►
From YouTube: .NET Design Reviews GitHub Triage
Description
-16:-57:-36 - Approved: Update HttpStatusCode enum with updates https://github.com/dotnet/corefx/issues/4382
00:04:27 - Approved: Add application/json to System.Net.Mime.MediaTypeNames.Application https://github.com/dotnet/corefx/issues/26201
00:06:23 - Approved: Unsafe API for comparing byrefs as pointers https://github.com/dotnet/corefx/issues/26451#issuecomment-359887092
00:28:53 - Approved: New APIs for Accessing RSS Optional Elements https://github.com/dotnet/corefx/issues/25718#issuecomment-359890337
00:40:44 - Approved: String-like extension methods to ReadOnlySpan(char) Epic https://github.com/dotnet/corefx/issues/21395#issuecomment-359906138
B
A
A
C
From
the
bottom
be
a
bunch
of
the
Cindy
sings,
but
we
should
talk
with
Eric
Earhart,
one
RV
together
yeah.
We
should
probably
do
like
a
good
overall
thing
in.
C
So,
let's
start
we
have
free
networking
claims
which
are
so
I
should
say
simple,
but
yeah
I,
almost
people
that
usually
can
go
down
on
Obama
number,
four
from
bottom
and
then
one
second
from
the
top
or
third
from
the
top.
There
are
very
basic
reading,
even
higher
up
higher
yep.
This
one.
C
C
A
G
G
I
D
A
D
A
C
Right
so
here
we
go
here
something
similar.
We
are
adding
constant
strengths
for
education,
Jason
application.
Somehow
we
have
text,
slash
XML
will
be
researched
or
research
bunch
of
things
that
might
be
useful.
The
application
JSON
is
actually
useful
because
there's
on
the
second
floor
question
this
confusion.
What
should
be
used
and
they're
like
you-
know,
four
four
or
five
variations.
This
is
the
ripe
on
based
on
I,
believe
RFC
or
something
this.
G
E
A
D
C
C
D
A
A
J
G
D
D
K
D
I'm,
nothing
like
how
they'll
mail
this.
This
is
not
the
right
API.
For
that
I
mean
you
could
argue
instead
of
having
a
property
to
just
make
it
method
like
get
highest
available,
but
then
I'm
not
sure
we
can
do
that
because
it
depends
on
where
it
should
be
version.
The
type
ships,
armor
implementation,
that's
right,
so
that
so,
let's
imagine
HP
version
is
a
thing
that
ships
in
box-
and
we
can't
change
it.
But
then
there
is
an
auto
ban,
it
should
be
client
or
something,
and
then
you
can
accept
the
HTTP
version.
A
C
Well,
we
they
are
overall
or
I,
think
it's
appropriate
I
believe,
but
in
general
they
don't
then
don't
set
it.
We
have
defaults
right
now
be
default
to
one
point
one.
So
the
point
is
like
hey,
you
know
one
phone
or
less
per
day.
All
2.0
is
faster,
so
let's
switch
it
to
2.0
and
with
that
we
want
to
give
us
about
people
opportunity
to
say
I.
Can
they
set
it
explicitly
to
say
like
yeah
yeah?
If
you
don't
care,
you
really
want
to
latest
greatest,
like
part
pastor
to
default.
That
was
like
use.
N
G
D
A
P
A
G
G
A
Used
for
query
and
if
I
like
you
know,
we
have
existing
ideas
that
return
versions
I
understand
that
they
need
to
change,
because
we
want
to
move
to
2.0.
But
if
I
explicit
this
like
call
some
API
and
that
I
pass
in
a
value,
I
think
I
would
like
I,
don't
understand
what
would
be
the
sign
for
I.
Don't
care
just
give
me
the
highest
if.
G
R
A
G
D
One
thing
that
is
problematic
with
like
having
central
ages
is
that
people
would
probably
end
up
comparing
the
sensitive
ages
themselves,
their
same.
Some
things
like
edge,
beavers
and
highest
supported,
equals
a
GP
version
version
2.0,
and
it
will
never
be
true
right,
so
we
probably
should
rename
into
these
two
something,
like
you
know,
use
highest
or
something
to
make
sure
that
it's
a
you
know
an
instruction
rather
than
an
actual
version
number,
but
maybe
exposing
something
that
they
do
itself
is
not
a
good
idea.
You
shouldn't.
D
D
Like
the
question
is
so
imagine,
instead
of
having
this
API
here
at
all,
like
whatever
takes
diversion
right
now
offers
the
different
property.
That's
basically
like,
instead
of
having
a
virtual,
you
just
say
you
know,
there's
another
thing
you
can
say
you
know
use
highest
equals
true,
or
something
or.
G
F
D
D
F
I
T
A
A
E
P
D
P
C
D
I
I
I
I
N
I
N
I
I
N
I
Be
honest
one
because
you
just
provided
the
pointer
right,
the
point
your
left,
which
we
already
have
an
API
for
that
and
I
tested
it,
and
it
gives
bad
performance
and
moderate
amounts,
the
reason
being
that
it
incurs
an
arithmetic
operation
where
we
really
just
want
to
direct
rehearsal.
So
what
does
the
LT
being
more
expensive?
The
sub
and
test
yeah.
G
G
I
G
I
Could
be,
however,
as
far
as
for
the
particular
scenarios,
where
we're
using
this,
it's
actually
okay,
because
once
an
area
we
know
for
a
fact
that
we're
working
with
the
same
buffer,
we
just
want
to
know
whether
we've
gone
past,
the
midpoint
of
the
buffer.
This
is
the
array
dot.
Reverse
example,
basically,
where
you
have
one
pointer
going
from
the
beginning,
one
pointer
going
from
the
end
eventually
they'll
meet
in
pass
each
other,
and
this
is
a
way
to
tell
if
they
pass
each
other.
D
Mean
like
the
discussion,
I
just
can
go
with
like
the
naming
discussion.
I
think
the
own
aesthetic
or
the
other
names
have
the
same
problem
like
I
mean,
if
you
say,
is
less
than
lady
one.
You
really
like
that.
It
still
gives
you
the
wrong
idea
of
what
the
API
does
probably
breath
value
warning
wrapped
up.
But
well,
that's
not
the
point.
The
point
that
Eddie
made
was,
if
you
read
it
like
this.
A
Your
head,
imagine
this
in
this
is
an
infix
operator.
Well,
except
I.
Don't
think
we
use
in
the
framework
terminal
below
and
above
use,
less
than
and
greater
than
no
sure
like
I
mean
they
are
synonyms,
and
why
would
we
not
have
to
convey
that
you?
What
you're,
comparing
is
the
address,
not
the
value
they
are
pointing
to
and.
D
D
D
X
X
L
Y
D
B
Z
Z
D
Z
Z
The
RSS
spec,
which
we
historically
never
had
implemented
it's
an
optional
field,
and
it's
like
how
often
do
you
have
to
refresh
this
so
like
if
the
feed
was
putting
out
new
articles
like
every
few
hours
they
might
time
to
live
to
be
you
know
an
hour
long,
but
then,
after
you
get
outside
the
publishing
hours,
they
might
put
it
like
up
to
eight
hours
or
something
like
that.
So.
Z
G
A
U
R
Z
Z
D
Z
D
H
A
U
E
D
D
A
Z
F
G
G
P
Z
S
Z
I
think
I
think
you
could
argue
varies
reason
to
allow
zero,
because,
as
its
defining,
how
long
we
can
cash
for
a
zero
value
could
mean
don't
cache.
Miss.
G
L
L
D
O
Because
they
acquire
the
dimension
machinery,
that's
listen
for
so
there
will
be
dead
core.
Specifically
they
can.
They
can
be
exposed
from
memory,
trations
short
well
yeah.
It
just
means
that
the
rest
subversion
doesn't
neglect
those
API
systems.
We
can
do
them
all
of
that
unless
we
essentially
copy
and
have
Pierce
lower
just
to
have
parity
for
portable
ApS
it
another
thing
somewhere
is.
The
memory
is
already
split
right
in
terms
of
fast
versus
worth
limitations.
Oh
you're
saying
it's
still
exposing
in
the
same
slot,
so
that
would
be
a
PRI
voice,
identical
right
levels.
A
O
A
P
I
A
A
A
A
X
O
L
O
K
K
D
A
N
V
D
D
D
Just
too
intelligent
you
just
you
know,
type
in
pad
left.
We
will
show
this
API
have
attributes
of
try
and
then
select.
This
is
more
argument
for.
Can
you
discover
using
the
old
name
as
you
can,
can
you
get
code
to
compile
it
continuously?
No,
but
that's
true
regardless.
So
in
that
sense,
I
would
argue
that
the
then,
as
long
as
the
name
is
included
somewhere,
I
think
it's
fine.
We
made
everything
out
there.
Try
and
yeah
I
think
more
consistent
consistency
before
rehearsals.
N
So
if
we
have
the
freedom
to
change
this
again,
so
I
would
suggest
like
like
yeah
try
to
place,
but
I
would
add,
also
ordinate
to
be
very
explicit,
because
this
is
very
confusing
in
the
string
today.
So
people
using
the
place
and
which
is
does
organize
stop.
If
you
didn't
pass
any
problem
and
we
will
expect
it
to
a
work
in
your
culture.
N
A
They
possibly
I'll
tell
you
my
concern
about
it.
If,
if
we
just
add
try
you
kind
of
like
learn,
it
start
using
span
eighty
ice
and
we
learn
that
they
try
very
hard,
not
allocate
their
first
versions
of
api's
that
used
to
be
without
trying.
Now
they
are
trying
returned
there.
So,
but
it's
kind
of
you
know
if
I
saw
a
method
of
this
type
and
it
was
called,
try,
replace
I
would
be
like
okay,
there's
just
a
version
of
replace
except
it
doesn't
allocate
when
I
see,
try
replace
already
now
I'm
thinking.
D
G
D
J
I
H
A
I
AA
O
O
I
O
O
A
O
O
O
O
Was
going
to
ask
somewhere
the
thing,
is
it
doesn't
have
any
of
the
other
states?
It
only
has
if
it's
not
not
enough
definition,
but
should.
G
It,
what
did
it
make
progress
yeah
instead
of
just
a
board,
should
you
be
able
to
continue
like
oh
I've
replaced
the
first
from
the
first.
You
know
100
characters,
but
then
I
ran
out
of
destination
space.
When
you
resume,
you
should
start
at
100
character
mark,
but
it
goes
should
make
progress.
It
has
bite
Regina
as
output.
G
It's
a
parameter.
We
have
any
of
our
span
tribe
things
that
return
false,
yet
still
did
work
and
promise
that
they
did
work.
I,
see
you're,
saying
they
all
return
off
in
shadows
when
they
are
promising.
I
think
the
ones
that
we
have
if
they
return
false
bytes
written,
is
zero
and
that
they
hopefully
didn't
actually
write
any
bites
before
saying
bye
to
zero.
L
G
A
A
I
U
I
Added
in
the
future,
run
if
first
name
or-
and
that
was
not
saying
that
this
has
another
decided
I
can
like
I-
can
see,
use
cases
for
to
lower
in
place
to
operations,
but
those
are
removed
by
the
way
some
amount
up
to
that
example.
Last
week,
I
think
the
German
SME
refer
are
we
sure
the
two
lowers
who
offer
will
always
preserve
the
number
of
Arthur's
numbering?
Yes?
Yes,
that's
one
on
Tiger,
two
dresses:
okay,.
N
N
I
Q
Q
AA
O
O
Expand
which
then,
how
do
you
fit
there
straight
or
super
checked
out
I
do
the
other
way.
D
O
AA
AB
D
AB
G
M
D
G
I
H
J
D
O
B
AA
O
O
A
B
A
B
G
L
L
E
A
A
O
A
A
But
now
we
just
decided
we're
going
to
have
them
all
in
all
places,
because
we
will
correct.
We
will
basically
have
three
cold
paths
in
there
in
the
kind
of
you
know.
If
it
doesn't
run
on
car
CLR,
it
will
try
to
get
span
out
of
this,
I
mean
try
to
get
the
string
out
of
this
one
and
then
call
the
existing
api's
or
it
will
just
copy
basic
power
source
that
to
string
or
whatever,
there's
stinking.
M
I
I
Code,
easy
as
I
was
thinking
back
to
the
scenario
that
you
have
earlier,
where
you're
doing
about
your
transforms
one
after
the
other,
so
you
have
your
source
and
your
destination
and
then
your
destination
becomes
your
source
for
the
next
country,
which
means
that
that's
actually
stand
up
to
you.
Instead,
yeah.
A
O
A
A
O
D
D
D
A
D
AA
L
N
I
A
A
N
N
B
N
D
I
know
I,
think
I.
Think
honestly,
the
way
it
happens,
this
people
write
code.
The
code
works
in
the
three
cultures,
they
happen
to
care
about,
and
that's
it
I,
don't
think
they
actually
understand
what
the
differences
are.
Now,
if
you
already
poured
the
code
to
span,
there's
a
million
things
you
can
do
wrong
anyway,
so
my
argument
would
be
I
would
be
more
with
Chris
up
and
say
for
this
API
service
I
think
they
would
be
better
to
say,
okay,
this
is
our
chance
of
faces
and
say
you
can
get
any
behavior
you
want.
D
AC
First
of
all,
there
is
a
fix.
This
is
suggested
to
give
you
culture
of
the
time
which
is
kind
of
a
pain
on
some
of
them
and,
second
of
all,
most
people
do
culture
ordinate
what
they
care
about
is
upper
case
if
he
solved
upper
case
and
lower
case
in
a
culturally.
That's
where
the
cultural
awareness
of
the
upper
case
and
lower
case,
but
that's
why
people
use
it.
He
believes
ordinal
when
they
don't
want
a
you
know
they
they
that's,
why
he
was
worried
and
so
there's
it's
very
better
case.
AC
E
A
D
D
D
N
N
AC
If
you
did
that,
instead
of
making
a
lack
of
it,
be
still
legal
but
means
something
different,
that's
what
I
was
going
I,
don't
like
that
to
say
that
if
you
allow
it
to
be
optional
and
then
the
default
is
different
than
it
is
in
the
non-spam
version.
That's
when
I
think
that
you're,
really
confusing
people,
if
you
now
require
it
and
default,
is
something
different.
I
think
it's
my
default
it
to
what
you
think
is
right.
As
long
as
it's
going
to
be,
there
I
think
that
would
be
okay.
D
N
D
V
D
Clearly
it's
years,
lugu
sticky
is
but
but
that's
something
I
think
as
developer.
I
would
be
like
I'm
I'm,
more
inclined
to
say
if
I
want
them
to
be
like
identical
I
understand,
maybe
there's
some
option
after
passing
to
make
them
at
people
like
medieval
assumption
would
be
it
gives
it.
If
the
memory
is
different,
then
I
would
expect
them
to
be
different.
I
think.
D
N
D
AC
D
How
this'll
so?
Ok,
let
me
tell
you,
I,
think
this
one
and
the
reason
it's
wrong
is
that's
effective.
What
at
ex-cop
tried
there?
Won't
you
every
single
time
it
didn't
pass
in
the
culture.
Guess
what
people
do
string
comparison
slightly
different,
a
culture.
You
say
that
but
like
in
the
end,
the
problem,
is
you
open
up
there,
the
intellisense
on
string,
comparison,
there's
like
six
values
and
the
thing
we've
seen
is
a
busy
set
time
and
time
again.
D
If
you
make
people
choose
like
the
thing
is
on
average,
what's
your
chance
that
the
person
typing
the
code
in
makes
it
better
decision
than
we
do
and
on
average
I
would
say
the
answer
is:
usually
they
don't
they
just
pick
a
random
value
at
some
point
like
I
would
say
if
we
can
make
a
default,
that
is
better
than
a
random
choice.
It's.
A
A
D
I
mean
I,
guess
the
other
thing
I
try
to
get
it,
which
I
think
I'm
not
doing
the
job
of
it's
like
take,
for
example,
the
the
semantics
we
have
with
all
the
parts
numbers
in
the
primer,
whether
you
pass
in
the
culture
right
so
unfortunately
made
the
choice
that
the
default
is
foreign
culture
or
to
string
same
thing
is
the
default
of
the
current
culture.
So
there's
two
kinds
of
bucks
you
can
have
based
on
the
default
right.
If
the
default
is
that
say,
invariant
or
ordinal
or
whatever
the
worst
part
you
can
have.
D
D
Then.
If
things
don't
look
right
on
screen
or
consumers
on
the
look
at
output,
they
don't
think
the
output
makes
sense.
But
if
you
return,
for
example,
two
things
or
two
strings
are
considered
equal
when
they
do
not
have
that
entacle
bytes,
it's
the
same
kind
of
boggle
sidewall.
They
should
have
been
equal,
but
we
said
they're
not
I,
think
there's
a
kind
of
bug
then
Oh
to
do
two
different
bytes
return.
True
because
it
may
you
know
school,
your
caching
or
whatever
else
you
do
when
you
actually
make
assumptions
about
that
I.
Don't.
AC
D
D
AC
You
have
a
scenario
where
well,
there
are
clearly
scenarios
where
it
would
be
better
to
have
the
default
to
be
ordinal
and
say
that
there
that
they
have
to
be
exactly
match
that
you
and
I
could
come
up,
particularly
you
I
focused
scenarios
where
it
would
be
absolutely
this
is
your
point.
You're
thinking
you
I,
you
think
what
the
user
sees.
The
user
goes
with
the
heck.
Do
you
talking
about
that?
Those
things
aren't
saying
in
fact,
they're
coming
out
even
looking
the
same
on
my
screen,
I
can't
tell
the
difference.
I'm
absolutely
broken.
AC
AC
Require
it
because
it's
a
freaking
mess
out
there,
it
is
any
default.
We
do
will
be
misunderstood
some
of
the
time
by
some
programmers
and
if
they
do
what
you
must
suggest,
they're
just
going
to
picked
up
one
and
intellisense
it'll
be
on
screen
in
their
face.
When
they're
looking
at
it,
it's
going
to
actually
be
in
their
code.
It's
not
going
to
be
a
default.
They
have
to
look
at.
So
what
about.
A
AC
A
A
A
bit
more
advanced
developers
who
look
for
performance
and
correctness
and
then,
if
we
kind
of
discovered
that
people
don't
like
the
situation,
we
can
always
add
that
involved
right.
But
at
that
time
maybe
they
will
already
like
people
who
use
the
ID
ice
will
learn.
Maybe
we
will
do
it
in
five
years
and
the
default,
and
by
that
time
everybody
will
know
that
you
know
well.
I
was
just
very
different
than
today
yeah.
What
I
would
say
like
we
don't
know.
R
A
AC
AC
D
Main
people
already
yelled
today,
I
mean
like
I
mean
this
is
like
super
popular
thing
that
hands
him
a
push
at
some
point
to
say
you
know,
string
equals
and
there
was
a
few
others,
but
I
said
I
don't
want
to
pass
in
the
thing
at
the
end,
because
it's
a
lot
of
text
like
people
yell
at
us
for
that
already
today.
So
I'm
I,
don't
think
like
five
years
longer,
will
give
us
any
more
insight
at
the
last
twenty
years
haven't
given
us
or
15
years
realistically,
because
Polly
previously
public,
but.
D
A
A
I
A
D
A
AC
A
S
N
N
AB
O
AC
I
I
A
I
would
say,
like
you
know,
people
who
don't
understand
it.
They
just
want
default
for
a
very
complicated
space
where
they
should
care,
they
don't
care,
they
should
just
use
string
idea,
it's
like.
If
you
hear
you
like.
How
did
you
get
this
pan
of
char?
You
probably
thought
jeez
I
want
none
allocations.
I'm,
like
you
know,
this
is
a
bit
of
more
sophisticated
software,
the
same
developers
who
care
that
deeply
about
birth.
They
care
about
correctness,
probably
a
bit
more
as
well,
and
they
want
to
be
explicit
about
what's
happening,
not
like
I.
N
N
AC
You
agree
that
they're
thinking
a
little
more
deeply
about
their
code.
They're
they're,
not
they're,
not
just
trying
to
get
down
the
fastest
thing
that
does
the
job
they're
truck.
They
are
thinking
more
deeply
about
code
in
general,
and
so
this
small
annoyance
of
picking
an
extra
thing.
If
it's
not
the
person
who's
like
doing
that,
you
know,
but
if
the
person
who's
actually
thinking,
what
is
my
code
doing?
How
is
it
doing
it?
I
AC
Two
there's
a
couple
of
overloads,
one
of
which
does
require
just
a
single
chart,
but
then
this
replaces
the
string
and
does
everything
it's
possible.
People
complain
and
ask
for
and
overload
that
they
don't
have
to
have
a
span.
They
can
just
pass
the
kick
to
char
in
which
is
what
they
do
for
one
of
the
overloads
right
now
they
might
ask
for
that.
So
so.
A
O
AC
A
It's
just
they
ignore
and
the
circuits
interesting
yeah
there's.
I
Otherwise,
correctness
well
otherwise
replace
will
always
always
always
refer,
not
enough
input
data,
because
how
do
you,
how
does
the
developer
ever
say
like
there
is
no
more
data?
After
this,
you
have
a
boolean
and
the
data
flag,
I'm,
not
sure
kinda
said
what.
So,
how
do
you
know
that,
like
the
source
side
passing
it
to,
you
isn't
some
is
part
of
like
you
know
the
collected
works
of
Jason
it
could
be,
but
like
it
would
be
silly
to
assume
that
so
we
would
never
return,
for
instance,
on
the
compiler.
I
I
Then
I
say
that's
dangerous
is
because
imagine
a
developer,
who
is
going
to
replace
a
larger
source
with
a
smaller
replacement,
so
he
allocates
his
destination
to
be
the
same
size
as
the
actual
incoming
data,
so
he
noticed
that
it's
never
going
to
fail
now.
He
just
calls
replace
without
checking
the
return
value,
because
why
would
it
return
anything
under
the
dock?
Well,.
A
So
we
should
maybe
we
should
talk
about
this
offline
because
we
spent
so
much
time
on
this.
The
other
alternative
is
just
not
correct,
so
you
feed
basically
chunks
of
spans
into
our
pipe,
and
it
will
give
you
incorrect
data
if
you
want
to
replace
a
B
but
a
BS
in
one
span
and
B's
in
the
other
yeah.
This
is
life.
This
is
basic.
A
Right,
that's
what
I'm
saying
it's
I,
don't
think
we
can
return
operation
status
and
not
return.
I
need
more
data
when
a
is
found
at
the
end,
but
I'm
searching
for
a
B,
because
it
will
be
basically
incorrect
code.
If
you
run
it
through
the
piper.
Do
you
agree?
Yes,
no,
no
I
never
saw
this
is
Python
as
I
saw
this,
then
their
shows
in
return
into.
A
Was
that
about
like?
No,
we
stopped
here,
plus
that's
exactly
what
the
API
is.
That
I
was
the
semantics
concept
tool,
because
there
is
a
Corral
yeah.
There
are
put
parameters
right,
so
it
will
basically
consume
up
to
the
last
a
and
it
will
say
this
is
how
much
in
consumed
I
cannot
make
a
decision
about.
G
A
We
don't
want
the
API
to
go
to
basically
be
used
in
and
with
a
very
usable
routine.
That
does
the
stitching.
So
if
we
don't
want
the
API
to
be
used
in
this
routine,
it
should
be
returned
operation
service,
because
basically
people
will
just
use
it
and
it
would
be
not
correct
and
it
will
be
not
correct.
Only
like
weird
edge
cases
when
a
and
B
are
at
their.
You
know
at
the
boundary
between
two
different
spans,
but.
O
A
Fine
with
going
back
to
boolean
and
basically
just
saying
this
is
an
API,
you
have
to
have
the
whole,
you
know
this,
it
doesn't
work
with
chunked
spans,
it
has
to
be
the
whole
span
and
the
destination
has
to
be
large
enough
for
the
replacement
work.
If
it
doesn't,
then
we
return
false,
but
will
still
returns
must
found.
N
O
O
N
N
I
A
A
AC
H
I
N
I
If
I
mean
yes
I'm,
you
can't
get
away,
so
you
need.
You
need
to
keep
track
of
how
many
different
items
you've
seen
like
where
all
of
those
indexes
line
up,
which
means
that
you
actually
have
a
separate
allocation
under
the
covers
string.
Don't
replace
today
already
has
a
separate
allocation
separate
from
the
actual
output
string.
That's
okay
and
the
reason
that
it
has
that
allocation
is
to
keep
traffic.
Weird
sync
data,
the
spam,
don't
replace
method
would
need
the
same.
I
D
A
H
I
A
Like
I
think
this
one
would
be
used
like
as
the
signature
is
today,
it
will
be
used
for
short
strings
like
I.
Have
you
know
like
20
characters
and
I
know
that
I
will
be
replacing
a
B
with
a
so
it's
going
to
shrink
so
I'm
gonna
just
allocate
destination
that
is
basically
the
same
length
as
their
source
and
I'm
going
to
call
it
and
not
done
anything
more
sophisticated
I
think
we
need
to
like
specifically
design
the
API
and
make
it
I'd
line,
makes
you.
I
AC
Y
A
A
Return
false
and
the
reason
for
I
kind
of
you
know
the
other
AP
is
it's
nonsensical
to
call
to
lower
as
a
smaller
bubble
this
one?
Yes,
the
mainland
scenario
is
what
I
discarded
it
will
not
grow,
but
there
are
some
other
scenarios
like
we
have
many
API
scene
already
in
the
framework
where
they
deal
with
small
buffers.
So
what
they
do
is
they
attempt
to
do
the
operation
and
if
it
they
cannot
complete,
they
return
false
and
then
you
just
double
up
the
buffer
size
and
it
will
succeed
next
time.
AC
I
A
I
D
U
AC
You
pull
that
out
and
you
don't
have
that
at
all
and
what
you're
getting
is
that
if
the
span
is
too
small,
the
destination
is
too
small,
then
they're
going
to
wind
up
with
an
exception
which
is
going
to
hurt
their
performance
and
because
tire
design
is
not
to
hurt
their
performance
to
Purvis
you
this
this
exists,
so
that
it
is
that's
why
this
exists
right,
but
we
can't
get
enough
information
that
that's
it.
Sir.
No.
I
AC
Don't
replace
make
sense
to
have
nothing
I'm
saying,
try,
replace,
doesn't
seem
like
it's
giving
you
much
the
confusion
and
like
I,
said,
I
read
it
wrong.
Nothing
think
it
was
doing
what
it's
doing
and
so
I
feel,
like
maybe
I'm
the
only
one
in
the
planet
that
will
ever
look
at
that
and
have
the
expectation
that
I
had.
That
was
completely
wrong.
It
is
possible.
It's.
AC
N
I
Like
the
pipeline
API,
because
in
the
case
that
you're
describing
you
want
to
be
able
to
do
partial
data
and
then
the
next
time
you
call
this
method
just
pass
in
like
a
different
window
of
the
input
string
at
a
different
output
buffer
and
then
you'll
just
manual
stitch
them
all.
Look
at
me
up
right
right.
A
This
is
what
I'm
seeing
adding
this
will
not
as
complicated.
So
so
what
about
this?
We
kind
of
ran
out
of
time.
What,
if
we
just
remove
it
and
basically
this
one
method
and
try
to
design
it
and
add
it
later,
because
it
seems
like
you
know,
it
would
be
good
to
support
the
pipeline
scenarios.
Maybe
have
some
measure
api's
for
some
other
things,
and
it's
not
like
you
know,
replace,
is
critical
to
be.
Anybody
opposes
to
this
plan.
Wes
is.