►
From YouTube: 20220215 remote labs
Description
Remote Labs and FPGA Stand-up meeting on 15 February 2022.
A
Other
boot
directories
for
choco
cat
and
move
that
as
well
to
the
ssd.
So
if
all
that
goes
well
and
if
just
moving
these
things
around
from
from
share
to
share
works,
then
we
should
be
back
to
being
fast,
but
I
don't
know
why
it's
as
slow
as
it
is
so
there's
a
possible
mystery.
That
may
still
need
to
be
addressed.
A
Salvatore
has
found
some
some
work,
arounds
some
using
kernel.
Caching,
maybe
you
can
tell
us
a
little
bit
about
that
and
what
you
discovered
and
and
what
it
takes
to
get
set
up
salvador.
Are
you
there.
A
B
I
don't
know
if
the
resolution
it's
kind
of
worked
around
so
at
the
end
it
comes
that
bekash
is
a
component
that
is
able
to
erase
only
at
the
sides
of
the
of
a
block
onto
the
onto
the
underlying
memory,
and
this
greatly
helps,
maybe
in
because
normally
ssd
anyhow
are
made
by
nand
memories.
B
Nand
memories
inside
are
controllers
that
are
able
to
do
a
so-called
garbage
collector,
especially
when
there
are
under
a
huge
load
and
a
huge
right,
and
if,
in
this
condition,
you
also
ask
the
memory
for
reading
and
doing
right
and
related
the
the
the
job
of
this.
This
task
is,
very,
it's
very
huge,
almost
impossible.
B
So
because
memory
has
a
huge
latency,
they
are
hidden
from
the
user
perspective,
because
there
are
a
lot
of
dice
in
parallel
programmed
in
parallel,
multiple
lane,
programming
and,
and
so
the
the
big
cash
seems
to
me
to
to
reconciliate
with
to
to
to
know
how
memory
then
are
good
to
work
with
and
just
do
what
memory
you
want
in
a
in
a
better
way,
and
this
seems
that
and
also
I
then
back
up
this
main
memory
with
the
with
the
cache
and
the
device
is
anything
in
linux.
B
I
actually
said
to
use
that
zero.
That
is
even
compressed
ram,
so
it
had
so
under
a
heavy
write.
The
there
are
two
there
are
better
locality,
because
then
data
are
compressed,
so
are
more
local,
also
from
the
cache
perspective
from
the
processor
and
obviously
the
so,
the
the
the
written
is
delayed
and
then
is
well
managed
by
doing
explicitly
an
array
first
and
then
the
right.
B
But
this
is
not
rocket
science.
I
know
I
just
tried
and
it
seems
really
impressive
how
easy
is
to
set
up
how
mature
is
linux.
I
was
really
wonder
about
that,
and
so
I
I
even
created
the
file,
and
this
is
not
optimal-
you
can
create
a
partition.
Maybe
it's
better.
I
created
the
file
of
some
gigabyte
and
I
was
able
to
to
have
this
kind
of
speed
in
booster,
so
no
more
errors
when
having
catalinx
built
and
it
seems
like
a
normal
device.
A
B
You
mentioned
that
the
same
drive
has
the
parity.
C
A
C
C
D
B
A
B
Please
thank
thanks
cardinal
developer
more
than
me,
and
also
I
asked
for
the
which
is
the
device
then
the
the
physical
device
we
are
running
on
on
this
big
partition.
Maybe
we
can
have
a
look
into
how
is
organized
internally
to
understand.
If
there
are
some,
we
can
explore
some
suggestions:
how
to
how
to
better
match
the
how
physical
devices
organized
this
something
should
be
transparent,
but
at
the
end
nothing
is
transparent
in
some
condition.
B
Our
transparent
and
microcontroller
are
working
hard
to
behave
like
transparent
to
to
boost
performance,
but
at
some
times,
especially
when
garbage
collection
come
in
the
the
performance
really
degradate
a
lot.
A
But
I
put
an
answer
on
the
slack
for
what
actual
physical
devices
do
hold
the
files
at
the
moment,
but
that'll
be
obsolete
by
the
time.
Choco
cat
comes
back
online
because
we
won't
be
using
that
for
for
tools.
Anymore
still
be
true
for
for
big,
which
is
going
to
be
always
a
trade-off
between
speed
and
size.
E
Paul
the
big
partition
where
we
have
our
home
directories
will
that
also
move
to
ssds
or
it
will
stay
yeah.
A
E
And
how
much
do
you
remember
how
much
space
is
allocated
for
our
home
directory.
A
The
total
for
everybody
for
all
of
the
all
of
the
vms
plus
other
uses
of
the
cache,
is
one
terabyte
and.
A
No,
that's
total
per
user
is
not
not
set,
there's
no
quota
per
user.
I
don't
expect
to
add
one.
Unless
we
really
have
trouble
cooperating
with
disk
space,
there
will
be
a
limit
per
vm
and
I
will
be
deciding
that
and
setting
that
sometime
in
the
next
couple
hours.
I
think
I'm
expecting
it
to
be
200
or
300
gigs
so
that
we
have
room
for
a
couple
of
them
and
still
have
some
cash
space
left.
A
If,
while
you
need
the
speed
and
then
if
you
just
want
to
save
it,
then
it
needs
to
be
pushed
back
into
the
you
know,
big
directory,
so
that
we
don't
over,
allocate
the
ssd
and
there's
a
further
option
which
remains
available
to
us,
and
we
may
have
to
go
to
that
if,
if
all
this
doesn't
work
out
adequately,
which
is
to
simply
buy
an
ssd
for
each
vm
and
and
dedicate
it
now,
we'd
have,
however
much
ssd
space.
E
Yeah,
because
I'm
coming
from
the
background,
like
I
have
my
full
development
environment
set
up
over
there,
so
it's
kernel
you
boot
this
hardware
stuff,
so
everything
is
in
there.
So
I
want
to
do
compilation
over
there.
In
my
home
directory
everything
and
moving
from
home
to
big
is
not
an
option
then
it
will
take.
It
will
consume
a
lot
of
time,
so
yeah.
A
E
A
A
E
A
A
C
D
Okay,
thank
you
very
much
all
right,
yeah,
I'm
looking
forward
to
that
because
the
so
we
we
have
the
the
matlab
install.
So
maybe
it
might
be
good
to
talk
very
briefly
about
that.
I
don't
yet
know
how
to
most
advantageously
install
our
matlab
license.
That
gives
us
access
to
the
tools
to
do
matlab
straight
all
the
way
through
to
hdl
so
hdl
coder
and
also
has
gpu,
coder
and
and
some
of
the
other
toolboxes
that
are
not
usually
available
and
are
definitely
not
available
to
the
home
license.
D
So
in
while
you're
working
on
this
maybe
look
and
see
if
there's
anything
that
we
need
to
do
in
order
to
to
properly
install
matlab.
I
was
thinking
about
installing
it
on
karapi
so
that
it
would
not
interfere
with
choco
cat
and
the
the
over-the-air
stuff
with
the
9371,
and
that
may
or
may
not
be
the
best
call.
So
so
anybody
that
has
an
opinion
about
that.
Let
me
know,
while
we're,
while
we're
in
the
process
of
configuring
and
reconfiguring
and
optimizing.
A
Yeah,
the
other
obvious
alternative
would
be
to
install
it
in
its
own
vm,
so
it
didn't
interfere
with
either
of
the
hardware
setups,
but
whether
that's
convenient
or
not
depends
on
exactly
what
sort
of
things
you're
trying
to
do
with
it.
If
you're
trying
to
go
smoothly
directly
through
the
construction
of
a
hdl
and
then
compile
it
with
vivato
and
then
load
it
into
the
processor,
then
it
would
have
to
be
on
one
of
the
yeah
hardware,
vms.
D
Yes,
vivado
is
necessary
for
this
because
that
it
bolts
up
to
vavato,
as
far
as
I
can
tell,
or
that's
what
they
say.
So
that's
at
the
brochure
level
and
what
we
need
to
do
is
move
past
the
brochure
level
you
know
past
the
contract
level
and
pass
the
deliverables
of
the
the
code
level
yeah
and
then
actually
see
what
happens
because,
as
we
all
know,
those
those
two
things
are
can
be
very
different
experiences.
D
D
D
Okay,
so
very
brief,
hamcation
update
this
is
all
good
news,
so
we
we
had
a
booth
at
hamcation,
which
is
a
amateur
radio
event,
so
it's
one
of
the
largest
conferences
in
the
united
states,
possibly
becoming
the
largest,
because
it
is
competing
with
ham
vention
in
ohio,
for
for
the
largest
this
this
year,
arl
the
american
radio
relay
league
selected
hamcation
in
orlando
florida
as
their
national
conference
as
well.
So
this
was
somewhat
of
a
combined
event,
lots
and
lots
of
people
there
and
for
an
event
coming
out
of
coved.
D
D
We
have
lots
of
people
that
were
very
interested
in
what
we
do
and
want
to
join
or
contribute
so
I'll
be
responding
to
all
of
them.
Today.
Lots
of
people
that
wanted
to
sign
up
for
the
mailing
list
and
I'd
say
a
majority
of
the
people
who
talked
to
in
in
detail,
didn't
know
about
us
and
didn't
know
what
we
did
and
then
maybe
30
or
40
percent
were
dimly
aware
or
had
heard
of
us
before.
D
Everyone
was
delighted,
so
we
had
a
solid
good
experience
talking
to
people
in
person.
Events.
This
is
this
is
really
why
you
do
them
and,
yes,
it's
very
difficult
to
to
pull
it
off,
and
there
is
that
you
are
taking
a
risk
still
with
kovid
to
do
it,
but
it
it.
It
was
very,
very
good
to
do
so
lots
of
good
news
from
there
and
you
may
see
some
new
people
on
slack
as
a
result,
or
at
least
on
the
mailing
list
so
word
was
word-
was
spread.
D
I
was
able
to
give
a
presentation
about
digital
technology
like
in
general.
Why
is
it
so
cool?
And
what
exactly
are
we
doing
about
it
at
the
arl
forum,
which
was
on
thursday?
It's
a
separate
event
before
hamcation
and
open
research
institute
sponsored
a
whole
day-long
forum
track
in
the
the
hamkation
forums.
D
So
a
variety
of
people
spoke
and
I'll,
be
writing
up
a
report
and
sending
it
out
to
the
mailing
list
and
posting
it
on
slack
with
some
details
about
that,
so
that
was,
it
went
well
better
than
I
expected
we're
all
super
tired,
because
it's
hard
and
after
not
having
done
an
event
like
this
and
not
doing
booth
duty
for
two
years
nearly
well.
It's
it's
a
workout.
D
I
think
people
are
going
to
be
blown
away,
so
we've
consistently
delivered
really
good
presentations
to
ham
expo
and
eric
guth
has
written
me
several
times
to
highlight
how
happy
he
was
with
with
the
previous
work,
such
as
on
shul's
presentation,
which
was,
I
got
a
very
high
rating
from
the
he
wanted.
Eric
wanted
me
to
know
this,
so
I'm
I'm
sharing
it
with
you
all,
but
but
angela
got
a
very
high
rating
people
that
watch
the
talk
were
were
dumbfounded,
that
they
understood
it,
which
is
so.
D
You
go
to
a
very
highly
technical
talk
and
you
get
somebody
that's
actually
explaining
it
in
a
way
that
can
be
understood
so
anyway,
that
that
feedback
eric
wanted
me
to
hear,
and
also
the
other
people
that
spoke
over
the
past
couple
six
months
ago
and
then
a
year
ago
were
were
complimented.
D
So
we
we
went
ahead
and
did
did
something
for
him
expo.
We
bought
a
virtual
booth
because
of
feedback
from
us.
Nonprofits
now
get
a
very,
very
good
discount
for
booths
at
at
ham
expo.
So
we
will
have
one.
It's
the
open
research
institute
booth,
so
people
can
hang
out
and
we
can
put
additional
videos
there
that
we
can
do
at
the
last
minute
or
up
to
the
last
minute
pretty
much
and
then
we
can
also
have
pdfs
and
and-
and
we
can
post
people
and
answer
questions.
D
So
I'll
put
more
of
that
out
very
very
soon,
since
we're
back
from
hemcation
and
then
the
next
biggest
in-person
event,
which
is
a
large
demo,
opportunity
is
going
to
be
late
in
the
summer
and
I've
forgotten
the
dates
already
for
defcon.
But
we
are
we've
been
accepted
as
a
part
of
rf
village.
So
this
is
the
largest
radio
frequency
related
village
at
defcon.
It's
the
site
of
the
wi-fi
ctf
and
the
people
that
run.
It
are
huge
fans
of
what
we
do
and
are
very
supportive.
D
D
You
know
enthusiasts
and
developers,
then
at
least
a
big
part
of
it
and
to
show
our
stuff
working
over
the
air
we've
been
given
the
green
light
for
m17
related
demonstrations
to
connect
up
different
villages
and
we've
been
given
the
green
light
to
record
or
broadcast
talks
at
rf
village.
They
have
their
own
speaking
track
and
to
transmit
it
using
our
software,
so
our
firmware
our
our
stuff.
D
So
if
you
the
things
that
we're
working
on
now
like
the
encoder
and
the
dvb
s2
and
s2x,
those
things
can
be
used,
live
to
transmit
part
of
the
show
and
anything
else
that
we
can
think
of.
Really
that's
within
the
realm
of
physics.
You
know
impossibility,
so
that's
about
six
months
from
now.
D
It
seems
like
it's
far
away,
but
it's
really
not
so
so.
We'll
just
continue
working
towards
providing
some
demonstrations
there.
Since
it's
within
driving
distance
from
here,
it's
a
lot
easier
to
bring
equipment
to
defcon
than
it
is
say
hamcation,
which
is
all
the
way
across
the
continent
for
me
and
requires
plane
travel,
so
lots
more
about
that
coming
up
in
in
writing
on
slack
and
the
list,
okay,
so
that
for
in
other
words,
for
for
both
remote
labs
and
for
the
fpga
development.
D
D
Reference
design
we've
tried
using
the
start
with
the
reference
design
and
then
incorporate
our
ip
and
then
we've
tried
a
really
kind
of
cool,
innovative
approach
from
swatu
where
he
starts
with
his
encoder
and
then
tries
to
pull
in
the
the
reference
design,
and
so
I'm
going
to
stop
talking
and
listen
to
any
updates
on
that
and
how
it's
going
and
how
we
can
get
it
to
work,
because
I
think
once
we
get
it
to
it,
be
incorporated
in
the
design-
and
this
may
just
be
held
up
by
the
tools
problems
that
we're
having
like,
when
you
said
two
hours
to
start
the
to
get
the
block
diagram
up
and
running
that
broke
my
heart
because
that
should
not
have.
D
F
Yeah,
so
I
I
put
the
yeah
essentially
I
am
cloning.
The
repository
and
running
the
standard
flow
until
the
block
design
is
build
up
by
the
tool.
F
Actually,
no
sorry,
let
me
correct
that
so
the
way
I'm
doing
is
the
the
dvb
encoder
has
a
tickle
script
and
and
and
names
I
think,
a
make
file,
and
so
the
mig
file
will
clone
the
analog,
device's
repository
and
then
sort
of
patched
eight
sql
scripts
to
say
you
know,
source
this
other
sql
script
and
then
you
know,
launch
the
regular
build
process
and
it's
yeah
that
works,
and
I
I
I
don't
remember
exactly
who
did
something
with
the
license,
but
the
it
worked
in
the
in
vivado
2021.
F
F
It
is
a
good
idea
to
maybe
start
with
you
know:
booting
the
card
via
sd
card
as
usual,
and
and
then
just
use
jtag
to
program
only
the
fpga
and
do
a
warm
reboot
to
see
if
and
unlock
devices
that
the
thing
doesn't
and
the
sd
card
image
doesn't
overwrite
the
bit
stream
and
that
is
sort
of
a
hacky
way
to
get
it
to
you
know
to
work,
but
the
thing
is
so
I
was
looking
at
so
this
is
one
thing
yeah
I
was
looking
at
pluto's,
sdr
and-
and
I
realized
that,
like
the
way
I
connected,
the
encoder
is
not
exactly
how
like
it's
not
gonna
work,
because
essentially,
when
we
configure
the
dma,
so
we
select
files
via
the
iio
oscilloscope
and
then
and
there's
two
channels
and
the
software
side
will
write
to
memory
and
there's
like
side
by
side.
F
F
But
the
problem
is
this:
inter
d
interleaving
and
also
converts
from
axi
to
non-oxy
and
so
on,
the
exercise.
The
data
is
interleaved
and,
on
the
non-axis
side,
there's
no
back
pressure,
so
like
the
encoder,
it
doesn't
buffer
anything
like
it.
F
It
needs
to
be
able
to
apply
back
pressure,
and
so
what
I'm
planning
to
do
is
looking
at
how
data
arrives
in
the
access
stream
side,
because
I
think
that
the
interleaving
is
not
hard.
So
it's
essentially
you
know,
we
have
an
interleaved
stream
and
we
didn't
leave,
apply
or
insert
the
encoder
in
one
of
the
sides
of
the
stream
and
then
merge
them
together
like
in
in
accident.
This
is
it
is.
Is
it's
not
hard?
F
F
So
let
me
see
I
have
on
so
when
choco
cut
is
back,
we
can
try
the
actually.
I
can
build
a
flat
out
unmodified
version
of
analog
devices
stuff
a
bit
stream
and
then
we
can
power
on
the
zc706
from
sd
card.
D
F
Then
program
the
fpga
in
warm
reboot
to
see
if
this,
the
new
stream
sticks.
D
F
Yeah
does
is,
does
he
see
706
with
near
you
michelle.
D
D
Yeah,
it's
next
door.
It's
it's
actually
at
paul's
house.
D
D
Okay,
yeah,
just
what
we'll
do
is
we'll
be
on
on
high
alert,
then
just
whatever
it
takes
to
get
it
working
and-
and
I
know
that
I
think
onshore-
found
a
lot
of
really
good
things
about
the
fpga
manager
for
the
in
the
in
the.
B
D
D
So
I
will
I'm
going
to
defer
to
onshore.
You
might
have
a
good
solution
for
this.
F
Yeah,
so
if
if
he
gets
this
to
work,
then
obviously
we
don't
need
the
warm
reboot
hack,
stuff
right,
yeah.
E
So
yes,
so
device
three
overlay
and
that
fbg
manager
is
needed
if
we
have
ps
running
and
while
the
ps
is
running,
we
want
to
program
pl
now
for
testing
stuff,
as
so
to
mention
so
the
workaround
that
I'm
trying
to
do
is
I
have
this,
I'm
building
my
own
kernel-
and
I
have
this
ready-made
bsp
from
adi,
but
I'm
building
my
own
kernel
so
with
adi
drivers
again,
the
kernel
that
has
been
provided
by
adi
so
doing
the
menu
conflict,
selecting
its
configuration
and
then
building
it,
and
then
I
want
to
load
through
jtag,
both
the
ps
and
the
bit
stream
simultaneously.
E
Bitstream
is
again
from
the
same
example
from
adi,
so
they
have
this
zc706
example
with
adr,
9371
and
jesd
link.
So
I'm
building
that
example.
Generally
I
generate
the
bit
stream
from
there.
Then
I
am
trying
to
load
both
bit
stream
and
the
ps
simultaneously.
That
simplifies
the
things
for
us.
E
We
don't
need
fpga
manager,
we
don't
need
overlay
and
we
don't
need
to
boot
from
sd
we,
I
can
put
it
up
from.
E
E
Okay,
then
it
should
be
good,
but
for
past
couple
of
days,
so
this
I
did
last
last
week
and
for
past
couple
of
days
I
have
been
involved
with
codec2,
so
I'm
getting
up
to
speed
over
there
talking
to
colleagues
over
there
and
going
through
the
talks
and
youtube
links
so
yeah.
That's
my
progress.
F
So
I
you
you
want
to
go
first,
it's
right,
salvatore.
F
So,
just
sorry
real
quick,
so
everest
said
so
he
has
and
pluto
sdr
and
it
might
be
so
like
there's
two
things.
One
is
like
getting
the
ii
oscilloscope
to
work
with
our
bitstream
and
there
is
the.
F
Putting
the
encoder
in
the
correct
data
stream
for
this
second
part,
I
think
we
can
use
pluto
sdr,
like
everest,
has
and
pluto
hdr
with
him
and,
like.
I
essentially
just
need
to
say,
look
at
what
the
files
look
like
see.
You
know
see
what
the
files
we
load
look
like
and
say
we
one
five
is
just
I
know
abc
abc
and
the
other
is
just
like
a
ramp
or.
F
Known
patterns,
so
I,
if
I
use
if
we
can
add
the
ila
to
the
that
part
to
that
bus,
I
can
see
you
know
how
data
looks
like
and
how
it's
split
so
that
that
should
be
easy
to
replicate
in
sim.
And
then
I
can,
you
know,
move
go
on
from
there
and
have
the
encoder
in
the
stream.
So.
F
I
I
think,
for
now
I
don't
know
I
don't
have
one.
I
was
going
to
rely
on
the
various.
C
D
We
have
a
couple
here.
I
have
one
that
I
was
using
for
for
trying
to
work
on
the
uplink,
the
axm0f243
chip
that
we
are
looking
at
using
for
for
the
uplink
and
so
there's
some
experiments
here.
So,
but
I
have
a
pluto,
that's
probably
a
little
bit
under
utilized.
I
mean
it
can
use
any
any
sdr
for
receive.
D
C
D
Then
we
can
add
it
but
yeah.
I
think
any
any
sort
of
leverage
that
we
can
get
from
every
success
with
the
pluto
and
then
look
at
that
like
okay,
here's
how
it
gets
working
over
the
air,
because
that's
what
we
want
then,
then
to
so
whatever
it
takes.
However,
I
can
help.
F
D
G
Okay,
no
problem:
if
I
can
help
you
on
the
on
the
pluto,
I
can
make
pluto
reachable
also
from
alive
from
my
side
and
well
flushing.
The
pluto
is
quite
easy,
so
you
don't
need
any
any
hand
to
switch
sd
card
or
something
like
that
all
could
be
reprogrammed
easily,
so
they
just
for
the
presentation.
G
I
try
today
to
integrate
the
encoder
on
the
on
the
pre-tool
and
it
seems
to
nearly
work
but
andres.
It's
not
okay,
because
we
need
well.
There
is
some
timing
issue,
but
I
could
try
to
share
you.
I
know
I
couldn't
share
my
screen.
Okay,.
D
D
G
It's
beautiful
but
it
doesn't
work,
but
it's
that's!
Okay!
It's
the
beginning!
Sorry,
for
that.
The
idea
is
to
first
to
see
if
the
block
the
uvs2
encoder
shoots
in
the
pluto
and
as
you
can
see
here,
we
are
at
eighty
percent
here.
Eighty
percent
of
usage
of
dsp.
C
G
I
need
to
remove
the
the
interpolator,
the
decimator
and
the
and
only
two
channels
to
the
computer,
because
else
there
is
no
room
left,
so
it's
really
really
tiny.
I
hope
to
it
could
be
interesting
to
have
the
this
running
on
the
pluto
sdr,
but
it's
it's.
There
is
not
a
lot
of
place,
the
other
the
other
way
I
mean
well,
I
begin
to
know
well
all
the
build
root
stuff
and
the
linux
and
well,
and
the
fpga
warm-up
and
etc.
G
It's
well
it's
a
knowledge
I
have
now.
So
if,
with
this
tiny
platform,
it
could
be
useful
for
you,
then
it
could
be
good
and
the
maybe
the
other,
the
other
ways
that
well
many
people
have
created
sdr.
So
if
they
want
to
replicate
replicated
that
it's
maybe
a
good
opportunity,
so
I
try
to
to
help
andre
and
if
he
wants
any
information
I
can
answer
easily.
G
I
tried
to
answer
his
question
and
put
a
platform
up
with
hdr
here.
F
F
G
Oh
no,
okay.
On
the,
as
I
said
you
on
the
on
the
fpga
side,
I'm
I'm
trying
to
well
the
learning
curve.
Is
it's
quite
long.
I
try
to
train
myself
and
because
there
is
not
a
lot
of
helpful
people
around
me.
F
F
D
Take
some
work,
it's
you
know
and
some
stuffing
and
some
cramming
and
some
stomping
on
it
to
get
it
in
there.
But
you
know
it's
in
the
realm
of
possibility.
If
we're
efficient
and
clever.
F
G
Here
I
have
two
two
solutions.
I
have
pluto
through
ethernet
and
and
through
usb
for
the
usb.
What
is
quite
helpful,
it's
a
dfu
mod,
which
means
that
you
can
run
your
bitstream
and
even
all
the
all
the
linux
plus
the
bit
stream
directly
on
the
ram.
So
you
don't
need
to
flash
it
before
trying.
So
it's
it's
very.
It's
very
quick.
F
Yeah
yeah,
we
can
it's
it.
It
should
be
as
easy
as
that
to
find
out
if
it
is
possible
to
look
at
the
bus,
because
essentially
the
the
jtag
allows
us
to
it,
has
some
control
and
it
will
insert
like
a
monitors
of
sorts
and
we
we
can
say
you
know
I
want
to
sample
data
when
the
valid
is
high
or
you
know
something
is
higher
certain
conditions
and
it
will
say
it
can
sample
like
one
thousand
two
thousand
samples
and
then
send
them
back
via
jtag.
G
Yeah
but
but
I
need
to
to
have
a
g
tag
attached
right.
F
I
sometimes
the
thing
is
also
like:
if
you
can
program
the
flash,
it
probably
uses
jtag
because
it
makes
sense
like
jtag
is
used
for
like
manufacturing
stuff
as
well.
So
when
you
know
when
they
they
are
manufacturing
the
thing
they
just
plug
into
usb
and
do
whatever
programming
and
checking
they
need.
F
Yeah
yeah,
but
so
the
jtag,
usually
we
use
like
a
usb
to
j
tag
and
bridge
and
jen.
Some
cards
have
like
a
beauty
this
built-in
in
the
card
and
some
cards
like
they
don't.
So
we
we
need
like
an
external
thing,
like
the
old
232
kind
of
stuff,
but
in
this
case
I
think
this
should
be
like
a
usb
switch.
You
know
with
the
regular
data
stream
to
communicate
with
the
the
ps
and
another
one
that
communicates
with
the
jtag
yeah.
I
can
look
it
up
and
I
will
let
you
know.
D
F
D
Well,
thank
you,
everyone.
It's
a
huge
and
very
generous
donation
of
time
to
do
these
sorts
of
very
difficult
things,
but
it
will
succeed
and
it
will
be
very
useful
and
a
wonderful
thing
working
over
the
air,
all
right,
we'll
we'll
be
at
it,
there'll,
be
updates
about
choco
cat
and
the
repairs
to
the
I
o.