►
From YouTube: #t20-gostin: Pre-hack meeting
Description
Transform2020 Hackathon project #t20-gostin.
Meeting 05 June 17:00 UTC: Pre-hack meeting.
A
B
Yeah
sure
I've
seen
that
except
Dominique
I've
met
everybody
in
person.
I
see
her
in
Washington,
so
I
was
good.
Hi
Dominica,
I'm
Miguel,
so
I
I
am
yeah.
Have
you
do
this
might
be
asleep
at
the
time
mystic
kind
of
doing
it?
At
the
same
time,
we
have
made
an
spin-off
already
from
the
Institute,
with
a
company
all
around
genetical
modeling
and
uncertainty
in
these
things.
B
So
at
the
moment
we
just
created
a
company
two
or
three
months
ago,
so
we
are
still
at
the
very
beginning.
Looking
if
we
look
for
investors,
clients
God
knows
how
all
these
things
works,
but
but
we
really
have
these
and
the
professor
is
part
of
a
company.
So
we
are
building
this.
This
type
of
a
hybrid
around
all
this
gem
pie
and
all
the
libraries
also
finally
I
know
they
was
written
by
my
professor,
that
is
for.
C
B
So
husband,
some
companies
already
asking
to
to
integrate
you
by
different
workflows.
We
are
also
in
a
huge
European
project
to
build
an
interface,
especially
in
that
materiality.
So
that
way
you
can
be
models
on
the
cloud
directly
and
they
display
it
either
in
your
desktop,
but
also
in
the
mind
for
exploration,
so
so
most
likely
is
going
to
be
open
core
and
so
using
the
back
hands,
always
building
it
open
source.
Why
we
are
building
specific
apps
on
front-end
for
specific
idea.
B
B
One
thing,
for
example,
that
we
have
already
said
a
couple
of
times
has
quite
a
lot
of
traction
is
an
of
materiality
sandbox.
So
I
don't
know
if
you
have
seen
it
in
some
museums
that
they
have
a
Kinect
on
top
and
they
display
the
geometry
basically
height.
So
what
we
are
doing
is
just
taking
that
height
and
passing
it
through
Empire,
and
we
display
the
charcoal
map.
So
this
type
of
tools
that
they
are
using
open-source
back-end
is
what
we
see
as
potential
to
generate
revenue.
B
B
See
as
we
have
to
just
help
people
to
so
do
more
consulting
here,
so
if
someone
has
started
to
make
up
model
which
Empire
and
they
get
stack
and
they
want
to
pay
as
to
help
them,
that's
also
fine.
I
know
also
specifically
probabilistic
model
in
a
so
so
we
are.
Most
of
our
research
is
in
Bayesian
statistics.
So
note
yes,
generic
one
model,
but
really
an
assemble
of
models
with
specific
likelihoods.
For
that
you
really
need
expertise,
so
I
think
that
that
could
be
an
expertise
that
companies
could
be
interested
to
to
hires.
B
So
we
are
in
conversations
with
seem
as
for
example,
that
they
so
jump
I
and
they
felt
this
could
be
used
useful
to
model
turbines
because
they
have
different
layers
of
coating
in
metals
and
then
they
they
look
for
for
impurities
and
cracks.
We've
we
tomography.
Basically
it's
like
wow,
that's
like
geology,
but
in
miniature.
A
B
I
guess
that
that's
going
to
be
a
common
thing
in
in
many
of
us,
so
the
input
of
jump
I
started
at
some
point
and
it's
really
a
surface
point
and
orientations.
So
it's
already
quite
creative
data,
so
I
have
the
feeling
that
we
really
need
something
like
subsurface
is
what
I
mentioned.
I
know
that
library,
that
somehow
is
the
common
denominator
between
all
our
libraries.
So
we
can
put
there
and
then
it's
very
clear
that
if
you
have
your
data
in
that
format,
you
can
use
it
across
the
plane.
B
B
And
I've
seen
that
that
is
something
that
we
are
really
missing,
because
at
the
moment,
so
let's
see
le
the
transform,
a
so-so,
dieter
and
I.
We
sit
down
together
and
actually
with
made
an
example
with
simple
ingredients,
it
was
and
with
a
3/3
dmg,
and
we
we
really.
It
was
not.
It
was
a
strip
lock.
So
we
started
to
hack
some
notebooks
of
how
could
everything
look,
but
in
the
end
that
is
one-to-one
communication.
So
every
single
time
that
we
want
to
communicate.
A
B
If
it
is
for
this
European,
brilliant
cation
that
has
funding
for
three
years,
we
promise
something
like
that
so
being
able
to
yes
import
data
in
open
source
and
then
visualize
in
having
to
reality
in
this
case,
so
I
I
will
have
ours
in
the
next
year's
to
really
put
work
on
that.
So
it's
not
like,
like
we
as
make
a
hack,
and
it's
going
to
back
there,
so
I
think
that
a
library
to
manage
in
a
bit
this
type
of
Jellicle
data
is
going
to
happen.
E
Can
do
everything
crisis,
beginning
yep,
yeah
I
really
agree
with
that.
I
think
when
the
the
what
makes
scientific
Python
so
powerful
is
that
you
can
go
from
so
you
can
load
your
data,
you
can
plot
it.
You
can
plug
it
into
a
machine
learning
model
and
then
you
can
save
it
all
with
the
same
underlying
data
format
yeah.
So
a
numpy
permeates
everything
from
from
x-ray
to
to
pandas
to
map.
Lot
live
just
like
it
learn
image.
So
you
can
you
don't
have
to
you:
don't
need
a
conversion
layer.
E
B
No
I
I
neveri
completely,
so,
for
example,
one
thing
that
I've
been
doing
in
the
last
couple
of
months
is:
is
yes,
14:10
pi
from
the
old
vtk
that
we
wrote
to
2
pi
this
time,
and
obviously
it
was
super
easy
for
me
to
two
years.
Take
the
data
structure
of
Jim
PI
and
put
it
into
my
mr.,
but
the
scale
is
a
few
lines
of
code
a
and
then,
if
you
start
adding
a
few
lines
of
because
we
are
using
numpy
so
in
that
I
agree
completely.
So
so
that
makes
things
super
easy.
B
D
D
So
I
mean
my
research
in
my
my
is
working
with
Gentile
and
data
exported
from
basically
Petrella
and
software
like
that,
because
I'm
analyzing,
a
lot
of
seismic
interpretations
and
it's
been
always
been
a
pain
for
me
to
just
I
mean
you
can
export
all
that
data
member
trail
easily.
But
then
you
have
you
have
so
many
weird
data
formats
and
grids
and
full
sticks
and
welds.
You
have
to
all
load
in
and
they
all
so
I
had
to
write
a
lot
of
custom
import
scripts
for
esoteric
data
formats
and
I
thought.
D
F
D
His
subsurface
module,
which
is
just
a
name
we
came
up
with
randomly
busy
put
all
the
GUI
o
in
there
that
doesn't
fit
into
others
like
in
segue,
already
exists
for
segue
data.
So
we
don't
need
to
touch
that
right,
but
to
have
everything
collected
and
a
place
where
people
easily
can
like
make
a
pull
request
with
this
import
function
and
then,
at
all,
like
everyone
who
converted
importance
like
fold,
stick
will
end
up
with
I.
D
Don't
know
the
same
data
structure,
numpy
array
or
Panos
data
frame
for
their
fall
sticks
in
the
end,
so
they
all
need
to
point
through
the
same
end
result
to
be
useful.
I
guess
that
was
well
I'm
still
hoping
for
that.
Even
though
I
only
have
one
year
left
in
my
PhD,
so
I
don't
know
how
much
time
I'll
really
have
to
keep
working
on
that.
D
D
E
Leo
most
of
you,
I
guess
know
me
already
I'm
at
the
University
of
Liverpool
and
to
Alex
this
point.
You
will
never
have
more
time
than
you
do
now.
A
time
only
disappears
faster
and
faster
weights.
I
started
here
this
year,
I
work
on
the
HMO
tahaafe
project,
since
2010
I,
guess
right.
What
I'm
really
hoping
here
is
to
like
have
this
better
bridge
between
the
other
different
projects.
E
Too
many
acronyms
pie
gimbley
right
where
you
you
have
a
unified
framework
for
solving
PDEs
and
that
allows
you
to
do
very
efficient,
very
fancy
joint
inversions.
But
it's
not
easy.
If
you
want
to
invert
for
the
geometry
of
an
interface,
for
example,
so
I'm
kind
of
leaning
more
towards
that
kind
of
problem
anyway,
yeah.
E
So,
but
mostly
what
I
would
hope
to
obtain
here
is,
as
I
said
in
the
chat
was
a
if
we
can
have
a
shared
environment
that
we
know
works
with
all
the
packages
because
I've
not
try
to
install
the
photo
packages.
Gem
PI
sin
peg,
discretized,
PI
Vista,
if
M
PI
mod
I
have
not
tried
installing
all
of
that
in
a
single
common
environment,
and
it
feels
like
you,
I
mean
it
might
just
work
out
of
the
box
and
yeah
we're
done
in
five
minutes.
E
But
there
might
be
some
crazy
conflicts
in
there
that
we're
not
aware
of
so.
If
we
can
build
that
or
just
make
sure
it
works
and
maybe
have
a
curated
meta
package,
like
the
NGO
stack
where
they,
they
have
a
meta
package,
that
curates
versions
of
the
libraries
that
they
know
work
well
together.
So
something
like
that
might
be
useful
so
that
you
could
install
like
the
the
geophysical
or
the
the
Geoscience
stack.
And
then
you
would
get
everything.
E
B
E
B
Are
using
people
that
jump
I,
we
try
to
see
that
for
all
the
libraries
well,
except
50k
was
recently
that
it
was
not
compatible
with
the
last
Python,
but
but
in
general
we
are
always
with
the
latest
versions
of
all
the
packages
which
I
don't
know
if
it's
good
or
bad.
But
what
I'm
really
try
to
in
the
moment
that
pandas
breaks
my
code
I
try
to
make
the
hotfix
as
soon
as
possible
to
yes
try
to
keep
it.
A
G
So
how
can
we
you
know
at
the
data,
ingest
and
stuff
of
your
workflow?
You
have
all
of
this
crawl
spatial
data.
We
need
to
visualize
that
stellastot
spatial
data
and
then,
when
you
do
your
pre-processing,
you
might
you
know-
maybe
that's
when
you
add
a
spatial
reference
or
other
things,
visualization
there
to
gain
insight
to
what
you're
doing
and
what
what
the
algorithms
and
the
black
boxes
are.
G
How
to
well-positioned
geoscientists
the
leverage,
those
technologies
and
having
some
sort
of
underlying
library
like
what
we're
talking
about,
would
really
enable
that,
because
we'd
have
a
sort
of
a
way
to
do
it
on
a
clear
interface
on,
you
know:
running
different
machine
learning,
algorithms
on
different
types
of
spatial
data,
or
things
like
that.
So
that's
my
my
quick
and
dirty
what
I'm
excited
about
here?
Cool.
D
A
Think
just
recently
be
removed
the
necessity
of
math
totally
from
district
eyes,
because
I
found
that
it's
quite
nice.
If,
if
you
do
your
computations
on
a
server
that
you
have
a
minimal
install
and
don't
need
to
install
all
the
library,
the
plotting
stuff.
So
in
that
sense
would
kind
of
exclude
PI
vista
from
being
the
the
core
and
more
like
an
X
or
a
thing,
so
that
the
only
need
the
plotting
stuff
locally
and
other
on
the
server.
But
yeah.
D
D
F
F
F
Is
both
I
guess
interoperability
between
all
of
these
packages,
but
then
also
thinking
about
what
are
some
of
the
current
sort
of
industry,
standard
geologic
modeling?
And
can
we
interface
to
those
because
I
think
that
that
could
actually
drive
for
their
adoption
of
open
source
software,
as
if
we
can
immediately
interact
with
what
folks
are
using
every
day
and
not
have
that
be
a
big
hurdle?
I
think
that
that
could
be
a
big
step
forward
and
then
one
other
package
and
I
sent
this
on
in
slack.
F
That
I
think
is
worth
like
looking
to
for
ideas
and
I
was
involved
in
some
of
the
API
design
and
sort
of
thinking
about
the
database
structure,
for
this
is
the
OMF
project,
so
it's
open
mining
format,
and
so
this
was
primarily
championed
by
Franklin
and
Rowan.
But
I
had
a
little
bit
of
a
pulse
on
on
that
project,
at
least
when
it
was
initially
created
and
their
goal
was
to
create
some
common
geologic
format.
At
the
end
of
the
day,
everything
C
realized
to
JSON
I
think
they
can
serialize
to
other
stuff
too.
F
But
that's
a
nice.
You
know
easy
plain
text
format
to
serialize
to,
and
so
we
thought
about
sort
of
block
models,
boreholes
lines,
surfaces,
those
sorts
of
things,
so
just
something
to
look
to
for
ideas
and
see.
If,
because
there's
a
lot
of
thought
that
went
into
that,
so
there
might
be
some
pieces
that
we
can.
We
can
leverage
and
sort
of
build
off
of
yes,
so
I
guess
in
terms
of
the
hackathon
in
the
next
couple
of
days.
B
F
B
One
thing
that
everybody
has
mentioned,
but
but
not
making
it
clear,
I
I,
really
think
that
if
we
do
this
type
of
library,
a
huge
part
is
the
parsers
like
being
able
to
read
on
the
commercial
software.
So,
as
you
are
saying
in
say,
but
also
being
able
to
export
to
any
format
that
you
want.
In
this
case,
the
open
mining
format
makes
a
lot
of
sense
yeah.
But
that's
this
for
long
term
storage.
F
F
F
So,
even
though
gask
wasn't
strictly
from
the
Pangaea
ecosystem,
a
lot
of
the
recent
developments
and
improvements
have
really
been
driven
by
that
community,
and
so
what
they've
been
doing
is
then
investing
an
open-source
software
but
then
deploying
Jupiter
hubs
with
a
stack
of
tools,
including
X
or
a
desk,
and
then
other
common
Geoscience
packages
on
the
cloud
next
to
big
datasets.
So
there's
datasets
like
cement
six,
which
are
on
the
order
of
I,
think
thirty
to
sixty
terabytes
or
something
like
that.
B
C
Yeah
I
guys
so
I
did
my
PhD
in
the
same
group
as
as
Lindsey.
It
was
definitely
on
true
fiscal
inversions
and
helped
out
would
be
addison
bag,
sniffing
library
right,
but
now
I'm
full-time
in
the
private
sector.
For
me,
rajo
science
I,
don't
know
if
you
guys
heard
who
they
are,
but
basically
they
were
consultants
and
retailers
of
go
CAD
for
probably
over
over
a
decade
and
now
they're
kind
of
trying
to
transition
towards
their
free
viewer
called
a
Jewess
lanced
analyst,
which
relies
on
on
ah5
a
structured
h5
file
format
that
is
open.
C
So
what
I've
been
doing
for
the
past
six
months
roughly
is
to
write
a
Python
API
to
be
able
to
to
read
and
write
to
the
geo,
h5
file
format.
So
obviously
I'm
kind
of
married
to
that
to
that
database
database
structure,
I
can
store
surfaces.
Black
meshes
like
to
read
your
holes,
all
that
stuff
and
so
yeah,
so
this
so
I
publish
a
package
represent
me
to
do
that
and
now
now
that
we
have
that.
C
I
would
like
to
just
be
able
to
have
a
proof
of
concept
that
I
can
I
can
create
the
very
simple
Giorgio
model,
with
Jen
PI,
to
be
able
to
digest
it
and
send
it
back
to
to
the
qh5
database
so
that
we
can
use
if
you
were
so
that
our
clients
and
our
sponsors
can
youth
in
3d
and
manipulate.
Oh,
so
I'm
not
too
married
to
like
you
know
for
it
to
end
up
in
gh
5,
but
it
will
have
to
end
up
in
into
H
5
at
the
end
of
the
day.
C
And
so
now
he
knows
that
this
transform
thing
is
happening,
so
I
get
free,
free
range
time
to
be
able
to.
You
know,
push
push
that
because
go
get
is
nice.
You
know
it's
like.
It
has
a
lot
of
power
power
to
do
very
complicated
things
but,
as
they
said,
we're
leaving
so
many
people
in
the
dust,
because
it's
way
too
complicated
to
do
the
modeling
with
hard
software's
and
if
we
go
to
open
source-
and
it's
still
there
still
like
it-
pretty
steep
learning
curve.
C
So
my
role
is
to
be
just
to
smooth
out
that
transition
and
just
give
like
simple
apps
that
people
can
use
jump
I
to
be
able
to
do
what
they
want
and
then
I'm
gonna
hook
up
same
peg
puts
little
simple
tools
with
some
pegs
so
that
you
know
people
can
just
have
one
or
two
buttons
that
do
exactly
what
they
want,
and
people
could
capture
as
many
people
as
as
possible.
So
that's
kind
of
that's
kind
of
my
role.
B
Actually,
in
two
months
ago,
just
before
the
pandemic,
we
were
in
Australia
and
I
was
with
Guillaume
and
we
wrote
a
part
writer
from
Jim
PI
to
to
go
cat.
Oh
and
the
other
way
around
I
think
I
had
an
italic
say
today.
I
have
a
snippet
code
here
there
that
I
don't
know
where.
To
put
so,
we
really
need
to
make
a
library
that
is
where
make
sense,
to
put
this
type
O
of
God,
because
I
didn't
spit
easy
to.
B
C
C
I
started
a
small
repo,
it's
on
there.
Obviously
the
New
Ager
science,
but
it's
on
guilt
right,
that's
gonna,
be
an
open
and
that's
gonna
be
the
role
the
goal,
basically
to
just
let
people
submit
their
own
little
apps
that
can
do
something
in
them,
but
I'm
sure
it
it
could
exist
somewhere
else
and
I'm,
not
I'm,
not
married
to
this.
To
this,
it.
A
F
A
A
Cool
I
click
finish
off
fun
on
DJ,
I.
Think
by
now
I
know
everyone
I'm,
probably
from
the
smallest
community,
because
it's
me
that's
why
I'm
probably
also
interested
in
the
interaction
between
the
different
communities
like
sim
pack
and
Tempe,
and
privacy
I'm
in
fatty
Ando,
which
is
I,
think
the
only
project
I've,
never
really,
I'm
always
observing
it,
but
I've
never
actually
done
anything
with
it.
So
maybe
this
hack
is
it's
the
time.
A
D
A
Priest
and
I
think
it
was
even
the
panel
last
year's,
or
we
had
this
in
browser
3d
movement,
which
was
pretty
cool
but
bein,
said,
there's
something
matrix
to
PI
star
coming
soon.
So
that's
even
easier,
because
I
think
at
the
moment
panel
is
not
working
anymore
right
at
PI
Vista.
All
that
notebook
doesn't
work
anymore,
yeah.
G
The
panel's
change,
some
things
and
they've
made
a
lot
of
work.
It's
a
lot
better
than
what
was
initially
there.
It's
worth
checking
out,
but
I
I,
don't
know
it's
there's
a
lot
of
people
doing
the
same
work
in
different
projects
and
different
ecosystems,
and
it's
it's
really
a
shame
right
now
and
I'm
actually
kind
of
working
to
bring
it
all
together
a
bit
so
I
think
in
time.
We'll
have
something
a
bit
more
mainstream,
whereas
right
now
everything's
kind
of
juice,
alpha
projects
being
developed
and
yeah.
A
So
my
my
intent
of
the
hack
was
a
little
more
to
finish
or
decided
last
year
to
have
this
interaction
Kampai
same
peg
and
maybe
fatty
Ando-
to
have
this.
Let's
model
something
realistic
in
Shanghai
and
the
model
gravity
data
model,
DC
data
empty
later
whatever,
and
even
more
so
was
jump.
I
put
out
many
more
fascinating
models
now
in
the
new
gallery,
which
which
are
really
cool,
we're.
A
That
was
my
idea
and
I
saw
that
you
have
much
higher
goals.
Loftier
goals
with
the
whole
subsurface
thing
and
I
think
both
things
fit
into
this
hack,
particularly
because
I
expect
that
that
will
come
people
that
have
probably
no
experience
at
the
idea
of
the
tools
and
part
of
their
hack
experiences
to
learn
those
tools
and
to
combine
them.
So
I
think
we
should
accommodate
those
and
that'll
be
cool.
B
For
maybe,
as
question
is,
is
of
the
others
how
much
they
want
to
get
involve,
because
that
project
first
I
mean,
if
you
convinced
me
that
makes
sense
to
spend
time
in
a
similar
project.
I
will
do
it,
but
but
I'm
going
to
put
hours
sooner
or
later
on
that.
So
if
Bane
and
Dominic
wants
to
start
doing
it
this
weekend,
so
we
can
brainstorm
together.
That
would
be
great.
B
C
F
This
is
a
suggestion.
It's
the
first
weekend.
We
focus
on
just
tying.
The
codes
together
and
that'll
be
a
chance
for
whoever's
engaged
to
just
play
around
and
sort
of
wire
things
up
and
sort
of
pack
round
get
a
bit
familiar,
and
then
it's
the
next
weekend.
Even
if
it's
just
short,
we
schedule
a
one-hour
call
to
start
thinking
about
like
NEP
I
design
together,
because
then,
hopefully
by
then
will
at
least
have
seen
seen
each
of
the
packages
presented
and
have
at
least
a
bit
of
a
flavor,
hey.
B
C
A
B
Yeah,
that's
that's
true.
I
was
also
thinking
that
we
could
really
try
to
do
just
think
of
some
area
with
more
hole
data
some
size
make
from
there
try
to
create
the
structural
model
in
all
these
steps,
already
visualizing
it
with
with
by
vista
and
eventually,
yes,
yes
doing
the
way
forward,
physics
or
whatever.
G
So
I've
done
some
like
preliminary
demos
of
this
kind
of
work
before
working,
unfortunately
outside
of
most
of
the
software
a
bit
with
system
tied
in
discretize,
but
also
with
other,
like
you,
statistical
libraries
and
using
pointers
to
as
that
glue
and
so
I
might
have
some
cool
data
from
the
Forge
geothermal
site.
I.
G
Think
that
we're
familiar
with
that
where
they
have
a
whole,
just
plethora
of
data,
its
aerial
imagery,
there's
geum's,
there's
actual
subsurface
models
that
were
created
from
like
you,
statistical
creating
of
sparse
data,
and
there
are
tons
of
well
logs
that
I
could
probably
dig
up
off
of
the
to
the
GDR.
Do
geothermal
data
repository
websites
like
website
and
so
and
I
I
think
their
seismic
there
too?
Who
would
have
to
look,
and
so.
G
Be
a
really
good
point
is
to
find
a
sort
of
case
study
where
we
have
a
plethora
of
data
of
all
types
of
data
types
and
then
sort
of
the
developers.
If
all
these
libraries
can
say
hey
what
what
can
jimp
I
actually
do
modeling
wise?
What
consent
that
do,
conversion
wise?
What
can
study
on
to
do
for
the
other
things
that
fatty
under
does
all
the
pre-processing
right.
F
G
G
E
D
E
Yeah
I
posted
a
link
on
the
on
the
hi
candy
so
on
the
I
found
this
on
the
geothermal
data
repository.
So
then
there's
a
tab
there
for
the
the
Utah
forge
and
it
has
links
to
all
the
data
sets.
So
they
have
the
the
3d
geologic
models.
They
have
microseismic.
They
have
that
they
have
gravity
magnetic
surveys.
They
have
yeah
everything.
C
F
F
A
D
E
E
E
C
E
G
F
Motivation
for
me
to
thinking,
but
it
is
a
separate
repo
is
this-
is
actually
I
mean
it's
gonna,
be
something
that
we
turn
into
like
a
joint
publication
or
like
turn
into
a
cool
sort
of
stand-alone
repo
that
has
a
demo
of
things
together.
Where
is
the
subsurface
might
be
better
just
to
sort
of
really
focus
in
on
the
API
in
the
package?
That'll
be
created,
yeah.
B
B
G
We're
making
a
bunch
of
Jupiter
notebooks,
it's
probably
good
to
not
Richard,
have
those
saved
somewhere
like
after
all.
This
is
over
without
you
know,
worrying
about
those
nerve
books
being
sort
of
hacky
in
their
nature,
clogging
up
space
and
of
what's
supposed
to
be
a
you
know,
actual
package
library
or
something
like
slips
like
this.
D
D
A
A
A
little
tonight
there
are
three:
no,
there
are
three
every
day
that
Filippo
does
so
the
first
one
is
too
early
for
the
Americas
and
the
last
one
is
quite
late
for
Europeans.
So
ok,
it's
just
to
cover
with
all
of
them,
but
yeah.
It
kind
of
officially
starts
when
Lindsey
on
plane
and
Dom
will
be
sleeping
soundly.
I
hope!
That's
why
I
hope
I
want
to
check
with
you
guys.
What
do
you
expect
and
what?
A
D
B
That,
but
for
how
can
I
think
that
that's
the
a
CSA
because
the
random
people
that
they
don't
know
what's
the
whole
story
about?
They
cannot
contribute
in
something
yeah
very
deep.
If,
across
the
hackathon,
we
start
to
have
to
work
in
subsurface
and
something
comes
out
of
it
that
we
can
really
start
thinking
of
individual
tasks.
Maybe
we
can
start
over
in
to
to
everybody,
but
but
I
think
that
in
general
should
be
yes,
look
to
the
tier
stack
choose
what
you
want.
E
A
B
B
B
A
These
there
are
please
or
three
zooms
continuously
going
right
from
its
assume.
It's
the
hackathon
cafe
and
it's
AIT's
the
transform
cafe
and
then
it's
the
transform
hackathon
they
are
always
on
and
they
can
just
login
or
logout.
So
that's
always,
we
could
always
say
the
meter
because
maybe
they
are
not
busy
at
all
and
so.
A
A
A
D
A
C
Wake
up
early
and
join
see
what
see
what's
happening.
I
think
if
we
take
that
day
to
just
parse
out
the
forge
data
sets
and
figure
out
which
one
we
want
to
tackle,
that
would
be
the
optimal
for
the
rest
of
the
week.
Is
then,
when
I,
do
the
training
for
a
jump
I
I
can
I
can
have
something
to
work
with
right.
C
B
G
Well
now,
no
geological
model
I
did
a
some
idea.
Statistics
work
don't
actually
help
me
with
doing
some
inversion
of
the
available
gravity
data
they
have
for
that
site.
All
did
not
turn
out
any
bit
good,
so
we
could
revisit
that
code,
but
I
have
done
a
lot
of
work
to
parse
a
lot
of
that
data
into
by
this
a
data
structure.
So.
B
G
E
G
C
G
C
A
E
Just
so
one
I,
don't
think
so.
I
remember
now
sigh
I
actually
been
in
contact
with
the
the
geophysicist
from
the
Utah
Geological
Survey,
who
was
running
the
forge
acquisitions
so
yeah.
If
we
come
up
with
anything
like
we
could
reach
out
to
him
and
and
I'll
see
if
he,
if
we
just
did
a
whole
bunch
a
nonsense
or.