►
From YouTube: Bisq developer session 05.03.2019
A
Hello,
everybody
to
this
week's
developer
session.
We
don't
have
a
bigger
agenda
or
nothing.
No
big
topic
for
this
call,
mainly
all
spit,
was
some
of
the
developers
cannot
join
like
Christophe
and
Greg.
I
think
still
has
health
issues,
which
is
back,
so
you
also
cannot
join.
Luckily,
Oscar
could
join
today
may
be
valid,
will
join
as
well,
but
yeah.
We
keep
it
a
little
bit
more
open.
A
Whatever
you
would
like
to
discuss,
we
can
handle
and
and
cover
as
well
I
have
a
small
again
that
has
probably
seen
I
would
like
to
cover
a
little
bit
the
reefs,
and
there
are
a
few
important
changes
and
if
you,
you
wouldn't
fix
the
small
stuff
main
may
be
about
good
to
know
for
everybody.
I
think
then
I
briefly
discussed
with
Lauren
about
options,
how
we
could
add
integration
tests
for
multiple
operating
system,
so
our
build
system.
A
Maybe
we
can
just
brain
some
a
little
bit
in
this
direction,
but
that
would
be
good
when
Bennett
would
be
here,
and
maybe
we
cover
it
more
in
deep
in
depth
when
Bennett
is
available
as
well.
Yeah
related
to
this
release
a
pure
gas,
something
about
locking,
locking
at
least
something
would
I
was
not
really
aware
of
and
which
is
important
to
know,
I
think
for
every
developer
and
when
we
still
have
time,
I
would
like
to
yeah
to
push
out
an
idea
for
how
we
can
improve
our
country
ting
and
maybe
there's
some
developer.
A
Who
is
not
super
good
bite
at
the
moment
already,
with
a
topic
maybe
like
Thomas?
Could
yeah
I
could
try
to
have
a
look
to
this
and
make
a
proposal
about
this
and
so
on.
But
let's
skip
this
at
the
end.
If,
if
there's
still
enough
time
so
I
maybe
make
a
screen
sharing
with
the
poor
requests,
so
we
can
have
a
look
together
over
yeah.
What
has
been
added
to
this
race
and
I'll?
Give
a
quick
overview
and
background
about
the
different
topics
make
the
screen
share
in
one
second,.
A
Do
you
see
my
do
you
see
my
screen,
yeah
yeah
yeah,
and
so
that's
the
closed
for
requests
or
I
think
there
is
no
pull
request
open,
which
is
planned
to
get
in
for
the
releases.
By
the
way
we
have
we
plan
to
have
a
new
release
either,
maybe
today
in
the
best
nah,
probably
not
today,
anymore,
probably
tomorrow,
and
the
main
reason
why
we
wanted
to
get
out.
This
release
was
because
I
found
a
solution
to
solve
this
problem,
with
not
being
able
to
spend
unconfirmed
change,
outputs
from
vsq,
which
is
usability
drawback.
A
So
when
you
made
it
be
as
few
transaction
and
you
had
only
one
you
take
so
we're
obvious,
you
were
sitting
and
then
you
have
to
wait
for
one
blockchain
confirmation,
because
bfq
is
only
verified
when
a
block
is
created.
We
are
not
verifying
unconfirmed
transactions
and
that
yeah,
when
you
wanted
to
create
multiple
proposals,
for
instance,
you
have
to
wait
them
to
the
next
block
and
so
on,
and
especially
for
traders.
It
would
be
very
bad
when
they
want
to
use
psq
who
are
paying
the
trading
fee
and
yeah.
When
you
create
one
tray.
A
If
then,
you
in
the
worst
case,
you
have
to
wait
for
another
block
until
your
you
tick
so
is
confirmed
and
you
can
spend
it
again,
it
would
depend
on
your
YouTube
so
set.
So
when
you
have
many
more
you
TX's,
then
it's
likely
that
you
have
other
remaining.
You
take
those
which
you
could
use,
but
it's
confusing
for
the
user.
Sometimes
you
could
do
it
sometimes
not.
A
So
it
was
a
high
priority
to
find
a
solution
to
make
it
possible
to
spend
your
own
change
outputs,
because
that
basically
has
no
risk
to
you
and
not
trying
to
cheat
yourself
with
creating
invalid
vsq
transactions,
so
it
can
be
considered
as
safe
I
mean
when
there
would
be
because
of
a
bug
or
whatever
an
invalid
transaction.
All
this
follow-up
transaction
will
become
invalid
via
skew
transactions
as
well
and
so
for
the
system.
It
doesn't
create
a
problem
from
usability.
A
Of
course,
it
could
then
screw
up
quite
a
little
bit,
but
that
is
not
expected.
We
don't
allow
spending
you
yeah
yeah,
you
take
those
which
are
confirmed
which
you
have
received.
So
when
I
send
you
some
esq,
then
you
cannot
spend
this
vsq
before
a
blockchain
confirmation.
So
we
have
not
changed
that
and
yeah.
That
was
the
main
reason
for
this
quick
release.
A
I
got
in
touch
with
power,
trader
and
Manero
trader,
who
gave
me
a
very
interesting
feedback
or
he
traded
really
a
lot
and
he
had,
unfortunately,
quite
a
lot
of
cases
of
failed,
failed
taker
take-off
attempts.
So
when
he
tried
to
take
an
offer,
he
got
the
time
out
and
lost
the
trading
fees,
and
it
was
roughly
10%
of
all
his
trades
or
yeah
have
been
failed
and
I
mean
we
have
this
reimbursement,
so
those
users
can
ask
for
reimbursement
on
github
support
repository,
but
it
also
I
think
in
the
last
one
or
two
months.
A
It
was
something
like
10
percent
of
our
revenue,
which
ended
up
in
reimbursement
for
reimbursing
people,
so
it
became
really
a
problem
over
the
last
month's
and
the
reason
why
this
happens
is
not
hundred
percent
clear.
Yet
there
I
suspect
that
it's
standby
behavior
on
different
operate
existence
and
probably
also
on
different
type
for
environment.
So
when
you
are
running
on
a
laptop
it,
standby
behavior
might
be
different
like
when
you
run
it
on
the
same
operating
system,
but
on
a
desktop
computer.
A
A
Whatever
you
define
in
your
preferences-
or
it
doesn't
matter
of
the
while
and
the
network
resources
get
very
low,
and
that
leads
to
their
bad
situation
that
your
offer
still
managed
to
get
republished.
So
it's
visible
and
off-the-book,
then
some
taker
try
to
take
this
offer
and
in
the
take-off
approaches
where
you're
exchanging
the
messages
which
appear
your
network
resources
to
load,
to
respond
in
time
and
the
other
peer
get
the
time
out
and
and
lose
the
trading
field
for
over
6.
A
A
It
seems
that
on
OS,
6
networks
and
it
was
tested
or
I
think
was
with
0
9
release
introduced,
and
we
have
seen
that
we
had
less
such
cases,
but
it
seems
that
most
of
our
users
are
Windows
users
at
the
end,
and
it
seems
that
we
still
have
issues
on
probably
another
from
these
other
operating
systems
from
this
particular
user
know
that
he
was
a
Windows
user
and
yeah.
We
need
to
investigate
stevan
started
a
little
bit
to
investigate
this
issue.
A
We
couldn't
reproduce
it
so
far,
but
I
assume
it
behaves
different
on
your
device.
So
when
you
run
it
from
a
laptop
by
Dennis
assumed
an
for
receiving
energy,
the
operating
system
behaves
different
when,
when
the
system
is
in
standby
mode,
so
we
need
to
go
on
to
investigate.
To
find
solutions
for
this.
One
possible
solution
would
be
to
make
sure
that
your
offers
get
removed.
So
when
you
don't
have
sufficient
network
resources,
then
you
remove
your
offer
yeah.
A
We
have
to
discuss
what's
the
best
strategy,
how
to
deal
with
it,
but
sort
of
underlying
problem
still
is
not
solved,
but
because
we
probably
cannot
solve
this
soon
and
quickly.
I
tried
another
solution,
which
is
that
we
are
delaying
the
probe.
Casting
of
this.
Take
your
feet,
transaction
just
to
the
time
before
the
taker
is
publishing
the
deposit
transaction.
So
in
the
trade
protocol
like
yeah
like
it
was
before
the
take
the
feed
transaction
was
published,
then
the
taker
send
a
message
to
the
maker.
A
It's
a
problem
for
when
some
whatever
chain
analysis
companies
want
to
spy
on
disk
users
they
to
basically
try
to
take
all
offers
and
then
fail
in
the
last
moment
and
by
publishing
the
take
a
fee
transaction
early.
We
try
to
yeah.
We
try
to
add
some
costs
to
this.
So
when
somebody
want
to
do
this,
it
becomes
expensive.
At
least
it's
not
a
very
strong
protection
and
it
yeah
at
the
end.
A
We
also
never
managed
to
really
make
the
verification
of
the
maker
side
to
only
sent
the
data
when
the
fees
comfort
is
publish,
it
would
be
problematic.
All's
because
it's
not.
We
cannot
wait
until
a
confirmation,
so
it's
a
pending
transaction
and
there's
no
guarantee
that
the
maker
will
see
the
transaction
and
network.
A
That
is
still
a
tiny
potential
that
in
this
last
moment
there
are
two
different
tasks
in
the
trade
protocol
and
when,
for
whatever
reason
it
would
fail
to
publish
the
deposit
transaction,
then
it
would
still
be
a
problem,
but
there
is
no
reason
why
it
should
happen
that
maybe
one
in
a
million
times
so
are
yeah.
That
was
a
very
important
feature
which
should
increase
usability
because
cause
it's
very
annoying
for
users
to
ask
them
for
reimbursement,
and
many
people
don't
know
it.
They
just
take
it
in
take
the
loss
he's.
A
A
A
So
that's
not
a
very
cheap
call,
then
all
that
get
executed,
even
if
you're,
not
on
that
lock
levels.
All
this
trace
and
debug
logs,
which
used
the
string,
concatenation
buoys,
plus
order
variable
to
string
concatenation
with
blasts,
were
a
problem
from
performance
point
of
view.
I
fix
this
now,
and
so
it
should
make.
Maybe
some
of
you,
and
hopefully
most
of
you
or
notice.
A
I
was
not
really
aware
that
there
is
a
performance
difference
between
that
method,
and
the
recommended
method
is
that
you
use
the
curly
brackets
and
the
comma,
and
then
you
can
add
yeah
whatever
wearables
you
want
to
add
there
and
then
the
expression
doesn't
get
evaluated
by
the
compiler,
and
it's
only
evaluated
when
you
are
on
the
lock
level.
Florine.
Do
you
want
to
add
something
here
or
maybe
make
it
a
little
bit
more
clear?
Like
my
explanations?
Oh
no,
no.
A
Think,
okay,
that's
good
enough,
it
yeah
so
yeah.
Basically,
we
should
care
and
we
will
take
care
in
the
refuse
for
new
pull
requests
that
we
don't
allow
I
mean
I,
try
to
clean
up
a
little
bit
of
the
old
stuff,
but
it
still,
of
course,
it's
along
a
task
to
clean
up
all
the
logging
that
we
are
not
using
our
pet
that
we're
using
the
preferred
method
and
maybe
also
remove
the
trace
and
debug
messages.
It's
because
mostly
they
are
not
really
needed
and
just.
A
What
else
then
yeah
edit
another
pull
request
yesterday
for
percentage
based
values
for
the
security
deposit?
Some
traders
complained
that
yet
it
was
a
fixed
amount.
So
it
didn't
matter
if
you
trade,
20
or
urine
Bitcoin,
or
if
you
trade,
500
euro
in
Bitcoin,
the
security
deposit
for
the
buyer
was
fixed
amount
in
Bitcoin.
The
maker
can
define
the
security
deposit
and
the
main
reason
is
to
avoid
kind
of
option
trading
or
a
future
trading.
A
We
get
reported
more
disputes
from
arbitrators
that
some
users
are
abusing
just
basically
and
they
they
prefer
to
lose
the
load
security
deposit
instead
of
continuing
the
trade
and
have
basically
more
loss
when
they
continue
to
trade
with
the
bad
price
for
them,
and
it's
cheaper
than
for
them
to
make
a
new
trade
with
a
better
price.
And
to
avoid
this,
we
keep
it
basically
to
the
to
the
maker
that
the
maker
can
define,
usually
they
know
or
how
volatile
the
market
is.
It's
not
a
perfect
solution.
A
Everyone
still
to
improve
this
that
basically
their
security
deposit
will
be
derived
from
the
volatility
from
the
trade
statistics,
but
that's
a
little
bit.
That's
not
a
task
for
another
Greece
and
a
short
solution
now
was
to
make
it
at
least
percentage
based
on
the
trade
amount,
because
if
you
trade
to
Bitcoin
in
monera
or
if
you
trade,
just
a
very
tiny
amount,
there
should
be
a
different,
a
different
amount
for
the
security
deposit
as
well
and
yeah
that's
implemented.
A
Now
we
also
increased
a
little
bit
the
default
values,
because
it
seems
that
yeah
we
get
too
many
disputes
where
people
are
using
this,
so
hopefully
with
higher
security
deposits
or
people
stopped
trying
to
cheat
invader.
The
trading
protocol.
It's
not
really
a
cheating,
because
it's
part
of
the
possibilities,
but
it's
not
the
intended
behavior
when
you
take
it
off
for
basically,
you
have
to
fulfilled
offer
and
this
yeah,
because
we
cannot
force
the
buyer
to
make
the
payment.
A
Of
course,
he
has
the
possibility
to
to
cancel
out
in
a
way,
but
it
should
be
so
expensive
for
him
that
usually
he's
not
doing,
even
if
the
price
moves
against
his
interest,
so
that
should
help
yeah,
because
one
of
those
feedbacks
which
was
very
reliable
from
these
traders,
which
is
a
pro
trader,
I,
wanted
to
find
out
yeah.
What
are
the
main
issues
for
pro
traders
with
the
current
dis
implementation?
And
what
can
we
improve
for
them
to
make
to
make
it
more
attractive?
A
And
one
of
the
main
thing
was
that
he
said
yet
to
have
the
funds
locked
in
very
short
time.
The
longer
they
have,
basically,
the
Bitcoin
or
money
are
locked
in
in
a
trade
the
worse.
It
is
because
yeah
they
have
only
a
certain
amount
of
of
resources.
What
they
can
have
on
on
both
occurrences
to
trade
and
when
70%
of
those
resources
are
locked.
In
and
trades
they
cannot
go
on
with
trading,
and
that
was
the
motivation
to
add
another
feature.
This
one
with
the
live
trading
live
trading,
it
may
be
not
really.
A
Their
life
was
struggling
with
the
right
term.
We
call
it
now
instant
trading,
it's
at
the
moment
only
for
all
coins.
So
when
you
create
an
altcoin
account,
there
is
a
new
jack
book
where
you
can
enable
instant
trading
and
with
that
their
trade
period
is
limited
to
one
hour,
so
you
have
to
fulfill
both
graders
have
to
fulfill
this
trade
in
one
hour.
So
only
both
can
trade
when
both
have
set
up
such
an
account.
A
So
you
need
basically
two
accounts
when
you
sometimes
prefer
to
fast
trading
and
sometimes
the
normal
trading,
and
so
you
kind
of
mix
when
you
have
set
up
a
normal
mony,
our
account,
you
cannot
trade
with
a
maker
who
has
created
a
offer
for
instant
trading.
You'll
need
to
create
a
new
account
for
this
instant
trading,
and
then,
when
you
take
this
off,
you
have
to
ensure
that
you
are
physically
online
and
you
can
fulfill
the
trade
in
one
hour.
A
Those
are
returned
to
your
wallet
and
you
can
spend
it
because
of
you
have
to
ring
after
ring
signatures
there
in
several
rounds,
and
this
time
you
can't
you
cannot
use
them
so
in
the
worst
case,
or
it
takes
a
half
an
hour
or
whatever,
but
the
malarious
has
to
wait
until
he
can
again
spend
the
money
and
of
course,
when
they
do
many,
many
like
trading
him
parallels.
It
would
get
to
the
same
problem.
A
A
Yeah
Florian
found
luckily
one
very
important
and
long
or
existing
memory
Niek,
which
was
mainly
whistle
only
on
the
seed
notes.
Whenever
could
see
it
on
normal
applications
on
the
way,
six
or
l2
from
users
about
not
really
reported
in
the
connection
class
yeah,
there
was
a
tiny
small
stupid
missing
line,
basically
where
we
were
collecting
for
checking
how
much
messages
you
are
handling.
We
were
collecting
those
and
when
you
had
yeah,
when
you
receive
too
many
messages,
then
you
stopped
was
when,
on
the
receiver
side.
So
when
appear
was
sending
you
too
many
messages.
A
You
disconnect
him
and
when
you
send
out
yourself
too
many
messages,
you
were
struggling
down
and
call
some
sleep
on
a
thread
in
between
and
there
was
a
list
which
were
collecting
all
this
messages
over
time
with
timestamps,
and
this
list
was
not
correctly
cleaned
up
when
it
was
hitting
a
limit.
So
we
were
only
collecting
us
in
2000
messages
there
and
then
starting
to
to
delete
all
messages,
and
there
was
we
are
missing
while
loop
to
do
delete.
Really.
A
All
messages
are
not
only
the
last
one
or
the
first
one,
and
that
should
fix
basically
the
main
issue
on
seed
notes.
That
memory
was
growing
on
seed,
not
we
still
and
approaches
to
check
and
verify
if
it's
was
the
only
issue
or
not,
but
at
least
it
was
for
sure
one
of
the
major
issues
so
I'm
happy
that
we
could
get
this
solved.
A
There
are
a
few
other
tiny
other
improvements
in
that
area,
especially
with
the
locking
and
so
and
some
other
small
improvements
in
the
peer-to-peer
network,
which
yeah
I
said
that
was
another
finding
with
feedback
from
disuse
and
also
from
other
a
little
bit
investigation.
We
we
have
seen
that
we
we
got
a
little
bit
more
lost,
European,
Network
messages.
A
Users
are
reporting
that
disputes
yeah
we're
not
handles
those
who,
because
they
are
permitted
open
dispute
message,
never
arrived
at
the
arbitrator
or
trade
messages
didn't
arrive
and
it's
probably
caused
because
subpoena
being
it
but
get
to
grout
it
already.
We
have
many
more
users
like
in
the
past
and
their
republishing
of
the
offer.
A
So
when,
when
the
maker
would
lose
network
connection
or
kill
his
application
without
without
the
plant
shut
down
a
graceful
shutdown,
then
the
offer
get
removed
us
at
the
time
to
live,
has
expired
from
every
peer
and
this
time
to
live.
I
think
it
was
seven
minutes
and
the
Refresh
mess
refresh
interval
was
five
minutes
and
it
seems
that
this
two
minute
in
between
was
not
enough.
A
But
it
was
probably
because
the
refresh
message
didn't
arrive
at
all
piers,
so
I
changed
this
parameters
for
the
refresh
and
for
a
time
to
live
increase
both
the
refresh
to
reduce
a
little
bit
the
network
load
and
the
time
to
live,
to
make
it
less
likely
that
you
get
in
this
problem
zone
that
you
didn't
receive
the
refresh
message
in
time.
I
think
that
probably
will
not
really
fix
a
lot.
It
should
be
just
a
tiny
or
a
small
improvement.
A
We
will
look
into
a
better
fix
by
patching
those
messages,
as
when
us
appear
receive
hundreds
of
messages.
We
have
seen
that
more
or
less
than
300
refresh
refresh
offer
messages
per
minute
that
something
like
or
five
messages
per
second,
which
is
quite
a
lot
and
we
try.
We
will
try
to
find
a
solution
for
batching
all
these
messages
and
pack
it
to
one
message
which
will
be
broadcast
to
the
other
peers.
A
So
you
are
not
sending
out
an
hour
the
worst
case,
something
like
20
or
unread
messages
per
second,
but
only
one
message
which
contains
and
all
this
message,
but
you
have
arrived
received
in
a
certain
time
frame,
but
that's
a
little
bit
more
complex
and
yeah.
We
discussed
with
Laureen
how
to
do
this
and
when
will
be
ready,
but
that
should
improve
alot.
This
network
contraction
issues
and
should
make
them
appear
appear
at
work
more
reliable
again,
then.
A
Another
important
feature
is
in
the
API
branch,
as
API
branch
is
still
separate
from
the
master
and
it's
still
not
ready
for
production,
but
I
think
we're
getting
closer
now,
because
a
very
important
to
important
features
would
have
been
implemented
from
Vinod.
One
is
a
security
framework
for
password
protection
with
a
token
for
API.
So
all
the
calls
for
API
require
token,
and
you
have
to
provide
your
pass.
What
today
API
to
receive
the
token
and
then
a
second
one,
the
connection
to
the
API
will
be
only
supported
over
tor
hidden
services.
A
We
were
considering
also
to
use
https,
but
it
would
have
had
the
problem
that
you
need
our
setup.
Let's
encrypt
to
get
certificate,
and
you
cannot
can
only
do
this
when
you
have
a
domain,
so
many
people
will
run
it
just
with
an
IP
address
without
the
domain,
and
you
have
a
problem
to
get
HTTP
support
for
this
and
HTTP
is
also
based
on
the
centralized
system.
Basic
is
not
so
super
great
fan
of
it.
A
I
mean
it
would
have
been
good
enough,
but
the
main
reason
was
the
others
in
this
limitation
that
you
need
a
domain
and
need
to
set
up,
let's
encrypt
and
so
and
tor
is
solving
all
these
problems
and
deliver
much
more
features
and
only
comes
with
the
extra
costs
that
it
costs
more
performance,
which
is
not
the
problem
with
API.
We
don't
have
yeah,
we
don't
have
a
performance
issue
there
that
you
are
calling
your
IP
I
wish
thousands
requests
per
second
or
whatever.
A
So
he
has
implemented
this,
and
it's
actually
a
quite
an
interesting
way
how
its
implemented
it
doesn't
use
a
socket
in
Java.
It
just
call
their
net
layer
library
that
it
created
to
create
a
new
hidden
service
which
is
only
used
for
the
API.
So
it's
completely
separate
from
the
normal
bisque
application
and
it
proxies
to
the
port
to
the
local
port
of
the
API.
So
it
it
doesn't
start
anything
in
Java.
It
doesn't
go
like
our
to
our
framework
at
the
way
how
we
use
it.
A
We
are
starting
a
socket,
and
this
socket
is
basically
our
local
proxy
in
their
application
are
for
API,
not
even
that
happens
so
I
think
from
performance
point
of
view.
It's
very,
very
cheap.
That's
really
done
purely
from
the
Tor
binary
and
proxying
hidden
service
to
the
local
port,
so
I
think
that's
quite
a
cool
implementation
and,
from
my
point
of
view,
a
synched
API
would
be
basically
ready.
Now
for
production
I
mean
there
are
there.
A
There
are
features
which
are
implemented
are
not
very
usable
for
I
think
it's
only
get
get
version
and
another
trivial
feature,
but
it
was
intentional
to
it,
basically
all
the
basic
flow
in
place
and
solve
and
then
start
to
implement
the
features.
But
beyond
that
is
working
on
this
and
maybe
in
the
best
case,
with
the
next
release,
we
can
release
publicly
the
api.
A
I
prefer
to
have
two
binaries,
the
normal
desktop
binary,
which
doesn't
include
anything
from
the
API
and
the
second
binary
desktop
plus
API,
where
those
users
who
want
to
take
advantage
of
this
feature
to
have
the
desktop
plus
API
in
one
application
they
can
install
and
on
this.
But
if
we
would
find
out
some
security
vulnerability,
it
would
be
limited
to
much
less
users
and
the
potential
damage
would
be
much
less
critical.
A
It's
a
little
bit
of
pain
because
yeah
we
need
to.
We
have
an
extra
application
built,
so
the
build
takes
a
little
bit
longer
and
so
on.
Maybe
we
can
find
some
kind
like
smart
solution,
that
by
default,
you
are
not
compiling
this
extra
application
only
further
for
a
release,
manager,
who's
doing
the
release.
He
will
then
compile
this
or
whatever
we
have
to
discuss
when
we
get
closer
to
this
and
just
looking,
if
there's
anything
else,
yes
sink
recover,
probably
most
of
the
important
features
so
I
stopped
the
screen
sharing
you.
A
If
not,
then
I
go
on
yeah
I
started
to
discuss
a
little
bit
with
Lauren,
just
before
the
call
possibilities
to
add
integration
testing,
especially
for
different
operating
systems
in
our
build
system.
So
when
you
make
yeah
when
we,
when
you
make
a
pull
request,
it
get
tested
on
Windows
on
Linux,
on
OS,
6
and
so
on
automatically.
A
It's
for
sure.
Not
the
easy
task
and
I
would
postpone
this
when
Bennett
is
available
as
well,
because
he
has
a
lot
of
experience
with
integration
tests
with
the
API
and
so
I
doubt
that
Travis
in
the
free
version
like
we
are
using
it
at
the
moment,
will
support
this
because
I
think
it
will
consume
much
more
resources.
So
probably
we
need
to
start
to
run
our
an
integration
test
server
and
handle
this
not
familiar
with
this
at
all.
But
maybe
Devon
can
help
that
error.
A
That
will
be
the
way
to
go
yeah,
because
the
main
issue
is
that
most
developers
only
are
developing
and
they
have
available
one
operating
system.
Even
if
you
have
like
myself
or
Christoph,
you
have
some
VMs
with
other
operating
systems,
it's
kind
of
like
painful.
To
start
this
way,
I'm
super
slow
and
at
the
end
it
always
behaves
a
little
bit
different
specially
when
you
want
to
test
some
more
complex
stuff
like
there
like
at
the
moment.
This
issue
is
the
standby
mode.
A
I
mean
that's,
of
course,
cannot
also
be
done
by
integration
test,
but
it's
limited
with
VM.
It's
not
really
the
same,
and
to
have
a
little
bit
more
realistic
system
would
be
good.
I
mean,
of
course,
this
integration
test
will
be
limited,
not
really
the
real
stuff,
but
at
least
it
will
help
us
to
spot
potential
issues
on
different
operating
systems.
Luckily,
we
we
don't
have
many
issues.
I
mean
one
of
the
very
few
have
been.
This
issue
is
the
seed
node
restore
problem?
It
seems
that
we
still
have
some
issues.
A
It's
a
terrible
problematic
because
at
least
he
can
recover
that
with
help
from
somebody
else.
But
it's
still
of
course
very
annoying
to
use
and-
and
it's
yeah
spent
for
a
reputation
of
bisque
when
such
a
critical
part
is
not
really
working.
Hundred
percent
safe
should
put
this
a
little
bit
on
the
our
priority,
so
maybe
Oscar
a
few
has
have
time.
Maybe
you
can
I
don't
want
to
interrupt
your
plan
at
the
moment.
I
know
mr.
A
A
I
will
assign
it
and
github
and
we
can
have
a
discussion
anyway,
when
you
have
time
private,
to
talk
a
little
bit
about
the
roadmap
and
so
another
issue.
What
we
have
spotted
over
the
last
weeks,
a
little
bit
more
like
in
the
past,
from
my
point
of
view,
is
our
startup
issues
we
store
or
heater
that
you
cannot
get
it
started
at
all,
and
it
seems
that
there
are
some
issues
and
experiences
myself
last
year
that
suddenly
I
couldn't
connect
to
tor
anymore,
with
different
new
on
and
resistance
on.
A
So
for
me
the
solution
was
to
use
VPN,
and
then
it
was
working
again,
and
so
that's
also
for
some
users
who
might
have
this
problem
they
could
try
to
start
over
with
a
new
on
in
address,
unfortunately
lose
the
local
reputation.
When
you
do
this,
but
yeah
that's
one
potential
solution,
I
mean
the
first
solution
is
to
just
clean
up
your
tour
directory
feed.
You
can
do
it
in
application.
A
Delete
all
the
files
excluding
their
hidden
service,
private
key,
that
you
must
not
delete,
of
course,
and
sometimes
this
helps
or
some
windows
it
seems
there
still
may
be
some
issues
with
some
locked
files
or
whatever,
and
the
next
point
would
be
to
try
to
start
over
with
a
new
hidden
service
and
I've.
Seen
this
all
in
the
past
sometimes
had
problems
that
one
hidden
service
didn't
work
and
when
I
created
a
new
one,
everything
goes
fine
again
and
yeah.
A
It
seems
that
sometimes
some
onion
addresses
also
get
kind
like
flagged
or
whatever
come
with
some
problems.
I,
don't
know
exactly
why
and
when
all
that
doesn't
work,
then
using
VPN
is
probably
during
the
last
solution
or
using
pluggable
transports
into
a.
But
then
it's
probably
more
related
that
you're
living
in
a
country
where
tourist
blocks
like
in
China
or
Iran
or
so
but
I
think
yet
is
tor
issues.
We
should
also
put
on
our
high
priority
list
to
get
this
soft,
because
it
seems
that
there
are
a
little
bit
too
many
reports.
A
I
mean
we
never
get
everything
perfect
and
when
there's
100
report
every
two
months
or
so
I
don't
care
really.
But
when
there
are
five
reports
per
month,
because
every
user
who
read
for
every
user
who
reports
we
have
ten
other
in
the
in
the
silence
who
are
not
reporting?
And
so
we
should
take
it
here.
If
you
take
it
serious,
when
we
get
repeated
reports
about
some
issues,.
A
B
A
Yeah
completely
agree:
it's
just
not
trivial
and
a
lot
of
effort.
That's
the
reason
why
they
don't
exist
yet,
but
luckily
that
has
already
done
a
lot
of
this
work
on
the
API
side.
So
we
say:
API
I
have
never
looked
very
close
to
what's
covered
exactly
and
how
everything
works,
but
I
think
our
Vista
API.
We
will
get
already
quite
good
coverage
from
integration
tests
because,
yes,
yes,
he's
covering
a
lot
of
use
cases
like
doing
all
the
trade
process
and
so
on
and
as
many
of
the
features
which
will
not
be
enabled.
A
As
in
the
API,
when
we
Reece
API,
we
will
have
a
program
argument,
flag,
allow
or
disallow
experimental
features
and
by
default
also
when
it's
called
allow
experimental
features
by
default.
It's
false,
so
a
normal
user
who
will
use
API
only
get
enabled
those
features
which
are
really
well
tested
and
reviewed
and
integration
tests.
You
need
all
the
stuff
with
ESRI
integrated,
so
basically
yeah
big,
a
big
scope
of
their
piss
features.
A
What
you
can
do
in
the
UI,
you
can
do
already
on
the
API
as
well,
but
it
will
take
much
more
time
to
get
it
already:
production
ready
and
refilled
and
test
it
well
and
to
to
not
be
blocked
too
long.
We
basically
go
on
with
one
feature
after
the
other,
from
starting
from
the
most
for
basically
creating
office
and
doing
a
trade
and
then
getting
further
and
further
to
create
a
payment
account
to
make
a
Bitcoin
transaction
as
on
and.
A
Yeah
for
the
user
perspective,
those
features
we
come
over
time
and
we'll
take
a
few
months
until
the
full
scope
will
be
available
for
users
who
are
not
willing
to
take
the
risk
to
allow
this
experimental
features,
but
for
testing
purpose.
We
can
use
this,
of
course,
because
there
doesn't
really
matter
and
yeah
hi.
Maybe
we
have
one
special
session
where
a
parent
is
a
little
bit
more
with
you
about.
Do
you
think
the
gray
shoot
has
what
how
it
works
and
yeah
I
explained
it
all
all
the
more
in
detail.
A
Recurring
yeah
with
the
bit
country,
maybe
give
a
little
bit
of
or
with
you
yeah
there's
a
new
release
done
now
with
segment
support.
So
that's
a
huge
release
which
took
two
years
or
so
so
it's
not
very
back
yeah.
It's
completely
different
to
the
latest
Greece
to
the
0:14,
but
we
are
using.
So
it
will
be
a
big
effort
for
us
to
move
over
to
the
new
release,
but
it's
basically
on
our
roadmap
for
the
next
few
months.
A
The
current
roadmap
is
that
we
get
as
we
are
using
the
0:14
for
release,
and
then
it
was
a
little
bit
and
you
were
released.
The
14-7
and
oscar
has
yeah
updated
all
the
code
through
the
14-7.
So
basically
that
should
be
I
think
it's
already
a
pull
request,
which
is
ready
for
review
and
merge
could
have
probably
for
the
next
release.
Now
it
just
didn't
want
to
get
it
in
into
that
release,
because
it
wanted
to
keep
risks
and
testing
effort
load,
but
probably
for
the
next
release.
A
We
will
basically
update
to
this
14
7
release,
which
doesn't
really
have
a
lot
of
changes
or
whatever
not
a
big
change.
For
the
user,
whatever,
but
then
a
big
task
will
be
to
really
test
well,
their
Bitcoin
tree
master
release.
So
maybe
we
create
their
own
bisque
branch
for
this
I'm
not
I,
wasn't
sure
how
to
deal
with
it,
but
it
would
be
good
basically
when
developers
would
work
on
this,
but
we
don't
have
any
master
as
long.
We
are
not
very
confident
that
there
are
no
bugs
there,
it
country
yeah.
A
We
don't
want
to
be
there.
The
first
alpha
testers
who
run
into
production
bugs
by
going
too
quickly
into
moths,
and
we
don't
leave
the
features
really
I
mean
we
don't
hardly
need
sacred
at
the
moment.
So
it
would.
We
have
personalized
so
at
how
many
people
request
it
more
for
political
and
principal
reasons
and
myself,
of
course,
I'm
also
a
big
fan
of
segments,
but
technically
we
don't
need
it
and
it's
not
used
in
a
trade
protocol
and
so
on,
and
it
will
be
for
sure,
because
of
many
API
changes.
A
It
will
take
some
time
to
adopt
all
the
biskits
code
order
for
this
master
release
and
also
then
to
add
the
new
possibilities
with
secret
and
so
on.
So
assume
that
take
minimum
one
months
work
and
then
probably
another
month
going
tensive
testing
and
so
on.
So
I
assume
maybe
a
realistic
time
scope
in
two
or
three
months
that
we
could
get
updated
to
a
bit
country
master
with
sacred
sword
and.
B
Having
say
we
support
me,
meaning
that
you,
you
will
have
like
said
with
a
versus
like
if
you
want
to
have
also
backwards
contact
compatibility,
meaning
that
you
can
receive
funds
from
wallets
that
cannot
sent
you
two
bears
3232
addresses
we
should
have
like
to
like
be.
We
should
enable
users
to
have
like
two
addresses.
One
is
the
all
others
and
the
other.
B
A
But
not
sure
I
understand
correctly,
as
I
didn't
have
planned,
that
we
are
moving
only
over
to
back
32
addresses
only
so.
Basically,
we
should
support
both
addresses
or
when
you
get,
that
you
consent
to
classical
Bitcoin
addresses
and
to
the
new
address
format
and
when
you
receive
funds
yeah
the
UI.
Of
course
we
need
to
to
have
some
changes
to
that.
The
user
can
decide.
When
you
want
to
receive
address,
you
can
decide.
A
You
want
a
classical
address
or
a
new
back
so
that
you
address,
but
that
and
of
course,
I'm
not
sure
in
the
trade
protocol,
how
exactly
to
deal
with
it,
because
you
can
fund
your
required
funds
at
the
moment,
only
with
classical
addresses
and
so
on.
So
I
would
do
it
basically
in
in
steps
and
start
only
with
the
minimum
that
you
can,
that
we
support
back
32
addresses
for
sending
and
receiving,
but
not
using
it
in
the
trade
protocol,
because
I
think
it
doesn't
really
add
any
any
big
benefit.
A
Technically
or
I
mean
it's
only.
Basically,
this
kind
of
like
sacred
support,
mentality,
which
is
more
a
political
thing
and
I'm,
also
not
100%
sure,
and
there
are
some
people
like
Luke
chair
who
are
against
using
back
or
a
sec
with
transaction
when
you
don't
need
it
for
lightning
Ursula,
because
basically
they
are
get
counted
in
this
additional
space
and
basically
create
helped
to
create
bigger
blocks.
So
when
more
and
more
people
are
using
psychic
addresses,
it
leads
to
factually
bigger
blocks.
A
I
mean,
of
course,
the
block
size
is
still
there
1
megabyte,
but
this
additional
block
space,
which
is
handled
differently,
is
in
reality.
Of
course,
you
need
when
you're
full
note.
You
need
to
download
this
state
as
well,
and
it
costs
you
more
resources
and
for
those
people
who
are
in
the
camp
like
Luke
death
share,
who
want
even
smaller
blocks
and
are
very
concerned
about
the
block,
size,
growth
and
resource
consumption.
A
I'm
I
just
found
it
interesting
all
these
comments
from
Luke
yeah,
because
there
was
not
aware
of
myself
earlier
and
people
like
masa
said
well,
a
new
part
of
73
from
passive
e-wallet
who
push
a
lot
to
sacred.
Only
I
have
not
discussed
with
him
and
I
I,
don't
know
exactly
the
background,
but
I'm
a
little
bit
skeptical.
A
If
that
really
is
such
a
good
idea
and
if,
basically,
if,
for
instance,
when
we
can
choose
to
use
only
psychic
transactions
in
the
trade
protocol
or
not,
if
it's
a
good
idea
or
not
because
we
don't
need
the
features
from
I
mean
transactional
ability
is
a
theoretical
issue
in
this
case.
Well,
but
it's
not
a
practical
issue
and
yeah,
but
that's
maybe
for
future
discussions.
But
I
would
say
in
the
beginning,
the
goal
is
only
to
make
it
available
for
sending
and
receiving
some
people
are
complaining
that
you
cannot
send
her
from
Mesabi
wallet.
A
For
instance,
you
cannot
send
funds
to
bisque
and
the
Vasavi
developer,
we're
even
trying
to
help
projects
who
are
not
on
track
with
yet
to
to
implement
it
and
so
on,
but
I
think
that's
now
obsolete
because
bid
countries
now
ready
and
they
are
not
Java
developers.
So
probably
they
are
not
really.
They
cannot
help
so
much
but
of
course,
it's
great
that
they
offer
this
help.
A
Yeah.
That's
a
little
bit
my
point
of
view
on
this,
but
we
can
have
maybe
a
our
self
discussions
word
and
when,
if
we
get
more
closer
to
this,
a
more
broad
discussion
about
the
strategy,
how
we
supported
and
what
we
support
and
the
roadmap
for
that.
Okay,
well,
I'm,
not
very
familiar
I
yeah.
With
this
there's
some
complexity
of
different
backward
couple
of
utility
with
addresses
it's
on
I
have
not
really
follow.
A
It
close-
and
maybe
you
gave
me
a
little
bit
of
update
about
this
before
we
were
discussing
this,
but
some
complexity
included
in
this
topic.
We
said
with
addresses
and
transaction
hands
on
that
yeah
and
I.
Think
older
from
the
wallet
side,
I
think
there
were
some
issues
that
we
say:
HT
wallet
there
was
so
some
non-trivial
complexities
included.
B
B
We
will
have
to
like
create
Q
key
chains
inside
the
wallet
to
have
both
say
with
addresses
and
all
other
C's,
because
if
we
don't,
we
won't
be
able
to
to
provide
users
with
legacy
addresses,
and
we
should
be
able
to
do
that
because
there
are
still
wallets
that
don't
have
that
will
be
implemented
and
those
words
won't
be
able
to
fund
this
Q.
So
we
need
to
be
able
to
support
both
and
but
that's
it.
A
So,
basically,
you
cannot
use
the
same
private
key
from
your
wallet
for
the
legacy
address
and
for
the
new
address
or
because
probably
bit
country
cannot
handle
it
and
probably
to
separate
or
to
merge
them
together
us
when
you,
when
you're
spending
funds
from
from
a
legacy
address,
you
cannot
use
the
private
key
of
this
address
for
sending
it
with
in
a
separate
transaction
or
or
the
other
way
around
other.
You
cannot
mix
it
in
when
you
have
only
one
key
chain
and
you
have
basically
four
yeah
one
private
key.
A
A
Okay,
so
we
will
have
basically
a
similar
system
like
with
peas,
cue,
that
we
have
same
support
route,
but
we
have
a
different
branch
of
the
HP
wallet.
Exactly
yes
and
that's
how
well
it
is
already
supported
from
the
bit
countries,
because,
actually
that
we
have
a
few
tiny
issues
in
with
this
concept,
because
it
was
basically
bid
country
was
not
developed
for
a
multi
wallet
usage
and
one
issue.
What
ask
you
discovered?
It
happens.
A
A
B
A
You
can
send
this
to
your
Bitcoin
wallet
and
that's
basically
trust
a
transaction
where
you
send
some
Satoshi's
from
your
psq
wallet
to
your
Bitcoin
wallet
and
that
works
without
problem
and
the
problem.
What
desk
you
had
was
when
you
were
sending
some
Bitcoin
from
the
Bitcoin
wallet
to
the
psq
wallet
and
I'm,
not
sure
why,
in
that
direction,
we
have
the
problem,
the
other
garage
not
not
very
long
or
children,
so
we
have
to
close
into
it.
They
arm.
Maybe
I
should
get
lost
too
much
in
such
details.
I.
A
Anything
else
anybody
want
to
discuss
I
mean
the
last
topic.
What
I
wanted
to
cover
regarding
this
ideas
for
outgoing
trade
improvements?
Maybe
I
postponed
for
another
call
because
it
was
mainly
I
would
like
to
have
Thomas,
are
in
a
call
then
as
well,
because
he
might
be
a
potential
developer.
Who
could
work
on
this
and
to
not
repeat
it
and
to
talk
directly
with
him
might
make
more
sense,
and
it's
already
late
to
the
right
one
hour.
So
maybe
we
wrap
up
here.
If
you
don't
want
to
add
anything
or.