►
From YouTube: Release Management Think Big #8
Description
Welcome to the Think Big Session on Release Management
https://about.gitlab.com/handbook/engineering/development/ci-cd/release/release-management/
A
A
A
A
B
Weird,
that's
why
I
went
to
the
city
centre
and
I
swear,
no
like
no
tourists,
no
bikes,
nothing!
And
it's
so
weird,
because
it's
just
this
area
in
the
city,
where
you
get
all
the
crazy.
You
know,
people
that
are
just
out
to
smoke,
weed
or
whatever
wider
stupid
souvenirs.
It's
just
so
hectic
all
the
time
and
it's
just
a
completely
different
world.
But
at
the
same
time
you
can
take
great
pictures
so
yeah.
A
C
Sorry
I'm
a
bit
like
guys
I,
just
when
I
saw
for
the
first
time
station
holiday.
D
D
Welcome
to
our
think,
big
number,
eight
for
release
management.
We've
had
a
couple
of
different
topics
to
discuss
here.
I,
wonder
if
Nathan's
going
to
make
this
meeting
I
might
ping
him
in
a
bit.
But
Shawn
looks
like
you.
If
the
first
went
up.
Do
you
want
to
kind
of
voice
over
what
you're?
What
you're
asking
here.
C
This
is
probably
some
features
right
and
this
is
a
release,
but
but
now
we're
also
now
associating
releases
with
milestones
and
in
and
now
we're
talking
about,
you
know
a
period
and
so
I
kind
of
I'm.
Just
wanting
I
guess
we
will
look
at
the
gitlab
release.
That's
looking
at
it
from
a
perspective
right,
like
it's
a
month.
Is
that
the
kind
of
model
we're
going
for
here
so.
D
Fire
release,
I
think
one
of
the
disconnects
that
we
have
is
that
a
release
can
be
a
single
commit
which
is
the
release
creation,
and
then
people
artificially
attach
information
to
that
release
because
they
can
link
anything
to
it.
They
can
attach
anything
to
it,
and
then
it
becomes
a
container
for
a
bunch
of
information
that
is
representative
right.
It's
not
that
Sara
Lee
truthfully
every.
D
So
I
think
when
we,
when
we
talk
to
you
Gabe
and
plan
his
perspective,
is
that
releases
should
be
more
about
the
merge
requests
in
the
commits.
So
we
should
be
focused
on
assigning,
maybe
a
release
version
to
an
issue
to
an
mr,
because
today
we
don't
really
have
a
great
sense
of
what
Mr
implements
an
issue,
because
an
issue
is
very
sprawling
and
in
mr
doesn't
necessarily
it's
not
a
one-to-one
relationship.
D
C
Okay,
that's
that's
the
nice
things
and
so
yeah
come
on
so.
D
C
D
Somebody
could
be
using
a
milestone
to
show
a
quarter
quarterly
release
schedule.
So
then
they'll
have
multiple
releases
assigned
to
a
milestone
because
that's
rolling
up
to
a
quarterly
release
schedule,
or
they
have
multiple
iterations
and
sprints,
and
a
milestone
is
representative
of
every
two
week
releases
and
then
the
release
object
is
where
they
upload
a
bunch
of
images
or
binaries
for
them
their
customers
to
consume
and.
B
C
Yeah,
we
think
I'll
actually
rephrase
it
slightly
because
I
said
commit,
but
I
mean
if
we
have.
If
we
have
a
release-
and
we
have
you
know
like
I,
set
period
for
the
release
and
then
we
may
have
merge
requests
well,
if
they're
both
tricky
actually
cuz
commits,
you
might
have
been
working
on
something
for
ages.
Right
I
might
be
on
a
branch
and
in
an
end,
and
then
the
merger
quest,
though,
is
when
you've
actually.
E
C
E
D
I
want
this
epoch
to
represent
a
given
period
of
time
and
not
be
inherited,
or
we
can
inherit
it.
So,
for
example,
I've
seen
three
use
cases
of
a
customer.
The
first
one
is
I'm.
Gonna
use
a
release
to
track
all
the
activities
are
run
a
deployment,
so
I
have
a
release.
That's
going
to
just
going
to
start
well
before
any
changes
are
made
to
production
and
then
I'm
going
to
have
a
date
that
is
representative
of
two
weeks.
D
After
all,
of
the
desired
changes
to
production
have
been
made
so
that
we
can
validate
defect
leakage
or
we
can
triage
issues
that
are
spun
out
of
the
release
version
tag.
So
that's
one
use
case.
Another
use
cases
I
just
want
a
representation
of
a
version
that
we
push
to
production
or
a
snapshot
of
our
continuously
deploying
application,
so
that
I
can
then
sample
when
I
get
audited.
D
So
that's
the
second
use
case
and
a
automated
100%
world,
where
a
date
is
going
to
be
pretty
arbitrary,
they're,
just
looking
to
take
a
regular
snapshot
of
production
so
that
they
can
then
surface
that
during
an
audit
or
compliance
need,
and
in
the
last
used
cases,
I
want
to
have
a
historical
record
of
planned
versus
released
changes
to
production.
And
that's
where,
like
the
release,
page
progress
view
is
made
and
designed
for.
D
C
C
D
C
D
Like
it'll
default
to
release
that
from
a
start
and
end
date
and
then
if
somebody
manually
specifies
a
start
and
end
date
that
will
take
like
that'll
be
the
new.
The
new
dates
but
default
will
be
released
at
because
the
assumption
is
they're,
creating
the
release
at
the
moment
of
deployment.
In
that
case.
So.
C
If
we
had
those,
maybe
I'll
just
share
my
screen
or
everyone
can
look
at
the
dock,
I
guess,
but
if
we
had
it
would
we
then
would
we
if
we
had
these
four
fields
so
collected,
that's
different,
but
if
we
had
style
a
starts
header
names
there
and
then
no
one
specifies
any
day.
Do
they
all
just
become
released
at
all?
Three
of
them
are
released
at
the
same
so.
B
D
We
should
we
should
allow
that
so,
for
example,
they
create
that
they
create
the
release
tag.
Actually
they
create
a
tag
and
a
bunch
of
people
are
making
changes
and
then
their
most
recent
tags.
They
decide
that
that
tag
is
gonna,
be
the
release
version
that
they
want
to
declare.
So
then
they
take
that
tag
they
might
update
it
via
the
API
in
the
future.
Hopefully
they'll
be
able
just
to
you
know,
create
a
release
from
that
tag,
and
then
it's
shows
up
on
the
release
page
like
today.
D
You
would
take
that
tag
and
you
would
add
release
notes
so
that
it
shows
up
on
the
release.
Page
like
I
said
somebody
would
do
that
new
Y
the
day.
That's
how
people
are
doing
it
in
the
UI
the
day
is
that
they
select
that
they
cherry
pick
that
tag,
and
they
say
I
want
this
to
be
my
release
version
and
then
they
start
uploading
arbitrary
asset
links
to
it
via
the
API
to
give
them
like
images
or
links
to
the
container
registry
or,
like
runners,
a
really
good
example
for
that
and.
D
C
Okay
and
then
to
the
Irate
and
also
the
reason
I
put
the
evidence
collected
at
or
near
asleep,
just
to
kind
of
highlight
that
you
know
when
the
evidence
we
record
that
when
the
evidence
object
is
actually
created,
so
it
could
be.
For
example,
you
know
half
a
second
after
the
release
is
created
or
if
you
had
a
future
release
data,
it
would
be
then,
but
it's
not
necessarily
precise
because
of
the
delay
in
background
processing.
D
C
D
It's
kind
of
a
weird,
because
it's
an
if
and
logic
to
kind
of
go
through
what
nathan
is
poking
it
out
right
now,
it's
overlapping
with
milestones
like
the
release,
may
not
be
associated
with
a
milestone.
So
then
it's
gonna
default
to
release
dot
until
somebody
puts
a
start
and
end
date
in
it
and
then
if
they
do
at
a
later
date
associated
milestone,
it
should
then
override
to
the
milestone
dates,
but
then
they
may
manually
change
those
dates
to
how
the
different
start
and
end
dates,
even
though
they
associated
a
milestone.
D
So
when
I
think
about
it
like
we
have
a
12.8
release
and
the
12.8
release
then
has
the
potential
to
have
a
12.1
release
for
security
patches
and
the
12.8
dot.
Two
four
or
five
for
security
patches
in
between.
So
the
release
cycle
isn't
just
the
seventeenth
that
the
first
months
to
the
22nd
of
the
next
month,
it
really
is
like
the
seventeenth
to
the
30th,
because
there's
all
these
patches
that
happen
afterwards,
so
the
end
date
will
be
the
30th.
D
D
I
can
also
very
open
to
creating
a
blog
post
to
and
we
should
potentially
like
talk
to
Moran
on
the
delivery
team.
Niveen
did
I
say
that
wrong
I'm
waiting
for
hi
on
as
I
brought
I
go
up.
She's
like
oh
this
white
person
saying
her
name
and
we
can
have
him
dog
food
it
with
his
team
and
give
us
a
feedback
on
like
if
we
should
be
prescriptive
or
not
says
the
risk
that
we
run.
C
D
And
I
don't
want
to
violate
our
like
cardinal
rule
of
no
enforce
for
clothes
and
it's
hard
not
to
do
that
for
releases,
because
releases
are
naturally
people
think
about
them
as
a
workflow
archetype
like
I
plan,
a
release,
I
do
things
really
release
I,
release
the
release
and
then
I
track
metrics
to
that
release.
So
very
it's
a
very
linear
process
for
many
companies,
but.
E
C
Okay,
so
I
mean
for
me
if
I
look
at
it
just
purely
from
the
backend,
it's
relatively
simple
right.
We
would
just
add
these
two
fields.
We
would
extend
the
API
so
that
so
that
when
you
create
a
release,
if
there's,
if
they're
not
provided
they
default
to
the
release
date,
which
is
also
a
default
itself
and
and
then
have
a
way
for
those
things
to
be
changed
through
the
API
and
then
I
think
the
risks
will
be
up
to
you.
Nathan
I
mean
I,
you
know
I
mean.
C
B
B
So
a
lot
of
people
are
confused
about
when
the
tag
would
officially
be
created
and
what
would
happen
with
the
evidence
for
this
release
that
has
a
different
start
date
from,
for
example,
the
milestones
that
are
associated
to
it,
and
also
about
visibility
of
all
these
assets
and
all
this
information.
But
mostly
what
is
my
tag
created
now?
It's
my
evidence.
B
C
An
interesting
question:
hi
Ana
because
yeah,
so
if
we
just
take
the
simplest
case-
and
we
just
put
a
future
date
well,
then
the
release
gets
created
there.
But
then,
if
we
have
created
the
release
and
now
we
want
to
change
the
dates,
what
we
do,
then
you
know
we
do.
Maybe
we
destroy
it
and
clear
again:
maybe
that's
the
soup
list
way
to
do
it.
Oh
yeah.
D
D
D
D
D
D
So
we
can
decide
to
say
that
if
you
have
a
patch-
and
it
is
correcting
something-
that's
already
been
released-
that
it's
a
new
release,
we
can
say
that
that's
how
we
support
releases,
but
people
typically
like
to
see
the
relationship
of
a
patch.
But
that's
where
we
recommend
semantic
versioning
is
just
to
show
the
major
release,
and
this
is
the
sub
release
of
the
major
release.
But
then
we
get
kind
of
conflated
with
the
release
dates
going
back
to
Nathan's
point
like
yeah.
C
D
So,
release
sorting
by
release
date
makes
that
weird,
but
we
I
think
if
I
was
to
look
at
the
problem
that
we
need
to
solve,
first
and
foremost
with
the
releases
page
and
these
dates,
we
need
to
allow
users
to
say
that
their
release
activities
are
going
to
be
outside
of
a
particular
deployment
window,
which
may
mean
that
a
release
may
end
in
the
middle
of
a
second
release.
That's
already
been
started.
F
One
question
you
had
mentioned
about
deleting
and
recreating
releases
in
my
mind,
that
kind
of
implies
that
you
would
be
either
creating
the
release
off
a
new
tag
or
you'd,
be
somehow
moving
the
tag
and
usually
moving
a
tag
is,
is
not
considered
best
practice
like
it's
time
kind
of
a
you
have
to
to
move
it.
You
have
to
like
force
push
the
the
tag
which
is
kind
of
a
generally
a
bad
thing.
C
D
Okay,
that's
a
good,
a
good
call-out
Nathan!
Thank
you
for
that.
So
let's
just
decision
point
is
update
the
evidence.
B
Users
say
that,
like
four
releases
custom
date
they
or
whatever
takes
it-
would
be
weird
to
have
assets
or
to
have
evidence,
attitude
and
upcoming
release,
because
only
having
assets
makes
sense
for
something
that
you
actually
working
with,
not
something
that
you
know
for
the
future.
So
maybe,
for
example,
your
tag
doesn't
exist
yet
or
which
does
make
sense,
because
you
need
to
have
a
tag
to
create
a
release.
B
But
if
you
don't
have
assets
to
add
to
your
release
just
yet,
why
would
you
create
a
release?
So
that's
also
one
of
the
open
questions
from
users
about
the
starts.
Not
that
that's
what's
confusing
to
me
and
also
some
people
were
the
start
at
and
they're
released
at
right,
so
they
they
kind
of
overlap.
And
how
would
we?
How
do
we
want
to
to
prescribe
this?
You
I
because,
for
example,
with
you
at
a
milestone,
the
milestone
would
automatically.
B
C
C
So,
for
example,
say
you're
a
you
know:
well,
I'm
a
maintainer,
an
open
source
project
on
another
platform,
and
and
so
what
I
do
is,
as
people
make
pull
requests
and
and
they
get
merged
to
master,
so
they
merge
to
master.
But
it's
not
released
right.
So
in
fact,
so
that
what
I'll
do
is
I
create
a
release,
but
it's
in
it's
in
a
draft
form.
C
So
as
the
maintainer
I
can
see
it
and
I
can
add
to
it
and
everything,
but
no
one
else
can
see
it
unless
they're
also
maintained
or
whatever,
and
so
then
I
can
put.
You
know,
for
example,
so
on
so
did
this
request
thanks
very
much
etc?
And
then
it's
you
know
it's
a
ruby
thing.
So
then,
when
it's
published
to
Ruby
gems,
that's
when
I
consider
it
released
and
that's
when
that's
the
you
know
they
use
the
bird
of
the
version
number
and
everything,
and
so
so
they.
C
So
that's
a
reason:
I
guess
we
would
want
to
have
the
ability
to
add
to
it
before
it's
actually
released,
but
it's
also,
it
also
kind
of
highlights.
I,
don't
know
if
we
want
this
feature
where
yeah
I
draft
release
that
that
isn't
visible
to
other
users
and
in
fact
I
was
adding
that
to
an
issue
this
morning
and
then
the
other,
the
other
use
case
might
be.
If
you
want
to,
if
you
had
some
other,
maybe
you
just
had
some
other
assets.
You
wanted
to
add.
D
This
is
a
toggle.
That's
at
the
project
level
hide
upcoming
releases,
because
I
think
the
intent
of
upcoming
releases
was
to
recreate
the
github
functionality
of
straffe
releases.
So
I
think
that
this
is
more
like
a
visibility
thing
that
we
should
allow
users
to
set,
which
is
a
configuration
item
not
really
into
that.
But
I
think
that
it
might
be
something
to
think
about.
B
And
we
also
had
some
people
say
that
they
would
like
in
the
future,
for
example,
so
you
just
specify
some
specific
users
that
would
get
access
to
that
release,
so
people
that
are
not
necessarily
involved
in
the
great
Cola
but
I
don't
marketing
people,
whatever
just
some
external
costumers,
just
someone
some
people
that
could
visualize.
Could
you
get
access
to
that
and.
F
Right
yeah,
it
gets
tricky
because
guess
technically
aren't
supposed
to
see
any
tag
information,
but
our
release
are
so
tied
into
tags
that
the
information
we
can
actually
show
them.
It's
very
limited,
yeah.
C
Well,
I
mean
it's
also
at
a
slight
sidebar,
but
so
apparently
so
what
no
apparently
but
on
a
public
project.
The
guests
can
see
both
the
code
and
and
releases,
but
on
a
private
project.
They
can't
see
the
code,
and
so
it
doesn't
make
any
sense
for
them
to
see
the
release.
I
don't
know
they
currently
can,
but
I
mean
there's
this
bug,
of
course,
but
aside
from
the
bug,
here's.
D
A
use
case
that
I
think
we've
seen
about
a
couple
of
customers
who
are
using
gitlab
for
inner
sourcing
exclusively
that
they
want
their
inner
sourcing
companions
to
be
able
to
see
that
this
has
been
published.
And
then
they
know
that
they
can
go
to
the
artifactory.
And
if
they
have
permissions
to
to
download
that
image
and
then
use
that
as
a
part
of
their
prescribed
image
for
their
inner
source
and
campaign.
But
they're
not
using
a
public
project
because
they're
usually
usually
restricted
to
contractors
or
customers
of
that
product.
C
C
D
D
Right
and
I
would
say
that,
unfortunately,
our
releases
feature
is
really
popular
in
three
camps.
One
of
the
camps
is
inter
sourcing
and
other
camp
is
embedded
systems
and
both
of
those
customers
or
benefit
from
published
private
route.
Private
projects
like
publish
releases
or
private
projects,
because
embedded
systems
would
take
that
image
and
deploy
it
to
a
radio
system
or
an
RV
system
software
and
then
ship
that
RV
to
the
customer
to
the
end
user,
RV
and.
C
So
maybe
adding
to
what
high-end
I
said
earlier
about
do
we
have
I
mean
it's
going
to
make
this
whole
feature
quite
complicated,
but
do
we
have
this
way
of
saying
the
easy
use
this
particular
user?
Can
you
know,
have
access
control
to
the
release
and
this
other
one
can't
we
did
what
we
were
saying
hang
on
yeah.
D
Think
we
should
talk
to
Dennis
again
hi
Anna,
because
he's
one
of
the
embedded
systems
folks.
So
he
would
be
able
to
kind
of
illustrate
how
he
was
using
github
and
we
can
probably
do
a
comparison
of
like
what's
his
expected
use
case
because
he's
really
gunning
for
us
to
be
able
to
upload
binaries
and
files
via
the
UI,
because
he
wants
to
be
able
to
publish
these
these
releases
so
that
his
customers
who
are
consuming
the
software,
can
download
it
onto
their
hardware
systems.
So,
but
he
doesn't
want
them
to
see
the
software.
C
D
D
So
we
need
to
be
able
to
reach
the
same
use
cases
and
solve
those
problems,
but
not
on
the
exact
same
way.
I'll
take
I'll.
Take
this
note,
but
Nathan
kind
of
reading
your
item
on
just
not
allowing
users
to
manually,
specify
start
and
end
date,
you're
just
saying
to
go
ahead
and
just
use
the
milestone,
dates.
F
Yeah,
just
maybe
it's
very
possible
I'm,
just
not
understanding
all
the
use
cases,
but
to
me
it
feels
like
release
is
all
about.
This
is
like
snapshots
of
of
when
your
code
is
released,
and
then
milestones
are
more
for
the
time
portion
of
that
like
time,
boxing
things
and
so
I
thought
that
was
kind
of
a
nice
different
like
when
we
associated
milestones
with
the
release.
That
was
a
good
way
to
do
it,
but
now
we're
adding
a
start
date
to
a
release
and
I
was
trying
to
think
what
that
really
means.
F
If
I
go
like
to
another
another
software
package,
summer,
I
don't
really
and
I
see
the
list
of
releases
I,
don't
really
know
what
it
means
if
it
says
this
release
started
at
this
point
of
time.
It's
like
this
snapshot
started
at
this
point
in
time.
I
can
see
like
with
milestones
how
like
this
is
kind
of
the
period
of
time
where
we
worked
on
this
release,
so
I
guess
yeah,
maybe
I'm
just
it's
possible
I'm,
just
not
thinking
through
all
the
use
cases.
No.
D
I
think
you
are
I
feel
like
the
milestone
is
gonna
catch
like
90%
of
our
users
needs,
especially
as
we
start
to
expand
time
boxes
in
the
planned
stage,
where
they'll
be
able
to
refine
what
time
boxes
they
want
and
even
specify
custom
time
boxes.
So
maybe
we
just
let
plan
handle
the
time
boxes
and
we
scratch
this
Shawn
yeah.
C
C
D
Are
telling
us
they
want
to
be
able
to
manually
specify
the
release
date,
but
I
think
what
they're
saying
is
they
want
to
be
able
to
create
a
custom
date
range
for
their
milestone,
so
I
think
that
that's
feedback
I
can
get
back
to
Gabe,
which
he
already
knows.
This
is
we're
very
bridging
the
gap
between
people
who
want
to
see
that
delivered
versus
planned
symmetric
again,
that's
pretty
much.
F
C
A
I
think
you
nailed
it
with
the
custom.
It's
like
folks
are
trying
to
set
custom
milestones.
That's
what
I
was
getting
the
disconnect
in
my
head.
Is
it
felt
like
start
date?
End
date
was
almost
a
subset
of
a
milestone.
Basically,
but
what
did
those
dates
actually
do?
Is
it
capturing
commits
within
that
time
frame?
That
was
what
was
confusing
me
a
little
bit.
D
B
Be
because
who
was
less
user?
Did
we
talk
to
you,
but
the
third
one?
He
say
that
out
of
these
milestones,
what
I
want
to
just
be
able
to
just
specify
that
he's
going
to
be
released?
He
said
I,
don't
care
about
this
dark
dates
because
releases
on
the
flight.
We
don't
have
a
patched
version
whatever,
but
I
want
to
be
able
to
tell
my
team
that
this
is
what
to
be
released
next
Thursday
and
just
add
an
end
date
for
that,
for
example.
So
in
that
case,
you
don't
necessarily
need
a
milestones.
B
D
F
Okay,
so
you've
kind
of
ready
answer
this
Jackie
about
is
just
asking
how
much
work
it
would
be
to
associate
group
level
milestones
with
a
release
and
just
because
that
would
allow
us
to
dog
food.
That
issue
summary
feature
on
Ghaleb
comm
to
be
really
cool,
especially
cuz.
We
do
a
really
good
job
of
associating
all
of
our
issues
with
milestones.
So
that's
like
the
perfect
use
case
for
that
that
feature
so
it'd
be
cool
to
dog,
treat
that
yeah.
D
D
F
D
F
All
so
the
next
one.
These
are
just
two
small
small
things.
While
I
was
working
on
that
script,
to
update
the
our
releases
page
from
our
beliefs,
posts,
there's
a
couple
things
I
thought
of
that
I
I'm,
not
sure
how
we're
handling-
and
there
might
be
solutions
to
this,
but
I
just
want
to
throw
them
out.
So
the
first
one
is
the
order
that
we
currently
list.
That
page
in
is
the
release
that
date,
it's
kind
of
when
they
were
created,
or
you
can
also
specify
the
date.
F
But
it's
not
always
quite
as
straightforward
as
that.
For
example,
we
release
security
patches
and
we'll
backport
that
release
back
to
several
major
versions
and
so
technically
we'll
release,
let's
say
12
9
and
then
we'll
have
a
security
patch
that
will
apply
all
the
way
back
to
version
10,
and
so
now,
if
in
the
releases
page,
would
look
like
that
version.
D
Have
this
other
issue
that
says
allows
sorting
my
different
criteria
for
releases,
so
we
have
created
a
updated,
a
due
date
and
start
date,
but
I
don't
know
if
we
actually
have
a
start
date
anymore,
because
this
issue
was
built
off
of
the
started
issue
that
we
just
kind
of
Chi
washed.
So
maybe
that's
milestone
start
date,
but
I'll
link
that
here
too,
as
well.
I
I,
agree
Oh,
whoever
just
linked
that
great
job,
Thank
You
Hana,
look
at
you!
So
yes,
I,
agree
with
you.
I
think
alphabetically
would
be
the
release.
C
F
D
Have
I
think
Jake
has
an
open,
mr
to
fix
that
grammatical,
weirdness
and
actually,
let
me
show
you
something
that
I
was
playing
around
with
yesterday,
because
it
is
kind
of
a
weird,
a
weird
use
case.
So
I
created
this
thing
and
then
Sean
and
Nathan.
You
guys
helped
me
associate
milestones.
I
created
this
tag
right
here,
converting
it
to
release
and
it
shows
up
in
the
middle
of
completed
past
milestones
and
future
milestones.
So
it's
like,
if
it
doesn't
have
a
milestone.
D
A
Wonder
if
it'd
be
cool
to
let
folks
sort
by
major
release
like
almost
let
them
sort
by
semantic
versioning,
so
you
could
kind
of
get
that
overall
major
release
snapshot,
but
then
you
could
also
kind
of
plug
in
any
patch
releases
in
the
order
they
were
actually
released.
So
you
can
see
both
of
them
because
I
think
Nathan
said
this,
but
like
1278
actually
came
after
twelve
nine,
so
some
folks
might
want
to
see
that
I
don't
know
just
thinking,
brainstorming
out
loud
a
little
bit.
That's.
F
A
A
C
D
A
F
D
That's
the
way
that
we
can
validate.
Let
me
add
that
to
my
customer
question,
repository
I'm
like
how
do
you
want
to
see
releases,
because
really,
what
we're
noticing
is
that
people
actually
have
a
very
picky
selection
that
isn't
supported
by
normal
filtering,
like
they
almost
want
to
have
like
a
multi-select
like
I,
want
to
see
releases
that
were
created
by
a
machine
versus
a
person,
because
that
kind
of
suggests
a
hotfix
versus
a
regular
release.
D
People
also
want
to
see
things
that
were
releasing
across
multiple
projects
in
the
same
group
in
a
certain
time
frame,
so
that
they
can
then
trace
back
defect
leakage,
so
it
kind
of
is
like
they
want
to
be
able
to
select
multiple
filters
across
different
projects
to
see
releases
in
a
compare
way
so
like
the
usability
of
the
releases
page,
is
really
low.
Right
now
for
people
who
are
trying
to
compare
and
filtering
it's
going
to
be
more
important
than
just
sorting.
D
D
F
Just
kind
of
support
that,
when
you're
talking
about
different
views,
even
for
ourselves
on
our
releases
page
like
not
our
project
one
but
the
one
that
we
currently
publish,
that
one
only
contains
major
versions
because
that's
kind
of
a
marketing
page.
So
we're
like
that's
the
use
case
for
that.
So
in
that
case
we
don't
want
to
show
patch
and
security,
except
just
be
noise
right.
D
And
I
can
see
where
people
who
have
multiple
approvals
for
their
releases,
so
they
have
manual
jobs
that
then
prevent
a
tag
from
being
created
until
somebody
runs
that
job
and
then
it
creates
the
tag.
After
all,
of
the
commits
have
been
approved,
which
is
what
we
see
some
of
our
financial
services
customers
doing.
They
would
want
to
only
publish
the
release
after
it's
been
approved
and
be
able
to
show
on
the
release.
Page
only
approve
releases
for
everybody
in
the
project
or
group.
C
B
E
F
It's
just
a
it's,
not
we're
not
using
any
component
to
render
it
as
a
list,
but
I
think
that
I
think
that
this
sorting
item
could
work
with
anything.
Is
my
thought,
like
this
sorting,
I'm
guessing
this
component,
just
kind
of
tells
you
can
just
like
kind
of
query
it
and
ask
it
what
it's
sorting
is
and
then
sort
your
list.
The
way
you
want
so
I
think
we
could
probably
make
use
of
it.
That's
a
good!
That's
good!
Fine,
Hanna,
I'm.
F
D
F
Worries-
and
we
actually
kind
of
already
talked
about
this
I-
was
thinking
through,
like
it'd,
be
cool
to
create
upcoming
releases.
So
we
could
see,
if
forget,
lab
comm,
so
we
could
see
like
that
progress
bar
getting
filled
out
as
we
go
through
a
milestone,
but
I,
don't
think
we
create
the
tag
that
we
need
until
we
actually
deploy
it.
D
Yeah
you're
you're
you're,
definitely
like
honing
on
on.
Why
I
hat
way
like
I,
have
a
reminder
on
my
calendar
that,
after
we
cut
the
the
release
for
12.9
to
like
create
the
tag
manually
because
I
couldn't
find
the
tag
for
that
release
until
after
it's
done.
But
today
customers
are
like
tagging,
the
first
commit
of
that
release,
and
then
they
end
up
just
attaching
like
links
to
the
mr-s
that
are
committed
and
the
subsequent
releases
or
in
the
subsequent
tags.
D
D
F
D
F
F
C
If
they
currently
I'm
just
added
at
a
technical
issue
that
level
they
are
currently
I,
think
I've
said
this
in
a
couple
of
posts.
The
way
we're
using
tag
is
almost
like.
It's
the
ID
of
the
and
and
if
so
we
have
an
issue
where
there
are
some
that
are
blank
and
they're
causing
kids.
It's
causing
problems
like
I
shouldn't
I
shouldn't
have
happened,
I'm,
not
sure
how
it's
happened
yet,
but
and
so
they
are
completely
tired
and
yeah.
So
if
we
want
to
break
that
apart,
we
could,
but
we
would
it
would.
C
E
Always
you
know
had
a
commitment
to
maybe
a
fuzzy
hand-wavy
commitment
to
two
or
three
releases
in
advance
as
to
what
may
go
in
those
releases
and
be
able
to
project
manage
and
see
what
issues
are
epics
may
go
in.
Those
releases
is
really
helpful.
So
if
that's
decoupling
it
from
a
tag,
I
could
see
where
that
would
be
a
really
helpful
use
case
for
a
project
management
standpoint.
Overly.
D
Alright
I
know
that
we
are
at
time
or
overtime,
really
sorry
else,
I
might
schedule
another
meeting
with
you
guys
next
week
with
you
all
next
week
to
kind
of
dive
into
upcoming
draft
releases
and
to
think
about
how
we
want
to
manage
that
it's
from
a
vision.
Side
like
it's,
definitely
something
our
customers
need.
They
need
to
be
able
to
plan
and
orchestrate
releases
without
having
their
final
version
that
they're
going
to
deploy
out.
C
I'm
sure
sure
Jackie
no
problem
can
I
just
add
one
last
thing
before
we
leave
yeah
and
so
something
we
brought
up
a
while
back
and
I'll.
Just
you
know
little
bit
of
a
bomb.
My
guess
is,
you
know
who
wears
Exodus
to
the
evidence
object,
and
we
talked
about
maybe
that
being
in
order
at
a
role
and
and
the
reason
I'm
bringing
it
up
now
is
because
there's
kind
of
like
stuff
swirling
around
about
access
to
evidence
in
the
issue
issue
sphere,
and
maybe
we
can
discuss
out
of
the
next
meeting.
You
know.
D
Actually
I'm
think
I'm
going
to
today.
I
was
going
to
crazy
user
stories
for
release
evidence
because
we
don't
have
them
and
I
think
that
will
help
define
what
people
should
be
accessing
them
and
if
there's
a
use
case
for
maintainer
and
owner
Steven,
I've
access
to
it,
because
it's
weird
when
I
look
at
the
use
case
and
who
it's
interested
in
adopting
this
it's
exclusively
for
the
compliance
and
audit
personas,
so
I
am
weird.
I
am
worried
about
that.
Aaron
is
using.
You
know
the
Aaron
Goldenthal,
who
kind
of
raised
this.