►
From YouTube: Ceph Docubetter Meeting 2020-10-14
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
Okay,
it's
three
minutes
past,
so
I'm
gonna
go
okay,
so
the
first.
The
first
thing
is
pull
request:
374-51,
which
is
fadm
faster
and
more
scalable
document,
which
I'm
just
mentioning,
because
it'll
probably
be
a
good
source
of
future
documentation
changes.
Nothing
has
to
be
done
right
now.
It's
just
a
thing
that
I
want
to
keep
at
the
top
of
the
meeting
in
mind.
A
The
next
thing
is
that
we
have
changed
the
behavior
of
github,
so
that
signed
off
by
is
no
longer
required
for
commits
that
happen
inside
the
dock
directory.
This
is
so
that,
if
you
have
a
change
made
through
the
github
web
interface,
it
won't
break
it
or
it
won't
it.
Won't.
The
signed
off
by
check
won't
stop
that
from
being
adjusted,
because
there
were
several
commits
that
people
made
and
then
other
they
would
try
to
change
them
through
the
the
web
interface
and
then
it
wouldn't
work,
and
then
they
would
just
abandon
them.
A
So
I
actually
don't
know
if
that
is
working
right
now,
because
then,
as
I'll
say
in
a
minute,
the
main
way
that
that
we
were
going,
the
main
reason
that
we
were
going
to
do
that
was
sort
of
obligated
and
we
ended
up
using
another
model
but
signed
off
by
has
changed
next.
There
is
a
cfdf
detail.
A
Pr
that
I
was
working
on
was
the
primary
thing
I
was
working
on
before
two
other
things
got
in
the
way
and
I'm
just
giving
an
update
to
say
that
I
expect
that
to
be
done,
probably
in
the
next
couple
of
hours,
but
maybe
tomorrow
and
now
the
the
meat
of
the
meeting
here,
the-
and
I
guess
this
good-
that
patrick's
here,
because
this
is
about
his
document
stuff
file
system.
A
So
josh
cullen
brought
to
my
attention
a
bug
from
maybe
four
or
five
months
ago,
that
complained
that
top
level
links
in
the
sidebar
went
nowhere.
There
was
a
little
hash
like
an
octathorp
or
a
little
little
hash
symbol
like
an
anchor
point
from
old
html,
but
they
didn't
go
anywhere.
I'm
not
entirely
sure
why
that
was
the
case.
But
in
order
to
address
it,
I
collected
the
sets
of
links
together
in
the
restructured
text
documents
into
talk,
trees
in
restructured
text
documents
so
that
there
would
be
targets
for
these.
A
And
I
removed
the
committed
out
html
that
had
previously
been
there.
I
don't
know
if
this
is
going
to
mess
you
guys
up
and
if
it
does
we'll
just
revert
it
you
know,
but
in
order
to
prevent
the
by
default.
When
you
put
a
talk
tree
in
a
restructured
text
document,
then
it
appears
in
the
text
on
the
page
included
in
the
page.
A
It
also
appears
in
the
sidebar,
but
it
appears
right
in
the
middle
of
your
document,
unless
you
use
an
option
for
the
top
tree
directive,
which
is
colon,
hidden
colon
and
then
that
hides
it.
So
I
added
colon
hidden
colon
to
each
of
these
included
them
in
top
level
four
step
file
system,
so
they
would
be
picked
up
by
index.rst.
A
I
I
included
those
talk,
trees
in
rst
files,
that
would
be
immediate
children,
direct
children
of
index.rst
and
as
far
as
I
can
see,
the
most
recent
pull
requests
or
the
most
recent
comment
that
I've
made
gives
us
everything
that
we
want
and
nothing
that
we
don't.
And
oh
I'm
getting
a
message
from.
A
A
Check
this
out-
and
let
me
know
if,
by
removing
the
commented
out
raw
html
sections,
I've
broken
something
that
you
guys
need.
I
I.
B
A
Think
so,
but
I
don't
like
I
wasn't
there
when
you
guys
built
that
in
there,
so
I
don't
know
if
I'm
taking
away
something
that
was
essential
all
right.
So
that's
it
for
for
that
item
and
then
the
thing
that
I
have
spent
the
most
time
on
in
the
past
week
and
a
half
is
anthony.
Diatri's
updates
he's
made
massive
updates
to
the
rados
documentation,
which
I
must
have
spent.
A
I
don't
even
know
maybe
two
days
of
10
hours
of
peace
reading
out
loud
and
then
sending
back
and
forth
comments,
and
so
the
radius
documentation
has
been
updated
significantly.
This
is
a
really
good
collaboration
and
the
sort
of
thing
that
we
should
encourage
more
of
he
has
specific
knowledge
of
this
stuff
and
he
made
a
lot
of
updates
and
we
were
able
to
get
it
into
the
docs
pretty
quickly
for
the
size
of
it.
A
So
that's
good,
that's
that's
something
that
that
I
would
like
to
do
more
of
and
then
finally
defio
is
out
of
sync
with
docs.seph.com,
and
this
is
something
that
I
had
complained
about.
I
don't
know
almost
a
year
ago,
because
when
I
was
first
hired
by
the
seth
foundation,
I
did
a
search
for
like
okay,
I'll
I'll
figure
out.
What
do
I
need
to
do
to
install
this
thing,
and
the
very
first
thing
that
comes
up
is
set.io
when
you
search
in
google,
the
docs.com
website
is
the
third
search
result
from
google.
A
So
if
you
click
on
the
first
search
result,
you're
sent
to
the
set.io
page,
which
is
older
and
which
doesn't
have
the
up-to-date
docs
and
mike
perez
had
written
to
me.
He
wrote
me
about
a
day
ago
and
he
asked
is
there
a
way
that
we
can
keep
them
in
sync,
so
that
I'm
assuming?
I
don't
want
to
speak
for
him,
but
so
that
he
doesn't
have
to
manually.
Do
this,
so
maybe
we
can
build
a
robot
or
something
that
does
it.
A
I
don't
know
anthony
you
just
missed
the
part
of
the
the
meeting.
When
I
talked
about
how
what
a
pleasure
it
was
to
work
with
you
on
the
radas
documentation.
A
And
that's
going
to
be
it
unless
you
guys
have
anything
that
you
want
to
throw
on
the
stack.
Otherwise,
I'm
going
to
go
back
to
finishing
off
the
ceph
df
detail,
pr
and
picking
out
a
few
more
bugs
there
on
the
stack.
A
Real
well,
let's
take
just
a
second
of
this
meeting.
Can
you
send
it
to
me
right
now.
C
I
was
just
going
through
the
documentation,
prs
or
the
pr's
label
documentation
on
github,
and
I
know
that
the
sign
off
by
check
was
really
blocking
a
lot
of
merges.
So
what
we
should
do
is
go
through
those
and
rerun
the
tests
on
those
and
make
sure
that
they
pass
and
then
I
think,
hound
any
ptl's
who
need
to
you
know
approve
them,
so
we
can
get
them
merged
yeah.
We
got
quite
a
bit
of
backlog.
I
see
the
older
this
one
is
from
july
of
last
year.
A
If,
if
this
doesn't,
if
the
signed
off
check
business
isn't
working
the
way
that
we
want
it
to,
maybe
we
can
just
create
new
pr's
for
each
of
these
and
push
them
through,
so
the
backlog
is
cleared
out.
C
Yeah
that
might
work.
C
I
know
re-running
tests
on
really
old
pr's
sometimes
doesn't
work.
The
way
you
might
expect
so
it
might
require
creating
new
pr's
and
some
of
these
people
never
come
back
to
fix.
How
many
fixes
we
request
sure
and
legally
speaking,
I
don't
know
whether
or
not
we're
allowed
to.
C
I
suppose,
if
you
just
if
you
just
use
their
commit
unmodified
and
then
make
another
commit
that
fixes
it
the
way
you
want.
That's
probably
fine.
A
Well,
the
the
thing
that
I
had
wanted
to
do,
and
you
know
I'm
not
a
lawyer
and
maybe
you're
a
bit
more
aware
of
this
than
I
am.
I
had
wanted
just
to
designate
these
stale
commits
as
like
designate
them
as
a
remote
in
my
origin
and
then
pull
that
branch
and
then
credit
them.
As
the
author,
which
I
know
is
like
it
might
be
a
little
confusing,
but
it
at
least
preserves
the
like.
C
A
Yeah,
that's
I
have.
I
have
specific
questions
about
this
for
my
own
get
education
but
they're,
not
exactly
apropos
in
the
middle
of
this
docs
meeting
but
yeah.
Please
please
send
me
that
and
anthony
I
see
37587.
So
if
you'll
give
me
just
a
second
to
check
on
that
sure,
it'll
be
the
end
of
the
meeting.
B
A
A
Or
wait,
maybe
I
have
to
re-request
review
on
myself
here
and
then
reload
and
yeah.
There
we
go
okay.
A
B
A
Yeah
I
had
to
learn
how
to
go
to
the
read
the
docs
rendered,
the
rendering
of
read
the
docs,
which
I
did
not
know
all
right.
It
changes
I'll
just
say.
A
Okay,
those
changes
are
good.
Are
you
good
with
me
going
ahead
and
merging
this
pull
request,
anthony.
B
A
A
That
would
be.
That
would
be
fine
with
me
right
now.
I
am
doing
this
ffs
thing,
which
I
think
is
works,
fixing
the
taxonomy
and
this
ffs
documentation
and
doing
another
thing
with
cfdf,
which
is
minor,
but
then,
after
that,
I'm
going
to
go
through
the
dev
guide
top
to
bottom.
Just
doing
basically
exactly
the
same
thing
that
you've
done.
B
Yeah,
I
I
have
a
note
that
I
think
in
the
I
think
the
existing
docs
for
the
the
pg
balancer
are
are
out
of
date,
but
but
that
was
more
than
I
was.
I
wasn't
prepared
to
get
into
so
it's
like
missing
the
highest
price
column
or
something,
but
that
has
changed
among
the
releases.
So
that's
a
little
hated
but
yeah
I'll
get.
A
A
And
the
other
thing
just
like
as
a
an
overarching
thing
pacific
will
be
out
in
march,
or
at
least
it
is
scheduled
to
be
out
in
march,
and
my
big
goal
is
that
when
pacific
is
released
in
march,
we
have
a
comprehensive
and
accurate
and
easy
to
navigate
documentation.
Suite.
B
A
Okay,
well,
unless
there's
anything
else,
so
I'm
going
to
call
the
meeting
at
20
minutes
all
righty.