►
From YouTube: Future of Attention Requests in GitLab
Description
We’ve made the decision to not launch the attention requests feature in GitLab. While there were many positive experiences, improvements and lessons learned; there were just too many outstanding concerns we did not feel we could appropriately address while still maintaining the bar of usability we have for ourselves in this area of the product. Thank you to everyone who contributed in both the inspiration, creation and feedback phases of this process.
Feedback about this can be left in https://gitlab.com/groups/gitlab-org/-/epics/5331
Slides: https://docs.google.com/presentation/d/1PVD-kowaW2X-DhAn1IDDuLhz8WyRxFZLu-w2gJE4j9E/edit?usp=sharing
A
A
So
we've
made
this
decision
to
not
launch
attention
requests
feature
at
gitlab
and
while
there
were
many
positive
experiences
improvements
and
lessons
learned,
there
were
just
too
many
outstanding
concerns.
We
did
not
feel
we
could
appropriately
address,
while
still
maintaining
the
bar
of
usability
we
have
for
ourselves
in
this
area
of
the
product.
A
First,
thank
you
to
everyone
who
contributed
in
both
inspiration
creation
and
feedback
during
this
process,
we'll
dive
into
this
a
little
bit
more
we're
going
to
cover
a
few
things
here
and
let's
start
with
the
first
one,
which
is
a
thanks
to
everyone,
and
I
just
want
to
make
sure
that
we're
very
clear
on
this.
I
want
to
pass
along
my
thanks
to
everyone
who
was
part
of
the
process
of
attention
request.
A
Thank
you
to
all
the
other
engineers
who
supported
phil
and
lent
to
hand
as
well.
Thank
you
to
our
community
and
the
wider
gitlab
team
who
provided
feedback
after
we
launched
the
feature,
and
especially
thank
you
to
the
giddily
team,
who
agreed
to
continue
dog,
booting
and
providing
feedback
on
the
future.
A
B
So
now
I'll
talk
a
little
bit
about
why
we're
not
shipping
attention
requests.
So,
first,
let's
talk
about
foundational
issues.
While
there
were
several
problems
with
attention
requests
once
it
was
integrated
into
the
product,
there
were
two
main
foundational
issues
that
we
didn't
feel
we
could
iterate
through
appropriately.
B
The
first
main
issue
has
to
do
with
notifications
and
the
way
to
do's
work
in
the
product.
The
to-do's
feature,
the
addition
of
attention
requests
emphasized
a
problem
with
many
similar
notifications
as
it
introduced.
Yet
another
notification
in
that
area
and
these
notifications
led
to
user
confusion
and
even
duplication
in
the
product,
as
attention
requests
had
their
own
special
section
as
well
as
showing
up
in
the
to-do's
area.
B
The
other
main
issue
is
that
attention
requests
were
not
well
integrated
into
the
existing
review
handover
moments,
while
the
initial
setting
of
an
attention
request
happened
automatically
other
actions
were
manual
or
hard
to
understand
for
users.
We
also
introduced
attention
requests
without
having
other
pieces
of
a
complete
review
flow
in
place
like
enhancements
to
our
batch
comments
feature,
and
that
would
have
allowed
us
to
make
use
of
the
feature
much
more
clear
and
effortless.
B
Then
the
little
things
there
were
also
many
little
things
that
didn't
work
quite
right
when
we
looked
at
attention
requests
in
the
larger
context
of
the
merge
requests
and
review
flow,
and
some
of
them
are
listed
here
and
we
believe
that
we
would
have
been
able
to
iterate
through
some
of
these
items.
But
still
they
highlight
how
challenging
it
is
to
get
all
of
this
right
and
the
gaps
that
existed
even
in
the
very
limited
run
of
the
feature
being
turned
on.
B
What
have
we
learned
then,
to
start
the
most
important
thing
to
remember
is
that
not
shipping
attention
requests
does
not
mean
we
failed
in
some
way
or
that
we
should
be
discouraged
about
the
work
we
did.
We
designed
research
built
and
got
feedback
and
we'll
repeat
that
cycle
to
make
the
best
things
for
users.
This
is
iteration.
B
But
if
we
don't
look
at
the
whole
picture,
we
can
miss
how
all
of
these
steps
fit
together.
For
example,
the
collision
of
attention,
request,
notifications
and
to-do's
are
a
perfect
example
of
this.
It's
also
important
to
think
about
the
order
of
iterations
that
we
delivered
to
our
users.
We
shipped
batch
comments
quite
a
while
ago,
but
then
then
follow
up
to
complete
that
flow
of
the
review
process.
B
This
makes
other
features
that
try
to
work
with
that
feel
disjointed,
because
we're
skipping
around
the
process
and
not
following
through
on
a
complete
vision
for
our
users
and
finally,
sometimes
we're
just
too
close
to
what
we're
building
for
large
projects
like
this
one.
We
may
have
worked
on
them
for
many
months,
and
so
we
know
how
we
expect
them
to
work
and
assume
our
users
have
as
much
institutional
knowledge
about
our
process
works
as
we
do.
A
We
also
want
to
take
a
look
at
the
end-to-end
review.
Experience
where
attention
requests
fell
short
is
that
they
weren't
well
integrated
into
existing
behavior,
separate
ui
controls,
quick
actions
for
assignment
some
automatic
behavior,
you
know
and
making
sure
we
address.
The
complete
review
flow
allows
us
to
be
more
deliberate
about
how
we
integrate
these
types
of
features.
A
A
Two
of
the
big
things
we're
currently
working
on
are
restructuring
of
merge,
requests
and
suggested
reviewers.
The
restructuring
effort
is
a
large
undertaking
to
really
address
some
of
the
organizational
and
information
architecture.
Issues
of
the
merge
request
you've
seen
some
of
that
in
the
beautifying
efforts
that
have
happened
already
and
there's
more
to
come
as
we
continue
to
work
in
this.
A
We're
also
focused
on
suggested
reviewers,
as
the
applied
ml
group
drives
a
solution
there
and
we
understand
how
that
fits
into
our
current
merge
request
experience
and
what
the
future
might
look
like
in
the
area
of
review
flow.
We're
also
working
on
better
rounding
out
your
reviews
with
the
ability
to
add
a
comment
and
an
approval
when
you
submit
your
review
as
well
as
taking
away
ui
elements
that
bring
people
out
of
their
flow.
A
So
that's
everything.
We've
got
going
on
for
code
review
currently
and
some
details
about
attention
requests
and
where
we
stand,
if
you've
got
any
further
questions,
you
can
find
us
on
slack
in
the
code
review
channel
and
then
you
can
also
leave
feedback
in
the
attention
request,
epic,
which
is
linked
in
the
slide
deck.
So
thanks
for
taking
the
time
to
talk
with
us
today
and
let
us
know
if
you
have
any
feedback.