irelia.
If
you
knock
out
versions
of
those
templates,
we
have
them.
B
A
A
B
B
B
B
Okay,
Frederik
asks:
when
do
we
get
easy
ways
to
install
tooling
that
gets
picked
up
by
the
SDK
attribute
in
the
MS,
build
project
node?
Yes,
we
call
that
SDK
acquisition
so
that
they
so
people
may
have
noticed
in
then
the
new
CS
projects
that
we
call
these
the
sdk-based
CS
proj.
There
is
a
attribute
on
the
project,
node
called
SDK,
and
it
points
to
one
of
two
things:
the
Microsoft
net
SDK
or
the
Microsoft
net
web
SDK.
B
What
we
would
like
is
that
you
could
put
an
SDK
name
in
there
that
you
don't
currently
have
on
your
machine
and
then,
when
msbuild
passes
that
cos
project
file.
The
first
thing
it
does
is
looks
at
the
SDKs.
Listening
goes,
oh,
you
don't
have
those.
Let
me
go
and
install
those
from
NuGet
for
you
into
some
global
location
on
your
machine
and
then
build
so
that
we
bootstrap
the
SDKs
for
you
automatically
as
part
of
the
build
the
project
building
process.
I
have
no
visibility
into
the
status
of
that
desire.
It's
no.
B
It's
been
talked
about
I'm,
not
aware
of
it
being
on
the
backlog
for
the
next
release.
There's
probably
an
issue
that
talks
about
it
in
the
msbuild
repo,
but
that
that
is
the
general
idea.
I
don't
have
a
status
for
you
on
when
that's
gonna
happen,
though
I'm
sorry,
right
now,
the
big
focus
in
in
sort
of
the
the
msbuild
style,
the
the
project
system
and
the
MS
build
stuff
for
the
next
releases
performance
like
we're.
Working
really
really
really
hard
to
drastically
improve
the
performance
of
multi
projects.
B
Cost
compiling
multi-project
solutions
in
package
restores
all
those
type
of
things
to
do
with
you
know
using
your
project,
that's
a
develop.
We
want
to
make
that
as
fast
as
it
can
possibly
be.
So
there's
a
little
work
going
into
that
right
now.
Hence
the
clubbing
PR,
which
was
external
to.
Is
that
embed?
What
else?
What
else
with
us?
Pausing?
Have
you
say
new
yeah
that
I've
jumped
over,
but
I
should
be?
Oh,
no.
B
Overall,
yeah,
all
right
so
at
the
end,
I,
don't
know
how
to
pronounce
his
name.
I.
Think
it's
your
vendor
Mia.
What
are
the
plans
for
identity
application
services,
so
that
is
in
scope
for
2.1
next
release
and
we're
still
figuring
out
exactly
what
it's
going
to
entail.
But
the
current
plan
is
that
we
will
likely
still
end
up.
B
We
will
likely
end
up
shipping
more
options
for
identity
out
of
the
box,
so
you
can,
you
know,
do
the
identity
options
we
have
today,
which
include
these
sort
of
in-app
identity,
which
is
just
a
database
of
staff
and
usernames
and
passwords
or
as
you'll,
be
to
see
or
windows
or
Thor
those
type
of
things,
but
will
add,
support.
Excuse
me
for
tokens
server
based
stuff
as
well,
where
you
run
your
own
identity
service
using
the
identity,
bits
that
we
already
have
plus
in
the
u.s.
B
yes,
sort
of
middleware
sort
of
endpoints
and
the
other
thing
we're
very
keen
on
is
making
it
easy
to
switch
between
one
and
the
other,
because
today
you
know
you
file
new
and
you
choose
identity.
If
you
do
in
app
identity,
you
get
a
500
line,
account
controller,
you
know
which
isn't
particularly
friendly,
so
we're
looking
at
ways
to
factor
a
lot
of
that
core
logic
into
effectively
a
module
right
that
that's
in
the
project
already.
And
then
you
can
sort
that
out
for
a
different
implementation.
B
If
you
want
to
and
extending
that
all
the
way
to
the
UI
as
well,
like
I'd
really
like
to
get
to
a
place
where
we
don't
have
to
have
all
the
identity,
views
or
pages
in
the
application
as
well
like.
If
you
just
want
the
default,
plug-in
experience
and
the
manage
experience,
you
just
get
all
that
without
having
all
those
razor
files
in
your
app.
B
And
then
we
have
a
customization
facility
that
lets
you
use
something
like
field
templates
or
entity,
templates
or
partials
or
or
just
spilling
all
the
pre-compile
views
in
your
app
as
the
last
resort.
If
you
then
want
to
go
and
customize
things
so
I
think
there's
a
whole
bunch
of
there's
a
there's.
A
lot
of
clicks
tops
on
the
on
the
on
the
dial
there
that
we
can
choose
from
and
we're
sort
of
working
through
a
lot
of
that
right
now
so
yeah
it
is
in
scope
for
2.1.
B
B
The
live
stream
we
could,
we
could
write
software
that
makes
that
work.
There
is
an
API
to
get
access
to
the
live,
chat
stream
while
the
broadcast
is
live,
and
so,
if
we
really
wanted
to,
we
could
run
the
little.
You
know
bot
that
comes
and
scrapes
it
out
and
preserves
it.
But
I
don't
know
how
we
would
you
know
present
I,
guess
we
would
just
save
it
as
a
chat.
B
B
Will
you
ship
default
support
for
Joe?
That's
from
Eric?
Yes,
so
that's
part
of
the
describe
of
the
attendee
service
staff
is
having
support
for
token
authentication
as
part
of
the
stack.
So
yes,
Richard
will
the
server
generate
GML
code
of
tight
quadruples?
He
sent
me
an
email
about
this
HTML
typescript,
CSS
and
c-sharp
ever
be
so
cute.
Like
the
pick,
your
framework,
J's
spar,
source
code,
type,
triples,
HTML
type
script
and
CSS
I.
Guess
it's
kind
of
like.
B
Will
we
ever
have
a
first-class
UI
module
where
you
can
very
easily
encapsulate,
maybe
a
chunk
of
razor,
some
client-side
logic,
some
server-side
logic
and
some
stylesheet
right
into
one
nice
little
chunk
and
then
use
that
I
don't
know.
I
can
certainly
tell
you
that
we
have
desires
and
lowercase
P
plans
to
continue
improving
razor,
especially
its
component
ization.
So
we
built
tag
helpers
and
we
built
view
components
in
a
spirit
core,
but
we
didn't
really
improve
the
composition
model
for
razor
and
that's
something
that
we
care
very
deeply
about.
I!
B
Think
what
you'll
end
up
seeing
we
don't
have
a
first-class
templating
support
and
razor.
You
can't
build
a
razor
template
from
razor
and
then
like
get
it
as
a
func
that
you
can
tell
you
that
you
can
execute
later
and
pass
in
the
data
to
the
closest
you
have
to
that
today.
As
a
partial,
you
can't
write
a
tag
helper
in
razor.
You
have
to
write
it
in
c-sharp
right
and
there
are
lots
of
cases
where
you
might
just
want
to
write
some
razor
and
how
that
appears
at
a
cobbler.
B
So
we're
thinking
a
lot
about
that
and
such
that
in
the
future,
we
think
we'll
probably
end
up
in
a
place
where
you
can
build
razor
components
as
a
new
thing.
We'd
have
a
thing
called
razor
component,
which
you
can
write
either
in
c-sharp
or
razor,
and
then
they
would
be
exposed
very
much
like
tag
helpers
are
today,
and
so
you
could
even
see
an
extreme
future
where
things
like
tag,
helpers
and
view
components,
kind
of
all
go
away,
and
we
just
have
this
thing
called
razor
components
which
have
a
whole
bunch
of
different
features.
B
So
yeah
there's
some
of
the
things
we're
thinking
about
and
I
also
hope.
It
would
then
lead
itself
to
getting
rid
of
things
like
layout
pages
and
sections
as
well,
because
the
composition
model
today
of
layouts
and
sections
with
your
content
page
isn't
particularly
nice.
Passing
data
between
them
isn't
first-class.
You
have
to
put
data
and
view
data
or
viewbag,
which
is
horrible.
It's
untyped,
whereas
I'd
much
rather
see
a
tight
component
based
model
where
my
page
can
set
data
on
the
layout
page
that
it
has
declared
it
uses.
B
Today,
if
you
factor
say
your
login
details
into
a
partial,
that's
used
by
the
layout
page,
but
you
have
to
get
the
current
user
or
some
detail
from
the
current
page
like
actually
you
have
to
bounce
it
through
all
these
layers
want
to
make
that
type
of
stuff
a
lot
more
first-class,
so
yeah.
So
things
will
improve
about
getting
all
the
way
to
things
like
their
typescript
and
CSS
with
that
as
well.
B
B
Really,
where
he's
David
is
hard
to
pin
down
it's
because
he's
yeah
book
him
far
enough
in
advance,
and
we
have
a
topic
for
him.
Of
course,
that
would
be,
though
I
could
get
him
to
talk,
come
in
and
talk
about
some
of
these
aspirations
for
the
underpinnings
of
Kestrel
and
stuff
he
has
this.
Is
it
project?
Bedrock
I
think
is
his
codename,
because
Barry
is
always
fun,
so
it's
big
game
we'd
be
good
to
have
a
topic.
B
Yeah
get
him
on
on,
so
we
we
we've
actually
started
a
schedule,
we're
gonna
slot
in
guests
with
topics,
and
once
we
have
a
bit
of
a
roadmap
for
that.
Well,
I
guess
we'll
make
it
public,
so
people
can
know
who's
gonna
come
up
at
a
certain
date.
What
I'm
thinking
right
now
was
about
the
alternate
between
sort
of
the
normal
format.
Show
like
this,
where
it's
just
me
crapping
on
about
something
and
answering
questions
and
then
and
then
the
next
one
we
would
do
a
guest
and
have
them
talk
about.