►
From YouTube: 2021-04-27 delta-rs open development meeting
Description
Agenda not yet determined
A
Wanted
to
mention
this,
so
we
had
a
live
kafka
delta
interest
demo
that
christian
and
mr
ran
last
week.
I
believe
and
tyler
mentioned
that
we
might
want
to
do
a
python
specific
demo
session
as
well
hosted
the
weeks
before
data
and
ai
submit.
B
B
Okay,
that's
correct
so
may
24th
to
28th
and
by
the
way,
since
a
bunch
of
us
are
committers
here
we
are
currently
trying
working
out
the
schedule
on
when
we're
gonna
have
a
delta
lake
journey
tech
talk
the
week
before
summit,
probably
the
18th,
I'm
gonna
finalize
the
schedule
next
day
or
two
so
I'll.
Let
you
guys
know
as
soon
as
and
invite
you
guys
to
just
I'm
still
trying
to
figure
out
the
logistics.
That's
all.
A
That's
cool,
so
tyler
has
a
python
demo
in
his
talk,
and
I
I'm
thinking
that
he
also
wanted
to
do
a
python
demo
as
well
before
the
actual
summit
happened.
A
I'm
going
to
move
to
the
the
next
item.
A
Right
so
I
think
the
the
next
thing
is:
there's
a
ongoing
pr
on
adding
vacuum
support
to
delta
rs,
which
I
recommend
everyone
to
go.
Take
a
look.
The
the
direct
the
vacuum.
Behavior
is
not
documented
in
details
in
the
official
spec,
so
we
have
to
go
through
the
reference
implementation
to
look
at
what
actually
is
working
should
delta
rs
do
when
like
vacuum
command
is
issued.
So
if
you're
interested
in
that,
please
help
join
the
discussion
in
the
pr.
A
C
Okay,
cool
now
I
was
just
asking
in
case.
You
know
the
specific
stuff
we
need
from
from
an
error
perspective,
because
ariza
pot
is
not
fully
there
and
there's
like
small
bugs
that
have
sort
of
been
chasing
on
and
off
that
when
you
get
like
exotic
combinations
of
structs,
especially
so
I
think
they
become
they
become
useful
when
you're
dealing
with
the
optimized
command
yeah.
It's
not
it's
not.
C
I
found
them
by
by
generating
random
data
and
sort
of
running
it
through
arrow,
as
well,
through
pacquiao,
the
packet
crate
and
then
checking
where
the
pi
arrow
doesn't
crash
or
anything,
and
I
found
something
that
I
don't
think
it's
a
pi
error
problem.
It's
a
it's
a
it's
a
rust
problem,
so
I'm
still
looking
into
that.
A
C
Nope
I'm
good
well,
I
would
like
to
I
think
I've
mentioned
it
before.
I
would
like
to
do
like
a
suppose,
a
small
demo
just
showing
how
well
what
about
what
I've
been
creating
in
like
in
my
spare
time.
I
guess
I
haven't
committed
anything
to
get
though,
but
is
a
delta
reader
on
the
browser.
There's
this
project
called
perspective.
C
It's
it's
a
it's
sort
of
like
a
data
table
library,
for
you
know,
angular
view
etc,
and
it
allows
you
to
read
error
data.
You
know
into
a
table
and
then
sort
of
you
can
pivot
that
etc.
So
I've
been
exploring
doing
that
with
e
with
with
with
delta
tables
from
from
storage.
So
I
think
that
would
be
like
an
interesting
one
where
you
can
browse
delta
tables.
Look
at
you
know
different
snapshots,
depending
on
whether
you
haven't
removed
the
the
time
travel.
C
C
I
can
get
it
ready
before
the
end
of
this.
What's
now
before
then,
before
the
end
of
may
so
I'll,
just
put
in
like
one
two
weekends
on
it
and
then
try
it
out
with
you
know,
s3
or
maybe
even
just
local
storage
channel.
A
Oh
that'll
be
interesting
to
see
the
demo
just
ping
us.
Whenever
you're
ready,
we
can
schedule
the
demo
for
one
of
the
meetings.
C
A
Awesome
all
right:
let's
move
on
to
interest
christian
and
misha.
Do
you
have?
What
do
you
want
to
discuss
about
it.
D
D
Neville's
working
on
that
with
me
at
the
moment,
trying
to
figure
out
a
couple
of
missing
pieces
and
after
that,
we'll
misha
is
making
good
progress
on
the
distributed,
locking
within
delta
rs,
and
once
he
rounds
out
some
tests
with
that
he's
gonna
come
over
to
kafka
delta
ingest
and
help
work
out.
The
concurrency
bits
that
we
need
to
have
multiple
consumers
handling
a
a
single
topic.
A
A
D
Yeah,
so
what
I've
done
right
right
now,
so
I
figured
out
how
to
get
the
typed
values
that
that
I
need
in
order
to
create
the
delta
version
of
the
stats.
Then
I
came
across
an
issue
where
it's
appears
that
for
structs
and
lists
the
statistics,
the
parque
statistics
aren't
populated
and
neville
is
we
have
a
thread
going
we're
nibbles
trying
to
help
help
me
out
on
that.
I
could
yeah
it's
actually
that
same
thread
that
you
replied
to
yesterday.
A
Looks
like
everyone's
happy
all
right
with
that.
I
think
we
can
end
our
meeting
earlier
today
and
looking
forward
to
naval's
demo
before
the
end
of
may.