►
From YouTube: NuPIC Sprint Planning - Jan 3 2014
Description
For Sprint 13: https://github.com/numenta/nupic/issues?labels=ready&milestone=7&page=1&state=open
A
A
Hopefully
I
can
fix
it,
but
when
you
create
an
event,
you
have
an
option
to
make
it
an
on-air
event,
and
but
once
you
do
that,
there's
no
way
to
create
that
on-air
like
that,
hang
out
from
the
event
you
have
to
create
a
new
event.
It
doesn't
make
sense,
but
anyway,
I'll
give
a
couple
minutes
see
if
scott
joins
subtitle.
A
B
B
To
go
to
the
event
on
the
original
page,
hopefully
that
works.
B
A
Spring
planning,
oh
great,
but
under
details
on
the
right
hand,
side
there's
a
link
says
for
sprint:
13.
nope,
that's
not
it
dude.
I
don't
remember
how
I
got
in
at
some
point.
I
saw
a
button
somewhere
that
said
join
and
I
hit
it
got
in,
but
it
was
after
clicking
around
for
a
little
bit.
B
A
Okay,
I'm
thinking
I'm
I'm
going
to
kill
this.
A
This
hangout
austin
got
back
in
all
right,
so
the
link
he
sent
out
doesn't
work.
It
just
gives
you
the
opportunity
to
do
a
q
a
so
you
can
watch,
but
you
can't
actually
join
in,
but
there
is
a
link
that
will
let
you
join
so
what
I
ended
up
doing
is
I
went
to
the
main
numenta
or
google
plus
page
at
the
top.
A
There
is
a
a
link
and
it
has
an
option
to
join
the.
A
A
There's
some
way
I
can
show
my
screen,
I
can
yeah
you
can
show
your
screen
there's
at
the
top
left
of
the
hangout
there's
a
green
screen
with
an
arrow
on
it.
That's
screen
share.
A
Okay,
so
this
is
the
minted
method.org
main
page:
there's
a
joint
hangout
button
right
there,
yep
okay,
whereas
if
you
somehow
find
your
way
to
this
one,
you
can't
actually
get
in
except
to
just
watch.
So,
if
anybody's
watching
this,
that
would
be
how
to
join
the
hangout,
which
is
confusing
once
I'll
scott.
This
yeah
scott.
A
Hey
thanks
for
showing
up
sorry
about
this.
I
will
the
the
hard
thing
about
this
whole
setup
is
the
only
way
to
practice.
It
is
to
do
it
because
everything
that
I
put
on
line
is
like
live
online
and
anybody
can
pop
in
and
do
whatever.
So
all
right.
So
we
have
a
few
people
in
let's
I'm
going
to
take
over
and
get
to
the
issues.
A
Okay,
what's
left
in
sprint,
12.
I've
already
moved
some
stuff
over
to
sprint
13.
That
was
just
obvious,
obviously
needed
to
be
continued
there,
but
the
things
that
I'm
I'm
not
certain
about
these
three
issues
were,
let's
see,
were
things
that
needed
to
be
kind
of
prodded
along
this
vector
fixes
was
hung
on
a
travis,
build
that
didn't
work,
so
I've
added
an
empty
commit
to
get
travis
to
run,
and
it's
running
right
now.
A
So,
if
that's,
if
that
passes,
I'm
gonna
ask
for
reviews
on
that
and
see
if
we
can
get
it
merged.
This
other
reduced
memory
footprint,
I
think,
was
in
the
same
situation,
so
I'm
just
kind
of
getting
getting
these
these
pr's
back
up
to
date
so
that
we
can
readdress
them.
They've
been
hanging
out
here
for
a
long
time,
and
this
is
in
the
same
state.
There
was
a
there's,
a
bad
travis
status,
so
those
builds
are
currently
running.
A
Yeah
yeah,
so
this
guy
passed-
and
I
think
this
is
just
unit
tests
and
this
is
blocking
another
issue.
So
scott
you've
already
commented
on
this.
If
you
could
take
a
further
look
at
this
and
see
if
it's
good
to
merge,
just
merge
it,
I
believe
it's
all
just
unit
tests.
A
It
looks
like
just
unit
tests.
Okay,
look
at
this,
the
c
plus
plus
11.
A
She
was
in
the
same
boat,
where
it
was
hung
on
a
in
a
bad
travis
state,
and
I
have
oh
and
it
can't.
I
tried
to
do
a
merge,
but
there
were
merged
conflicts,
so
I'm
gonna,
I'm
gonna,
rely
on
mark
to
take
a
look
at
that
before
I
move
ahead
with
it.
A
The
utility
encoder
is
ready,
ready
to
be
merged.
It's
just
a
new
encoder
matt,
just
a
couple
quick
comments
so
for
the
utility
coder,
I'm
not
sure.
If
that's
something
that's
gonna
be
generically
useful.
I
need
to
look
into
it
a
little
bit
more,
but
did
you
have
you
had
a
chance
to
look
at
it?
A
I
haven't?
No,
I
don't.
I
don't
know
if
I
it
would
be
useful
or
not
either
and
then
the
unit
test
one.
My
comment
on
that
is
basically
it's
just
printing
out
a
bunch
of
stuff.
It
doesn't
actually
it's
not
actually
like
a
unit
test.
It's
not
like
asserting
things
and
failing
when
something's
wrong,
it's
just
printing
stuff
out.
So
that's
not
really
it's
fine,
if
he's
using
that
for
debugging
things,
but
I
don't
think
that's
an
appropriate
test,
including
the
project.
B
I
see
okay
so
for
this,
for
the
you're
talking
about
this
math
functions,
unit,
test
thing
yeah
and
I
think
he
was
he
wanted
to
make
a
change
to
use
an
existing
function
rather
than
one
that
we've
implemented,
and
so
he
was
using
that
I
believe
to
test
that
he
was
getting
the
same
results
which
is.
A
Fine
but
yeah
so
so
I
was
waiting
to.
I
don't
know
if
he
wants
to
keep
that
around
or
not,
but
I
don't
think
we
should
merge
that
and
it's
going
to
stay
you're
talking
about
the
unit
test,
one
more
to
that
one.
Okay.
So
so
we
need
to
to
broad
mark
to
do
the
whole
proper
unit
test,
recording
format,
no!
So,
okay,
so
he
he
knows
that.
But
in
this
case
he
wants
to
make
a
change
to
it,
and
then
he
wants
to
test
that.
It's
the
same.
A
Why
so
it's
more
of
just
like
a
transition
thing?
It's
not
necessarily
a.
C
Test
that
we
would
want
to
keep
around
after
that,
oh
I
see
so
I
just
think
it's
probably
just
something
we
won't
be
merging
okay,
so
it
relates
to
this
other
pull
request
correct
with
the
boost
speed
up
and
this
one's
a
little
tricky,
because
it
looks
like
one
of
the
things
he's
changing,
isn't
exactly
the
same
so
unclear
if
we
want
to
make
that
change
or
not
so
so,
to
move
forward
on
these,
we
need
to
have
some
continued
discussion
with
mark
on
on
how
this
is
going
to
work.
C
So
can
can
you
can
I
what
one
thing
I'd
like
to
do
is
if,
if,
if
a
contributor
outside
of
demeta
gets
something
into
a
state
where
they're
relying
on
us
to
either
approve
it
or
provide
feedback
or
something,
I
think
I'm
I'd
like
to
change
the
assignment
from
him
to
one
of
us
who
should
do
that
work.
A
Sure
and
matt,
maybe
one
of
your
tasks
for
the
sprint
is
to
make
sure
that
you
have
resources
in
the
future
to
assign
these
things
to
just
because
I
haven't
really
been
left
much
extra
time
to
work
on
new
pick
stuff,
so
yeah,
I
know
so
at
least
I
need
to
at
least
figure
out
a
way
to
keep
track
of
these
and
and
get
them
in
a
queue.
A
That's
properly
prioritized
yeah,
maybe
maybe,
if
you
think
with
me
in
the
future,
before
the
grok
sprint
planning,
so
I
can
take
new
good
stuff
into
account.
That
might
be
good
yeah.
That's
a
good
idea.
C
I
I
will
do
that
perfect
so
next
time.
C
A
Sprint
planning
will
sync
up:
okay,
okay,
these
these
other
two,
these
other
three
things
are
simple
things.
Anybody
can
do
they're,
not
necessarily
extremely
important.
A
Okay,
so
I
can
run
I'll
run
through
my
tasks.
I
guess
I've.
I've
got
an
appointment
already
to
look
at
a
venue,
one
venue
next
week
at
linkedin
or
next
hackathon,
and
I
have
some
other
things
that
that's
documentation
this
one.
A
So
this
is,
we've
got
a
lot
of
encoders
and
they're
they're
pretty
well
docked,
they're,
just
not
in
docsis
format.
So
if
I
had
them
in,
if
I
had
the
doxygen
format,
we
should
get
some
decent
stuff
in
our
automatic
api
dock.
So
I
thought
that'd
be
a
good
thing
to
do
and
it's
a
continuous
process
improving
the
the
network
and
region
et
cetera,
api
documentation,
there's
still
a
lot
more
stuff
in
there.
That
could
be
done.
So
I'm
going
to
continue
working
on
that.
A
A
couple
of
these
issues
came
up
over
the
past
few
days
and
they
are
build
issues
and
even
though
there's
been
some
discussion
about
them
and
people
are,
I
think,
making
some
progress.
I
want
to
address
each
one
of
these
problems.
People
are
having
building
or
running
tests
or
failing
tests
on
their
environment
and
see,
if
there's
any
any
actionable
things
that
could
be
done
to
improve
that
process
or
build
for
each
one.
So
I'm
assigning
those
to
me
and
calling
them
build
bugs
I'm
trying
to
continue
doing
that
in
the
future.
A
This
descriptive
page-
I
don't-
I
don't
know
if
you,
if
you
notice,
if
you
go
to
nimitz.org
docs,
you
get
this
empty
page.
So
I
want
to
put
something
here
that
describes
what
what
this
is
and
eventually
have
some.
You
know
a
little
tutorial
usage
example
stuff,
like
that
that
points
specifically
to
some
of
the
classes
that
are
documented
elsewhere
in
the
docs.
A
So
that's
what
this
is.
This
is
an
administrative
thing,
I'm
that
I'm
working
on
I've
we've
got
all
our
aws
assets
in
a
grok
controlled
account.
I
want
to
move
it
to
a
new
controlled
account.
This
guy
is
a
is
a
big
job
and
I
really
haven't
identified
exactly
what
it
what
it's
going
to
entail.
I
think
this
is
going
to
take
more
discussion
within
the
community
about
where
what
our
api
is
going
to
be,
how
we're
going
to
spec
it
out
and
what
direction
we're
going
to
take
with
it.
A
So
that's
a
could
can
be
a
bit
of
a
a
hard
discussion
to
have,
and
then
I
want
to
have
you,
scott
and
subita
involved
in
that
discussion,
and
I
know
everybody's
busy
right
now,
but
it
probably
it
needs
to
happen
with
this
whole
api
documentation.
C
A
To
occur
yeah,
you
can
schedule
something
with
me
whenever:
okay,
good,
okay
cool,
so
those
are
those
are
my.
D
Things
subatai
has
a
few
things
that
he's
created
and
he's
started
work
on
some
of
them,
so
I've
already
moved
those
over
into
sprint
13..
This
is
joe
block.
Is
our
sre
he's
he's
gotta
help
me
get
the
proper
admin
rights
on
our
aws
account.
D
These
are
some
of
the
sort
of
marks
issues.
This
one
should
be.
D
There
we
so
we've
already
reviewed
some
of
this
stuff.
A
Let's
see
some
of
these.
B
The
ones
that
are
in
the
ready
state
we'll
go
over
those
next
in
case
we're
missing
any.
We
talked
about
the
utility
encoder.
B
A
few
of
these
are
newbie
tasks
that
are
that
are
not
assigned
like
this
one.
For
example,
if
anybody
wants
some
low-hanging
fruit,
and
especially
if
you
know
travis,
this
is
probably
just
a
build.
A
travis
configuration
change
that
needs
to
be
done
because
we're
currently
uploading
duplicate.
A
Artifacts
to
s3,
we
don't
need
to
so
some
of
these
things,
I'm
just
letting
hang
there
and
in
case
anybody
wants
to
do
a
little
work
on
new
pick.
They
don't
have
to
understand.
You
know
the
algorithms.
They
can
just
pick
up
some
things
to
help
out.
So
anything
that
look
like
that
is
labeled
newbie.
B
A
Months,
what
is
the
current
state
of
this
yeah?
So
it's
it's.
Basically,
I
think
ready
for
a
re-review,
and
I
don't
I
don't
know
exactly
what
state
this
is,
but
this
was
probably
a
significant
amount
of
work
to
review
scott,
so
yeah
take
a
look
understand.
I
just
I'll
try
to
you
break
up.
Well,
all
right!
It'll
take
a
little
bit
of
time.
I
need
to
pull
the
code.
What,
because
there
are
some
things
I
wanted
to
take
a
look
at.
D
I
don't
know
if
I
can
give
an
exact
eta
with
traveling
and
stuff
this
weekend,
but
I
guess
early
next
week.
Hopefully,
okay.
D
A
A
I
I
asked
fergal
if,
if
he's
ready
to
work
on
this
contributor
listing,
he
said
he
may
or
may
not.
If
he
doesn't
I'm
going
to
pick
it
up
and
probably
get
it
done,
hey
jeff.
Are
you
online?
I
thought
I
think
I
saw
you
jeff
full.
A
A
A
So
I
I
mean
we
could
just.
We
could
have
a
bunch
of
different
solutions
out
there.
There's
you
know
like
a
full-blown
virtual
machine
that
you
can
just
download
through
bittorrent
and
or
some
other
location
and
and
it'll
run.
The
docker
solution
seems
good
to
me
because
it's
a
little
more
lightweight,
but
I'm
not.
I
don't
feel
like.
I
should
be
the
one
saying
this
is
what
we
should
use.
A
A
So
I'm
I'm
okay
with
with
docker
I
mean
any
one
of
these
should
give
us
an
environment
where
we
would
be
able
to
update
the
source
code
and
recompile.
Is
that
true,
yeah?
Okay?
So
as
long
as
that
is
the
case,
I'm
okay,
if
people
want
to
add
their
own
tools
on
top
of
it
once
they
get
the
environment
set
up,
that's
kind
of
their
business.
In
my
opinion,
what
I
would
like
to
see
here
jeff
is
a
a
wiki
page
with
instructions
that
I
can
come
in
as
a
dumb
user.
A
Who
knows
nothing
about
it
and
kind
of
walk
through
step
by
step
and
get
something
working
and
if
once
once
we're
in
that
state,
this
can
be
assigned
to
me
and
then
I
can
kind
of
validate
it.
That's
the
best
case
scenario.
I
think
okay
wiki
page
on
github
yeah,
that's
fine,
okay,
okay,
cool!
So
I'll
leave
this
assigned
to
you
to
get
some
instructions
up,
just
reassign
it
to
me
and
I'll
take
over
from
there
all.
A
Right
this
new
sp
implementation,
subut,
is
still
working
on
and
this
sequence
11
we
already
talked
about.
A
So
let's
look
at
the
ready,
pr's
and
let's
look
at
them
across
all
milestones.
A
Okay,
so
these
are
things
that
are
that
are
marked
to
be
ready
for
review
or
ready
for
review
and
merge.
There
is
one
of
seven
of
them.
Some
of
these.
We
just
talked
about
this.
This
arm,
one
is
getting
pretty
old.
This
is
from
the
hackathon
getting
things
running
on
arm.
So
I
really
don't
know
what
this.
What
the
state
of
this
is
at
this
point,
but
it
can't
they
can't
merge.
A
It
needs
a
little
bit
of
tlc
so
and
it's
currently
assigned
to
hari's,
I
think
his
name's
ari,
so
I'm
gonna,
I'm
just
gonna
ping,
him,
let's
say
yeah,
okay,
great
and
see
if
you
can
make
any
more
progress
on
that.
A
These
other
reddies,
I
think
we
we
already
talked
about
and
I'll.
I
will,
since
I've
got
pretty
much
full
time
to
to
work
on
this,
I'm
going
to
try
and
continue
to
put
pressure
on
these.
I
know
that
the
our
grok
employees
are
are
pressured
already
from
getting
a
product
out
the
door.
A
So
it's
it's
going
to
be
hard
to
to
wrap
these
all
up
within
the
sprint,
but
I'll
at
least
try
and
rob
them
along
the
road
until
they
get
either
merged
or
closed
or
some
type
of
resolution
on
them.
A
So
that's
where
we're
at
for
sprint:
13
we've
got
27
issues.
I
don't
expect
all
of
them
to
get
closed,
but
I
like
to
have
this
many
many.
At
least
these
are
the
high
priority
issues
that
we're
working
on
and
anybody
else
can
come
and
pick
up
if
they'd
like
to
there's
some
that
are
unassigned,
and
if
somebody
wants
to
work
on
them,
they're
already
signed
to
somebody
else,
just
check
with
them
and
say:
hey.
Can
I
take
this
over?
That's
perfectly
fine
and
that's
where
we're
at
so.
A
Does
anybody
want
to
open
up
any
any
discussion
about
this
process?
I'd
like
to
say,
first,
we
had
a.
A
We
had
a
very
light
sprint
because
of
the
holidays,
so
we'd
get
a
whole
lot
of
work
done,
but
at
the
next
sprint
I'd
like
to
kind
of
try
and
do
a
quick
retrospective
at
the
end
of
spring
planning
to
to
get
for
you
guys
to
give
us
any
feedback
that
we
might.
That
might
be
valuable
to
us
to
see
how
we
did
in
the
last
sprint
what
we
could
do
better.
A
A
A
A
Well
talked
about
it,
scott.
Do
you
know
any
details
about
that?
I
know
the
new
pick
one.
You
know
back
with
the
previous
algorithms
they
did
have.
I
believe
they
did
have
a
windows
release.
A
A
Well,
we've
had
a
lot
of
work
on
on
cmake
and
I
thought
there
were
some
windows,
some
possible
avenues
to
work
on
windows
if
we
made
it
compatible
with
cmake,
because
who
was
it
was
david
trying
to
use
a
like
a
net
ide
or
something
with
it.
I
thought
that
was
part
of
the
reason
he
wanted
to
use.
Cmake.
A
Okay,
yeah,
I
don't
think
cmake
is
required
to
get
it
working
on
windows.
I
think
we
we
may
or
may
not
have
the
same
issues.
I
don't
know.
If
that's
going
to
solve
issues
well,
I
can
at
least
say
dean
that
the
direction
that
we're
trying
to
go
with
new
pick
we
will
have.
A
We
will
go
down
that
road,
probably
this
year
at
some
point,
to
get
something:
buildable
run
and
compile
in
windows
just
just
because
we,
the
plan,
is
to
extract
this
c
plus
plus
core
and
provide
language
bindings
for
different
environments.
So
we're
we're
on
the
path
in
that
direction,
but
I
don't
think
we're
definitely
not
there
yet
it's,
I
don't
think
anybody's
actually
compiled
and
run
on
windows.
Yet,
okay!
A
Well,
I
I
might
just
on
my
own
pick
up
that
and
see
what
I
can
do
with
it
and
see
what
I
can
draw,
I'm
looking
for
a
way
to
begin
contributing,
and
that
might
be
a
way
that
I
can
get
my
feet
wet.
That
would
be
a
great
way.
I
would
appreciate
it
very
cool.
Okay,
thank
you
enter
and
you
I
assume
you're
on
the
mailing
list,
so
any
yes
feel
free
to
log
your
process.
Ask
questions
anything
like
that:
really,
okay,
guys
any
anybody
else
have
any
comments.
A
Okay,
great.
Thank
you
for
attending.
See
you
on
the
on
the
mailing
list,
so
feel
free
to
hop
on
irc.
I've
been
on
irc
a
lot
more
often
lately.
So
it's
hash
new,
pic
or
hash
new
for
the
development
section,
which
is
usually
just
me,
so
all
right
take
care
thanks
a
lot
thanks
thanks.