►
From YouTube: 15.10 Kickoff - Global Search
Description
Learn more about the Global Search groups plans to work on for our 15.10 release. Our planning issue is here: https://gitlab.com/gitlab-org/search-team/team-tasks/-/issues/130
A
Hello,
everyone
and
welcome
to
the
global
search,
15.10
kickoff
video
I'm
joined
by
CZ,
our
engineering
manager,
who
will
help
me
walk
through
and
discuss
what
we
are
thinking
of
working
on
for
the
coming
Milestone,
we're
really
excited
and
so
I'll
be
taking
us
through
the
features
section
cesela
going
through
the
maintenance
and
bugs
and
then
I'll
be
picking
up
some
of
the
ux's
items
and
then
we'll
wrap
up.
As
always,
you
can
go
to
the
planning
issue
as
well
as
our
Direction
pages,
to
learn
more.
A
But
with
that
we'll
jump
into
what
we're
working
on
the
first
item
here
is.
We
want
to
continue
mainly
finishing
out
some
work.
You
have
in
progress
here
for
these
first
two
items,
the
first
one
is.
A
Similarly,
we're
also
working
to
finish
up
the
option
to
use
advanced
search
to
search
for
users.
One
thing
we
are
doing
is:
we
are
working
to
turn
on
advanced
search
regardless
of
namespace
and
license
basically
we'll
have
the
option
to
to
utilize
advanced
search
if
elastic
is
configured
for
all
namespaces
that
have
it
available,
and
this
way
it
just
it
just
mainly
will
offload
the
postgres
database
and
will
also
return
results
more
quickly.
So
no
real
functionality
difference
aside
from
performance
and
then
from
there
we
will
be
working
on
our
new
navigation
experience.
A
We're
really
excited
about
this,
and
what
we'll
do
is
we'll
have
a
new
command
palette
option
within
the
new
navigation
and
search
can
then
easily.
Let
you
drill
down
further
if
you
actually
do
completed,
search
from
that
command
palette
window,
and
so
you
can
see
here
some
of
what
that'll
look
like
in
our
designs.
A
And
ultimately,
you
know
mobile
here,
for
example
and
tablet,
but
Global
search
will
play
a
pretty
Central
role
overall
in
the
in
the
new
navigation,
and
we
want
to
make
sure
we
have
this
available
for
folks
when
they
want
to
start
doing
some
opt-in
testing
in
the
next
couple
of
releases.
I
was
looking
forward
overall
to
this
change,
really
I
think
it'll
have
a
big
impact
on
the
ability
for
users
to
navigate
around
gitlab
and
find
what
they're
looking
for
as
well
we're
also
continuing
our
Zoot
production
work
here.
A
Overall,
we
want
to
move
away
from
elastic,
to
Zoot,
primarily
to
fix
a
number
of
the
key
Limited
stations
that
we
have
today,
which
result
in
essentially
code
results
returning
less
results
than
they
should
or
would
be
expected
to.
This
is
things
like
exact
substring
matching
and
more
so
zuk
solves
those
problems
for
us,
and
so
you
can
see
here.
A
We've
got
a
epic
rolling
and
with
our
iteration
plan
and
we
are
working
towards
our
first
MVC
of
having
it
work
for
a
very
small
set
of
namespaces,
as
we
continue
to
in
the
future
work
on
things
like
our
projects,
overall
performance
and
scaling
and
other
such
results,
but
I'm
looking
to
get
started
quickly
here
with
a
first
iteration
to
deliver
some
value
and
start
testing
out
Zoot
in
production
and,
of
course,
we're
excited
to
make
progress
and
ultimately
provide
a
much
more
reliable
coaches
experience
in
the
future
and
last
we're
also
working
on
making
some
initial
cost
estimates,
just
as
we
help
to
Think
Through,
Global,
search
and
advanced
search
in
the
future.
A
So
we'll
be
doing
some
of
those
for
our
SAS
Services
to
inform
those
decisions.
With
that
I'll
hand
it
over
to
CZ
who
can
take
us
through
maintenance.
B
Thank
you
Josh,
so
other
than
the
exciting
features
we
also
plan
some
work
for
maintenance
and
Bug
fixes
for
the
maintenance.
The
first
one
we
are
going
to
work
on
is
to
implement
our
proposal
to
fix
and
repair
other
index.
If
the
projects
are
missing,
we
eventually
planned
it
for
59
release
but
dot
post
to
1510
and
due
to
other
ongoing
priorities.
B
The
second
one
is
recently:
we
introduced
the
bulk
Chrome
worker
parallel
processing
and
to
increase
our
throughput
of
our
indexing
jobs.
So
we
now
need
to
refactor
our
search
index
indicator
code
to
match
the
parallel
processing
change.
We
also
plan
to
clean
up
some
of
the
feature
Flags
to
fully
release
a
couple
of
features
for
bug
fixes
the
first
one
we
received
a
our
user
report
that
the
project
ID
is
missing
from
the
return
of
our
other
subscription
apis
So.
Currently
there
is
actually
a
awesome.
B
Community
contribution
on
this
issue
will
help
to
review
the
code
and
and
fix
the
issue
eventually.
The
second
one
is.
We
also
found
a
performance
issue
with
our
past
filters
in
advanced
search.
B
We
have
consulted
with
our
electric
consultant,
and
now
we
have
a
solution
for
that,
we'll
be
working
on
it
in
the
1510
release
to
fix
the
problem.
The
last
practice
that
we
plan
here
is
to
address
an
issue
that
the
global
search
does
not
match
on
the
full
project
path:
yeah.
Those
are
the
plan
for
maintenance
and
Ducks.
Now
I'm
handing
over
back
to
Josh
for
the
rest
of
the
plan.
A
Yeah
thanks
Gizzy,
so
next
up
is
our
ux
design
work.
One
of
the
items
you
want
to
work
on
is
having
a
design
for
providing
code
results
when
there
are
multiple
matches
within
a
single
file
right
now
we
don't
do
this,
and
so
we
want
to
go
ahead
and
extend
our
design
such
that
when
we're
able
to
with,
for
example,
Zoot.
We
have
a
good
idea
of
how
we
want
to
present
this
to
users
to
represent
that
there
are
again,
you
know,
multiple
results
in
a
single
given
file.
A
A
We
want
to
make
sure
we
go
ahead
and
have
a
process
for
adding
it,
and
then
we
also
want
to
go
ahead
and
audit
the
theaters
dictionary
and
make
sure
that
we
can
continue
to
make
sure
that
we
have
all
those
things
was
documented
and
how
they
function.
So
looking
forward
to
that
and
supporting
our
database
group
in
that
effort,
as
it
relates
to
the
tables
for
Google
Search.
So
overall,
thank
you
so
much
and
thank
you
for
joining
me.
A
We're
really
looking
forward
to
the
future
work
that
we're
planning
here
right
or
all
the
work
actually
and
15.10,
and
thanks
for
watching
and
as
always,
if
you
have
any
questions
or
ideas,
please
find
us
on
the
issues
I'm
going
to
put
a
new
one
and
we'll
engage
with
you.
There
thanks
everyone.