►
From YouTube: 2021-12-28: Object Storage Working Group - APAC
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
B
B
C
B
Good
thanks
yeah
is
it
just?
Is
it
just
the
two
of
us
today
yeah,
I
think
so,
okay?
Well,
I
guess
I
was
kind
of
hoping
that
gregash
would
be
available
to
go
through
the
requirements
he
added,
but
we
can
just
do
it
next
week
as
well.
If
he's,
if
he's
not
in
today,
yeah,
I.
B
Like
he's
still,
oh
actually
he's
okay
he's
actually
on
pto
starting
today,
so,
okay,
okay,
by
the
way,
I'm
not
sure
if
this
is
recording
correctly
or
where
it
says
it
is
recording,
and
unless
you
make
me
a
co-host
because
he's
out
so
I
hope
it's
going
to
end
up
in
his
cloud
recording
somewhere.
B
Let's
see
well,
I
mean
I
can
like.
I
only
have
one
update,
which
is
basically
come
from
alessio
as
well,
which
is
that
we
reopened
the
the
describe
the
object,
storage,
buckets
issue
because
there
were
a
couple
items
that
fell
out
of
some
old
documentation
unless
you
stumbled
upon
so
like
this
does
not
come
up
when
we
first
asked
around
the
company
like
what
teams
are
currently
you
know,
engaged
with
in
terms
of
features
that
use
object,
storage,
so
he
extended
the
initial
table
by.
B
I
think
two
entries
for
that,
and
I
can't
speak
exactly
to
what
this
is,
but
it's
now
linked
for
from
the
document
and
he
also
started
to
look
into
because
we
had
talked
about
the
possibility.
It's
not
the
society
yet,
but
the
possibility
of
replacing
or
removing
carry
away.
So
he
started
to
look
into
which
of
those
features
actually
use,
carrier
wave
and
there's
a
second
table
now
that
he
added,
which
lists
all
of
those
integrations.
B
So
maybe
this
will
is
a
lot,
it's
quite
a
few,
so
yeah.
Maybe
maybe
this
will
help
us
decide
how
complex
that
would
be.
If
we
were,
you
know
going
down
that
that
route
that
might
actually
tie
nicely
into
the
spreadsheet
gregosh
added
to
score
these
different
kind
of
buckets
of
things
that
we
might
start
looking
into
and
like
how
how
much
work
they
would
be
potentially.
A
Okay,
yeah,
I
don't
have
anything
we
have
been
discussing
a
bit
on
the
issue
about
removing
carrier,
carry
weave,
but
it's
high
level
for
now
and
yeah.
This
list
will
be
quite
useful
to
know
if
that
is
possible
or
not.
B
So
what
the
what
the
idea
be
to
like
write,
a
replacement
or
when
we
say
we
remove
it?
What
would
that
entail
exactly.
A
B
A
B
B
A
So
no
more
temporary
locations
like
we
have
now
right,
yeah,
that's
that's
crazy
stuff,
so
workhorse
would
upload
the
the
fireplace
to
the
to
the
right
location
and
then
just
confirm
that
it
has
been
uploaded,
which
is
the
final
last
call,
and
so
during
the
finalist
call,
you
would
have
the
attachment
record
created,
and
this
gives
us
some
advantages.
A
A
We,
at
least
in
the
package
team
we
have.
We
are
users
of
like
massive
removals
of
files
from
object,
storage
and
well.
We
we
can
just
issue
a
delete,
all
sql
query
and
that's.
A
A
C
A
Yeah,
so
if
we
can
use
active
storage,
it
provides
a
lot
of
shortcuts
because,
like
the
tables
definitions,
it's
there,
you
have
scopes.
You
are
even
preload
functions.
If
you
want
to
like,
you,
want
to
load
a
bunch
of
business
models
with
their
files.
You
have
a
preload
function
to
eagerly
load
the
attachment
and
the
blobs.
A
B
A
A
It
would
be
better
to
have
attachments
and
blobs
tables,
packed
together
per
team
or
feature
set
so
that
if
the
database
decomposition
that
happens,
it's
just
a
question
of
moving
those
two
tables,
and
so
right.
This
is
against
active
storage
approach
where
everything
is
in
two
tables.
So
can
we
somehow
put
some
options?
A
Polymorphic
associations
are
not
really
recommended
from
the
database
team
reading
their
guidelines
so
yeah.
I
guess
we
would
need
to
make
a
case
to
yeah.
This
is
needed,
but
perhaps
there
are
other
impact
that
I
don't
see
from
the
polymorphic
association,
but.
A
Thing
is
that
the
benefit
is
really
really
high
in
the
sense
that
if
you
want
to
add
a
new
business
model,
and
you
need
to
attach
a
file
to
it,
you
wouldn't
need
to
create
an
additional
table
yeah,
because
the
the
relationship
is
recorded
on
the
attachment
side.
So
if
the
table
is
there,
that's
it
you
you're
done.
Basically
it's
just
a
statement
on
the
model
and
that's
it.
B
Do
we
have,
I
mean,
even
without
answering
all
of
these
questions,
I'm
just
wondering.
I
think
it
would
be
good
if
we
get
into
the
scoring
soon
that
we
can,
because,
if
you're
just
looking
at
this
requirements
table,
that's
it
it's
great.
I
mean
it's,
it's
very
comprehensive,
but
that's
a
lot
of
work
right.
I
I
like,
I,
don't
think
we'll
get
to
all
of
these
things.
So
I
wonder
if
the
next
two
steps
would
be.
B
I
mean
it
would
be
nice
if
we
need
everyone,
of
course,
to
like
chime
in
on
this,
but
like
I
was
wondering
if
the
next
two
things
should
be
to
start
scoring
these
individual
tasks.
Some
of
them
will
be
very
large
in
themselves,
but
also,
I
wonder,
could
it
make
sense
to
have
some
kind
of
like,
like
linear
linearize,
these
different
buckets,
because
I
suspect
that
they
somehow
depend
on
each
other
right?
Maybe
not
all
of
them
do,
but
maybe
it
makes
more
sense.
Let's
say
we
do.
B
You
know
work
on
task
x,
but
then
another
task
why
it
might
you
know,
might
make
more
sense
to
do
first
right
because
maybe
it
makes
it
easier
to
then
to
then
work
on
x
right.
So
I'm
wondering
if,
if
there's
some
kind
of
dependency
graph
between
all
of
these
things
and
then
we
could
look
into
which
chunks
we
would
commit
to
and
then
in
which
order
we
would,
we
would
do
them
at
all,
because
this
is
something
I
just
don't
quite
understand.
B
Right
now,
thing
is
a
good
example
right
and
I
feel
like
some
of
these
are
almost
circular
in
nature.
You
know,
because,
okay,
so
there's
a
problem
because
of
carrier
wave,
and
then
that
leads
to
some
other
problem.
But
then
the
question
is:
why
do
we
have
carrier
waves?
Does
it
only
exist
to
support
something
else?
You
know
in
the
interaction
of
how
we
upload
files
and
then
the
question
is
like:
where
do
we
start-
and
it's
like
in
this
like
kind
of
cycle.
A
Yeah
yeah,
the
thing
is
that
I
only
have
my
well
the
the
team,
where
I
am
vision,
which
is
our
bucket,
and
I
see
the
other
team's
bucket
and
so
to
me.
There
are
already
quite
well
split
but
yeah.
Perhaps
we
have
some
features
that
depend
on
each
other.
B
I
mean
I
mean
I
don't
just
mean
the
different
buckets
so
much
but
like
like
the
different
kind
of
work
streams
more,
you
know,
because
carrier
wave
is
also
affects
a
bunch.
B
Features
right,
so
I'm
not
not
so
much
thinking
about
this
kind
of
like
kind
of
the
horizontal
relationship
between
different
features,
but
rather
more
like
the
in.
If
you
think
of
it
in
technical
buckets,
you
know
like
there's
the
carrier
wave
integration,
then
there's
this
extra
work
developers
have
to
go
through
to
even.
A
B
Direct
upload,
so
I
I'm
not
sure
if
there
are
dependencies,
maybe
there
aren't
a
right.
Maybe
it
doesn't
matter
where
we
start,
I'm
just
thinking.
If
there
are
any
and-
and
this
is
maybe
something
only
someone
can
answer
who
has
really
touched
all
of
these
different
parts
of
the
system-
and
I
haven't
so
I'm
I'm
not
totally
sure,
but
maybe
maybe
alessio
has
or
maybe
gregor,
I'm
not
sure,
but.
A
Well,
at
some
point,
I
guess
we
will
need
to
feel
the
scoring,
and
that
would
be,
I
guess,
pretty
much
the
output
of
the
working
group.
If
I'm
not
wrong
and
then
the
next
step
is
okay,
the
best
solution
is
this
one?
How
do
we
start
on
this
and
how
we
create
an
iteration
plan
to
make
it
happen
yeah,
but
we
need
the
scoring
first
yeah.
A
B
B
B
Okay,
I'm
actually
checking
when,
unless
he
was
back
and
select,
doesn't
tell
me
because
it's
truncating
the
text,
but
okay
he'll
be
back
early
january.
So
so
maybe
we
can
just
then
check
back
in
okay,
all
right
yeah
did
you
have
anything
else.
Otherwise
I
think
we
can
just
wrap
it
up
early
today.