►
From YouTube: SimPEG meeting March 19
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).
B
A
A
D
D
F
H
G
D
J
L
M
Hi,
my
name
is
Tebow
I'm,
a
PhD
student
here
with
Doug
and
singing
our
Georgia
code
and
it's
equal
constraint
and.
J
E
N
N
Kenya
came
down
here
to
Carrasco
my,
so
my
interest
is
new
version,
primary
potato
field
that
mean-
and
that's
thank
you,
what
brought
us
to
this
meeting
today.
So
thanks,
everyone.
A
Just
so
you
know
UBC
I
just
needed
you.
We
were
getting
a
bit
of
feedback
so
Doug.
If
you
do
want
to
speak,
please
just
unmute
yourself.
It
doesn't
actually
look
like
I
can
unmute
you
sorry,
so
Dahmer
Tebow
can
jump
in
and
do
that
yeah
with
just
with
this
many
people.
Sometimes
you
get
these
these
feedback
loops.
So
Tebow
are
you,
okay,
helping
dunk
with
mute
versus
unmute.
A
Hurry
UBC's,
good,
you're,
still,
muted,
okay,
so
just
on
me,
if
you--if
you
want
to
jump
in
so
what
I'm
thinking
we
could
do
is
first,
just
do
sort
of
quick
updates
on
things
that
we
want
to
make
sure
the
whole
group
is
aware
of
or
interesting
things
you've
been
working
on
and
then
from
there.
We
can
move
on
and
go
more
into
some
discussion
on.
How
do
we
support
them?
A
A
Okay,
so
yeah,
don't
let
me
don't
mind
just
like
going
back
and
forth,
then
we
can
try
and
get
him
up
and
running,
but
not
totally
stole
okay.
So
one
quick
update
from
my
end,
I
was
at
the
Jupiter
2
meeting
last
week.
That
sort
of
I
didn't
really
get
much
done
on
the
simple
side
of
things.
I've
been
pushing
a
little
bit
more
in
the
district,
wise
document.
K
P
Begins
here
so
last
week,
I
I
was
telling
you
guys
that
John
and
I
at
the
asteroid
look
working
on
the
magnetic
radiometry
kernel
function,
so
we're
we're
really
close
to
be
a
definite
testing
on
this,
but
I
didn't
want
to
have
to
wait.
So
can
you
see
my
screen
all
yeah?
It's
all
good,
then
zip
ya
screen
is
kids,
so
we're
starting
to
thinking
in
terms
of
like
being
able
to
have
multiple
components
on
the
on
the
magnetic
survey.
P
So,
instead
of
only
having
TMI
as
usual,
so
we
can
have
also
the
gradients
on
the
object,
and
so
we
I
just
coded
the
first
implementation
other.
So
that
works
really
well
and
then
right
away,
because
everything
is
already
hooked
up.
We
will
do
the
person
versions
to
be
able
to
do
nvi
inversions
on
Greta
on
Meg
radiometry,
so
yeah
everything
works.
Works.
Great
here
is
a
nice
collaboration.
This
is
a
hundred
hundred
thousand
dollar
picture
right
there.
This
is
the
vtk
viewer
of
bang
using
Jose
Jose
tree
mash.
P
N
A
A
Okay,
all
right!
Well
then,
let's
start
the
conversation
a
bit
about
what
we
can
be
doing
to
for
CSM.
So
what
challenges
have
you
run
into
you
so
far,
and
then
maybe
we
can
try
and
come
up
with
some
strategies
for
like
a
productive
back-and-forth.
So,
as
you
run
into
challenges,
we
can
try
and
help
fix
them,
but
also
make
sure
that
that
stuff
is
documented
so
that
in
the
future
other
people
can
can
learn
from
that.
A
A
F
Worse,
so
I'm
I'm,
trying
to
so
use
control,
search,
lush,
magnetics
and
Mt,
so
I
would
like
to
characterize
anisotropy
parameters,
so
preferably
I
would
like
to
have
a
full
3x3
and
I
saw
trippy
tensor.
In
order
to
do
this,
however,
as
far
as
I
am
aware,
simpang
only
has
a
scalar
and
I
saw
to
be
input
at
this
point
or
sorry
scalar
conductivity.
It's.
A
N
A
Yeah
especially
research
applications
like
if
you're
trying
to
do
sort
of
a
straightforward
inversion.
We
have
a
fair
number
of
examples
for
that,
but
as
soon
as
you
sort
of
start
going
into
something
that
is
a
bit
out
of
the
ordinary,
a
lot
of
that
stuff
is
not
documented.
Okay,
okay,
so
that's
good
to
know
and
I
saw
Tripathi.
It
is
that
for
both
the
FTE
and
are.
N
Cool
and
then,
if
I
can
just
jump
in
one
comment
is
in
one
of
the
paths
that
we
think
now
is
empty
going
through
is
fancy
means
mastering
paste,
inversion
forth
and
Strabane,
and
that
goes
parallel
with
two
methods
that
antique
gives
commentary
information
both
enstrophy.
So
that's
general
direction
is
going
cool
and.
A
A
It's
it's
a
1d
code,
but
in
it
might
be
a
good
starting
point
as
well
and
we're
actually
so
feeder
who's.
The
main
developer
there
he's
in
this
new
Pegg
slack,
okay,
mm-hmm
and
we've
worked
with
him,
and
we
have
a
first
pass
of
an
inversion
of
EMP
mod
working.
So
if
you
want
to
play
around
with
like
1d
inversions
with
that
code,
that's
an
option,
but
right
now
that
code
is
like
sort
of
it's
in
development.
It's
not
actually
well
documented.
So
that's
something
that
we
can
be
in
touch
on.
A
A
A
N
That's
I'm
thinkin
that
that
sounds
great
Christy
I
mean
waiting
for
the
fall,
and
it's
rotating
me
and
also
exploring
the
possibility
of
using
poisonous
costly
in
which
Joe
has
worked
out.
He
said
in
case
all
those
reading.
That
kind
of
will
be
part
of
that
whatnot
to
his
research
is
so
it's
one
that
clear
that
there's
an
expectation
that
all
buildings
on
fire
some
bactine
to
get
out
things
down
at
this
point.
As
long
as
andrea
has
gift
for
modeling
capability
up,
then
she
can
proceed
with
the
research
eventually
put
some
discussing.
A
C
F
The
end
goal
is
actually
to
see
why
anisotropic
properties
we
can
actually
recover,
and
that's
why
I'm
I
want
to
do
the
joint
inversion
with
the
MT
and
CS
to
see
what
I
can
recover
using
those
methods
and
have
an
evaluation
of
those
parameters
and
how
well
we
can
recover
them
using
those
that
type
of
door
inversion.
So,
but
the
end
purpose
will
be
for
geologic
interpretation.
N
Yeah
I'll
come
down
make
is
that
it's
pretty
rare
for
had
to
have
the
principal
directions
of
the
I
was
tropic
whose
the
property
alignment
was
using
the
coordinate.
They
were
rare
occasions
we
might
see
more,
but
as
soon
as
you
have
a
couple
of
people
to
a
lot
of
units
that
clear
principal
directions,
so
much
more
lined
designing
in
the
same
cooling
system
that
works.
E
E
D
G
That
the
data
that
I've
collected
is
over
an
oil
field
for
co2
monitoring.
It's
a
time-lapse
dataset,
there's,
there's
controlled
source,
iam
and
Magneto
lyrics,
but
that's
yeah
at
the
monitoring
co2
injected
into
a
reservoir
at
about
1400
meters
depth
and
the
survey
scale
is
on
the
order
of
five
kilometres
by
five
kilometres
or
so
and.
G
That
setting
the
reservoir
is
overlain
by
a
lot
of
shale,
which
is
going
to
have
at
least
vertical
and
horizontal
ice,
not
repeat,
and
then
in
addition
that
that
fail
isn't
necessarily
you
know
perfectly
layered.
You
know
it's
not
a
perfect
1d
model
either.
There's
we're
kind
of
at
the
at
the
edge
of
an
anticline.
G
Yeah,
so
I
would
say
that
my
project
is
kind
of
a
jumping
off
point
for
Andreea.
But
the
specifics
of
that
of
my
data
set
are
the
only
motivation
for
her.
But
we
have
the
data
and
it's
a
good
good
opportunity
for
her
to
take
advantage
of
it,
and
we
kind
of
move
into
the
things
that
she
was
really
interested
in.
N
Were
gasses
were
the
crossing
data
indicates
entities,
so
data
indicates
the
ratio
and
from
the
mission
in
the
mobility
can
be
up
to
32
one,
that's
huge.
So
in
our
case
the
organ
see
no
to
our
table
so
that
a
model
apartment
longer
is
interesting
with
and
present
fracture
ones.
That
would
help
me
Joe
Slovak
preventive.
What
also
sushi
was
that
electrical
activity,
so
we
eventually
interest
is
not
necessary
to
our
and
it's
probably
back
to
conductivity
what
apply
that
one.
Can
we
get
you
in
terms
of?
We
are
like
all
the
primers
hey.
E
Understood
so
I
again
just
kind
of
close
the
loop
you're
thinking,
you're
thinking
about
a
forward
modeling
in
situations
like
this
I
think
it's
entirely
the
place
to
concentrate
ones.
Efforts
just
can't
figure
out
how
to
forward
model
something
of
the
scale
that
you
are
interested
in,
and
seven
transmitters
look
at
what
what
the
data
are
and
to
get
some
sense
above
what
that
relationship
is
so
I
think.
E
J
F
A
Okay,
yeah,
that's
totally
fine,
just
kind
of
want
to
get
a
sense
of
like
if
we
send
you
a
Jupiter
notebook,
yeah,
a
good
starting
point:
okay
yeah
one
of
the
things
there's
a
couple
of
things
that
we
can
try
and
think
about
for
starting
points
and
then
also
for
perhaps
how
this
like
feeds
back
into
the
this
impact
ecosystem
for
us,
helping
onboard
other
researchers.
G
A
Sorry,
can
you
say
it
again:
yeah
1d,
he
sort
of
cut
out
yeah.
G
A
A
N
A
For
empty
as
well
I
would
suggest
we
pick
I
would
suggest
we
pick
just
one
to
start
with
and
get
familiar
with
the
API
and
just
like
we
yeah,
we
can
sort
of
learn
a
bunch
of
things
with
one
example
and
then
move
to
the
other.
So,
whichever
you
prefer
to
start
with,
we
can
yeah.
A
E
A
Yeah,
you
read
one
of
the
things
that
would
be
helpful
to
do
along
the
way,
and
you
can
do
this.
However,
you're
comfortable
one
of
the
things
that
I've
seen
when
people
are
getting
on
boarded
with
a
new
code,
is
that
I'm
just
sort
of
keeping
a
log
of
like
their
thought
process
and
what
challenges
they
run
into
along
the
way?
So
if
you
run
a
command-
and
you
get
a
strange
error
that
confuses
you
just
like
keeping
track
of
those
things-
is
really
valuable
information
for
us
and
to
help
you
as
well.
A
If
there's
a
space
like
you,
can
either
do
it
in
like
a
github
issue
or
if
you
want
to
just
keep
a
running
Google
Doc
or
something
like
that,
that
we
can
share
back
and
forth.
It's
yeah.
It's
helpful
because
often
like
it's
not
as
straightforward
as
debugging
a
simple
function,
because
there
are
so
many
layers
to
simple.
Sometimes
there
are
non-intuitive
things
that
happen,
and
so
we
sort
of
know
how
you're
thinking
about
the
problem.
We
can
try
and
help
diagnose
that
a
little
more
quickly.
A
Are
you
comfortable
with
like
git
and
github
comfortable
in,
like
the
sense
of
have
you
cloned
a
repo
and
things
like
that
before,
or
is
that
a
new
concept
I
have
not
cloned
a
repo
or
anything
like
that
before?
Okay,.
A
Okay,
because
one
of
the
things
that
might
be
useful
and
we
can
give
that
a
try
and
see
if
it
works
for
you
and
if
not
that's
fine,
but
we
have
a
simple
research
organization
where
we
just
sort
of
have
notebooks
that
are
like
research
projects.
So
I
put
all
my
notebooks
from
my
thesis
in
this
organization,
so
we
can
start
a
repository
of
notebooks
or
scripts
or
things
like
that.
Then
that
way
we
can
share
things
back
and
forth
and
it
helps
us
keep
track
of
issues
and
things
like
that.
Okay,
so.
J
A
A
Andy
well
we're
on
the
cult.
Do
you
have
questions
or
things
like
with
impact
that
we
could
supporting
Luis
right
now
or
is
it
sort
of
just
more
keeping
a
pulse
on
on?
What's
going
on
in
the
project
or
yeah.
A
Q
Q
C
I
just
got
a
one
comment
to
sort
of
grounded
source
I
think
we
we
have
a
line
current
source
like
that,
can
actually
emulate
the
ground
this
worse
and
that
tensor
mesh,
but
what
we
don't
have
that's
the
same
function
for
the
oxy
code,
so
it'll
be
kind
of
helpful
to
get
some
help
from
Joe,
because
I
I
think
we
haven't
really
used
that
octree
mesh
for
the
grounded
source.
So
it'll
be
a
good
test
case
anyway,
and
really
neat
that
Semyon,
like
it'll,
be
a
good
feedback
loop.
So.
A
C
Make
an
issue:
do
you
know,
do
you
know
what
that
would
require
I
think
it's
mostly
like
a
searching
function
like
a
like
the
word
nodes
in
a
Cell
and
yeah,
mostly
like
how
like
that
searching
part,
is
organized.
So
it's
a
it's
like
a
like
a
like
a
numerical
integration
code.
That's
required,
then
most
of
the
basics
are
done
in
tensor,
so
I
mean
that's
basically
the
same
but
yeah.
We
just
need
a
like
a
couple
of
functions
for
octree
I.
Q
J
P
Q
Q
E
And
then
here,
Lindsey
I
was
just
wondering
there's
a
few
things.
In
addition,
you
could
potentially
talk
about
with
OCS
and
group.
In
particular,
you
know
maybe
looking
at
gravity
to
ringing
corrections
and
seeing,
if
that's
a
point
of
connection
here,
but
Caitlin
is
here
from
sjb
and
a
lot
of
the
conversation
that
we
just
had
carries
over.
Like
first
week,
yeah
started
out
very
interested
and
electromagnetic,
large-scale,
loops
and
yeah
the
whole
business
of
trying
to
get
involved
with
SIF
a
and
you
know
even
do
the
forward.
O
L
N
O
And
we're
just
getting
our
feet
wet
now,
so
so
we've
got
some
slack
accounts
just
this
week
and
Google
hangout
we've
tried
that
out
and
we
tried
it
out
in
on
Firefox
earlier
this
week
and
guess
what
it
worked.
And
then
we
we
tried
to
join
the
meeting
today.
It
says:
oh,
don't,
doesn't
work
on
Firefox
only
Chrome
I,
don't
know
if
you've
had
that
before
anyway,.
O
O
I've
just
been
playing
around
with
some
code
this
this
week
in
experimenting
a
bit,
but
that's
the
extent
of
it.
I
guess
you
know
you
guys,
know
goodne
and
he
worked
for
us
for
a
little
while
and
he
was
our
simpe
guy
and
then
he
up
and
left,
which
was
really
unfortunate
because
now
we
want
to
take
this
to
the
next
step
and
he's
not
around.
O
O
That's
what
I
have
to
say
right
now:
I
mean
I'm,
really
keen
on
you
know,
making
comments
about
what
I
find,
and
you
know
fixing
the
odd
thing
here
and
there
and
and
and
you
know,
putting
that
back
right
away.
So
I
don't
know
what
the
the
best
way
to
proceed
from
here
is
I
mean
I've
got
like
a
one-liner
bug-fix,
so
I've
never
done
out.
O
A
pull
request
in
in
github
before
so.
Maybe
that
would
be
a
good
good
idea
to
try
that
out
there.
It's
very
minor-
and
you
know
when
I
was
going
through
some
of
the
examples,
and
some
of
the
code
was
obviously
copied
from
say
that
grabbed
to
the
mag
or
vice
versa,
because
a
lot
of
the
comments
and
they're
still
refer
to
the
beg
manometer
when
it's
actually
the
graph
revenue
example
and.
A
Well,
if
you,
if
you
have
fixes
to
contribute,
we
can
have
you,
so
you
never
really
have
you
used.
It
hope
before
not
much.
Oh
I.
O
O
A
Excellent
and
then
for
the
the
large
loop
modeling
that
you're
doing
is
it
time
domain
or
frequency
domain
time.
A
O
A
D
E
Not
really,
except
that
I'm
I'm,
just
wondering
now
for
the
point
of
view
of
Kalyan
actually
trying
to
put
together
something
for
doing
the
forward.
Modeling
and
testing
is,
is
that
are
you
planning
to
do
that
all
in
that
link
in
a
jupiter
notebook
so
that
you
know
it
could
be
passed
on
and
other
people
could
kind
of
take
a
look
at
it,
and
maybe
you
know
work
with
it.
I'm
thinking,
especially
as
you
as
things
are
set
up
in
an
octree
and
then
you've
actually
got
you
know
a
plate
in
there.
E
O
D
E
H
O
E
I
guess,
as
a
esta
comparison,
I'm,
not
sure
this
is
a
good
idea
when
Z
but
yeah.
It's
just
cross
my
mind.
So
if
the
cylindrical
code
aid
with
it
with
a
1d
background,
and
so
if
you,
if
you
take
a
loop
and
you
put
a
B
field
receiver
at
the
center
you're,
going
to
get
virtually
the
same
result
as
if
you
have
a
square
root-
and
you
know
normalized
by
the
by
the
area
of
that
movie,
you
should
start
to
get
pretty
much.
E
The
same
result
from
the
Selena
Koko,
which
you
have
really
good
accuracy,
because
you've
got
one
period
dementia
and
then,
when
you
go
to
both
when
you
go
to
the
an
octree
code
and
your
discretizing
things
with
lockrey
I
mean
just
it's
just
a
way
of
you
know
generating
an
example.
It
gives
you
something
and
then
you
could
use
that
as
a
check
for
your
next
step
of
sophistication
in
the
modeling.
A
A
C
I
think
that's
it
that's
a
kind
of
really
good
point.
That's
outside
of
the
like
good!
Well,
not
like,
not
sure
what
is
outside,
but
like
other
people
can
contribute,
because
the
sin
that
code
is
let's
say
am
code,
for
instance.
Test
is
for
a
certain
cases
like
a
certain
applications,
but
like
a
not
really
tested
for
a
whole
bunch
of
applications.
C
A
That
is,
that
a
term
that
resonates
or
no
I'm,
not
that's
very
so
API-
is
application
programming
interface.
So
what
that
means
is
basically
like
the
layer
of
code
that
you,
as
a
user,
are
interacting
with
so
you're,
not
necessarily
like
being
every
single
line
of
code
but
you're
calling
functions
and
so
the
way
that
we've
organized
and
named
those
is
an
API.
A
O
O
A
Doug,
did
you
want
to
bring
up
questions
or
ideas
of
a
gravity,
or
do
you
want
to
table
that
for
later?
We
still
have
a
few
minutes
so
well,
maybe
it
just.
E
Open
that
up
a
little
bit,
the
reason
that
it
kind
of
surfaced
was
that
the
he'd-
and
he
could
talk
a
little
bit
about
the
problem-
is
that
doing
some
gravity
blocking
mostly
gravity
inversion
on
the
geothermal
deposit
in
Iran
and
so
he's
got
the
gravity
data,
and
then
he
was
doing
a
terrain
correction
using
a
couple
of
different
programs
and
getting
out
different
results
and,
of
course
the
inversion
results
are
different.
You
know
like
what
do
we
believe?
E
Yes,
yesterday,
where
you're
doing
terrain
Corrections
using
your
quadtree
formulation
and
vertical
cylinders,
it
seems
it
since
it's
a
problem.
That's
an
interest
here.
Right
now,
it's
a
problem
that
you
guys
at
CSM
are
expert
on
that.
It
would
be
really
advantageous
for
a
number
of
perspectives
to
think
about
this
as
a
connection
project
where
we
develop
something
within
a
Python
framework
within
sempai
that
actually
does
terrain
Corrections
and
perhaps
does
it
as
well
or
better
than
anything
else.
E
That's
out
there,
so
I
think
I
just
wanted
to
kind
of
throw
that
out
as
as
a
possibility
for
kind
of
working
on
things,
because
now
we've
got
something
involved,
strong
countries
and
you
know
real
topography
and
if
we
could
do
something,
be
interesting.
I'm,
not
sure
that
there's
enough
new
stuff
there
to
actually
write
a
paper.
E
But
what,
if
sitting
around
lunch
today,
we're
realizing
that
over
Iran
you
don't
really
have
great
resolution,
digital
terrain
data
and
right
now,
but
he's
using
basically
250
meter
cells
I,
think
we
can
get
90
according
to
Tom,
but
okay,
there's
a
lot
of
Carter
and
frankly,
I'm,
not
sure
what
the
best
way
of
handling
that
is.
When
you
only
got
it,
you
know
poor
terrain
model
or
what
we
could
do.
E
There's
there's
stuff
here.
There's
there
scope
doesn't
play
it's
in
the
end.
It's
a
tool
we
readily
available
would
be
interest
to
people,
and
so
I
don't
know
that
the
code
that
you
have
Yahweh
you
said
you
distributed
it.
You
know,
do
people
know
about
it.
Are
they
using
it
or
as
a
God
where
what.
N
A
stupider
to
the
sponsors
of
GMO
see
is
not
widely
distributed.
No
is
it
open
source
at
this
point?
Actually,
your
Frank.
We
are
open
to
to
do
whatever
we
can
and
that's
something
that
the
Joe
I've
been
talking
about
is
actually
fun
to
read
some
of
its
project
books
under
Python
and
I'll.
Let
Joe
coming
on
that
shortly,
but
back
to
the
forum
itself,
yeah,
so
I'm
leaving
you
ran
in
low
flattery
area.
N
You
know
the
mid-latitude.
There
should
be
30
meter,
srtm
data
not
is
kind
of
really
two
very
low.
That's
you
that's
one
thing
and
another
is
that
with
the
code
we
have
a
gas
emissions
today
we
use
the
flat
talked
to
prison,
I.
Think
if
we're
gonna
implement
this,
we
should
really
go
to
the
next
step,
make
it
slow
control
top
prison.
N
That's
one
thing,
and
other
part
I
want
to
suggest
is
that
if
you
have
really
extreme
terrain
and
if
it's
still
possibility,
someone
really
could
just
go
back
to
those
stations
kind
of
measure
using
a
Konami
made.
A
few
data
points
run
those
vehicle
stations.
You
can
use
that
supplement
as
artyom
data
and
if
someone
want
to
get
a
paper
out
of
that,
that's
doable
and
no
I,
don't
think
anyone
has
combination
of
the
MIR
station
measurement
where's.
The
DM
data
in
general
that
there's
possibly.
J
N
Q
I,
don't
you
know
I
can't
comment
a
little
bit,
so
it's
like
in
for
our
stuff
that
we
have
its
I've,
been
working
on
having
most
of
our
Fortran
slash,
close
or
Fortran
and
C
stuff
wrapped
into
Python.
So,
like
our
sensitivity,
matrices
and
just
for
modeling
gravity,
gravity,
grady
and
all
those
things
to.
Q
Just
try
to
celebrate
them
and
being
able
to
you
know
both
take
advantage
of
open,
em
key
parallelism,
which
you
can't
really
do
with
Python
as
easily
as
you
can,
with
C
or
site,
or
also
use
it
with
syphon
as
well.
So
most
of
things
are
just
basically
seeing
sideline
wrappers
for
compiled
totes.
My
thoughts
are
to
be
are
that
they're,
currently
kind
of
built
around
a
bare-bones
in-house
meshing,
rectilinear,
mesh,
netting
and
I
thought
would
need
to
switch.
Q
J
N
Yes,
so,
within
that
context,
depending
on
how
the
grouping
collective
decide
to
proceed
with
the
trend,
correction
was
go
ahead,
was
putting
together
a
native
set
of
simple
code,
and
now
that
I
can
see
a
possible
possible.
One
is
just
wrap
the
foreign
code
that
we
Joe
has
been
doing
to
make
it
somewhat
part
of
the
sim
bag.
That's
what.
Q
I
thought
would
be
there's
a
way
that
you
can.
You
could
possibly
provide
like
a
Python
native
version
and
then
also
provide,
at
the
same
time
a
compiled
version
that
would
call
the
same
codes
and
way
you
can
write
it.
You
can
write
it
in
such
a
way
that
if
you
import
the
accelerated
version,
that
would
replace
those
codes
within
simple,
so
it
wouldn't
necessarily
do
anything
like
to
the
end
user.
It
they
would
just
be
able
to
natively
take
advantage
of
it
if
they
have
it
installed.
E
Great
appeal
from
a
whole
bunch
of
perspectives,
first
of
all,
just
kind
of
getting
you
know
a
project
that
GSM
and
UVC
the
rest
of
us
are
all
kind
of
interested
and
also
it's
it's
generic
right
I
mean
the
things
that
are
done
here
from
the
methodology
point
of
view
then,
can
get
extended
to
other
other
problems.
So
I
I,
don't
know
it's
just
got
the
right
feel
about
I.
Really
quite
like
this.
E
A
That
sounds
good.
Are
there
any
so
we're
just
a
just
over
the
hour
now?
Are
there
any
last
questions
comments
things
that
people
would
like
to
bring
out
today.
E
Do
you
have
anything
else
to
offer
and
at
the
end
of
our
conversation,
the
other
day,
you
also
said:
there's
a
there's,
a
programmer
SJD
that
you
know
we
could
potentially
get
to
help
that
there
was
some
problem
of
kind
of
mutual
interest.
Is
that
still
available,
and
does
that
what
kind
of
a
project
you
you
have
anything
in
mind
that
might
actually
be
useful
at
this
point,
given
that
you're
interacting
with
good
me
and
you
sort
of
see
what
he
was
doing.
O
I,
don't
know
I
guess
we'll
have
to
see
what
kind
of
resources
you
know
we
can
a
lot
to
to
get
him
involved
with
this
a
little
bit
he's
not
a
geophysicist
or
anything
like
that
by
any
means.
So,
but
you
might
be
able
to
help
out
with
some
programming
stuff
I.
E
E
Well,
it's
just
something
to
kind
of
keep
it
background.
We
don't
have
anything
particular
at
this
point,
but
it
it's
just
nice
to
know
cool
out
there
within
you
know.
The
group
has
got
capabilities
and
if
there's
problems
that
come
up
then
sometimes
maybe
you
can
just
ask
that
particular
person
to
contribute
or
some
ideas
and
go
for
a
lot
more
efficiently.
O
E
A
Who
is
interested
in
getting
hooked
up
to
github
slack
sort
of
those
community
pipelines
just
stay
online
and
that
otherwise
we
will
connect
again
next
week,
and
so
we
always
circulate
the
notes,
we'll
try
and
maybe
get
a
bit
better,
both
circulating
them
a
day
or
two
beforehand
and
add
things
that
they
want
to
talk
about
the
the
following
day.
So
but
it's
every
Tuesday
and
you're
always
welcome
all
right
music
thanks.
So
much
thanks.
Everyone.