►
From YouTube: Hummingbot Weekly Developer Call 004 | 10 August 2022
Description
The Hummingbot Weekly Developer Call is a technical and general discussion intended to gather active community developers who contribute to making the Hummingbot codebase better through Hummingbot Improvement Proposals (HIPs) and Bug Fixes.
Meeting Info:
- Schedule: August 10, 2022 : 10:00 AM EST / 2:00 PM UTC / 10:00 PM HKT
- Duration: 60 minutes
- Location: Hummingbot Discord at https://discord.com/channels/530578568154054663/996769563054186536
Agenda:
- Review Bot Orchestration PRD https://www.notion.so/Bot-Orchestration-fcac18bd90d74b0ebca9b260617522f0
- Discuss Indicators PRD https://www.notion.so/Indicators-a227377028954b8283417a2b75136d03
- Discuss Include drivers for Postgres and MySQL/MariaDB by default (PR 5452) https://github.com/hummingbot/hummingbot/pull/5452
- Community QA
#hummingbot #developer #algotrading #cryptotrading #liquiditymining #marketmaker #trading #bottrade
A
Well,
I
think
that
we
can
start
now
so
well.
Mike
probably
is
not
going
to
join
today,
so
the
explanation
of
the
normal
hack
will
be
next
week.
He
will
be
the
one
that
will
explain
that
so
well,
for
today
we
have
in
the
in
the
developer,
like
like
in
developer,
things
that
waters
to
us
are
like
two
big
projects
that
we
are
facing,
and
one
of
them
is
the
water
orchestration
and
that
the
whole
structure
actually
alpha
naida
are
here,
are
the
two
most
involved
devs
in
these
in
this
project.
A
This
is
the
one
of
the
the
things
that
work
that
was
voted
before
in
this
community
call
is,
there
is
to
include
some
indicators
or
hlc's
in
hummingbirds.
So
probably
we
are
going
to
review
the
notion
page.
I
will
paste
that
into
the
community
called
chat,
and
you
can
understand
you
can
put
your
opinion
or
join
to
the
stage.
If
you
want
to
say
something
about
that,
and
the
last
thing
is,
I
don't
know
if
stood
sayer
is
here,
but
well,
one
of
the
one
of
the
topics
was
to
discuss
the.
A
If,
if
you
want
a,
we
can
start
with
this
topic,
because
it's
very
simple
we
can
create
a
pool
in
in
the
community
called
chat
kalito.
If
you
can
help
me
with
that,
create
a
pool
that
says
yes
or
no
with
two
buttons,
so
the
people
can
click
on
the
button
so
like
if
they
are
agree
or
not
with
this
idea,
and
probably
they
could
say
why
they
are
agree
or
not
with
this
idea.
A
Building
your
custom.
Docker
image
or
if
you
are
running
it
from
source,
you
have
to
install
the
dependency
in
your
current
environment.
A
user
calls
suit
sayer.
That
is
not
here,
but
is
the
one
that
creates
that
issue
on
the
pr
it
says:
hey.
I
always
deploy
hummingbots
in
different
machines
and
every
time
there
is
a
new
version
of
hummingbirds.
I
have
to
build
my
own
image
and
also
deploy
that
image
in
the
different
machines.
So
it
will
be
much
more
easy
if
you
can
add
the
dependency
to
the
source
code
so
well.
A
A
A
It
will
be
much
more
easy
to
go
ahead,
so
one
of
the
main
to
to
explain
different
opinions,
one
of
the
opinions
from
one
dev
from
queen
alpha
was-
I
don't
think
that
iron
dependency,
that
probably
is
not
used,
is
a
a
good
idea
because
you
won't
be
using
it.
The
dependency
the
and
this
guy
should
tell
yourself
yeah,
that's
true,
but
normally,
if
you're
running
multiple
boss,
you
want
to
use
yeah
the
whole
rusher.
A
I
I
said
the
same
thing
that
you
that
create
different
images,
but
they
told
me
that
he's
like
over
complicating
the
things
and
yeah
well,
they
dismiss.
My.
My
idea,
too,
is
like
is
that
in
the
in
the
issue,
the
threat,
but
well
the
the
the
point
of
saying
was
like
this:
this
driver
was
it
consumes
like
too
low
memory,
but
I
don't
know
what
do
you
think.
A
B
A
I
have
to
explain
what
they
are
talking,
because
this
is
not
doesn't
have
video,
so
the
whole
roger
said.
I
think
that
it
will
be
good
to
have
two
different
docker
bills
on
the
docker
hub,
slim
and
full.
I
guess,
which
will
include
sql
and
gateway
and
obviously
are
already
used
to
choose
between
latest
and
development
in
the
docker
scripts
no
steam
and
fat
will
have.
A
A
B
A
A
A
A
A
A
Yeah,
I
think
that
currently
in
the
creator,
search
sh
is
asking
for
which
version
you
want
to
use,
so
you
can,
you
can
do
it
as
a
with
the
current
script.
The
first
question
that
is
asking
is:
which
version
do
you
want
to
use
by
default
latest
and
if
you
check
the
create
sh
script,
is
using
that
type
to
select
the
image
that
you
want
so
currently
is
working
with
that
a
as
you
mentioned,.
A
Now,
but
just
to
understand
that
script
is
pulling
the
image
from
the
docker
hub.
So
if
you
have
the
the
docker
hub,
if
we
we
have
to
have
all
the
images
already
built
in
the
in
the
docker
hub,
but
well,
what
what
are
you
saying?
Tdc
is,
instead
of
selecting
the
tag
select
like
latest
on
which
database
do
you
want
to
use
mysql
postgres?
B
A
A
There
is
a
video
in
our
youtube
channel
where
emerald
air
coding,
probably
kalito.
We
can
announce
that
in
the
announcements
too,
but
he
creates
a
an
explanation
of
how
to
add
the
the
dependency
in
to
build
a
custom,
docker
image
with
the
things
that
you
want,
but
that
will
be
the
whole
rusher.
That
would
be
the
option.
Number
three:
the
do
nothings
on
the
user,
how
to
add
the
dependencies.
A
A
We
should
create
a
new
image
for
postgres
for
in
mysql
with
nothing
or
we
should
do
nothing
and
the
people
have
to
create
their
own
image.
So
that's
the
that's
the
the
solution
that
that
we
are
trying
to
to
the
the
question
that
we're
trying
to
answer
so
based
on
the
on
the
rotation
right
now
we
have
three
well.
Is
it
my
my?
A
B
A
A
A
A
Currently,
that's
not
possible.
So
the
thing
is
what
should-
and
it
was
a
there-
was
a
pr.
There
is
a
pr
that
kalitha
mentioned
before,
so
what
we,
what
what
we
are
discussing
now
is,
is
correct
to
include
these
dependencies
in
the
requirement
txt,
you
will
be
the
driver
for
the
postgresql
and
driver
for
the
for
mysql.
A
That's
the
option
number
one
that
probably
from
a
perspective
of
how
many
libraries
you
are
including
in
the
code
base.
Probably
you
are
not
going
to
use
that
libraries,
but
the
point
of
superior
in
the
in
the
discussion
was
hey.
I,
and
these
two
dependencies
won't
increase
the
size
of
the
image
too
much
and
it's
like
an
easy
fix
for
this
solution.
A
B
A
Like
my
perspective
is
like
my
first
solution
was
to
create
multiple
texts,
but
in
the
thread
was
like
they,
they
told
me
that
we
will
have
like
a
lot
of
tags
and
they
should
build
their
own
image,
but
then,
once
the
discussion
starts,
we
define
that
the
best
way
if
they
have
this
solution
is
to
ask
the
community
which
solution.
Do
you
think
that
is
more
is
better.
A
B
A
Well,
kit
emerson
says
think
the
size
think
the
size
increment
is
not
so
important
if
you
include
all
dependencies.
Okay,
yeah,
that's
the
point
that
should
say
says
in
the
in
the
in
the
pr2,
and
I
think
let
me
see
the
pr,
but
I
think
that
he
also
put
which
is
a
size.
I
think
that
it
was
1.3
megabytes.
A
A
A
A
But
it's
like
we
have
to
make
that
kind
of
the
the
the
point
that
sayo
was
mentioning
was
like.
There
is
a
bias
trade-off.
This
is
a
trade-off
between
having
things
practical
or
not.
Well,
for
me
is
that
we
should
provide
a
solution,
because
if
it's
the
documentation,
we
said
that
postgre
and
mysql
is
used
is
a
is
an
option.
We
should
provide
that
solution.
A
A
A
If
you
have
the
your
custom
image
or
your
custom
trashy,
you
will
probably
create
your
own
image.
You
won't
use
the
base
image
that
humming
mode
offers.
So
don't
worry
about
that
here
you
can
add
the
dependencies
that
you
want
there,
and
also
the
video
from
emerald
is
very
useful
to
add
the
drivers
that
you
want.
A
He
creates
and
install
sh
file
that
lets
you
add
the
dependencies
that
you
want
well
next
topic
of
today
will
be
the
bot
orchestration
idea,
and
I
will
invite
kia
panagi
and
the
holy
rusher
to
explain
this
solution
that
you
are
trying
to
accomplish
these
guys.
Are
we
work
together
in
the
solution
because
they
have
like
they
work
in
parallel
in
different
parts
of
the
system?
A
The
us
very
that
was
like
the
most
voted
improvement
that
the
community
wants
in
this
community.
Calls
so
we
have
a
discussion
in
this
course.
We
have
a
discussion
in
discord,
the
ones
that
were
interested
in
developing
the
solution
where
panagia
and
the
hover
rusher.
So
as
we
thought
that
it
was
like
not
fair
to
select
just
one
of
them,
because
they
did
a
lot
of
work
in
their
own
by
their
own.
A
So
we
ask
them
if
they
want
to
work
together
as
two
community
members
are
showing
their
efforts
to
combine
their
solutions
and
bring
the
best
solution
to
the
community.
So
do
you
want
to?
If
you
want
to
ask
a
for
the
world
to
talk
the
horoscope,
you
have
to
touch
the
bottom
that
is
in
the
community
calls.
A
B
A
For
me,
just
by
some
words,
I
am
very
excited
about
his
idea,
because
I
think
that
when
you
are
managing
multiple
bots
is
very,
very
useful.
This
tool
that
you
are
trying
to
build,
and
also
I
like
that
you
can
work
together
in
order
to
accomplish
this,
I
pass.
I
paste
in
the
community
called
chat,
the
the
link
to
an
ocean
page
yesterday.
We
have
a
call
with
them.
The
we
try
to
have
this
work
done
in
three
phases,
so
I
don't
know.
A
If
you
want
to
talk,
help
and
again
the
horoscope
about
the
solution
a
little
bit,
then
I
will
go
and
explain
the
faces
and
then
the
community
can
say
I
can
have
your
ex
put
your
opinion
in
the
chat
or
whatever
you
want.
If
you
want
to
go
into
a
stage
you
can
go
to
so
I
will
let
you
talk
about
the
solution.
B
Yeah
so
I
mean
we've,
I
know
some
people
have
been
using
my
roger
that
solution
already.
So
my
solution
is
only
been
designed
to
work
with
trading
views
so
far,
but
I
don't
know
how
to
say:
kill
panagi,
he's
kind
of
done,
a
similar
thing
to
me,
but
for
slightly
different
purposes,
but
there's
a
lot
of
our
code
base.
That
is
probably
going
to
be
the
same
when
we
end
up
reviewing
them.
B
B
A
Okay,
perfect
well,
regarding
which
will
be
like
the
faces
we
are
trying
to
accomplish
in
this
task.
We
split
the
work
like
in
three
phases
so
now
I
would
like
to
that.
The
community
explain
or
put
your
your
opinion
in
the
chat
about
what
you
think
so
the
phase
one
will
be
like
having
like
basic
stuff.
A
We
will
have
here
like
the
in
the
conf
client,
the
the
configuration
of
the
of
the
broker,
so
that
will
be
like
the
phase
one
phase
two
here
we'll
have
like
we.
We
are
going
to
include
some
parts
of
the
prd
that
we
are
going
to
discuss
next
of
the
indicators
and
it
will
be
to
create
an
interface
that
lets
the
strategy,
listen
to
multiple
topics.
A
So
for
now,
like
the
phase,
one
will
be
like
start,
stop
and
config
restart,
undo
and
execute,
like
all
the
the
commands
in
the
bot
and
check,
which
is
the
status
of
the
vote
right
now.
The
second
phase
will
include
like
integrate
signals
inside
the
bot,
so
we
will
have
like
two
types
of
signals:
one,
like
the
current
trading
view,
integration
that
when
you
receive
the
signal,
a
method
is
triggered
in
the
strategy.
A
So
that
will
let
you
I
don't
know
suppose
that
you
receive
a
signal
that
from
trading
view
that
is,
sell
all
your
assets
that
will
trigger
the
method
and
execute
that.
Well,
the
other
way.
The
other
kind
of
signals
that
we
are
going
to
have
in
this
in
the
in
this
phase,
2
will
be
to
receive
signals
from
an
external
source
suppose
that
we
are
receiving
a
prediction
of
a
machine
learning
model
in
in
from
a
for
a
topic,
and
the
module
will
store
the
the
predictions
in
in
an
object
in
a
class.
A
This
class
will
be
the
same
class,
but
will
have
the
ability
to
trigger
events
in
the
strategy
or
to
store
the
the
signals
and
when
he
encode
the
logic
inside
the
strategy,
and
also
it
will
have
the
interface
to
have
the
the
task
of
deploying
new
instances
of
hunting
bots
in
a
single
machine
using
docker
or
from
source.
So
I
think
that
that
would
be
a
very
a
very
good,
a
very
good
feature,
because
this
will
let
that
new
users
run
new
instances
of
hummingbit
from
the
ui.
A
Then
the
ui
has
to
be
able
to
import
new
configuration
files
and
instances
to
the
bot
and
to
finish
the
phase
two
once
we,
the
code
is
shipped
for
the
phase
two.
We
are
going
to
setup,
send
a
poll
to
the
community
to
define
which
do
you
think
that
will
be
good
to
have
or
which
indicators
which
metrics
do
you
want
to
have
from
the
bots
so
and
also
which
extra
commands
that
already
they
handled
always
has
are
good
or
not
to
include
in
this
web
application.
A
A
Well,
the
the
fair,
the
the
last
thing
that
will
be
like
the
phase
three.
It
will
be
the
ability
to
run
hummingbird
headless.
This
is
a
feature
that
a
lot
of
users
that
are
more
like
technical
or
developers
or
if
you're
running
like
40k
of
bots,
a
4k,
now
40
bucks,
sorry
48k,
you
will
be
like
a
king
of
market
making,
but
if
you
want
to
run
multiple
instance,
instances
of
the
bot-
probably
you
won't
be
entering
in
all
the
uis.
A
A
Will
let
you
create
clusters
and
not
only
run
in
this,
in
the
single
machine
that
you
have
the
web
server,
but
this
could
raise
like
different
docker
instances
in
different
machines.
Well,
this
will
be
like
the
world-class
solution
at
the
end.
That's
why
it's
phase
three,
and
also
we
will
include
that,
like
the
both
management,
the
adopter
compose
that
will
be
like
in
a
simpler
way,
instead
of
having
a
having
this
setup
of
kubernetes
or
micro
k8
tests
and
then
for
finishing
this.
A
A
So
well,
that's
that's
like
the
like,
the
like
the
plan
to
run
to
run
this
idea,
the
work,
construction
idea
and
the
ones
that
will
be
working
on
this.
It
will
be
kyo
banai
and
the
hoover
rusher
and
obviously,
if
anyone
of
the
community
wants
to
out
there
to
share
their
opinions
or
add
some
ideas
they
could.
They
could
add
that
so
well,
anyone
wants
to
share
an
idea
share
an
opinion
about
this.
Where
email
says
this
is
a
very
nice
idea.
Sorry
was
a
wait
for
a
bit.
A
I
really
wish
I
had
this
when
I
started
yeah.
I
think
that
a
lot
of
people
will
love
this
idea
and
also
will
let
a
lot
of
new
users
try
hummingbit
and
it
will
be
like
I
don't
know
if
you
know
the
platforms
like
three
commas,
the
you
have
to
put
the
things
that
in
three
commas
you
have
the
agreed
strategy
on
the
dci
boards,
but
the
things
that
you
have
to
share
your
api
keys
there.
A
So
it's
like
probably
there
are
a
lot
of
users
that
wants
to
keep
keep
all
the
keys
inside
their
systems.
That's
why
honeywell
is
very
popular
for
that,
but
that
will
let
users
deploy
bots
like
in
an
easy
way
like
is.
I
will
be
like
using
three
commas
so
that
that
will
be
like
the
the
north
that
we
are
trying
to
work
on
so
well.
A
A
You
are
receiving
the
gifts
of
the
other
book.
You
will
be
receiving
all
the
divs
from
the
other
book
from
the
markets,
connector,
the
the
markets
or
the
component.
These
market
components
will
send
that
messages
through
mqtt
to
all
the
different
bots
and
all
the
bots
will
update
the
order
books
locally.
So
you
will
be
like
having
one
connection
to
update
all
the
objects
of
all
the
bots.
A
B
A
That's
what
I
answer
now.
These
shows
that's
really
awesome.
Emerald
coin,
er
csm
says
cool
initial
deployment.
Automation,
including
starting
dependencies
on
a
new
instance,
be
done
using
ansible.
Perhaps
something
similar
you
want
to
add
new
dependencies.
Okay,
probably
we
can
merge
some
efforts
of
what
emerald
air
coding
have
done
before
so.
A
A
Okay,
then
alkalifa
blockchain
says
I
have
a
question:
how
you
guys
were
able
to
figure
out
accessing
hamilton
instance
from
the
outside.
In
other
words,
I
would
like
to
write
script
access
time,
with
instance,
without
facing
the
ui
to
type
commands
for
the
docker
instance,
how
we
can
do
that
yeah,
that's
the
implementation
that
they
are
trying
to
accomplish.
A
Okay,
the
whole
rusher
answered
that
so
no
yeah,
so
that's
what
they
are
trying
to
build.
Probably
they
are.
They
will
be
using
mqtt
or
some
distributed
system
in
order
to
to
that
and
send
messages
to
the
bot
so
that
that
will
be
covered
in
the
phase
one
in
the
start,
stop
send
all
the
commands
to
the
bots.
This
shield
says
one
other
question
with
all
bots
be
sharing
the
same
order
book,
perhaps
a
gross
shared
memory
or
issues
yeah
for
that.
A
If
that
would
be
like
an
phase
four
details
of
this
project,
but
probably
we
have
to
discuss
that
a
little
bit.
I
think
that
we
will
have.
We
will
have
one
or
the
book
for
each
bottle.
First,
I
know
the
message
will
be
will
be
forwarded
for
all
the
bots
and
they
will
update
the
other
book,
but
well
that
that
would
be
like
the
the
phase.
A
The
phase
four
or
phase
five,
maybe
another
hip,
will
be
that
because
probably
it
will
be
nice
to
have
because
right
now
is
like
having
bots
when
you
are
sending
a
trade
or
you
want
to
buy
or
sell.
That
is
a
promise
that
will
be
executed
once
all
the
the
the
things
of
the
strategy
finished.
So
probably,
if
we
are
thinking
like
more
a
modular
way,
we
could
have
like
a
data
feed
module,
a
trader
modules,
so
the
bots
can
sense
the
trades
that
they
want
to
execute
directly
to
well.
A
Maybe
we
have
to
discuss
a
lot
of
this
a
lot
of
this
there.
Well.
Thank
you.
So
much
for
the
support
says.
Alcaferia
perfect
well,
nulla
haishak.
He
says
what
is
the
timeline
on
the
development
of
the
phase?
One,
that's
a
good
question.
That's
the
same!
That
kaffarya
said
yeah,
that's
very
good
question,
probably
the
whole
russian
can
answer
that.
I
think
that
they
have
developed
like
a
lot
of
of
that,
but
first
we
should
have
the
hip
approved.
A
A
B
Yeah,
I
think
the
same
with
it
could
take
like
a
few
months,
maybe
to
the
creation,
or
maybe
some
even
a
month.
I
would
say
to
this,
but
it's
up
to
the
availability
of
us.
So
when
the,
if
the
proposal
is
likely
to
pass
early,
it
will
be
easier
for
us
to
like
to
schedule
our
programs,
so
I
would
suggest
that
we
we
have
this
requirement
the
proposal
like
the
time.
A
Then
we
can
start
thinking
about
the
phase
two,
the
time
of
the
phase
two,
but
probably
we
can
say
like
one
month
for
phase
one
well,
the
timeline
is,
it
has
to
be
defined
by
you,
but
well
maybe
we
can.
I
think
that
one
month
you
can
accomplish
the
task
in
in
one
month,
because
I
saw
like
your
solutions
separate,
like
the
different
solutions
are
already
are
usable.
So
I
would
like
to
support
in
testing,
says.
A
A
Rice
sierra
says:
do
you
force
any
timing
issues
when
sharing
streams
with
multiple
bots?
Do
the
bots
needs
to
guarantee
that
they
are
all
get
shared
data
at
the
same
time,
a
array
syria?
Currently
they
are
not
sharing
anything
the
bots
and
we
are
going
to
develop
that
this
feature
or
or
the
community
wants
to
develop
the
feature
in
the
future.
A
A
Just
to
mention
the
last
minute
that
we
have
for
this
is
the
indicators
are
pro
is
a
model
that
provides
like
ohlc's
to
be
used
in
the
strategies,
because
currently
we
don't
have
that
we
are
using
external
solution
like
trading
integration,
but
there
are
some
depths
that
wants
to
code
their
solutions
directly
in
python.
So
well,
maybe
we
can
discuss
this
in
the
next
community
call
next
week.
A
Another
point
to
discuss
here.
Thank
you
very
much.
The
horde,
rusher
giovanna
to
join,
explain
the
different
ideas
and
I
think
that
this
way
of
working
as
a
whole
community
trying
to
achieve
these
new
improvements
in
the
system
is
a
way
to
first
of
all,
put
all
of
us
in
the
same
page
and
try
to
work
in
the
things
that
matter
for
all
of
us,
and
this
will
let
us
improve
the
high
involved
software
a
lot.
So
I'm
very
excited
about
this.