►
From YouTube: GitHub Quick Reviews
Description
Powered by Restream https://restream.io/
A
B
So
that's
basically
to
bring
api
parity
with
the
other
dialogues.
We
have
so
client
guide
allows
you
to
associate
a
specific
paths
when
you
open
a
dialog
and
the
initial
directory.
Again
it's
something
missing
in
the
folder
browse
dialog
that
exists
in
the
file
dialog.
A
C
B
A
B
Yes,
I
I
think
the
use
case
and
that
api
was
proposed
by
external
contributor
joseph
master.
I
believe
in
dotnet
five
time
frame,
if
you
have,
if
you
reuse
the
dialogue,
because
you
have
a
single
instance
of
a
dialog,
but
you
want
to
open
different
like
folders
when
you
open
the
dialog,
you
can
associate
a
good
with
every
target
folder
and
whenever
you
recreate
a
dialogue,
you
set
that
guit.
That
automatically
sets
that
target
folder.
D
There's
one
other
issue
that
you
can
that
you
can
mitigate
with
this
and
that's
when
you're,
using
the
folder
browser
dialog
more
times
than
once
in
an
app
then
you're,
just
preventing
from
showing
the
second
time
the
same
starting
path
that
you
did
before,
which
is
really
annoying.
So
it's
not
that
the
instance
are
automatically
fall
back
to
a
certain
start
folder.
It's
just
that.
You
always
see
the
the
last
usage
of
the
of
the
browser
dialog
as
a
starting
part
of
that
app.
That's
really
annoying.
B
I
need
to
look
up
the
the
behavior,
so
I'm
a
bit
unprepared
for
this,
but
the
use
case
was
you
use
the
folder
dialog,
but
you
use
for
for
different
operations
like
say
opening
files
or
as
well
as
saving
files,
and
you
want
to
open
files
from
one
location
that
save
files
to
another
location,
and
you
want
to
distinguish
between
two
operations,
but
you
don't
know
which
locations
user
is
doing
things.
B
D
B
A
E
E
A
Have
called
like,
I
don't
know,
you
know,
save
file,
dialog
or
whatever
right.
You
just
have
two
instances
of
the
folder
browser
dialog
in
your
component,
tray
use
one
or
the
other,
and
then
each
of
them
basically
have
their
own
good
because
they
are
logically
their
own
instance
right,
but
yeah
I
mean
either
way
it
works.
I
guess
it
just
means
that
people
always
have
to
set
it
right.
That's
that's
the
downside.
B
Well,
no,
like
you,
don't
have
to
use
it
right.
It's
an
optional
field,
right,
it's
just
an
enhancement
for
user
experience
where
you
want
to
serve
your
clients
better
and
like
I
traditionally
don't
in
my
experience
with
windows.
Phones,
I
don't
tend
to
drag
into
the
component
tray.
I
typically
prefer
to
create
it
at
runtime,
so
it's
different
styles
of
implementation
of
the
spouse.
E
B
And
a
lot
of
customers
we
we
do
come
across.
They
are
conscious
of
memory
and
by
having
too
many
duplicate,
controls
and
or
components,
meaning
that
they
would
may
create
the
single
dialogue
and
tend
to
reuse
it
across
multiple
operations.
So
right.
B
B
A
All
right,
then,
let's
look
at
the
other
one.
This
is
the
click
environment.
B
They
introduced
a
new
markup
for
hyperlinks
and
how
you
supply
hyperlinks,
how
you
engage
hyperlinks,
how
you
clear
hyperlinks
and
what
we
found
in
git
extensions,
migrating
from
dotnet
4.6
point
something
into
net
coin
and
net
5,
as
well
as
some
other
users
reporting
on
winforms,
that
they
lost
ability
to
have
hidden
text
and
subsequently
to
retrieve
hyperlink
which
been
clicked
migrating
functionality.
B
So,
whilst
we
fix
that
the
proposal
comes
to
help
users
to
have
better
information
about
that
hidden
portion
of
the
text
for
the
hyperlink
when
we're
migrating
from
the
old
into
new
world
and
expose
these
additional
to
properties
when
link
starts
and
when
link
how
long
the
link
part
is.
B
A
B
Length,
I
suppose,
that's
totally
valid
feedback
and
I
think
we
can
do
that
so
ultimately,.
D
This,
what
is
again,
what
is
the?
What
is
the
default
when
we
have
a
selection
like
like
in
the
text
box,
we
also
have
selection,
start
and
selection
length.
Is
it
the
same?
Is
its
selection
start.
B
D
B
You
just
scroll
down
a
little
bit,
please
emo
just
go
past.
My
comment
is
a
comment
from
tobias
with
an
example
of
the
actual
change
yeah.
So
we
have
too
much
that's
that
green
one
yeah.
So
we
already
have
that
information
which
is
provided
to
us
by
the
underlying
control,
which
is
don't
expose
it
to
the
end
user.
A
A
Yeah
yeah,
I
think
it's
not
about
that.
It's
more
about
the
the
you
know
for
the
default
values
right,
because
there's
already
is
the
notion
of
the
selection,
start
selection,
length
and
win
forms,
and
if
you
don't
have
a
selection,
the
question
is:
what
are
the
values?
Are
the
values
minus
one
and
zero?
Are
they
both
just
zero.
D
D
A
A
I
guess
the
thing
in
wind
forms
is,
as
you
said,
right
selection
is
basically
double
purpose
right
like
if
you,
if
nothing
is
selected,
it's
the
position
of
the
carrot,
but
I'm
not
sure
that
makes
sense
for
a
link
right,
because
the
link
clearly
has
a
position
and
that
position
is
either
known
or
it's
not
known
right.
But
if
it's.
A
E
D
D
A
Right
then,
I
think,
unless
there's
more
questions
that
looks
good.
D
All
the
non
involved
things
done.
C
D
Okay,
so
this
this
idea,
this
this
proposal
was
actually
kicked
off
when
we,
when
we
worked
on
the
web
api,
when
we
worked
on
the
on
the
webview
on
the
new
webview
control
and
the
windows
guys
asked
us
for
a
little
bit
of
support
because
they
had
problems
to
actually
access
resources
in
design
time
and
which
the
control
should
not
do
so.
D
So
this
is
the
control
itself
in
design
mode,
which
is
obviously
you
can
just
check
for
design
mode,
and
that's
that's
okay,
but
then
there
can
they
can
be
user
controls
that
are
actually
hosting
user
controls
and
those
are
in
the
designer
and
that's
the
problem,
because
now
the
control
itself
doesn't
get
cited,
so
it
can
never
actually
find
out
if
it
is
in
design
mode
or
not.
But
what
it
could
do
is
it
could
ask.
Is
one
of
my
ancestors
in
design
mode
and
then
react
in
the
same
way?
D
So
it
does
not
use
the
runtime
resources.
However,
if
then,
there
is
no
designer
at
all,
then
that
that
would
be
the
next
picture.
Then
you
would,
you
would
actually
have
the
control
to
use
the
resources
and
then
it
can.
It
can
determine
that
it
is
in
design
mode
and
there
are
countless
discussions
on
stack,
overflow
and
and
all
those
places
practically
as
long
as
wind
farm
exists.
D
What
the
right
way
is
to
to
actually
you
know,
find
that
out
and
what
the
what
the
right
things
to
do
here
and
there
are
two
to
two
fronts-
to
parties
actually,
the
one
that
wants
to
use
this
for
licenses
purposes
so
to
just
check
if
a
control
is
hosted
in
something
that
makes
it
necessary
to
have
a
a
design
time
license,
and
then
the
other
party
is
just
for
like
reason
for
functionality.
D
This
is
not
for
licensing
reason.
This
is
just
for
finding
out
if
what
willforms
defines
as
the
design
mode
is
actually
valid
from
the
controls
perspective
and
then
of
the
view
of
of
its
own
ancestors,
so
that
it
can
act
accordingly
and
and
also
the
reason
that
this
is
not
for
the
design
mode
is
that
you
will
never
be
able
to
detect
the
design
mode,
and
that
has
nothing
to
do
with
this
proposal.
D
This
was
always
like
that
that
you
will
never
be
able
to
detect
the
controls
design
mode
in
the
constructor,
where
many
control
vendors
want
us
to
already
give
them
a
way
to
detect
if
they
need
to
check,
for
you
know
for
a
license
or
not
right.
So
this
has
nothing
to
do
with
that,
and
I
wrote
kind
of
a
disclaimer
at
the
beginning
of
that
proposal,
and
I'm
saying
this
is
not
about
that
because
half
of
the
discussion-
and
it
is
really
a
long
discussion-
was
about
that-
and
I
was
always
saying
guys.
D
This
is
not
about
licensing
that
we
need
to
come
up
with
a
different
thing
for
licensing
we
are
not
there
yet.
This
is
just
for
don't
establish
a
connection
to
a
sql
server
when
you're,
when
your
controller
is
dealing
with
sql
server
connections
at
runtime.
You
don't
want
to
have
that
connection,
because
it's
set
up
with
your
properties
at
design
time
and
also
this
control
that
is
hosted
in
another
control
and
that
control
is
in
design
time.
D
Please
also
don't
do
don't
open
a
connection
to
the
secret
server,
but
if
you
are
running
actually,
then
you
should
open
a
connection
to
the
sql
server
the
web.
The
webview
guys
had
that
had
that
problem
that
they
started,
rendering
the
url
that
was
given
in
the
property
explorer
already
when
they
were
in
design
time
and
it
was
crashing
left
and
right,
so
that
was
actually
the
the
trigger
to
yet
again,
you
know
deal
with
this
and
then
I
think
we
finally
should
come
up
with
something
that
where
we
can
say
hey.
A
So
that
makes
sense
to
me
I
mean
my
only
question
is
like
the
way
you
propose.
It
is,
basically
you
have
a
property
that
only
answers
whether
one
of
you
enters
stresses
in
design
mode
like
do
you
actually
care
about
that,
because
it
seems
like
what
you
actually
care
about
is
am
I
displayed
on
the
design
surface?
D
That's
the
discussion
that
we
have,
but
I'm
not
really,
I'm
not
really
seeing
what
what
the
actual
point
is.
So
if
I'm
part
of
that
of
the
hierarchy,
then
clearly
I
am
in
a
design
mode
in
the
weight
in
the
meaning
of
wind
forms
for
the
design
mode
is
you
could
always
argue
that
this
control
can
get
again
cited,
there's
a
context
where
the
actual
ancestor
is
has
kind
of
a
designer
property.
D
If
you
will
and
then
you
would
report
back
a
wrong
result,
and
that
was
also
part
of
the
discussion
that
we
had.
But
this
is
not
for
the
for
the
contrary
to
the
side,
because
we
have
a
clear
definition:
what
the
design
mode
is
the
design
mode
is
when
you're,
when
you
are
so
when
the
designer,
with
the
design.
No,
when
the
control
is,
is
cited,
and
then
the
site
reports
the
design
mode
back
to
you.
D
So
if,
in
that
case,
one
of
the
ancestors
is
cited
and
then
that
site
reports
back
it
is
in
design
mode.
You
would
take
that
into
account.
Or
not
I
mean
it's
not
that
we're
breaking
anything
it's.
We
need
to
make
really
clear
for
what
purposes
that
is,
and
we
should
also
point
out
that
there
can
be
wrong
reports
about
the
design
mode,
but
this
is
the
same
thing
with
the
design
mode
that
we
have
already
right.
No.
A
I
think
I
think
you're,
you
know
you're
missing
my
point
like
it's
it's
not
about,
but
the
design.
What
is
the
correct
thing
or
not?
What
I'm
saying
is
that
the
way
you
propose
the
property
is
that
it
basically
only
reports
the
parents
design
mode
right,
because
basically,
you
get
the
you
basically
ask
the
control
side.
A
If
it
has
a
parent
and
then
you
ask
for
the
parents
control
side
right
and
what
I'm
asking
is
that,
basically,
your
property
would
return
false
if
I
am
designing
the
user
control
itself
in
the
designer
right,
because
at
that
point
I
don't
have
a
parent,
so
the
parent
would
be
no.
But
I
am
myself
cited
and
I
am
myself
in
design
mode
and
I'm
asking.
Is
that
a
meaningful
separation?
D
Well,
I
mean
you
can
never
well,
actually
a
custom
control
will
ever
will
always
have
a
apparent
because
it's
hosted
inside
of
a
top
level
control,
so
a
user
control
can
be
well.
It
would
be
counting
for
a
user
control.
That's
true,
but
if
you
have
a
custom
control,
then
that
custom
control
would
always
part
of
some
some
top
level
control,
but
you're
absolutely
right.
I
mean
it
can
be
the
case
that,
from
the
perspective
of
a
user
control,
well,
no,
actually,
no,
because
also
user
control
cannot
live
alone.
D
But
we'll
always
have
yeah
in
the
designer
yes,
but
it
will
always
have.
D
D
D
Like
for
a
user
controller
to
drive
you
you
would
you
would
test
the
design
mode
and
then
the
answers
design
mode
as
well,
and
then
you
get
the
right
thing,
but
this
is
only
for
user
controls
then
actually
be
hosted
in
forms,
which
is
also
possible.
A
Yeah,
that's
kind
of
what
I'm
proposing
is.
I
would
basically
just
have
a
property
called.
I
would
even
go
further
and
say
is
in
design
mode.
I
would
just
literally
call
it
like
that
and
say
returns
true.
If
the
control
itself
is
in
design
mode
or
one
of
the
emphasis
is
in
design
mode
and
just
tell
everybody
use
this
property
instead,
because
that
will
generally
give
you
the
correct
answer,
and
if
you
really
really
need
to
know
whether
you're
yourself
in
design
mode
or
one
of
the
answers
in
this
in
design
mode,
that's
fine!
A
On
which
con
on
which
on
control
well
actually
on,
is
it
on
component
or
in
control?.
B
D
D
So
it's
component
design
what
property
and
and
control
is
inherited
from
component,
and
so
there
you
have
it.
A
D
Also,
the
things
that
the
that
components
are
not
getting
in
the
car
are
not
getting
hosted
in
the
the
control
in
the
control
collection,
so
you
wouldn't
be
able
to
actually
go
to
the
top
of
that
with
that.
So
that
is
also
the
reason
that
this
property
should
be
on
control.
Another
component.
B
Thought
of
different
sort
of
names
we
organized
on
this
and
like
ease
design
mode,
would
be
good,
but
because
there
is
design
mode,
it
immediately
creates
that
duality
and
like
developer
would
be
asking
questions
like.
What's
the
difference
all
right?
That's
why
so
klaus
chose
chose
to
go
for
more
like
the
bose,
but.
A
I
would
just
yeah,
I
think,
ignore
the
name
for
a
second
right.
I
guess
all
I'm
asking
is
that
it
seems
weird
that
you
would
ask
people
to
basically
say
you
have
to
check
for
design
modes
and
you
and
you
have
to
order
it
with
its
ancestor
site
and
design
more
to
get.
The
correct
answer
like
I
would
just
propose,
add
a
single
property
that
basically
does
the
or
already
for
you
and
basically
answers.
E
C
B
B
Or
any
or
any
of
my
ancestors
in
design
mode,
and
we
just
need
to
come
up
with
like
a
better
name
for
this.
A
Yeah
just
come
up
with
a
name,
but
I
think
you
want
the
the
union
property
right
well,.
B
A
B
Let's
start
with
that
one,
because
that
would
totally
segue
into
the
so.
The
idea
here
is
to
introduce
these
three
additional
overloads,
which
would
allow
us
to
simplify
in
the
way
we
write,
begin
invoke
and
invoke
methods
which
does.
C
C
B
So
like
like,
if
you
scroll
below
these,
there
are
a
few
examples,
the
the
current
way
of
invoking
or
beginning
blocking
be
cumbersome,
because
you
need
to
use
a
delegate
leave
that
method
in
boca
or
you
need
to
create
a
new
action,
we'll
create
a
new
font,
etc.
But
if
we
create
a
action
overload
or
func
of
t
overload
for
invoke,
we
can
just
use
that
lambda
expression
and
they
just
would.
D
A
A
I
I
have
no
idea
what
this
thing
does
under
the
covers:
it's
possible
that
it
already
allocates
it
may
not
matter,
but
the
problem
is
right
now
you
would
basically
say
we
don't
need
the
the
params
one,
because
you
know
you
just
do
a
closure
right,
but
you
wouldn't
be
able
to
bind
to
aesthetic
method.
This
way
right
and
this
this
done.
This
thing
doesn't
take
any
arguments.
B
A
B
A
B
B
C
A
A
Pretty
short,
wpf
dispatcher
has
a
similar
issue,
but
that's
somebody
else's
problems
off
right.
Then
then,
I
assume
you're
all
good
with
this
api,
then.
D
So
I
know
we
had
that
discussion,
but
if,
if
we
are
so
so
what
happens
in
the
visual
basic
compiler
cafe,
maybe
that
you
know
that
if
we
now
add.
D
If
we
now
add
an
overload
where
the
compiler
would
resolve
it
in
the
in
the
same
way,
does
that
could
you
think
of
anything
where
there
is
now
a
problem
with
ambiguity
in
the
visual
basic
compiler,
where
the
compiler
doesn't
know
any
longer
which
message
overload
to
take,
because
it
could
as
well
be
the
delegate
and
an
action?
D
E
Right
so
I
can
think
of
some
nearly
pathological
cases
that
might
happen
if
the,
but,
but
I
actually
can't
even
get
my
head
around
the
completion
of
that,
but
I
really
would
like
alexi
or
somebody
else
who
is
seriously
an
expert
on
the
vp.
Compiler.
Definitely
should
look
at
this
before
we
do
it.
So
perhaps
you
could
start
by
attacking
jared
and
alexi
on
this
already.
E
Well
I
mean
I
I'm
gonna,
I'm
I'm
going
to
say
that
we
need
to,
you
know,
make
sure
compiler
is
happy
and
then
do
some
testing
and
I'm
not
sure
what
testing
that
we
need
to
do.
I
I
don't
know
what
our
schedule
is
to
get
this
in,
but
you
know
a
little
bit
of
preview
on
this
wouldn't
hurt,
although
I'm
not
quite
sure
who
would
preview
it
in
the
vb
community
and
just
to
check
one
more
thing
that
I'm
afraid
that
I
didn't
get
in
full
clarity.
E
Does
this
affect
framework
and
core.net
flash.net
five
applications.
E
Okay,
okay,
all
right
yeah,
so
I
would
definitely
have
heartburn.
If
that
wasn't,
the
case
is
breaking
lower
level.
Apps
is
exactly
what
we're
trying
to
not
do
with
bb
so
yeah.
I
think
I'm
comfortable
with
us
doing
it.
You
know
I'm
just
trying
to
play
a
compiler
in
my
head
and
I
can't
find
a
scenario
where
it's
broken.
E
Alexi
definitely
knows
more
about
the
compiler
than
jared,
so
we
we
might
still
want
to
run
it
by
him.
But
if
you
put
the
link
in
chat
I'll
just
reach
out
to
him
on
teams
and
make
sure
he
doesn't
go,
oh
yeah
there's
this
thing
so.
D
E
C
E
E
E
We
have
tens
of
those
we
have
tons
and
tons
of
those
yeah
I
mean
that's.
That
is
that.
Is
I
mean
this?
I
I
can
tell
you
how
to
build
classes.
That
will
will
do
that.
I
think
in
general
we
don't
happen
to
fall
into
those
cases
in
the
in
the
bcl,
but
any
time
you
have
overloads
that
are
the
overloads
in
the
base.
Classes
are
north,
then
they're
not
ignored
to
be
b
and
therefore
you
have
a
different
code
path,
cool.
A
All
right
so
then
I
guess:
let's
move
on
from
this
one.
I
have
a
heart
stop
at
five,
so
hopefully
we
can
get
this
done.
Yeah.
A
E
B
E
Let's
do
that
we'll
write
we'll
write
an
email
to
him
to
get
together
and
that'll,
be
good,
just
reach
out
and
we'll
make
sure
that
that
we
get
that
confirmed
where
we're
happy.
D
So,
for
that
one,
this
one,
I
don't
really
know
how
big
it
actually
is
after
I
got
the
first
comment
on
that:
it's
rather
fresh.
D
Basically
this
one
I
I
came
across
also
when
I
was
doing
the
sample
app
for
the
web
view
to
control,
and
the
problem
was
that
I
needed
to.
Basically
I
want
to
be
able
to
await
an
asic
method
while
at
the
same
time
marshaling
the
call
for
that
method
to
the
ui
threads.
D
E
D
As
you
know,
it's
I
know
it's
a
little
bit
it's
an
edge
case,
but
it's
it's
less
and
less
than
edge
case.
That's
my
point
because
now
we
are
having
with
the
five
we
having
especially
good
for
wind
forms
the
option
to
to
project
all
the
uwp
apis
into
our
realm.
So
to
say,
and
now
we
are
dealing
with
with
async
method,
calls
more
and
more
and
more,
and
also
the
rescue
control
that
we
have
is
also
doing
stuff
asynchronously.
D
So
I
have
that
a
couple
of
times
now
and
I
actually,
I
constructed
a
sample
when
which
you
you
know
scroll
down,
you
will
see
all
those
things
and
and
what
happens
if
you
do
and
if
you
do
not
to
to
get
async
stuff
running.
So
this
is
not
really
a
stupid
example
like
if
you,
if
you
come
back
from
a
from
a
threading
timer,
for
example,
when
you're
not
on
the
ui
thread,
and
if
you
then
want
to
await
into
something
that
is
running
on
the
ui
thread.
D
You
basically
need
this
invoke,
async
it.
The
this
is
the
argumentation
for
it.
For
that
proposal
is
rather
is
it
worth
to
explore
that,
and
I
think
it
really
is.
We
need
that
in
wind
forms,
because
this
is
coming
more
and
more
and
the
only
thing
to
to
to
get
this
done
actually
is
to
use
begin
invoke
and
the
task
completion
source
and
this
scheduling
on
on
the
thread
pool
so
that
the
callback
comes
in,
and
you
have
to
do,
I
I
you
know
practically
whenever
they
did
the
sample.
D
I
I
also
implemented
a
you
know
rough
version
of
that
already
to
see
how.
How
would
I
do
it
if
I
made
it
without
it,
and
it's
just
something
that
I
think
is
not
typical
to
do
for
our
winforms
audience
for
our
typical
winforms
audience.
So
we
really
need
a
way
to
do
this
in
a
more
discoverable
way
and
in
a
more
with
an
easier
approach
and
by
the
way
wpf.
I
think
has
that
too.
D
So
there
is
a
there's,
an
invocation
method
on
the
dispatcher,
but
I
don't
think
it
it
makes
sense
to
to
resort
to
something
else
than
than
with
forms,
so
they,
the
wps,
guys,
know
what,
where
to
go
mainly
under
this
there's
such
a
class
that
that
holds
all
that
functionality.
For
that,
and
we
have
that
on
control
forever,
so
in
bulk
and
beginner
and
focus
on
control.
We
should
you
know,
have
that
on
control.
A
Yeah,
so
I
I
mean
it
makes
sense,
I
mean
I'm
not.
The
only
concern
I
have
is
so
right
now
we
are
working
on
something
related
to
this
in
the
runtime
repo.
I
can
just
open
this
guy
here,
which
is
this
issue
here,
which
basically
is
about
making.
You
know
the
basic,
giving
more
options
for
how
things
can
be
awaited.
A
So
I
think
we
churned
the
design
a
million
times,
but
basically
it
boils
down
to
something
like
you
know,
configure
a
weight
where
people
can
actually,
for
example,
pass
in
the
a
cancellation
token
that
is
not
for
the
operation
you're
waiting,
it's
more
for
the
awaiting
itself
that
you
can
cancel
and
also,
we
also
think,
maybe
we
add
support
for
for.
A
Timeouts
directly
in
there,
so
we
should
just
reconcile
these
two
things
and
make
sure
that
what
you
propose
here
with
you
know
directly
taking
the
time
of
direct
taking
cancellation
token
makes
sense.
Other
than
that
I
mean
I
don't
really
have
a
concern.
I
mean
clearly
when
you
you
know
begin
invoke,
is
terrible
as
an
api,
as
you
said,
so
you
want
something
that
is
actually
truly.
You
know
natively
async
in
the
sense
that
you
can
just
await
it.
So
that
seems
reasonable.
A
I
mean
there's
probably
some
implementation
concerns.
I
have
no
idea
what
beginning
you
know
how
we
met
begin
invoke.
I
suppose
it's
the
normal
task,
completion
source
that
we
there
that
we're
doing,
but
we
should
just
make
sure
that
that
all
adds
up.
So
I
I
have
no
concern
with
the
api,
but
I'm
also
not
not
a
setting
guy.
So,
like
all
I'm
saying
is.
We
should
involve
stephen
tube
and
make
sure
that
he
looks
at
his
design.
A
D
Oh
no,
I
would
definitely
you
know,
ask
people
that
that
have
more
knowledge
about
this
than
than
I
have
about
this.
It's
it's
it's
just
if
it's
worth
to
pursue.
You
know
actually
do
this
and
enhance
it.
E
I
would
like
I
would
like
to
hear
stephen's
take
on
this,
because
the
way
I'm
looking
at
this
right
now
this
is
the
very
pm
forgetting
what
I
know
about
the
compiler
for
a
minute,
just
a
very
pm
comment,
which
is
that
if,
if
we
believe
certain
things
about
our
one
forms
cohort,
including
that
they're
they're
very
smart
people,
but
they
don't
necessarily
want
to
think
a
lot
about
the
mechanics
of
how
things
are
working
is
putting
this
on
control.
E
E
So
there
is
precedent
for
that,
but
I
would
just
like
somebody
who
is
you
know
deep
into
this
stuff
that
says:
oh
yeah,
that's
going
to
work
because
it's
just
going
to
be
a
fragment,
that's
going
to
run
back
on
the
ui
thread
and
we'll
try
to
train
people
not
to
do
extensive
things
on
that,
and
you
know
that's
kind
of
as
it
and
one
of
the
reasons
I
think
that
is
that
there's
nothing
in
this
name
that
says
I'm
going
back
to
the
ui
thread.
Make
sure
you
make
this
fast.
D
Going
to
the
usb
we
have,
but
but
people
know
that,
because
we
have
controls
of
invoke
to
go
to
the
ui
set
and
we
have
control
dot
invoke
required
to
detect
if
we
need
to
go
to
the
ui
thread
for
the
next
all
right
for
updating,
and
that
is
the
reason.
I
think
that
you
should
also
have
the
just
consistently
the
invocation
on
the
stay
in
the
same
group
of
of
functionality.
So
we
would
have
control
dot,
invoke
required,
control,
dot,
invoke
and
control
dot
invoke
async.
E
Yeah
so
yeah,
I
I
think.
That's
probably.
I
think
that
I'm
I
will
agree
with
with
people
like
stephen
and
this
that,
if
they
say
sure,
but
if
they
say
there's
these
ways
for
people
she
just
feed
out,
then
I
think
that
we
want
to
take
another
look
at
it
to
see
if
we
can
find
ways
to
protect
it,
and
we
might
also.
This
is
one
of
the.
I
don't
say
this
on
everything
I
promise
we
might
want
to
try
to
do
a
user
study
on
this.
E
I
worry
about
us
getting
the
right
user
study,
but
will
users
use
it
correctly
if
there
is
a
correct
and
incorrect
way
to
use
it?
So
as
far
as
find
out,
if
there
is
a
correct
and
incorrect
way
and
things
that
people
need
to
avoid
doing,
and
then
we
can
see
whether
we're
guiding
them
to
the
success
which
I
think
we
need
to
do
here,
so
that's
just
where
my
thinking
is
coming
from.
E
I
think
we
probably
should
go
ahead
with
it,
but
not
without
a
little
bit
more
effort,
research
on
possibilities,
so
I
mean
you
should
go
ahead
and
explore
it.
A
Yeah,
I
think
I
agree
with
everything
you
said,
except
I
I
would
I
don't
know
I.
I
don't
think
you
need
the
user
study
here,
because
I
think
the
like
this
thing
is
not
introducing
new
complexity
right.
The
the
whole
complexity
about
marshalling
back
to
the
ui
thread,
the
fact
that
it's
called
invoke
the
fact
that
it's
a
method
on
control-
I
mean
that
is
basically
in
winform
since
day.
One
right
and
wpf
has
the
same
concept
there
you
they
use
it
by
the
dispatcher.
A
That's
a
fairly
well
understood
thing
to
me:
it's
more
about
the
the
the
the
async
and
the
awaits
notion
here
like
because
we
also
talked
about,
for
example,
making
our
synchronization
primitives,
like
you
know,
semaphore
mutex,
add
a
weight
method
there
as
well,
but
there
are
concerns
around
whether
you
can
deadlock
because
of
the
way
the
infrastructure
works
right
so,
like
I
think,
conceptually
from
a
from
a
from
an
api
standpoint,
I
think
the
apis
is
that
proposed
makes
sense.
It's
just
that.
A
You
know
somebody
with
actual
you
know,
deep
understanding
of
how
the,
if
the
you
know
the
state
machines
work
and
how
the
whole
underlying
plumbing
works,
especially
with
the
ui
threats
and
the
message
cues.
That
is
something
where
I
would
just
you
know,
defer
to
stephen's
expertise
and
say
like
does
that
make
sense
to
you
or
not
right,
but,
like
I
feel
from
a
I
don't
know
if
we
would
just
ship
the
api
and
presumably
it
would
work,
I
think
it's
it's
a
fine
api.
A
E
The
thing
is
two
questions
for
steven
one
is:
are
there
things
we're
not
saying
about
people
shooting
their
feet
off,
which
was
my
concern
and
your
concern,
which
is
technically?
Is
this
going
to
work
right,
or
are
we
just
setting
ourselves
up
from
a
technical
level
to
be
creating
undesirable
situations
regardless?
Well,.
D
The
thing
is
that
currently
we
are,
we
are
having
problems
with
deadlocks
with
with
async
and
awaits
in
in
so
many
windforms
apps,
because
people
tend
to
use
it
in
a
alternative
way.
D
If,
if
a
name
of
a
method
is
ending
in
a
string,
they
know
they
need
to
await
it
and
then,
if
they
cannot
await
it,
because
the
the
signature
is
not
compatible
compatible
to
the
means
they
have
to
call
it
which
is
invoked
currently,
what
do
they
do
so
if
they
then
put
it
put
the
whole
thing
into
an
into
an
invoke
asic
and
it
fits.
E
It
shouldn't
work,
so
you
basically
first
part
of
what
you're
saying
there
is
that
the
pedophilia
is
easy
to
fall
into.
Right
now
is
doing
trying
to
do
synchro
async
in
this
case,
because
the
call
that
you're
going
to
the
invoke
delegate
is
already
a
it
needs
to
be
available
and
the
signature
you.
D
D
I
said
I
my
only
concern
was
for
this
for
this
session.
Just
should
we
should
be
consider
it,
and
this
is
what
I'm
thinking
and
I-
and
I
want
to
make
a
point
that
I
think
it's
really
necessary
to
have
it,
but
I'm
open
for,
of
course,
I'm
open
to
to
any.
You
know
suggestion
I
can
get
how
to
implement
it,
and
maybe
I'm
not
even
implementing
it,
because.
B
I
am
you
didn't,
did
what
we
have
like
the
the
last
sentence
of
the
previous
comment
from
to
beers
and
that's
something
I
completely
overlooked.
Some
reason.
I
think
we
should
consider
adding
cancellation
tokens
to
the
begin
invoke
methods
we
already
have.
B
So
we
should
consider
adding
a
few
more
overloads
for
our
invoke
to
have
begin
invoke
action
method,
cancellation,
token,
big,
invoke,
delegate
method,
cancellation,
dog,
etc.
So,
like
we
offer
that
cancellation
functionality
as
well,
which
we
can
only
done.
A
Yeah
yeah
that
yeah,
I
would
make
this
a
separate
one,
because
it's
like
it'll
be
my
same
answer,
because
we
already
have
this
work
item
here.
That
basically
adds
it
generically
to
the
to
be
able
to
cancel
in
the
weight,
which
might
be
enough
for
your
case
like
if
you
really
need
to
plumb
it
through.
Then
that's
a
different
thing,
but
the
problem
is
here
that
the
previous
api
source
is
just
syntactic
sugar
over
the
ones
you
already
have.
B
A
So
then
yeah
I
need
to
run
now,
but,
like
you
know,
if
you
know
I
would
not,
I
cannot
approve
it
right
now
because
it's
like
we
would
need
stephen
to
look
at
this,
but
like
yeah,
just
just
work
with
stephen
on
this,
and
if
he's
okay
with
it,
then
that's
it
cool
all
right
then,.