►
Description
Today I learned how @gonzoyumo uses Issue Boards to manage Engineer level tasks and overall team capacity. We also discussed some of the underlying problems and how the sub-issue convention could solve them.
Here is the original issue that sparked the discussion - https://gitlab.com/gitlab-org/gitlab-ee/issues/2036
A
A
So,
either
way
explaining
why
we
are
starting
today
with
the
issue.
Birth
is
engineers
principle
for
completing
the
work,
which
means
tracking
the
changes
up
to
the
deployment
and
production
with
a
new
CD
workflow.
So
when
we
have
multiple
separate
pieces
in
the
same
issue,
it
makes
it
really
hard
to
do
this
tracking
and
Dragonite
a
day-to-day
workflow,
because,
for
example,
when
I'm
pushing
for
that,
my
engineers
are
that
kind
of
issue
bolts.
A
A
They
can
go
check
if
this
has
been
deployed
and
it's
working
on
prediction
as
I
can
check
for
stuff
in
review
if
they
need
to
think
main
10
years
to
see
all
the
review
is
going
on
something
like
that
in
there
what
they
are
focusing
on
and
and
so
on,
based
on
the
different
stages
on
the
other
issue.
So
if
you
have
one
single
issue
for
cross-functional
people,
you
cannot
do
that.
B
Well,
II,
why
can't
you
I
mean
like
I
guess?
If
so,
if,
if
you
have
conditions
for
when
an
issue
can
move
from
one
workflow
stage
to
another,
I
guess
like
the
question
is:
if,
like
you
have
something
in
dev
right
and
you
you
sit
there
and
you
know
that
it's
in
dev,
it
has
to
have
UX
done
before
it
can
move
out
and
then
it
has
to
have.
You
know,
like
the
engineer,
do
the
work
that
the
engineer
needs
to
do
so
you
have
like
two
people
that
are
working
on
it
and
I.
B
A
The
current
status
like,
for
example,
here
I
think
I.
Don't
remember
where
is
this
is
the
main
issue.
The
main
issue
is
in
open,
so
it
probably
should
be
in
dev,
but
they
can't
see
the
whole
picture
like
what's
the
current
state
of
the
overall
feature.
But
what
also
is
the
current
state
of
their
own
sub
tasks?
They
are
only
the
only
one
responsible
for
them,
and
you
cannot
do
that
without
digging
into
this
the
main
issue,
if
you
are
using
tax,
because
you
cannot
have
its
formation
from
an
issue
but
level
what.
B
Use
the
whole
view
right
here
for
tags
but
like
if
you
have
two
people
assigned
to
an
issue
and
they
both
have
tasks
assigned
themselves
and,
let's
say
you're
filtering
on
this
view,
if
there's
some
sort
of
indicator
that
the
person
who
offers
this
view
it
belongs
to,
they
have
a
task
that
they
stop
to
complete.
On
that
issue,
yeah
it's
like
as
a
visual
indicator.
You
know
this.
A
Is
what
part
of
my
reminds
whether
you
can
have
written
very
far,
but
main
of
the
limitation
that
we
are
today
if
you
can
put
that
in
an
issue?
I'm
perfectly
fine
dealing
with
just
one
issue
and
it
will
be
easier
simpler.
This
solution
convention
is
just
a
workaround.
The
fact
that
we
need
more
per
se
feature
on
an
issue
like
the
issue
waits
the
word
for
tracking.
So
if
T,
if
the
workflow
tracking
is
a
native
property
available,
that's
fine
like
if
you
have
a
per
se
me
issue.
A
Wait,
that's
fine,
because
the
reason
we
want
also
issue
eight,
a
separate
issue
to
have
separate
issue
weights
is
because
yielding
such
as
me.
Boards
I
can
quickly
a
good
grasp
of
my
my
team
members
capacity
for
a
given
release,
so
we
are
not
using
the
issue
weight.
Currently
in
my
group,
this
is
upcoming,
but
you
can
see
that,
for
each
column,
I
have
the
the
total
work
for
a
given
as
uni
and
I
can
see
the
weight.
A
So
it's
easy
for
me
when
we
get
more
predictability
about
the
velocity
of
the
team
that
if
this
person
is
a
way
for
two
weeks
during
the
iteration,
the
way
should
not
be
that
much.
And
this
is
super
helpful
for
the
engineering
manager
to
to
be
faster.
When
doing
the
scheduling
of
the
next
iteration.
That.
B
That
makes
a
lot
of
sense
and
I,
like
is
a
park
manager.
I've,
had
to
work
with
other
teams
where
I'm
the
one
who's
responsible
for
resource
allocation
and
making
sure
that
we
have
the
right
capacity
for
what
we're
planning
and
committing
to
so,
like
I
kind
of
understand
that
problem
I
think
the
thing
that
I
wanted
to
try
to
solve
from
a
more
purity
standpoint
is
this
sub
issue.
B
Convention
adds
a
whole
lot
more
workload
and
issue
management
instead
of
salt,
like
it
solves
a
problem,
but
it
doesn't
solve
it
as
effectively
as
like,
maybe
some
of
the
stuff
before
we
talked
about
like
if
there
are
subtasks
and
you
can
roll
up,
you
know
all
the
issue
weights.
You
know
per
assignee
across
all
the
issues
that
they
have.
You
know
and
just
count
those
or
look
across
you
know
I
think
there's
different
ways
to
potentially
solve
it
that
require
fewer
artifacts.
If
that
makes
sense
and
that's
what
I
want
to
explore
together
again.
A
I'm
totally
agreeing
video
that
just
said
that
during
the
past
months
and
with
all
the
load
that
all
of
the
teams
have
a
github
I
see
this
herb
issue
MVC
to
be
more
reachable
goal
than
having
all
those
features.
Natively
supported
in
a
single
issue
and
the
reason
I'm
pushing
for
this
is,
with
the
upcoming
feature
of
negative
futures.
You
can
get
rid
of
that
cluttering
of
having
more
issues,
because
you
say
we
have
moisture
to
manage.
Yes,
this
is
true
globally,
there
are
more
issues,
but
this
is
still
per
individual
issues.
A
So,
instead
of
having
one
issue
being
maintained
by
two
people,
they
are
their
own
issue
to
maintain.
And
yes,
it's
true.
The
main
issue
is
a
share
responsibility,
because
we
still
won't
have
cross-functional
people
working
together
and
being
together
as
principle
for
the
value
to
be
shipped
for
the
collection,
but
three
to
go
to
go
back
to
what
I
was
seeing
if
I'm
looking
at
what
I
would
call
a
project
manager
board,
oops,
sorry,
the
system
and
the
link
so
product
manager.
But
we
may
look
like
that.
A
Maybe
you're
using
that
kind
of
a
year
to
to
know
what
the
Robles
stretch
IHM,
that
I've
been
open
or
closed
during
the
micelle
and
see
the
progress
and
three
more
example.
What
other
feature
that
I
closed?
That
I
can
announce
in
my
monthly
release?
Blog
post,
you
could
just
add
a
filter
like
a
negative
filter
to
remove
said
issues,
and
you
would
have
a
clean
view
about
what
are
they
the
issue
that
are
about
shipping
value
to
the
users
and
nothing
about
technical
implementation
details?
A
B
You
just
won't
be
able
to
go
backwards.
If
we
do
the
sub
issue
thing,
we
can't
like
deprecated
sub
issues
and
they
become
like
a
permanent
solution,
even
if
they're
not
the
ideal
solution,
if
that
makes
sense
and
I
think
it's
like
worth
considering,
but
then
like
there's
a
lot
of
other
things
that
I'm
thinking
through
to
of
like
we're
we're
of
the
product
that
the
applications
overly
reliant
on
filters
right
now
to
solve
workflow
problems
and
I,
don't
want
to
keep
solving
them
with
more
filters.
B
A
Yeah
and
to
be
honest,
I
I,
don't
think
actually
I
mean
if
we
are
reaching
using
that
convention
and
once
we
have
the
negative
filters.
I,
don't
think
we
need
the
native
service
you
functionality,
because
today
what
we
are
doing
is
I'd
say
good
enough
for
us,
it's
May.
It
needs
a
bit
more
manual.
Work
like
we
are
explicit
in
the
served
issues
in
that
part,
which
is
an
implementation
plan.
So
here
you
can
have
what
we
could
have
as
child
issues
here.
You
have.
A
B
Yeah
I'm
gonna
put
this
through
the
validate
some
validation
cycles
and
kind
of
do
a
little
bit
more
exploratory
research.
The
other
reason
why
I'm
kind
of
you
know
interested
in
the
better
solution
in
the
long
run
is
because
eventually
we're
gonna
want
to
redo.
Like
the
to-do
list.
You
know
the
individual
to
list
and
we
can
either
solve
that
by
showing
a
bunch
of
issues
or
we
can
work
in.
B
The
concept
of
tasks
is
like
kind
of
a
more
first-class
citizen
across
issues
and
across
things
so
that
you
can
track
your
tasks
in
one
place.
So
you
know
your
engineer
when
it
comes
to
down
the
morning.
They
don't
go.
Look
at
a
ward.
They
just
go
look
at
their
their
their
individual
task
list
or
to-do
list
up
in
the
top
right.
You
know
I
think
ultimately,
that's
where
I
would
like
to
get
to
just
because
I
want
to
tailor
the
experience
and
not
force
people
to
go
through.
B
You
know
a
bunch
of
filters
through
specific
or
but
just
to
go
like
here's.
The
work
I
have
to
do
today,
I'm
going
to
sit
down
and
see
it
right
away
and
I
think
you
know
the
open
issues
is
great
but
kind
of
trying
to
mirror
the
function
of
the
issue
boards
and
take
you
some
that
in
on
a
task
level
for
the
individual.
So
that's
just
some
of
the
bigger
picture
of
limb
pieces
that
I'm
thinking
about
as
we're
exploring
this
together,
yeah.
A
I
really
like
that
I
mean
I
was
a
from
a
user
of
JIRA
and
it
makes
things
way
easier
to
manage.
Let's
be
honest,
I
like
that-
and
you
know
we
already
like
for
the
poor
in
this
issue-
Elise
you've
already
a
summary
of
the
tasks.
So
if
you
could
have
something
similar
bubbling
up
at
the
issue,
bot
little,
which
this
could
be
achievable
instead
of
using
sabisu,
so
I,
really
like
that
approach,
cool.