►
From YouTube: 20230621 ORI Office Hours
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
A
So
this
is
office
hours
for
today.
If
anybody
needs
anything
any
resources
or
or
help
with
any
roadblocks,
then
then
please
go
ahead
and
speak
up,
but
just
to
catch
us
up
and
and
have
an
office
hours
and
and
talk
about
all
the
adventures
and
Misadventures
over
the
past
couple
of
weeks.
A
B
Repair
schedule
I'm
actually
we're
doing
good
here,
because
I
just
finished
finals
for
the
previous
term
last
week,
so
I
finally
got
a
break
and
I
can
focus
more
on
fpega,
okay,.
A
Congratulations
on
on
completing
another,
a
big
chapter
of
school.
That's
a
big
deal
to
go!
Okay,
oh.
A
How
great
that's
that's
fantastic
news,
good
to
hear
yeah
fpgas
yeah
plenty
going
on
there.
We
have
the
9009
and
the
9002,
which
are
the
radio
cards.
The
9009
is
on
a
zc
106,
which
is
a
very
nice
Dev
board.
The
7000
series
zinc
on
it.
That's
one
station,
that's
the
transponder
station
is,
is
what
we've
started
to
call
it
intended
for
for
transponder
work
and
then
the
other
station
is
a
zcu
102..
A
We
also
have
a
zcu
106
sitting
to
the
side,
but
the
zcu-102
directly
supports
the
adrb
9002,
which
is
a
another
radio
card
by
Analog
Devices,
and
that
radio
card
gives
you
access
to
a
chip,
that's
designed
for
low
power
and
mobile
applications.
So
that's
the
one
for
the
Neptune
Project,
the
Drone
project,
that'll
be
the
first
customer
and
the
zcu
102.
The
U
is
kind
of
a
hint.
It's
an
ultra
scale
chip.
A
A
You
go
from
Matlab
and
simulink
to
HDL,
and
it
also
has
some
fpga
in
the
loop
and
test
bench
functionality
and
they
recommend
using
their
version
of
Linux,
which
is
called
build
root,
and
that
has
been
quite
the
Odyssey,
which
we've
documented
in
a
couple
of
different
places
internally
and
also
externally,
on
the
MathWorks
community
site
in
a
long
write-up.
And
then
we
have
a
immense
thread
with
some
very
helpful
people
at
MathWorks
is
unexpectedly
difficult
to
to
get
the
the
MathWorks
version
of
build
root
built
and
there
they
advertised
this.
A
This
version
of
Linux
embedded
Linux
to
go
on
the
platform
as
efficient
and
easy
and
simple,
and
it
will
make
your
life
better.
So
our
feedback
has
been
that
it's
not
done
that.
For
us
at
least,
we
are
using
a
supported
combination
of
the
development
station
and
the
radio
card,
but
putting
in
the
radio
card,
which
you
have
to
put
it
in
with
the
through
the
device
tree.
So
you
have
to
modify
the
device
tree
to
include
the
the
radio
card
that
fits
into
the
fpga
development
station
that
part's
almost
totally
undocumented.
A
We
do
have
some
experience
with
some
successful
builds
of
petal
Linux
on
on
the
other
station
and
that's
right
before
IMS
2023,
the
international
microwave
Symposium,
which
was
last
week
so
right
before
that
we
went
ahead
and
tried
to
build
petal
Linux
for
the
9002
station
as
well.
So
that's
where
we
were
and
through
lots
of
feedback
we
we
went
into
IMS
2023
week.
It's
a
week-long
microwave
set
of
conferences
and
I
was
a
volunteer
and
organizer
for
the
conference
so
spent
the
entire
week
at
the
conference.
A
A
You
know
some
of
the
the
process
and
the
workflow
and
tools
xilinx
wasn't
there,
but
there
was
a
lot
of
people
with
familiarity
with
the
xilinx
there,
and
the
good
news
is
that
we
did
get
all
the
way
up
to
getting
a
a
ticket
entered
into
the
into
Analog
Devices
internal
system.
To
try
to
resolve
our
weird
version,
SKU
problem,
yet
another
problem
with
the
with
the
stations,
so
so
that
happened.
A
It
was
very
good
to
talk
to
these
folks
in
person
and
to
be
able
to
go
from
one
booth,
just
a
few
rows
down
to
the
other
and
try
to
harmonize
all
these
different
versions.
A
So,
like
any
sort
of
complicated
system
where
you
have
development
cards
and
and
products
from
different
companies,
it
multiplies
in
in
the
complexity
and
the
main
challenge
that
we
have
is
understanding
which
versions
of
What
tool
are
compatible
and
there
has
to
be
a
Venn
diagram
where
at
least
where
all
this
stuff
overlaps
with
the
in
terms
of
compatibility,
it's
been
probably
more
effort
than
it
should
be
to
figure
out
exactly
how
to
how
to
set
the
stations
up.
A
We're
really
really
close
and
and
kind
of
the
the
good
news
about
all
of
this.
The
good
news
is
that
lots
of
progress
was
made
talking
to
people
in
person
and
there's
some
interesting
gossip
that
I
got
the
the
good
news
is
that
pedal
Linux
actually
built
and
and
ran
for
both
stations,
so
I
stopped
working
on
the
MathWorks,
build
route
right
before
IMS
and
got
got
to
build
a
pedal
Linux
for
the
Neptune
station
as
well.
A
I
think
we
should
go
ahead
and
try
to
use
pedal
Linux
with
HDL
coder
and
the
fpga
thing
in
the
loop
and
just
go
ahead
and
try
it,
and
it
may
be
that
we
have
to
build
the
MathWorks
version
of
Linux
to
get
all
of
their
functionality
working.
But
when
I
try
to
pin
them
down
to
tell
me
exactly
what
was
special
about
their
version
of
Linux.
A
They
couldn't
tell
me
so
it
could
be
that
they
it's
just
not
invented
here
syndrome
and
that
they
would
really
like
for
you
to
use
their
version
of
of
Linux.
But
it
is
not
as
well
supported
by
far
it's
not
as
well
supported
as
xilinx
pedal,
Linux
or
The
Analog
Devices
version
of
embedded
Linux
called
Kuiper.
A
That's
it
from
my
end
and
if
you
have
any
questions
about
all
this
complicated
stuff,
you
know
I'm
very
sorry,
it's
lots
of
acronyms
and
versions
and
stories
and
and
and
potential
paths
you
know,
but
we're
just
trying
to
be
sort
of
Relentless
about
this
and
solve
these
problems
and
get
all
the
tools
and
the
tool
flow
and
the
new
stations
set
up.
So
we're
trying
to
do
that
as
quickly
as
possible
and
I
think
we're
we're
pretty
much.
A
Almost
there
we're
there
with
petalytics
and
and
the
only
outstanding
question
is
profile
mismatch
with
the
software
from
Analog
Devices.
In
order
to
control
the
transceivers,
the
API
versions
have
to
match
in
Linux
and
in
their
utility
program,
and
they
don't
and
sure
enough.
The
profiles
don't
load,
that's
a
big
problem
and
then
the
other
big
problem
is
trying
to
build
whatever
Special
Sauce
MathWorks
has
in
their
version
of
build
root,
which
I'm
starting
to
suspect.
There
isn't
really
a
functional
difference.
It's
all
live
iio
calls
underneath.
A
So
that's
that's
the
giant
ball
of
wax
as
as
concisely
as
I
can
try
to
explain
it
and
then
I'm
gonna
I'll
hand.
It
over
to
Paul,
he
I
think
he
has
at
least
one
story
about
the
outage
that
we
had
at
the
lab,
which
I
think
is
pretty
interesting
and
then
anything
anything
else.
He
has
to
to
talk
about.
D
Okay,
first
on
compatibility
and
profiles
have
come
to
understand
that
there's
there's
a
transition
underway
from
an
old
way
of
doing
things
to
a
newer
way.
We
seem
to
be
caught
in
an
awkward
place
right
now,
where
you
can't
use.
I
I
o
exclusively,
because
it
doesn't
give
you
control
of
all
the
setup
and
all
the
details
of
what
you
have
to
do
to
get
the
setup
to
work,
sample
rates
and
frequencies
and
stuff
like
that
filter
coefficients
those
things
are
not
exposed
by
the
iao
API.
D
As
far
as
I
can
tell
they're
exposed
at
a
lower
level
in
a
proprietary
API,
which
is
where
the
versions
have
to
match,
and
so
you're
stuck.
If
you
want
to
use
iio,
which
is
the
more
modern
way
of
doing
things,
you
still
have
to
do
your
configuration
with
these
profiles
and
the
profiles
can't
really
be
generated
by
hand.
D
They
have
to
be
generated
by
this
application
that
Analog
Devices
provides
because
it's
the
only
thing
that
knows
all
the
rules
and
if
you
give
it
a
very
vanilla
set
of
parameters
like
the
defaults
and
a.
E
D
If
you
do
anything
wrong,
then
it
generates
a
profile
that
it's
just
not
accepted.
There's
no
error
messages.
It's
just
no,
and
this
is
not
very
helpful,
very
difficult
to
debug,
so
I
think
we'll
be
stuck
with
tiny,
incremental
changes
and
hope
we
can
get
from
the
default
to
where
we
need
to
be
in
a
reasonable
sort
of
way.
D
If
you
give
it
a
set
of
invalid
parameters
for
clock
rates,
for
instance,
it
will
tell
you
that
such
and
such
a
clock
rate
has
to
be
greater
than
that
one,
and
you
know
between
this
value
and
that
value,
and
typically
in
in
our
case,
what
we're
trying
to
do
something
relatively
narrow
band
on
a
wideband
radio.
What
we
want
is
outside
the
boundaries.
A
A
I
think
that
yeah,
the
9009
is
very
similar
in
capability
and
sort
of
style
to
the
9371,
which
is
a
very,
very
high.
You
know
this
is
a
very
capable
radio
card,
but
the
as
far
as
I
can
tell
the
lowest
sample
rate
that
we
can
well
that
sample
of
the
lowest
bandwidth
it'll
do
is
30
30
megahertz.
You
know
it's
like
a
you
know:
30
30
samples
per
second,
something
like
that.
So
we
were
aiming
for
10
and
or
10.
A
You
know
10
ish
that
and
that's
a
a
pretty
big
difference,
but
that's
that's
very
true
of
the
9009.
We
may
just
want
to
get
it
working
at
a
higher
bandwidth
and
Sample
rate,
and
you
know
first
and
then
and
then
work
on
on
narrowing
it
down
later
or
or
whatever.
A
But
if
we
get
good
advice
on
on
this
from
from
people
with
good
systems,
knowledge,
then
then
we'll
take
it,
but
over
on
the
Neptune
station
that
the
that's
a
it's
intended
for
to
support
an
airband
modes
and
it
so
it'll
go
down
to
12
kilohertz.
So
so
this
problem
is
is
not
as
acute
on
the
Neptune
station.
D
That's
good.
My
my
mission
at
the
moment
was
to
do
do
the
Uplink
demo
on
the
9009,
and
that
was
not
a
particularly
critical
path
thing:
it's
not
something
we
even
need.
It
was
just
a
I
thought
it
would
be
an
easy
thing
to
try
on
the
on
the
9009
station
transmit
side,
and
if
it
turns
out
it's
just
plain,
not
then
I'll
have
been
in
that
and
go
back
to
the
Pluto
for
demonstrating
the
narrow
band
stuff
which
is
actually
closer
to
what
we
need
for.
Were
they
Uplink
anyway.
C
D
So
that's
where
that
is
in
in
other
remote
Lab
news.
There's
a
server
software
upgrade
pending
now,
which
we'll
probably
want
to
adopt
at
some
less
disruptive
moment
than
than
more.
When
exactly
that
is
I'm,
not
sure
we,
we
did
go
through
some
pain
and
agony
due
to
a
disc
full
situation
which
has
all
sorts
of
nasty
consequences,
we're
able
to
sort
ourselves
out
of
it
by
moving
some
stuff
and
deleting
some
stuff
that
was
no
longer
needed
had
enough
Headroom
to
to
put
it
back
into
a
working
condition.
D
D
So
at
some
point
we'll
take
the
system
down
and
hopefully
for
a
very
short
time
and
bring
it
back
up
on
6.12.
I'll
probably
do
that
on
the
lab
South
machine
first,
because
it's
not
under
heavy
use
and
when
that
works
hopefully
should
be
straightforward,
then
try
it
on
the
remote
lab
West
system
and
hopefully
it
won't
disrupt
the
workflow
for
for
very
long
I
I
think
I
covered
did
I
cover
the
story.
You
wanted
me
to
tell
I
wasn't
entirely
sure.
A
Which
one
you
were
yeah
the
well
there's.
There's
lots
of
stories
plenty
of
good
stuff
going
on,
but
no
the
yeah,
the
the
I
thought
the
outage
was
pretty
interesting.
The
the
disc
full
situation
because
it
kind
of
caught
us
by
surprise
in
the
middle
of
a
build.
So
everything
just
completely
froze.
A
Full
stop.
You
know
no,
no
mouse,
no,
nothing!
No!
So
so
the
fear
was,
oh
dear.
You
know
this
yeah,
but
but
it
did,
it
turned
out
to
be
a
disc
full.
We
thought.
Maybe
there
was
a
a
disc
failure
and
how
how
it
got
full
is
still
a
mystery
to
me,
because
it's
it's
not
obvious
when
you,
when
you
kind
of
run
the
command
line,
Diagnostics
that
that
it
was,
it
was
too
too
full.
So.
D
I
I
do
not
know
that
in
all
the
ins
and
outs,
but
I
think
what
happened
is
that
we
had
overfilled
it
by
doing
repeated,
builds
on
the
system
disk
essentially,
and
we
caught
ourselves
real
right
realizing.
D
This
was
going
to
be
a
problem
and
deleted
some
builds,
which
should
have
given
us
Headroom
for
quite
a
while,
but
because
of
the
way
this
particular
file
system
stack,
Works,
you
end
up
with
blocks
allocated
that
stay
allocated,
even
if
you
delete
the
files
that
they're
holding-
and
this
is
one
of
the
problems
with
this
file
system
that
we're
currently
using
for
for
the
Cash
Drive,
and
one
of
the
reasons
why
we
want
to
do
the
overall
system
upgrade.
D
A
You
said
you
you
wanted
maybe
to
to
use,
maybe
try
this
on
the
at
remote
lab
South
to
to
Pilot.
It.
D
C
C
F
I
I'm
sitting
here
listening
so
my
my
background
is
in
software
engineering
or
software
development
I,
don't
have
a
degree
and
citizens
Administration
and
I'm
wondering
do
we
have
like
first
of
all
was
that
an
inode
starvation
issue
it
sounded
like
maybe
you
were
saying
the
blocks
weren't
getting
freed,
but
also
do
we
have
like
monitoring
on
critical
systems
so
that
we
can
catch
those
things
before
they
bite
us
and
if
not,
is
there
an
appetite
for
setting
up
like
moon
or
something
to
keep
track
of?
D
F
Yeah
I,
don't
know,
I,
don't
know,
EXT2
was
I
nodes,
I,
don't
know
what
exe3
and
and
xfs
and
the
rest
of
them
use,
but
I'm
sure
they
have
a
similar
like
abstraction.
Underneath
that.
D
F
Yeah
definitely
as
I've
heard
of
ZFS
at
least
and
and
have
installed
it
I've,
never
I'm,
looking
at
butterfs
now
but
cool
yeah.
If
there's
anything
I
can
do
to
help
out
like
I
said,
I
have
a
background
in
systems
programming
systems
Administration
almost
all
Linux,
so
I'm
happy
to
help
out.
If
I
can.
A
F
F
Still
talking
about
I
knows
in
2023,
but
the
concepts
I
think
translate
so
and
and
I
have
I
have
friends
that
are
still
working
actively
in
Industry,
so
I
at
least
have
resources
to
pull
from.
If,
if,
if
we've
got
questions
so
anyway,
glad
to
help
if
I
can
thank
you.
A
Yeah,
let's
see
we
have
what
else
is
going
on
in
Labs
any
Uplink
any
Uplink
progress
over
past
couple
of
weeks.
It's
been
been
since
we,
since
we
last
talked
anything
about
opulent
voice
that
you
want
to
share.
D
D
Can
wanna
you
know,
I
want
to
create
something
more
and
more
like
a
real
transmitter.
You
know
push
to
talk
and
actual
interfaces
to
microphones
and
and
such
instead
of
you
know,
file
the
file
and
simulation
work.
So
we
can
find
out
what
the
new
problems
we
don't
know
about
are,
hopefully
that'll
all
just
work.
Once
we
get
past
the
this,
the
slugging
yeah
slogging
through
getting
it
all
built.
F
All
right,
I
did
want
to
ask.
You
were
talking
about
configuration
files
for
I'm,
guessing
analog,
one
of
The
Analog
Devices
ICS.
Is
that
a
one
and
done
thing
or
is
that
something
you
have
to
generate
every
time
you
start
up
or
like
what
is.
A
Oh,
it's
a
really
good
question,
so
they
are
they're
files
that
you
generate
with
their
utility
and
depending
on
the
IC,
you
may
have
a
single
human,
readable
text
file
or
you
may
have
a
Json
file
and
then
a
binary
file.
The
binary
file
will
go
to
the
arm
and
the
text
file
gets
read
in
to
this
the
system
directory
through
the
system
to
directly
to
the
devices
with
a
cat.
You
can't
redirect
it
to
this
pretty
neat.
A
It's
a
little,
you
know,
so
it
feels
like
you're
mucking
around
with
the
central
nervous
system,
and
that
happens
when
it
when
it's
as
far
as
I
can
tell
that
happens
when
it
starts
up,
and
so
it
is
sort
of
one
and
done
and
that
you
figure
out
your
your
clock
scheme.
Your
sample
rates,
your
bandwidths
there's,
there's
some
gain
settings
in
there.
Adc
settings
whatever
that
particular
IC
needs,
there's
a
pretty
long
list.
This
stuff's
documented
in
the
the
the
user
guide.
Sorry.
F
A
Yeah
so
yeah
to
create
this
profile,
which
then
lives
inside
your
your
your
deployed
system.
You
know
to
create
these
files,
you
use
their
utility
and
and
all
of
these
settings
all
these
registers
are.
They
are
all
documented
in
the
the
programming
guide.
So
it's
not
super
secret,
but
it
is
an
enormous
amount
of
you
know,
configuration
the
configuration,
configuration
files,
the
ones
that
are
human
readable.
When
you,
when
you
look
at
it,
they
sort
of
look
to
me
like
they
mirror
the
device
tree
so
they're.
F
B
A
Conflicts
and
yeah
the
version
conflict
is
between
the
the
what
the
Linux
version
that
you
have
and
the
API
in
Linux
and
then
the
API
that's
buried
in
the
utility
tool,
which
is
proprietary
and
it
comes
from
Analog
Devices,
and
the
reason
why
this
has
been
a
very
tricky
thing
is
because
Analog
Devices
only
publishes
the
absolute
latest
version
of
this
utility.
A
However,
we
are
using
2021.1,
paddle
Linux
and
the
2021.1
version
of
the
HDL
reference
design
from
Analog
Devices
in
order
to
use
the
transceiver
toolbox
from
Analog
Devices
and
a
lot
of
other
stuff.
So
there
it's
weird
that
this
you
know
well,
I,
guess
it's
not
too
weird.
If
you've
ever
actually
worked
at
a
company
where
there's
a
significant
amount
of
people
and
there's
different
departments,
you
do
know
that
sometimes
they
don't
talk
to
each
other.
Oh.
A
Then-
and
that
is
what
I
was
told
at
the
booth-
is
that
these
are
two
teams
that
are
not
communicating
well
during
this.
You
know
because
it
used
to
be.
There
was
no
version
really
the
tez,
the
transceiver
evaluation
software
produced
a
file
for
a
particular
radio
chip,
and
that
was
the
file
that
you
use
the
utility
to
create
the
file
and
if
you
in
the
different
versions
of
tez
all
produced
at
least
the
same
format,
but
what
they've
done
is
they've
started
changing
the
format
between
these
versions.
A
A
It
is
a
nightmare
and
I
I
use
that
word
at
the
booth
politely
to
the
nice
Fae
that
that,
after
drilling
down
for
a
couple
of
layers
of
very
nice
people
at
The,
Analog
Devices
Booth
the
pit
somebody
that
was
familiar
with
their
this
particular
part
of
their
product
line
and
and
he
he
was
extremely
empathetic
and
understood
what
I
the
problem
immediately.
F
A
Yeah,
never
lose
your
copy
of
tez
that
you
managed
to
get
distributed
or
go
so
we
went
to
the
internet
archive
that
was
Paul's
idea.
He
says
why
don't
we
just
look
in
the
internet
archive
and
see
if
we
can
find
this
version?
Well,
we
found
every
version
before
and-
and
we
found
the
version
after,
but
the.
A
One
for
early
from
early
2023,
so
this
was
the
last
it
was
available,
was
it
looks
like
around
January
2023
was
not
snapshotted
by
internet
archive,
so
I
put
the
appeal
out
on
LinkedIn
and
and
on
our
list
and
everything
so
somebody's
got
to
have
it
in
an
install
laying
around
it's
Windows
only.
So
it's
just
a
little
tiny.
A
A
You
would
I
I
agree
with
you,
so
I
attempted
to
Earl
hack
at
Analog
Devices
for
about
a
half
hour
and
and
and
make
it
regurgitate
it
somehow,
maybe
through
their
you
know,
through
all
of
their
different
directories
and
I've
written
several
people
that
I
know
and
the
answer,
and
also
posted
on
engineer
Zone,
which
is
where
you're
supposed
to
go
for
all
the
tech
support
and
it
has
not
hasn't
worked
yet.
A
You
know
but
I
I'm,
Gonna,
Keep,
chipping
away
and-
and
you
know,
I
think
that
Analog
Devices
once
they
get
through
this
transition
and
kind
of
realize
that
they're
they're
you
their
own
product,
like
the
transceiver
toolbox,
only
works
with
2021
era.
Stuff,
like
that,
has
a
limit.
It
doesn't
work
with
the
latest
and
greatest
of
of
pedal,
Linux
or
yeah
for
a
variety
of
reasons.
It's
a
since.
It's
targeted
for
math
works.
You
know
so
it's
their
own
product
that
now
won't
work
with
with
their
own
product
and.
A
Yeah
I
think
you
know
with
with
people
like
us
speaking
up
about
it
and
I
I
have
to
believe
that
they
have
profit.
You
know
for-profit
commercial
customers
that
are
going
like
okay,
yeah
yeah,
this
sort
of
insistence
of
only
using
the
absolute
latest
version
of
every
of
everything.
You
know
when
some
of
this
stuff
updates
twice
a
year
like
xilinx
updates
twice
a
year,
Matlab
updating
twice
a
year
with
if,
if
Analog
Devices
can't
keep
all
of
their
products
in
lockstep
ever
you
know
moving
forward
with
each
of
these.
A
Got
to
give
right
so
we
just
like
Paul,
said
I
think
we're
caught
in
a
transition,
an
ugly
transition
and
we
just
gotta
you
gotta,
stay
persistent
and
consistent
and
keep
asking
and
bringing
it
up.
And
you
know
eventually
this.
This
sort
of
the
persistence
will
pay
off
and
we
could
just
say
well,
let's
just
like
we
could
just
back
off
to
a
much
older
Hardware,
but
then
we'd
be
kind
of
giving
up.
C
F
F
A
Agree,
the
their
official
line
is
that
the
only
way
to
produce
these
profiles
is
with
the
tez
Tes
utility
software,
and
you
know
what,
if
it's,
if
you're
going
to
make
all
the
versions
available
that
are
all
compatible
with
all
the
different
varieties
of
of
Linux
and
hdls
and
la
la
you
know,
they're
fine.
But
if
you're,
if
you're
going
to
restrict
it
to
exactly
one
version
in
the
latest,
then
you're
just
asking
for
somebody
to
try
to
reverse
it.
The
trick.
A
A
A
That's
a
good
point,
because
the
binary
file
is
just
an
arm,
firmware
build
it's
just
off,
it
goes
to
the
arm
and
and
that's
a
it's,
a
it's
a
fairly
compact.
It's
a
small
file
and
we.
A
Might
just
well
yes
and,
and
we
we
know
what's
in
it,
because
we
know
what's
not
in
the
Json
file,
so
you
can
look
at
the
Json
file
and
then
compare
that
to
the
older
full
text
files
they
used
to
all
be
human
readable,
so
you
can
see
which,
which
things
have
been
moved
off
into
the
into
the
arm.
Firmware,
build
and-
and
you
know
it's
for
arm-
and
you
know
which
arm
it's
for
right.
A
So
this
makes
it
much
easier
to
kind
of
analyze
if
you
have
to
and
I
don't
know
that
that
actually
was
the
FAE
at
the
booth
at
IMs
eventually
said.
If
all
else
fails,
you're
just
going
to
have
to
hand
edit
the
profile
that
you
have
from
the
old
from
the
newest
tool
and
and
just
you're
just
gonna
have
to
do
some
detective
work
with
the
manual
and
it's
like.
Okay,
that's
that
takes
care
of
the
human
readable
profile.
It
doesn't
doesn't
necessarily.
C
A
Care
of
the
binary,
so
I
said,
is
the
binary
the
same,
because
is
it
just
you
know
is?
Is
it
the
same
as
as
it's
you
know
for
all
these
different
versions,
where
it's
been
a
split
and
he
said
he
thought
so
so
we'll
we
we
worst
case.
We
could
hand
edit
the
profile
somehow
and
I
just
I
just
would,
but
it
like.
A
A
You
know
that
we're
not
the
only
people
it
was
like.
Okay,
so
tell
me
you,
you
guys
tell
me
what
versions
of
all
the
the
five
different
Frameworks
that
have
to
work
together.
You
tell
me
what
versions
We
we
need
to
use
and,
and
they
were
they
were
like.
Oh,
oh,
no,
you're
right,
you
know,
all
of
this
stuff
is
correct:
you're
limited
in
the
in
the
version
of
of
Linux
that
you
have
because
of
our
own
transceiver
toolbox.
You
know
and
you're
limited,
because
xilinx
only
has
these.
You
know
these
versions
of
the
HD.
A
You
know
the
the
stuff
that
they're
pulling
into
their
HDL
reference
design,
there's
limitations
there.
So
it's
been
quite
the
adventure
and
I'm
really
looking
forward
to
solving
this
and
getting
the
stations
live
and
working
on
radio
problems,
because
we
have.
We
have
plowed
into
this
and
done
and
have
gotten
so
far
and
worked
so
hard
on
this.
That
I'm
pretty
sure
our
momentum
will
just
we'll
just
carry
us
through
into
actual
on
the
air
problems,
people
it.
F
Was
just
thinking
if
it
was
like
if
it
was
a
process
thing
where
you
had
to
you
know
the
app
loaded,
the
the
byte
stream
or
the
you
know
whatever
that?
Maybe
there
was
a
way
around
it,
but
it
sounds
yeah.
It
I
mean
it
sounds
like
you
know
what
what
the
problem
is,
and
it's
it's
just
a
supplier
problem,
I'm
just
trying
to
ask
questions
to
get
a
better
handle
on
it
to
maybe
find
a
solution.
So
yeah.
D
A
F
B
F
A
It's
been
quite
the
adventure
so
and
we
I
I
got
good
feedback
from
the
MathWorks
Booth,
like
so
I
showed
up
and
and
Robin
Goetz
recognized
me,
and
you
know,
because
we
worked
together
in
various
communities
to
to
put
on
events
and
to
get
things
done.
It
was.
It
was
very
happy
to
see
me,
which
is
good.
A
You
know,
because
I've
written
lots
and
lots
of
email
to
MathWorks
recently
about
problems,
and
he
pulled
me
aside,
and
he
told
me
that
the
experiences
that
we
had
were
directly
leading
to
to
revisions
in
their
in
their
product
that
well.
A
Yeah
I
didn't
ask
which
product
but
I
have
to
assume
that's
their
build
root.
You
know
and
I
was
like.
Oh
that's,
cool
and-
and
you
know
great,
you
know
I
think
that
that
was
good.
So
it
was
good
to.
It
was
good
to
hear
that
they
were
they'd
taken
a
kind
of
a
hard
look
and
because
of
our
experiences
trying
to
use
their
stuff
that
they
were,
they
were
going
to
put
some
some
time
into
making
it
truly
easier
to
use.
C
A
That's
yeah,
it
did
make
me
wonder:
I
mean
these
there's
products
like
the
the
xylex
chipsets
and
the
and
Analog
Devices
ICS.
This
is
like
the
main.
These
are
the
main
go-to
combinations
for
lots
and
lots
of
telecommunications
work.
Surely
we
can't
be
the
only
people
that
have
had
this
problem?
This
makes
me
think
like
what
are
what
is
everybody
else
doing?
What.
F
You
probably
are,
though,
right
like,
if
you
think
about
it,
most
people
aren't
building
new
designs.
Most
people
are
maintaining
existing
designs,
so
either
they're
not
changing
the
configurations
or
they've
got
tool.
Chains
set
up
from
2000
21
or
whatever
they've
got
VMS
running
or
you
know,
I
mean
that's
the
first
thing
back
in
the
day
when
you
could
actually
own
software
like,
if
you
did
a
project
for
a
client
or
at
least
I,
would
have
a
zip
file
with
the
installer.
F
For
whatever
thing
it
was
like,
you
know,
kai
CAD,
whichever
version
of
kai-cad
or
whichever
version
of
protel
or
whatever,
plus
all
the
files,
because
you
you
don't
want
to
be
stuck
looking
for
a
website.
That's
been
gone
for
10
years.
When
you
know
in
2030,
you
need
to
make
another
one
of
those
boards.
So.
F
Can't
do
that
now,
because
you
know
you
don't
own
any
of
your
software
anymore,
it's
all
in
the
cloud
and
licensed
and
yeah
so,
but
it
sounds
like
you
know
in
this
case.
Probably
everyone
who's
using
these
is
you
know,
people
who
are
using
or
who
are
working
on.
You
know,
finished
designs
that
are
in
production
or
whatever,
and
so
they've
got
their
build
systems
set
up
whenever
the
next
thing
comes
out
and
all
the
new
tools
for
that
come
out.
F
They'll
all
you
know
set
up,
build
environments
and
and
then
never
upgrade
those
that's
the
thing
that
kills
me
is
that,
like
you're
talking
about
a
Linux
system,
so
there's
a
driver
in
this
Linux
system,
and
if
you
update
the
driver
now,
which
I
would
assume
happens
when
you
update
the
underlying
OS
to
you
know
address
any
other
vulnerabilities
well
now,
suddenly
your
config
files
don't
work
anymore.
So
you've
broken
your
product
by
doing
what
everyone
wants
you
to
do,
which
is
keep
your
stuff
up
to
date,
exactly.
A
D
C
A
D
F
Or,
like
all,
those
lines
are
now
making
cheap
Farms
or
something.
E
D
F
A
A
A
A
A
Yes
and
like
would
not
phased
out
in
a
gentle
way,
yeah
like
the
plug
is
pulled.
B
A
It
was
never
there.
This
happened
with
smart
cities,
the
the
particular
embedded
processor
that
was
used
for
the
smart
city,
there's
a
little
like
little
surveillance,
pods
that
are
on
street
lights.
C
A
A
It's
like
it.
Yes
and
it's
you
know
they're
meshed.
They
they
speak
six
low
pan.
You
know
they
have
a
back
haul
over
machine
to
machine
iot
and
everything.
So
this
is
a
great
it
actually.
A
It
felt
exactly
it
fills
a
hole
in
terms
of
the
like
iot
Market.
Smart
cities
is
one
of
the
largest
fractions
in
terms
of
like
Hardware
shift
and
people
working
on
it
and
so
San
Diego
bought
in
big,
and
we
went
to
the
the
hack
Fest
that
they
threw
to
try
to
get
people
excited
about
right.
It
wanted
to
have
their
own
App
Store
for
apps
for
Smart
City.
A
And
the
the
Baseline
you
know
the
brains.
The
operation
was
a
and
blanking
on
the
name
of
the
processors
embedded
processor
for
Intel,
named
after
an
inventor
like
the
Edison
or
something
like.
F
That,
oh,
the
the
yes,
the
Arduino
thing
that
they
did,
that
they
partnered
with
Arduino.
It
was
like
an
fpga
weird.
It
was
like
a
weird
I-32
thing,
I
think
like
there's
some,
it
was
some
re
recycled
thing
that
they
were
trying
to
sell
as.
A
A
Vaguely
remember
it's
like
an
inventor
sign
or
something
like
that:
yeah,
it
was
like
a
name,
a
famous
inventor's
name.
It
was
also.
This
part
was
also
used
in
one
of
the
Defcon
badges
I.
Think
the
skull,
the
cool
Skull
Terminator
badge
used
it
and
we
set
up
the
development
environment
to
take
a
look
at
it.
So
that
was
that
was
the
part
that
was
in
the
Smart
City
pods,
and
so
we
show
up
at
this
hack
Fest,
the
very
first
part
of
it.
A
They
do
the
food,
the
drink,
the
the
big
you
know
inspiring
speech
and
then
there's
a
couple
of
tables
set
up
and
one
of
them
was
the
Intel
people
were
there
to
kind
of
like
with
the
brochures,
and
this
is
what
the
part
that's
in
it,
and
here
is
an
actual
street
light
sitting
here.
You
know,
exposed
and
everything
that
day
Intel
had
announced
this
part's.
Just
like
perfect.
A
It
must
be
a
strange
experience
standing
here
talking
about
all
this
future
work
for
a
part
that
has
been
canceled
and
he's
shoulders
sagged.
He
took
a
deep
breath
and
he
was
like
yeah
it's
a
weird
day
because
they
said
they're
pretty.
He
was
pretty
sure
that
he
would
not
be
coming
back
to
a
job
that
he
would.
F
A
Why
would
anyone
buy
one?
Why
would
anybody
buy
one
yeah?
It
did
say
he
had
actually
been
on
a
canceled
project
before
at
Intel
and
they
they
do
kind
of
catch
their
own
people.
So
he's
hoping
like
like
a
catch
and
release
process
that
you're
thrown
out
of
work
and
then
they
Court
you
back
and
I,
wished
them
very
well.
But
like
everybody
there,
when
the
word
kind
of
spread
around
you
know
people
are
looking
on
their
phones.
Saying
look.
This
park
got
canceled,
you
know
and
are
you
going
to
then
have
a
new
part?
A
Is
this
the
Smart
City
people
were
totally
caught
off
guard
they're
like
what
do
you
mean
it's
canceled
and
they're
like
well?
This
is
a
two-year-old
design,
so
you
know
it's
infrastructure,
they're,
not
gonna.
It's
not
updated!
There's
no
plans
to
change.
You
know
that
Hardware
team's
long
gone
so
yeah.
This
is
what
is
in
all
of
these
smart
City
pods,
but
finding
people
to
write
code
for
it
suddenly
became
a
whole
heck
of
a
lot
harder.
A
A
A
Get
paid,
you
know,
someone
will
pay
them
to
go
fix
the
bank
Mainframe.
You
know
it
was
so
it
was
a
very
interesting
hack,
Fest
and
and
the
timing
could
not
have
been
more
wacky.
You
know
that
was
literally
that
day
earlier.
F
A
A
It
was
I,
don't
know
market
share,
maybe
or
should
you
know
not
not
enough
momentum
or
whatever
they
call
it
when
you
scale
up
like
if
it's
like
I.
F
So
I
worked
at
Salesforce
until
January
I've
been
there
for
13
years
and
I
got
I'm
very
pleased
to
have
been
laid
off
as
it
turns
out,
but
I
saw
it
there.
All
the
time
like
we
would
get.
F
You
know
60
of
the
way
into
a
project,
and
you
know
a
big
project
like
you
know,
automating
capacity
ads
right
things
that
took
a
team
of
you
know
60
people
four
months
to
do
we
wanted
to
get
down
to
you
know
we're
we're
just
shipping
Hardware
like
white
label
Hardware
in
and
you
know,
standing
up
new
capacity
daily
and
we
we
were
getting
there
right.
We
had
Solutions
two
different
times
and
two
different
times.
F
They
canceled
our
projects
and
restarted
using
entirely
new
technology
Stacks
with
entirely
new
architecture
teams
and
as
far
as
I
know,
like
the
the
solution
ended
up
being.
Let's
just
move
everything
to
AWS
wow.
A
F
F
You
know
it
was.
You
know,
10
database
servers
that
were
8U
each.
It
was
you
know,
60
apps
servers.
It
was
you
know
a
network
Stacks
I
mean
it
was
a
ton
of
stuff
that
had
to
all
be
set
up.
You
had
to
to
check
all
the
hardware.
F
You
had
the
inventory
to
make
sure
all
the
peripherals
that
were
supposed
to
be
in
each
device
were
there
for,
like
the
file
servers,
you
had
to
make
sure
that
the
the
wwn
on
the
on
the
transceivers
was
all
you
know,
matched
what
you
expected.
F
You
know
you
had
to
make
sure
that
all
memory
worked
like
you
had
to
you
had
to
test
every
piece
of
hardware
and
inventory
every
piece
of
Hardware.
Then
you
had
to
Cable
it
according
to
a
patch
plan
for
that
particular
piece
of
infrastructure,
so
everything
had
to
get
cabled
to
the
network
stack
for
that
pod
and
then
that
pods
Network
stack
had
to
get
cabled
into
the
data
center.
F
F
You
know
proxy
skus,
like
some
some
number
of
configurations
for
Hardware
that
covered
all
of
our
needs,
and
then
we
would
just
predict
basically
how
much
Hardware
we're
going
to
need
over
the
next
year
and
we
just
start
placing
orders
and
as
they
come
in,
they
get
racked
and
cabled,
and
what
we
wrote
was
I
I
repurposed,
the
red
hat
CD
installer
image
to
instead
of
kicking
off
an
install,
it
would
go
through
an
inventory
all
the
PCI
devices.
F
It
would
start
up
lldp
and
watch
the
network
and
figure
out
what
switch
it
was
talking
to
and
basically
inventory
itself
and
its
connections
and
then
connect
to
a
reporting
server
and
tell
the
reporting
server
hi
here.
I
am
this.
Is
you
know
this
is
everything
you
need
to
know
about
me
and
then,
when
we
would
add
Hardware
or
ad
capacity,
we
would
have
a
a
bomb,
a
bill
of
materials
for
like
we
need
60,
app
servers
and
10
database
servers
and
whatever,
and
it
could
just
go
and
pull
all
that
inventory.
F
Out
of
you
know
all
these
things
have
checked
in
it
could
assign
vlans
to
the
ports
that
they
were
on
to
make
sure
everything
was
networked
properly
and
so,
basically,
as
long
as
you're
racking
Hardware
as
fast
as
you
can
order
it
and
get
it
shipped
in,
we
could
keep
up
with
demand
and
we
were
getting
there
like.
We
had
the
inventory
stuff
written.
The
reporting
server
was
almost
done,
I
was
I,
I
worked
on
the
data
collection
reporting
side
of
it
and
we
were.
We
were.
F
You
know
60
percent
of
the
way
there
maybe
70
of
the
way
there
and
they
puppet
the
team
that
did
puppet
came
out
with
something
I
can't
remember
what
it
was
called
now,
but
it
was
basically
the
same
idea
right.
F
It
was
a
network
boot
image
that,
on
you
put
on
your
default
VLAN
and
when
something
booted
up
it
would
inventory
itself
and
check
itself
into
a
a
repository
and
you
could
use
their
language
to
configure
sort
of
what
hosts
you
needed,
but
it
didn't
do
everything,
and
so
they
were
going
to
have
to
go
through
and
anyway
they
canceled
our
project.
They
started
this
one
and
at
about
the
same
point,
three
months
later,
they
canceled
that
one
and
started
something
else
and
yeah
I
mean
the
problem,
never
got
solved.
Wow.
E
F
Was
yeah
this
was,
you
know,
2012
2013.,
so
you
know.
Eventually
we
moved
a
lot
of
stuff
to
the
cloud
and
and
that
solved
a
lot
of
the
problem.
But
you
know
it's
no
different
really
than
Intel.
You
know
it's
like
we
we're
not
seeing
results
fast
enough.
So,
rather
than
like
it's
it's
a
it's.
A
sunk
cost
fallacy
fallacy,
oh
right
right,
like
they're
they're
like
well,
we
don't
want
to
throw
good
money
after
bad.
Well,
it's
like
you,
haven't
given
the
good
money
time
to
work
right
right
like
it's.
F
You
you
haven't
gotten
to
the
to
the
finish
line
yet
and
there's
just
this
fear
of
like
oh
well.
If
we
don't
stop
and
restart
now,
we'll
never
get
done.
Yeah.
A
B
A
F
C
A
C
A
Yeah
and
then,
like
a
lot
of
us,
pay
attention
to
the
distance
covered,
you
know
so
the,
but
that
second
derivative
is
what
is
what
people,
maybe
in
in
the
Executive
Suite
or
or
that's
the
only
thing
that
they're
looking
at
and
the
distance
cover
doesn't
matter.
If
the
second
you
have,
the
Velocity,
isn't
and
then
I'm
sure,
there's
there's
even
other
ways
to
to
slice
and
dice
it.
You
know
I.
F
F
A
C
F
A
I
think
we
do
Ken
did
you
have
you
have
any
questions
or
comments
for
today?
A
Yeah
welcome
to
the
welcome
to
the
Meetup
to
Ken,
he's
a
very
experienced
Asic
and
a
designer
and
interested
in
getting
more
involved
in
fpga
work,
so
he's
going
to
be
working
on
all
sorts
of
HDL
and
you
know
all
the
things
that
we've
been
talking
about
with
the
embedded,
Linux
and
and
helping
us
out
with
maybe
some
of
the
harder
harder
logic
and
math,
but
yeah
Ken.
If
you
had
anything
that
you
wanted
to
nope.
A
F
Thanks
so
it's
it's
tangently
related
to
ori's,
Mission
I.
Think
I
mostly
wanted
to
bring
it
up,
because
I
thought,
if
it
wasn't
something
or
I,
was
interested
in
directly.
Maybe
you'd
have
some
some
thoughts
as
to
where
I
could
take
it.
But
I
have
a
friend
who,
a
few
years
ago
started
a
private
internship
thing.
F
He
found
someone
who
was
interested
in
learning
how
to
how
to
code
he
hired
them
as
an
intern,
taught
them
the
basics
of
what
they
needed
to
know
and
then
contracted
with
them.
Once
they
were.
You
know
skilled
up
and
had
them
work
on
side
projects
with
him
and
I
thought.
The
idea
was
fascinating,
like
just
the
idea
of
like
you
know,
lots
of
people
have
side
gigs
or
they
have.
F
You
know
hobby
projects
that
they
want
to
turn
into
a
real
thing,
or
you
know
lots
of
places
where
having
an
extra
person
would
be
useful,
and
you
don't
necessarily
need
someone
who
knows
all
all
the
things
already,
especially
if
it's
like
a
hobby
project
where
your
schedule's,
flexible
or
whatever
so
my
thought
was.
It
would
be
a
really
cool
thing
if
there
was
a
an
organization
that
could
help
put
people
on
a
path.
F
So,
let's
say
you're,
you
know,
you've
worked
in
construction
or
in
you
know
some
manual
field,
and
you
want
to
write
software
so
we'll
point
you
in
a
direction
of
like
free
code,
Camp.
Well,
first,
we'll
start
with
like
what
are
you
interested
in?
What
do
you
want
to
do?
Do
you
want
to
make
web
pages?
Do
you
want
to?
F
You
know,
build
electronics
that
have
software
on
them
and
once
we
figured
out
what
your
interests
are,
we
point
you
at
self-paced
resources
for
getting
started
and
for
evaluating
your
progress
and
then
once
you're
to
a
point
where
having
one-on-one
interactions
would
be
useful.
We
pair
you
with
someone
who
needs
an
intern.
Basically
it
could
be
an
organization
like
Ori,
where
you
know
during
the
summer
trying
to
you
know,
put
in
some
work
on
one
of
the
fpga
projects.
F
You
take
an
intern
or
it
could
be
like
my
friend
Andrew,
who
you
know,
wanted
to
do
a
good
thing
and
had
some
development
projects
that
he
was
going
to
need.
Help
on
so
it's
kind
of
a
skill
up,
plus
matchmaking
HR,
and
also
providing
training
for
mentors
right
so
for
people
who
want
to
do
this,
who
want
to
have
an
intern
or
who
want
to
provide
mentorship.
F
Even
if
it's
unpaid
right
like
if
it's
just
mentorship,
not
it's
not
interning,
you
know
how
to
Be
an
Effective
Mentor,
how
to
be
how
to
effectively
manage,
and
in
turn,
that
kind
of
stuff.
So
both
sides
of
the
equation,
helping
people
make
career
transitions,
maybe
go
and
Bug
the
companies
that
are,
you
know
turning
profits
by
laying
off
people
and
get
them
to
help
people
find
new
jobs.
F
I
think
there's
a
lot
of
opportunity
there
to
do
some
really
cool
stuff
and
I
think
there's
probably
a
lot
of
people
that
would
have
an
interest
in
taking
advantage
of
this
from
one
side
or
the
other.
So
it's
just
the
bare
bones.
You
know
skeleton
of
the
idea
at
the
moment,
but
I
don't
know
what
to
do
with
it
from
here.
So.
A
I
think
this
is
sounds
fantastic
and
we
have
done
chunks
of
this
or
swaths
of
this
pretty
much
across
the
board.
We
we
do
take
on
people
for
professional
development
and
and
in
a
lot
of
cases,
the
various
projects
that
we
have.
That's
that's
the
motivating
sort
of
driver.
A
You
know
from
everything
from
RF
bit
Banger
to
rub
it
to
hyperia,
to
and
so
on.
What
you've
done
is
you've
you've
thought
about
this
and
and
with
a
with
some
very
observant
skill
here,
you've
you've
laid
out
like
a
a
more
formal,
specific
or
explicit
path.
You
know
to
to
kind
of
meet
people
where
they
are
and
take
an
inventory
and
then
provide
advice
on,
like
those
the
you
know,
there's
a
whole
lot
of
self-paced
and
free
things
out
there
and
I
can
give
it.
A
I
can
give
two
examples:
there's
an
excellent
self-guided
tutorial
from
gnu
radio
for
getting
into
SDR
very
quickly.
There's
also
some
some
excellent
free
courses,
many
many
hours
worth
of
courses
from
from
MathWorks
to
to
get
good
at
dealing
with
with
Matlab
to
use
Matlab
and
simulink.
A
A
So
those
are
the
things
that
we
routinely
recommend
to
people
if
they
ask
like
well
how
so,
how
do
I
learn
and
preferably
for
free,
you
know,
there's
a
lot
of
free
resources,
but
then
the
third
part
was
to
like
to
pair
up
to
to
make
a
match
to
to
recommend,
and
we
do
have
some
success
here.
We've
we've
tried
to
apply
for
Google
summary
of
code
to
as
a
project.
It's
very
competitive.
We've
never
been
able
to
get
approved
as
a
Google
summer
of
code
project.
A
Funding
for
for
their
for
that,
which
is
usually
not
a
whole
lot
since
it's
it's,
it's
all
saw
it's
supposed
to
all
be
software,
but.
C
A
They
do
provide
some
some
amount
of
funding
for
a
Google
summer
code,
but
we
on
the
flip
side,
we've
also
directed
students
to
apply
as
a
Google
summer
of
code
student
volunteer
and
then
supported
them,
and
you
know
kind
of
given
them
a
little
shove
to.
C
A
Involved
with
Google
some
of
our
code
there's
a
lot
of
of
potential
interaction
that
we
it's
a
lot
of
interaction
that
we
had
in
the
past
with
IEEE,
where
finding
people
some
sort
of
volunteer
role
or
matchmaking,
essentially
for
for
IEEE.
A
A
A
Most
part,
actually
like
the
meetings
and
and
section
level
stuff,
and
you
know
we-
we
have
a
lot
of
good
interactions
with
with
some
of
the
projects.
F
That
all
sounds
really
cool
like
and
I'm
I'm,
actually
I'm
really
interested
in
sort
of
the
experience
of
working
with
other
organizations
to
manage
those
relationships
or
to
you
know
to
find
those
opportunities,
the
the
sort
of
differentiator
that
I'm
I'm
thinking
about
like
when
you
say
intern,
you
think
of
college.
You
know
it's
college
kid
I'm,
45
years
old,
you
think
you
know
a
young
adult
who's
graduating
from
college
soon
and
I.
You
know
there
is
infrastructure
in
place
for
that
there.
You
know
companies
do
that,
all
the
time
yeah.
F
Is
which
is
cool,
and
it's
it's
good
like
having
actual
industry
experience
I
think
is,
is
crucial
because
I've
I've
worked
with
you
know:
college
grads
who've
never
been
in
industry
and
it's
it
can
be
frustrating
for
everyone
involved,
but
I.
So
what
Andrew
did
is
he
found
someone
who
was
not
a
student
who
was
like
I
I,
don't
know
what
their
background
was,
but
it
was
not
in
Tech
at
all
right.
C
A
The
vast
majority
of
the
people
that
that
are,
that
we
I
guess
you
would
consider
them
to
be
sort
of
interns
or
technical
volunteers,
is
that
you
know
when
they,
when
they
come
to
us
and
want
to
work
on
a
project.
The
vast
majority
are:
are
not
students
they're,
not
right
out
of
college,
so
they're,
not
the
the
traditional
sort.
F
F
A
F
Yeah
that
somehow
high
school
education
and
I've
been
and
I've
managed
to
work
in
this
industry
for
25
years
right,
it's,
but
it's
hard.
It's
it's
and
it's
a
lot
harder
today
than
it
was
when
I
started
right,
like
I.
Couldn't
do
it
today,
I,
don't
think
I
was
fortunate
enough
to
know
people
to
get
my
first
couple
of
jobs
and
then
from
there
I
could
leverage
my
past
experience.
But,
like
you
know,
someone
who's
been.
F
You
know
on
auto
mechanic,
since
they
were
18
and
now
they're
they're
30,
and
they
want
to
make
a
career
change.
Well,
they
can
either.
You
know,
go
spend
forty
thousand
dollars
on
a
boot
camp
that
may
or
may
not
get
them
a
job
and
isn't
really
going
to
prepare
them
very
well,
for
you
know
anything
in
particular,
or
they
can
try
and
find
a
open
source
project
to
contribute
to,
but
that
because
it's
open
source
because
it's
all
volunteer
it
can
be
very
tricky.
F
F
A
F
It
attracts
it
attracts
a
couple
of
very
specific
personality
types,
and
you
know
that
can
work
really
well
or
it
can
work
really
yeah.
It
can
be.
It
can
be
challenging,
but
yeah
I'd,
love
to
see
and
I'd
love
to
be
involved
in
whatever
way.
I'd.
B
F
To
see
something
that
tackles
this
problem
that
you
know,
people
who
you
know
are
like
my
daughter
is
autistic,
and
you
know
she
has
an
interest
in
technology,
but
like
she's,
never
gonna
go
to
college
right
like
she's,
not
that
that
environment
is
gonna
would
destroy
her
and
you
know
so.
I'm
teaching
her
the
stuff
I
can
teach
her
and
and
she's
going
to
be.
She'll
be
okay,
but
there
are
you
know
how
many
kids
on
the
Spectrum,
whose
dad
isn't
a
software
developer,
who
doesn't
have
those
resources?
F
You
know
what
do
they
do,
so
you
know
again
not
just
that,
but
everybody
who's.
You
know
trying
to
trying
to
do
something
better
than
than
working
at
McDonald's
or
or
you
know,
or
is
just
not
happy
doing
what
they're
doing
and
wants
to
try
something
else.
Like
I'd
love
to
see
people
have
those
opportunities.
I've
long
held
a
belief
that
there's
nothing
special
about
software
right
like
it's
just
problem
solving
and
most
people.
F
If
they
sit
down
and
work
at
it,
they
can
solve
problems,
and
the
rest
of
it
is
just
knowing
how
to
explain
the
solution
to
the
computer
right.
If
you
break
it
down
into
those
terms,
I
think
anybody
can
do
it.
It's
just
more.
People
need
the
opportunity,
so
I
think
this
is
a
way
to
do
it.
I
just
don't
know
how
to
how
to
make
it
happen.
Like
I've,
I've,
never
I've
been
involved
in
non-profits.
F
As
you
know,
a
member
I
was
actually
president
of
a
non-profit,
which
meant
I
ran
two
meetings
once
a
year
once
a
year
for
two
years,
and
that
was
my
experience
there.
I
just
don't
know
where
to
go
from
here.
So
I
I
was
hoping
that
you
know
if
it's
something
that
Ori
that
fits
the
Ori
Mission
that'd
be
awesome
to
see
it
happen
here.
F
If
it's
not,
then
maybe
you
know
where
to
go
to
give
it
a
chance,
and
you
know
if
I
can
be
a
part
of
of
seeing
it
move
forward.
That'd
be
awesome,
but
really
I'd
just
like
to
see
it
happen.
Okay,.
A
I
think
from
I
think
you're
hired
so
yeah,
let's,
let's
try
to.
Let's
try
to
make
this
happen,
because
this
is
all
of
these
ingredients
are
things
that
were
already
at
pretty
much
they're
in
the
water.
This
is,
this
is
just
how
we
we
do
and
the
I
could
not
agree
any
more
strongly
with
your
the
inclusive
and
positive
attitude
that
you've
very
eloquently
expressed
here
today.
It
it
is.
That
is
absolutely
the
truth,
and
it's
why
we
tell
people
that
you
do
not
have
to
be
an
expert
to
join.
A
A
Hard
to
sell,
we
have
had
some
very
negative
reactions
to
this
sort
of
approach
we
have,
and
we
know
that
those
organizations
and
Foundations
are
not
our
friends
and
that's
okay,
we'll
just.
C
A
A
favor
by
revealing
you
know
their
their.
C
A
Know
the
overall
I
think
that
we
have
I
have
it's
just.
This
is
my
observation.
Over
over
the
years,
we
have
narrowed.
The
definition
of
you
know:
we've
made
gatekeeping
and
and
narrowing
the
definition
of
it
like
what
is
an
acceptable
student.
What
is
what
say,
what
is
a
successful
worker?
What's
a
successful
person,
what
what
sort
of
it
we've
made
it
much
narrower
we're
much
less
willing
to
accept
the
broader
yeah.
A
Think
so
I
was
I
was
part
of
the
last
class
as
an
undergraduate
I
got
a
degree
in
something
called
engineering
technology.
So
it's
not
a
double
e
degree.
It's
very
related
and,
and
the
theoretical
work
sure
felt
just
as
hard.
The
difference
was
that
all
of
our
classes
required
a
lab
and
in
a
traditional
sort
of
double
e
degree,
you
might
get
some
Labs,
you
know,
and
so
I'd.
A
It's
it
I
at
the
time
it
was
a
compromise
and
I
thought
that
it
was
it's
considered
a
lesser
degree
bye
by
by
most
The
Double.
A
And
there's
more
lab
the,
so
the
engineering
Tech
degree
is
not
as
popular
and
in
my
class,
but
graduating
class
was
the
last
one
that
was
allowed
to
sit
for
the
engineer
for
the
professional
engineer
process.
So
so
you
get
a
you
get
an
EIT
which
is
the
engineering
training
certificate
and
then
you,
after
a
couple
of
years
of
of
working
as
a
as
an
engineer,
you're
allowed
to
sit
for
the
professional
engineer,
exam.
A
And
that's
how
it
used
to
be.
It
was
if
you
could
pass
the
exam.
You
got
to
call
yourself
of
a
professional
engineer
and
it
changed
from
now.
You
had
to
have
a
college
degree,
which
is
an
enormous
step
forward
in
gatekeeping,
and
then
my
my
year
was
the
last
year
allowed,
and
then
it
was
only
double
E's
or
Emmys.
B
B
A
B
A
Still
keep
my
my
engineering,
my
engineering
intern
certificate
I
still
keep
it
current
and
at
some
point,
I've
I
ever
so
often,
I
get
a
attempted
to
go
through
the
process
of
of
like
sitting
for
the
PE
exam,
get
it
filling
out
all
the
paperwork
to
get.
F
A
You
it's
just
it's
just
it's
a
it's
just
kind
of
a
pain
to
to
do,
but
yeah
it
was
I
was
like.
Why
are
they
they're
they're,
basically
saying
my
entire
this
entire
degree
program,
my
entire
University
is
garbage
and.
A
Never
fun
and
yeah
I
think
that
we
have
this.
Trajectory
has
continued
over
the
years
and
and
we're
the
right
repercussions
seem
pretty
harsh
to
me,
but
I
guess
gatekeeping
is
profitable
for
somebody.
You
know
it's
making
sure
that
the
credentials
are
are
rare,
means
that
that
you
know
the
people
that
have
them
salaries.
A
Think
that
that
all
you
know
once
you
once
you
make
it
lucrative
for
someone
to
to
be
kind
of
part
of
a
system
that
does
the
gatekeeping
there's
there's
a
lot
of
that
at
all
different
levels,
and
you
know
sure
some
some
amount
of
of
enforcing
repercussions
and
gatekeeping
is
necessary
in
order
to
to
have
things
be
basically
safe,
but
it's
it's
like
anything
else
as
a
balance
and
there's
a
lot
of
stuff
in
Tech,
that's
out
of
whack
right
now,
with
the
you
know,
with
the
way
that
we
do
things
and-
and
it's
you
know
this,
this
process
that
you
that.
A
A
C
A
Yeah
it's
very
similar,
except
that
with
Engineers
General
we
were
gonna.
We
offered
to
employ
people
like
to
for
contractors,
but
this
sounds
like
it
would
be
less
work
really.
Well,
it's
still
it's
the
good
work.
You
know
to
to
support
and
build
and
and
to
get
people
you
know
involved
and
engaged
and
and
moving
forward,
but
then
we
wouldn't
we
wouldn't
be
the
eventual
Target
for
the
for
the
contractor
having
to
manage
it.
D
D
A
Hard
they're
they're
very
difficult.
You
know,
like
my
degree
program,
that
I
that
I
come
from
and
Keith
Keith
wheeler
came
from.
We
we
graduated
I,
think
almost
the
same
year
and
all
my
a
lot
of
other
friends
that
are
that
work
like
Stephen
O'connor.
He
works
for
a
for
a
big
contractor
and
in
in
Arkansas.
A
All
of
us
are
highly
successful.
Technical
people
do
engineering
work,
and
it
wasn't
very
many
years
after
that
that
my
undergraduate
institution
was
kind
of
cut
off
from
from
letting
people
go
forward
with
the
EIT
PE
thing
that
they
completely
changed
the
entire
technical
program,
so
that
that
engineering
technology
degree
was
sunseted
and
they
they
fired
up
a
systems
engineering,
what
they
called
a
systems
engineering
degree
which
was
a
four-year
degree
in
I.T.
So
it
was
computer
systems
and
computer
Administration.
A
D
F
I
wouldn't
think
so,
but
like
if
they
it's
crazy,
that
they
replaced
your
your
track
with
with
that,
that's
just
ridiculous,
so
someone
was
showed
up
looking
for
Mark
thinking,
they
were
supposed
to
be
at
my
house
to
cut
out
my
air
conditioner
and
he,
oh
dear,
had
planned
to
just
go
back
and
start
working
on
it.
Oh
God,
but
it's
running
and
so
he'd
come
to
like
just
be
like
hey.
Can
you
shut
off
the
air
conditioner,
so
I
can
cut
it
out,
I'm
like
no?
F
No,
no
I
can't
wow,
so
that
was
yeah.
There
was
there
were
some
phone
calls
we
figured
it
out.
He
was
I'm
at
103,
North
and
I
thought
he
wanted
103
South.
It
turns
out.
He
wanted
108
North,
so
that
was
almost
but
yeah.
He
was
also
looking
for
someone
named
Mark
with
you
know.
It's
just
crazy
coincidence.
Anyway.
A
A
Yeah
we
were
yeah,
we
were.
We
were
been
talking
about
this,
this
very
interesting
yeah.
This
is
foolishness
yeah,
it's
tackling,
like
you
know,
Paul
says
it's
tackling
a
pretty
severe.
You
know
cultural
social.
You
know
problem,
but
you
know
I
think
that
we
we
have
already
made
progress
in
this
area
and
with
a
with
a
sort
of
a
a
I
hesitate
to
say
formalization,
you
know,
because
that
tends
to
scare
people
off.
A
You
know
we
just
we
like
getting
things
done
and
we
do
like
documenting
so
I
think
doing
some
documentation
here
and
kind
of
it
helps
crystallize
the
the
things
that
we
already
are
doing
in
a
very
decentralized
like
per
project
manner.
But
but
this
is,
this
largely
is
our
mission,
and
this
is
what
we
we
feel
strongly
about
doing.
You
know
this
is
cool
the
sort
of
the
vocational
work
that
we
that
we
want
to
to
do.
We
want
to
engage
people,
we
really
want
to
be
accessible
and
you.
F
Know
so
welcome
my
my
sort
of
like
initial
thought
of
where
this
might
go
was
kind
of
like
a
if
it
was
a
standalone
thing,
like
a
non-profit
kind
of
you
know
both
for
people
who
need
help
and
want
to
provide
mentorship
and
all,
and
also
for
people
who
are
looking
for
mentorship.
If
it
did
it
within
Ori
I
think
it
would
be
interesting
from
a
process
perspective
to
like
look
at
sort
of
what
what
makes
the
most
sense
from
those
perspectives.
F
F
It
grew
out
of
a
project
at
GitHub
and
it
was
learning
languages
and
they
had
like
40
different
languages,
and
they
had
exercises
along
with
lessons
so
that
that
gave
you
kind
of
the
the
self-evaluation
checking
your
progress,
kind
of
thing,
maybe
Partnerships
with
lightly
code
or
one
of
the
the
coding
test
providers
and
coding
challenge
companies
would
be
useful
and
again,
my
expertise
is
in
the
coding
in
the
software
side
of
things.
F
Another
thing
that
I
think
would
be
kind
of
neat
for
people
who
have
a
background
in
software,
doing
transitions
to
either
embedded
systems
or
into
HDL.
F
C
F
Like
I
get
that
never
goes
away
actually,
but
I
yeah
I've
been
writing
software
for
too
long
to
write
Hardware
without
without
some
major
effort.
It
is
still
something
I
want
to
do,
and
I
I
like
building
Electronics,
I
love,
doing
and
I
can
I
understand
that
but
like
having
having
sort
of
not
necessarily
like
predefined
paths
but,
like
you
know,
if
you're
looking
to
move
from
this
type
of
work
to
this
type
of
work,
these
are
the
sorts
of
things
you
might
try.
F
You
know
I
think
there's
a
lot
of
opportunities
there
within
sort
of
the
areas
of
expertise
of
Ori,
specifically
but
yeah.
If,
if
we
trial
it
here
and
it
turns
out
to
work
I'd
love
to
see
it
become
a
bigger
thing.
Okay,.
A
A
Accessible
yeah:
let's,
let's
see
what
we
can
do,
then
I
think
this
is
great.
F
Yeah,
let
me
let
me
know
how
I
you
know
how
I
can
help
like
I'm
I
will
write
up
the
ideas
that
I've
got
and
send
them
to
you.
Okay,.
A
Yeah,
let's
do
that,
let's
collaborate
on
some
writing
and
get
like,
maybe
I
I,
don't
know
sort
of
white
paper-ish.
You
know,
but
at
least
like
a
an
essay
or
yeah
capture
capture
these
things
in
a
in
a
in
a
document
and
yeah.
If
you
want
to
kick
it
off
as
like
a
shared
Google
doc
or
anything
that.
F
At
least
yeah
they
so
so,
while
at
Salesforce,
they
I
don't
know
if
they
bought
or
if
they
developed
in-house
a
suite
called
quip
q-u-ip,
and
it
was
basically
Google
docsplit
at
Salesforce
and
they
they
made
us
use
them,
and
it
was
terrible.
So.
A
F
C
F
Like
I
said,
I'm
happy
and
happy
that
I
got
laid
off
I'm
taking
my
Severance
and
the
savings
that
I
had
and
I
am
trying
to
change.
Industries
I'm
I'm,
actually
trying
to
move
into
making
custom
Woods
woodworking
stuff.
F
But
also
moving
into
making,
like
you
know,
small
furniture
and
like
keepsake
boxes
and
that
kind
of
stuff
and
and
doing
software
for
fun,
so
yeah.
F
But
yeah
so
I,
I'm
I
will
write
that
up
and
we'll
go
from
there.
Thank
you
very
much
for
for
taking
the
time
to
listen
and
for.
A
This,
oh
sure,
no,
this
is
I'm
actually
inspired
and
heartened
I,
think
you've.
You
have
expressed
it
so
very
well
and
really
clarified
the
this
process.
You
know
this
sort
of
the
tendency
or
sort
of
vocational
work
that
we
do
and
made
it
clear
and
understandable
and
and
it's
it
really
does
fit
with
with
how
we
do
business
so
I.
F
Who's
the
inspiration
for
this
like
I
I,
just
like
was
thinking
of
it
as
like
private
internships
or
or
you
know,
private
mentorships,
but
I
don't
have
I,
don't
have
anything
catchy
yet.
Okay,.
A
Y'all
I'll
I'll
assign
some
some.
A
A
F
Yeah
I'll
think
about
it
as
well.
I,
since
leaving
my
job
I've
been
filling
notebooks
with
ideas
and
names
for
things
is
probably
the
worst
part
yeah.
A
A
Okie
dokie
well,
I
will
be
on
on
slack
and
I'll,
be
in
town
here
in
San
Diego
through
the
27th
of
June
and
I'll,
be
traveling
from
the
27th
to
the
2nd
of
July.
A
I
will
be
in
northeastern
Mississippi
working
my
day
job
and
should
be
able
to
to
stay
online
for
that.
But
but
just
let
you
know
we'll
be
out
of
town
cool.
Okay,
see
you
guys
on
slack,
so
you
see
everybody
on
slack
and
thanks
so
much
for
the
really
wonderful
Meetup
today,
looking
forward
to
to
a
meeting
back
up
again
next
week
and
I'll
also
try
to
to
put
together
another
meet
up
later
in
the
week
for
for
the
evening
hours
and
on
the
on
the
west
coast.
A
So
we
can
catch
some
other
folks
that
couldn't
be
here
today,
all
right!
So
thanks
everybody
see.