►
From YouTube: Meshery Development Meeting (August 24th, 2022)
Description
Meshery Development Meeting (August 17th, 2022)
Agenda:
[Debopriya] meshery.io redesign UX program proposal
[Calcote] LFX Fall'22
Meshery: User Interface Overhaul: State Management w/Apollo/GraphQL
Meshery: MeshMap - Visual Infrastructure as Code
Meshery: MeshMap - Visual Infrastructure as Code
TAG Network and Observability: Kubernetes ontology and subgraph module design
Meshery: Integration of Open Policy Agent (OPA) and Meshery
Service Mesh Performance: Adaptive Load Controller
Service Mesh Performance: Convergence of Network and Graph topologies:
[Subedy] Components to Doc queries
[Ashish] Enhance notification center with better event handling #91
[Cyrine] [Helm] Consolidate Helm chart sets Draft PR #6005
[Nishant] Fix problems with url, pattern handling in mesheryctl onboard #6085
[Alpha/Ruturaj] Implementation of “app import” and refactor of “app onboard” #6061
[Pankaj] Add loader to mesheryctl pattern apply sub-command #6079
[Harkirat] Docker Extension approach updates for #5782
[Harkirat] mesheryctl user friendly error message wrapped #5661
[Harkirat] Feedback for the mesheryctl pattern/filter/app #5928
Join the community at https://layer5.io/community
Find Meshery on:
GitHub: https://github.com/meshery
Twitter: https://twitter.com/mesheryio
LinkedIn: https://www.linkedin.com/showcase/meshery/
Docker Hub: https://hub.docker.com/u/layer5/
A
Okay
but
yeah
so
hello,
everyone
welcome
to
the
message
development
meeting
on
Wednesday
so
yeah.
Today
we
will
be
talking
about
the
development
people
like
so
yeah
as
a
convention.
We
we
just
get
introduced
to
some
newcomers,
join
joining
our
community
and
get
to
know
each
other.
You
know
very
well
so
that
this
relation
can
do
long,
so
I
think
pankaj
is
new
to
this
call,
or
probably
the
first
time
on
this
machine
development
meeting
right,
pankaj.
B
Sure,
thanks
Abhishek,
so
I
joined
the
community
for
some
10
15
days
ago,
because
I
wanted
to
take
part
in
the
Linux
Foundation
mentorship
program
under
layer
five
and
about
myself.
I
am
a
financial
student
of
Information
Technology
from
Mumbai,
India
and
I.
Have
some
experience
in
developing
with
go
and
kubernetes
from
my
internship
and
I
am
looking
forward
to
connect
with
more
of
you
over
many
years.
C
C
D
A
C
Hey
Abhishek,
I,
hope
you're
having
a
great
evening
I'm
others
and
I
recently
joined
a
message
just
like
15
years
ago,
and
so
far
I've
managed
to
get
like
two
small
pull
requests
done
and,
coincidentally,
just
after
I
joined
I
just
got
to
know
that
there
was
LFX
there's
an
LFX
opening
at
measuri
regarding
the
Redux
State
overall
project
and
the
team
is
pretty
nice
so
and
I'm.
Looking
I'm
looking
forward
to
take
up
more
issues
on
the
front-end
side
of
things,
so
yeah.
E
Am
devanshi
yadav
I
am
a
final
year
undergrad,
currently
studying
at
the
University
of
Mumbai,
so
I
am
also
from
Mumbai
like
like
monkeys
and
yeah.
I
have
also
been
you
know.
I
just
joined
the
community
last
week
and
I
have
some
experience
working
with
JavaScript
and
typescript
and
stuff
so
I
have
been
mostly,
you
know,
looking
to
contribute
to
the
front-end.
E
B
F
C
A
F
Yeah
so
recently
we
had
a
request
to
redesign
the
meshire
on
isi
and
we
reached
out,
and
we
had
an
overwhelming
amount
of
response
from
a
lot
of
UI
ux
designers
and,
like
the
I'm
gonna
talk
about
the
plan
that
we
have
for
this.
I
am
laying
down
a
certain
set
of
instructions
and
a
feature
requests
that
we
need
in
the
new
site.
So
the
main
purpose
for
this
redesign
is
to
include
hand
showcase.
F
All
of
the
new
machete
features
that
all
of
the
new
features
that
measuring
supports
to
list
a
few
like.
We
have
moved
from
service
mesh
Centric
to
Cloud
native
Centric.
Now
we
need
to
focus
more
on
communities
and
all
of
the
latest
Machinery
features
like
arm
support,
multi-cluster
support
application
management
and
so
on.
So
our
current
site,
the
this
is
like
the
major
purpose
of
the
redesign
and
I'm
laying
down
some
other
pointers,
which
I
will
share
with
the
participants
of
this
UI
ux
program.
F
So
the
main
like
for
the
next
few
weeks,
starting
from
next
week,
we
have
a
two-week
design
period
where
all
the
participants
they
are
open.
It's
open
for
everyone
to
start
designing
the
site
in
the
way
that
they
want
after
the
two
week
period,
we'll
have
initial
round
of
shortlisting
to
select
a
few
members
and
then
we'll
have
a
detailed
review
process
in
one
of
the
websites
called
and
after
that,
since,
like
we
want
the
best
possible
design,
we
won't
be
looking
at
like
one
particular
winner
sort
of
a
thing.
F
We
will
look
for
collaboration.
We
might
take
some
ideas
from
one's
One
Design,
some
ideas
from
the
other,
but
at
the
end
of
it
we
will
look
for
the
top
designs
and
look
for
collaboration
amongst
all
the
people,
and
we
have
had
a
couple
of
UI
ux
members
contributing
actively
in
the
community,
so
I
will
seek
out
to
them
for
helping
in
the
reviewing
process
and
everything.
F
D
D
So
so
that's
I
State
this
to
to
say
that
that's
in
large
part
why
the
mastery.io
website
needs
to
be
rearranged.
A
bit
there's
some
parts
of
the
site
that,
like
just
make
you
cringe
and
like
I,
are
like
particularly
painful
to
to
look
at
I,
have
I
have
nightmares
over
the
the
programs?
Page
I
think
it's
just
outdated,
terrible!
D
There's
videos
of
me
out
there,
which
you
know
I
watch
to
fall
asleep
at
night
and
but
the
other
parts
of
the
page
is
need
to
like
prominently,
articulate
the
value
of
meshery
out
of
context
of
a
service
mesh
and
so
yeah.
So
I'm
I've
got
my
fingers
crossed
for
the
million
dollar
pot
at
the
end,
for
whoever
whoever
gets
that
I
don't
know.
Of
course
the
P
those
people
have
to
have
their
names
registered
in
the
attendees
list.
Otherwise
we
just
won't
know
who
they
were
so
the
so
devopria.
D
Thanks
for
taking
that
on
I
know,
there
were
well
well
over
a
hundred,
maybe
150,
folks,
that
expressed
interest
signed
up
it's
too
bad
that
we
don't
have
150
sites
to
build.
We
do
have
like
how
many
sites
do
we,
how
many
web-based
applications
do
we
have
now?
We've
got
like
six
sites
or
something
mystery.
Ui
we've
got
the
play.meshery.io.
I
was
just
meeting
with
with
sivo
earlier
this
morning
on
potentially
something
to
do
there.
D
Anyway,
we've
got
a
lot
of
sites,
a
lot
of
front-end
things
going
on.
As
a
matter
of
fact,
if
Mr
Cambogia
is
on,
Yash
he's
been
routinely
telling
the
community
about
an
upcoming
rewrite
in
measure
UI,
so
the
next
release
of
meshri
there's
a
refactoring.
That's
in
flight
now,
there's
new
designs
that
are
need
to
be
made.
There
devil.
D
We
might
do
well
to
allude
to
measure
uis
like
new
ux
there,
as
as
a
separate
or
different
or
related
project,
that
the
uxrs
might
want
to
try
their
hand
at
it's
a
lot
more
difficult
because
it
requires
context
for
who
the
users
are
and
what
tasks
they're
trying
to
accomplish,
but
but
it's
very
much
needed
so
anyway,
on
the
second
topic
there
we
do
LFX
all
the
time
we
do,
internships
all
the
time
we
were
doing
Halifax
before
it
was
called
LFX,
I.
Guess:
I!
Guess
not
that
matters
matters
yeah.
D
We
do
internships
all
the
time,
Doom
direct
with
layer
five.
We
do
them
through.
Well,
there's
a
couple
of
G
soccers
on
the
call
there's
some
Halifax
there's
on
the
call
there's
some
direct
interns
on
the
call.
There's
some
folks
who've
interned
and
are
still
here
and
that's
very
much
so
the
goal
is
to
engage
with
people.
Long
term
is
to
build
out
a
great
set
of
capabilities.
D
There's
some
people
who
are
quite
excited
about
the
capabilities
that
we
have
there's
some
companies
that
you
might
recognize
that
are
quite
excited.
Just
had
a
partnership
meeting
with
Intel
last
week
on
the
mesh
mark
related
to
measury
measure,
implements
meshmark
I,
think
there's
an
internship
focused
on
service
mesh
performance
that
might
touch
on
this
another
partnership
meeting
with
hpe
they're
thinking
of
integrating
meshery
into
Green
Lake,
which
is
their
well.
It's
not
just
their
kubernetes,
offering
it's
well,
it's
it's
it's
their
their
core,
offering
from
hpe
right
now.
D
D
As
a
matter
of
fact,
if
you
show
up
and
you're
doing,
if
you're
consistently
showing
up
you're
doing
compelling
or
like
impactful
contributions,
which
don't
necessarily
need
to
be
code,
you'll
probably
get
one
shoved
in
front
of
your
face
like
we
just
want
to
support
those
that
are
engaged
and
doing
well,
we
want
to
not
only
make
software
that
changes
people's
lives.
We
want
to
have
internships
or
experiences
that
change
people's
lives,
and
that
is
actually
why
some
people
who
do
internships
are
still
here
because
their
their
lives
are
still
being
improved.
D
D
That's
it
for
me.
If
you
have
questions
about
internships
or
the
love
of
God,
ask
those
questions
publicly.
Please
I
can't
tell
you
how
many
hundreds
of
people
ask
and
inquire
and
suggest
that
they're
the
person
well
in
advance
of
them
having
even
gotten
familiar
with
the
projects
it
to
to
be
forthright
like
it
usually
takes
about
three
months
of
contributing
before
your
oh,
like
awarded
an
internship.
Internships
are,
are
earned.
D
D
G
Okay,
as
stated
in
the
previous
meetings,
I
have
been
working
on
to
making
all
the
components
and
its
Integrations
visible
on
the
measured
docs.
So
talking
about
my
current
progress,
I
have
been
able
to.
You
know
populate
these
fields
like
number
of
components:
service,
mesh,
total
number
of
versions
automatically.
G
So
so,
if
by
saying
automatically
I
mean
that
we
are
the
workflow
for
this
is
that
we
are
running
an
automated
workflow
on
measuri
on
different
adapters
repository
from
there
all
the
components
that
are
present
in
those
in
that
repository
we
are.
We
are
copying,
those
to
measuri
machine
repository
inside
the
docs
folder,
so
inside
the
docs
folder.
We
have
all
these
Integrations
and
all
their
like
all
the
Integrations
of
all
the
components
and
versions
stored
and
from
there
we
are
populating
the
dot.
G
The
measuring
documentation
website,
so
yeah,
so
currently,
I've
been
able
to
like
replicate
all
of
these
with
code,
the
Integrations
number
of
components,
total
number
of
versions,
I,
had
one
questions.
So
currently
I
have
one
question
that
how
are
these
latest
component
version
service
mesh
version
different
because.
H
H
H
Okay,
so
I
think,
okay,
so
the
file
name
and
the
okay.
So
this
can
you
find
an
issue
for
this,
but
this
this
shouldn't
be
the
case.
The
mesh
version
should
match
the
file
version
so
somewhere
in
the
code,
while
generating
these
components.
There
might
be
some
discrepancy
between
what,
basically,
what
version
we
are
assigning
to
the
so
yeah.
That's
not
fault
at
your
part,
so
you
can
remove
that
extra
column.
H
H
H
G
D
A
H
Okay,
so
just
to
get
a
give
a
little
bit
of
context,
we
have
this
notification
center
that
uzair
and
me
have
been
working
on
and
currently
whatever
operation
you
perform
any
service
misleaded
operation
you
perform
when
that
operation
completes
the
error
or
the
info
or
the
warning
or
whatever
the
result
of
that
operation
was,
is
streamed
back
to
this
notification
center
So.
Currently
there
are
a
few
bugs
around
this,
and
also
the
kind
of
thing
that
we
get
back
is
pretty
non-descriptive.
H
Indeed,
so,
like
you
can
see,
there
is
not
much
data
here
right
now.
Even
the
errors
do
not
carry
them
with
themselves,
the
probable
cause
and
the
first
remediation.
The
kind
of
mesh
kit
Paradigm
that
meshkit
editor
Paradigm.
That
you're
following
across
are
reposed
through
messages.
Adapters
and
user
would
be
talking
about
this
thing
after
me.
So
I
will
not.
You
know
talk
about
his
stuff.
H
I
will
talk
about
what
I've
been
working
on
okay,
so
this
notification
center
has
been
pretty
fragile,
so
I
have
very
basically
in
this
document
that
I'll
share
in
the
zoom
chat.
I
have
laid
out
a
rough
architecture,
diagram
of
how
it
would
look
like
after
my
few
few
of
my
PR,
so
it's
kind
of
a
position
slash,
you
can
say
proposal,
so
you
on
a
high
level
overview.
We
have
like
these
producer
and
these
producers
are
event
producers.
H
These
can
be
our
adapters,
they
produce
events
so,
for
instance,
when
I
install
a
service
mesh,
that's
a
service
mesh,
specific
operation,
the
adapter
is
the
event
producer
and
the
event
would
be
that
something
has
been
installed
or
something
has
been
removed,
XYZ,
etc,
etc,
or
this
can
be
meshity
itself.
So
when
we
perform
any
kubernetes
specific
operation,
we
do
not
have
a
kubernetes
adapter.
We
have
the
logic
that
does
all
the
kubernetes
stuff
is
inside
of
mesht,
so
in
that
case
there
would
be
a
routine
inside
of
meship,
which
would
be
the
producer.
H
So
we
have,
we
will
have
these
bunch
of
producers
and
they
will
still
have
a
buffer
in
which
it's
kind
of
a
circular
queue
in
which
we
will
store
events,
and
this
event
should
be
streamed
back
from
a
grpc
call
if
we
are
talking
to
adapter
or
a
normal
function,
call
if
it
is
within
the
same
process
or
if
these
are
measure
events
and
for
each
request
for
each
user
request
or,
as
I
can
say,
for
each
client.
So
in
this
case
each
of
these
tabs
one
can
be
thought
of.
H
As
you
know,
one
client,
so
each
client
will
form
a
graphical
subscription
with
the
message
server
and
for
each
of
that
graphql
subscription.
We
would
go
reach
out
to
all
the
producers
and,
first
of
all,
copy
all
their
events
from
the
buffer,
throw
it
back
to
the
client,
and
then
you
know,
keep
open
that
this
stream
to
keep
streaming
back
all
the
new
events
just
to
give
a
short
demo.
So,
for
instance,
like
right
now,
if
I
try
to
let's
say
if
I
go
back
on
mesh
map
and
yeah,
this
is
mishma.
H
I
I'll
try
to
deploy
this
test.
Namespace
and
if
I
deploy
this
okay,
so
this
design
deployed
successfully
was
the
result
of
the
normal
HTTP
call
and
the
other
the
note,
the
notification
that
you
saw
about
successfully
deployed
k8's
component
test
and
if
I
go
back
to
all
of
these
other
tabs.
This
same
notification
will
appear
so
all
of
the
clients
could
get
the
notification,
no
matter
whether
they
initiated
the
event
or
not,
and
if
a
new
client
connects
so
just
to
simulate
like
if
a
new
client
connects
to
the
same
HD
instance.
H
They
would
also
get
back
all
these
notifications,
so
the
last
I
guess
they
will
take
the
last
n
notifications
from
all
of
the
producers
and
then
it's
stream
them
back
here
and
yeah.
So
that's
kind
of
the
summary
of
what
I've
been
working
on
and
we
said
you
can
you
talk?
Can
you
take
it
from
here
about
your
enhancement
of
error
messages
and
all.
H
D
D
I
just
learned
something
new
that
that
you
can
show
that,
like
10
different
notifications
fit
on
the
screen
at
the
same
time,
I
didn't
bother
so
anyway,
it's
really
good
to
either
pop
up
pipe
up
in
this
meeting
or
in
the
discussion
forum
or
in
the
slack.
You
won't
be
you'll
only
ever
be
accidentally
ignored
that
you
won't
and
that
which
will
almost
never
happen.
It's
like
never
the
case
that
someone
doesn't
get
their
question
answered.
D
C
D
Yeah,
you
should
be
it
whether
you've
deployed
meshery
in
a
container
or
you've
built
Mastery
and
you're
running
it
locally.
If
you
invoke
an
action,
that's
performed
by
an
adapter
like
this
is
the
istio
adapter.
If
you
invoke
one
of
these
actions,
whether
the
operation
is
successfully
completed
or
not,
you
should
get
back
a
notification.
A
H
So
right
now,
okay,
so
right.
The
end
goal
is,
of
course,
a
graphical
subscription
right
now
that
slash
API
interest
endpoint,
which
you
know,
does
not
terminate
the
TCP
connection
and
keeps
streaming
back
the
events
using.
So
we
are
right
now
using
that,
but
that,
as
a
last
step,
that
would
be
like
a
very
easy
process
to
take
all
that
logic
and
to
have
a
schema
defined,
a
graphical
schema
and
then
basically
yeah.
That
would
be
very
less
code
change,
because
the
log
we
wouldn't
have
to
change
the
internal
Logic.
H
For
this
thing,
the
only
the
interface
would
change
and
we
can
have
a
new
subscription.
So
after
we
have
made
a
decision
on
the
back
end
side
of
these
things,
so
our
like
how
we
want
to
stream
back
events
from
from
this
end,
once
we
have
once
we.
Basically,
this
whole
thing
is
finalized
and
we
get
all
those
PR's
merged,
and
then
this
can
be
a
last
small
step
to
Port
it
to
a
graphql
subscription.
B
I
H
H
Okay,
so
right
now
also
yes,
each
of
the
clients
would
create
a
grpc
connection,
but
because,
if
10
events
are
produced,
we
need
to
send
those
10
events
to
all
n
number
of
consumers
that
we
have.
So
in
this
case
this
is
a
per
user
request.
Instance,
you
can
imagine
so
if,
for
each
request,
if
there
are
like
10
producers
with
all
of
those
10
producers,
one
grpc
connection
would
be
formed,
so
there
would
be.
If
there
are
n
producers
and
one
client
there
would
be
ngrpc
connect.
H
Now
it
doesn't
have
to
be
a
grpc
connection,
as
I
said
it,
it
can
be.
Basically
it
can
be
a
grpc
connection
or
a
normal
function
call
right
depending
upon
whether
we
are
remotely
calling
this
producer
or
whether
this
producer
logic
is
local
or
reside
within
the
same
process.
So,
depending
upon
that
this
can
be
a
different,
but
yes,
so
it
would
be
you
you're
talking
in
in
terms
of
adapters.
H
The
reason
I'm
also
talking
about
the
normal
function
call
is
right
now,
when
I
deployed
the
namespace
here
and
you
we
got
this
successfully
deployed
this
case
component.
This
was
a
function
called
and
not
a
grpc
call,
because
the
logic,
the
event
production
of
deploying
kubernetes
stuff
resides
within
the
adapt.
We
recite
within
the
message
server,
also,
if
I
deploy
an
application.
In
that
case
also
our
producer
would
be
Maxi
server.
H
So
in
that
case,
that
would
be
a
function
call
and
yes,
that
there
would
be
this
this
one,
a
normal
function
call
would
be.
You
can
imagine
like
a
blocking
call
that
returns
a
channel
and
stuff
is
streamed
back
on
that
channel
go
Channel
and
in
grpc
also
we
keep
calling
dot
receive,
and
essentially
it
just
keeps
streaming
back
stuff
on
the
channel,
but
it's
on
wire
instead
of
in
a
single
process,
so
yeah
there
would
be
n
function,
calls
to
end
producers
for
each
client.
H
H
Yeah
actually
yeah,
that
the
reason
that
it
was
giving
on
one
of
the
client
is
because
that
what
was
happening
is
that
there
was
no
buffer
or
there
was
no.
So
the
struct
I
have
created
a
PR
that
actually
creates
this
struct,
which
does
this
thing
a
little
bit
elegantly.
What
we
were
doing
before
this
is
that
if
you
have
like
one
client
the
what
this
but
the
okay,
so
here's
the
thing
at
the
producer
side,
if
n
clients
are
connected
to
these
producers,
there
has
to
be
ngrpc
calls
right.
H
But
in
previous
case,
even
if
there
are
ngrpc
calls
once
an
event
is
sent
to.
One
of
those
calls
like,
if
imagine
these
pipes
to
be
like
n
pipes
like
these,
if
one
event
is
sent
like
this,
it's
not
being
sent
in
N
minus
one
other
thing.
So
basically
we
were
using
go
channels
to
do
broadcasting
with
in
an
improper
way.
We
were
not
broadcasting,
we
were
actually
just
sending
to
one
random
client,
and
that
was
debug,
but
in
this
case
the
struct
that
I
have
introduced
here.
H
I
E
E
How
and
what
is
the
remedy
for
that,
and
also
I
have
moved
the
the
notification
messages
to
into
a
deduct
variable
so
that
it
is
easier
to
handle,
and
it
includes
the
changes
in
adapter
library
and
protoper
files,
and
there
are
more
things
need
to
be
done
like
whenever,
whenever
we
you
deploy
an
application
or
pattern
file,
the
you
just
receive
a
notification
like
it
is
deployed.
You
don't
receive
all
the
process
that
happens
like
when
you
deploy
a
pattern
file.
The
the
request
gets
sent
to
the
server.
Then
it
gets
verified
like
all.
E
We
have
a
pattern
engine,
so
it
verifies
that
the
pattern
file
is
correct
or
not,
so
these
events
should
also
be
streamed
to
the
client
that
he
he
should
know
that
your
request
is
being
processed
and
not
like
it
is
lost
or
is
conveyed,
so
this
this
also
needs
to
be
incorporated,
so
it
will
be
done
in
next
iteration
and-
and
the
next
thing
is
also
like.
There
are
two
kinds
of
events.
E
The
there
are
first
type
is
when
the
the
notifications
that
are
that
are
received
when
the
user
performs
some
operation
like
deployment
or
provision,
and
there
are-
and
the
second
type
is
when,
when
you
like,
when
you
have
kubernetes
cluster
connected
and
due
to
some
error
of
fault,
the
cluster
gets
disconnected.
So
it
is
not
an
event
initiated
by
user,
but
this
is
an
important
event
and
needs
to
be
notified
to
the
user
So.
E
Currently,
we
are
not
sending
that
event,
so
this
is
all
this
also
needs
to
be
streamed
and
probably
properly
segregated,
like
operation
and
notification.
So
here,
if
you
will
see
like
there
is
only
one
type
notification,
so
some
design
change
will
some
design
changes
will
be
required
here
as
well,
and
so
yes,
this
this
was.
E
This
was
the
this
was
what
I
was
working
on
and
I
was
I
was
also
working
on
catalog,
measuri,
catalog
and
I
did
some
changes
here
like
there
was
a
bug
with
pagination
in
the
application
application
patterns
and
filters.
I
have
made
changes
for
that
and
added
queries
for
graphql
queries
for
catalogs.
C
A
D
A
E
Yes,
so
like
I
I
was
working
on
like
fixing
some
problems
with
the
message:
CTL
on-board
command.
So
basically
there
were
two
problems.
The
first
one
was
that
recently
in
the
API
of
creating
a
application
earlier,
we
were
not
like
converting
an
application
yaml
to
a
pattern,
but
recently
we
started
converting
into
a
pattern
and
storing
it
and
in
mystery.
E
City,
before
deploying
we
used
to,
we
still
are
like
every
time
we
get
and
get
the
application
from
the
back
end
and
then
convert
again
to
a
pattern.
So
it
is
so
because
we
already
have
started
converting
it
to
patterns
and-
and
we
are
again
converting
into
a
pattern-
it
is
giving
errors,
so
I've
removed
that
logic
from
here.
C
D
E
D
D
E
I
think
in
this
case,
like
in
this
gate,
walker
like
as
far
as
I-
remember,
we
don't
use
GitHub
base.
We
blowing
the
complete
repository
into
a
temporary
folder
and
then
we
Traverse
the
folder
so
we're
not
using
itabies.
There
is
another
Walker
GitHub,
Walker
I,
think
that
uses
a
GitHub
apis,
but
this
doesn't.
C
E
Yeah
yeah,
it
was
like
skipping
some
files,
so
yeah
that
was
used,
okay
and
so
yeah.
This
is
what
I
wanted
to
discuss,
that
it
was
not
working
for
files
because
in
this
root
path,
you
were
entering
this
complete
path
of
the
file,
and
this
should
be
a
directory,
not
a
file,
so
it
was
not
working
but
the
current
we
have
modified
it.
E
A
I
A
A
C
B
So
this
is
the
pr
I
created,
so
the
issue
was
that
no
progress
or
loading
or
any
context
is
shown
to
the
user
when
measuring
CTL
pattern
applies
used
or
there
are
many
such
apply
commands
or
deploy
commands
that
take
a
lot
of
time
and
there
is
no
feedback
given
to
the
user.
As
of
what
is
happening,
so
this
issue
was
to
add
a
loader
and
even
add
the
pattern
name
to
it.
So
I
created
a
pull
request
for
that,
wherein
I
can
simply
show
this
demo.
B
H
By
the
way,
before
this
PR
after
this
basically
synchronously
waiting
on
the
terminal,
the
log
had
this
okay
pattern
successfully
applied
and
along
with
that,
if
the
patent
has
had
n
components,
it
used
to
say,
like
book
info
deployed
this,
this
component
deployed
this
component
deployed.
So
are
those
logs
being
removed
from
somewhere
or
did
when
you
deployed
that
pattern
so
yeah,
can
you
promote
that?
No.
B
B
B
But
while
the
test
case
is
running,
it
is
actually
erasing
another
line
which
is
part
of
the
output
of
the
test
case.
So
that
is
not
supposed
to
happen
for,
for
that,
I
should
actually
try
and
shift
up
the
cursor
or
down
so
I'm,
trying
to
figure
out
which
kind
of
ANSI
escape
sequence,
I
should
use
or
which
combinations
that
the
output
should
match
with
the
golden
file.
B
That
is
there
and
also
earlier,
had
asked
me
if
there
are
some
other
commands
as
well,
which
I
can
identify,
which
would
use
this
loader
functionality,
and
so
performance
apply
for
running
a
performance
test
for
deploying
a
mesh
and
app
on
board
app
off-board,
as
well
as
the
voice
and
filter
deployed.
These
are
the
different
commands
that
I
have
identified,
which
can
make
use
of
such
a
logo.
H
C
H
C
B
H
Okay,
okay,
that
actually
we
I
know
by
that
why
you
would
get
an
error
that
gamble,
Dot
and
Marshall
is
Marshall
to
work,
that
we
have
a
function.
Recursive
map
string
interface,
something
like
that.
You
can
find
that
function.
If
you
use
that
function
that
basically
goes
on
and
before
yaml
dot
Marshall.
So
it
has.
H
It
is
logic
built
upon
yaml,
dot,
Marshall,
so
the
bugs
yamlor
and
Marshall,
so
the
bugs
you
are
getting
in
yaml,
Dot
and
Marshall
won't
measure
if
you
use
that
function
after
the
call
I'll
tell
you
where
the
function
is,
but
if
you
directly
call
into
that
function,
you
can
actually
basically
that
function
was
built
for
this
thing,
that
you
can
unmarshall
anything
to
a
patent
structure.
I'll
point
you
to
that
function
and
instead
of
the
ml
dot
unmaster,
you
can
call
that
function.
C
H
Then
the
SEC,
the
second
question
was
the
function
that
you're
calling
for
loading
is
that
generic
enough,
that
other
command
would
just
call
into
that
or
there
are
some.
What
are
the
arguments
that
that
loading
function
take
or
whatever
the
struct
it
has
how
generic
it
is
to
be
used
for
other
commands.
B
So
a
status
can
be
added
to
the
struct
that
can
take
care
if
the
response
is
not
200,
so
there
will
be
a
different,
there
will
be
a
tick
mark
or
there
will
be
a
cross
or
some
something
else
that
will
be
shown
in
place
of
the
loader,
so
that
can
be
added
to
the
struct.
Apart
from
that
I
think
for
now,
if
a
loader
is
to
be
added,
then
a
message
will
suffice
for
so
that
the
log
message
can
be
passed
into
the
spinner
and
it
will
be
simply
loaded.
D
D
That
would
be
all
the
better
does.
It
doesn't
necessarily
have
to
be,
but
it's
not
nice,
just
another
food
for
thought,
there's
something
you've,
probably
already
thought
of,
but
just
I'll
press
upon
the
point
and
then
I
thought
that
was
one
other
thing.
No
great
yeah,
like
any
just
making
a
great
ux
revolves
a
lot
around
consistency,
and
so
thank
you
so
much
for
taking
a
look
across
the
rest
of
the
commands
for
where
else
there
might.
D
D
This
progress
indicator
this
loader.
This
progress,
like
actually
showing
you
a
percentage
or
something
like
from
zero
to
I,
don't
know
that
that
you
can
know
in
advance
what
100
is
it's
actually
in
part
to
my
comment
about
how
timely
disenhancement
is
of
what
uzair
and
Ashish
were
showing
is
is
very
relevant.
D
So
you
know
just
be
patient
and
wait,
and
so
so
to
the
extent
that
the
messaging
framework,
like
the
the
portion
of
measure
between
meshery
server
and
its
adapters,
to
the
extent
that
their
their
intermediate
messages
as
things
are
being
provisioned,
because
sometimes
it's
not
just
a
simple.
Sometimes
it's
like
a
multi-work.
D
But
but
it's
a
good
kind
of
point
of
discussion
inside
of
what
Ashish
are
doing
is
a
good
question
to
be
asking
ourselves.
Are
we
designing
this
system
architecting
the
system
in
such
a
way
that
eventually
we
might
be
able
to
to
actually
let
the
user
know
hey
if
you've
been
watching
this
spinner
for
five
minutes?
Yes,
you
should
keep
watching
it
for
another
10,
because
it's
only
at
the
25
more
because
they
might
be
watching
it
for
15
going
well,
I,
guess,
I!
D
H
By
the
way
bunker,
just
one
more
thing,
I
have
added
the
link
to
the
notification
center
design
spec
in
the
chat
you
can
go,
go
through
that.
Go
look
at
that
and
if
you
dig
into
the
code,
just
two
keywords
for
you:
inside
of
mystery
server,
look
at
slash,
API
events
instead
of
adapters,
look
for
stream
events
functions
and
while
things
would
make
sense,
if
you
will
look
at
the
code.
C
A
C
When
we
used
to,
like
the
user,
have
the
option
unchecked
and
we
were
not
able
to
show
the
containers,
but
I
came
up
with
the
approach
like
with
this.
Now
we
can
also
get
the
the
containers
if
the
option
is
unchecked,
so
this
error
need
a
little
formatting
and
it's
good
to
go,
and
then
another
update
I
have
is
regarding
the
user
trending
machine.
So
it's
wrapped
for
now
and
we
are
now
working
on
the
daily
test.
C
I
think
Nishan
would
be
also
working
with
me
in
the
unit
test
for
this
particulars
of
commands
as
well,
and
the
last
is
something
related
to
very
easy.
So
the
last
thing
is
like
I'm
asking
feedback
for
the
for
this
stuff
regarding
yeah.
So
when
we
have
to
view
a
pattern
based
on
this,
like
ID,
we
are
not
able
to
use
the
pattern.
Id
is
too
long.
C
So
when
we
use
the
pattern
this
and
we
get
a
particular
idling,
so
we
are
not
able
to
give
it
so
I
need
some
feedback
so
over
this
yeah,
so
Dockery
or
something
like
this
like
they
have
also
have
this
long
ID.
But
when
we
have
to
inspect
we
can
by
the
by
the
short
ride,
we
cannot
change
that
second
energy.
So
this
is
all
on
my
side
under
a
minute.
D
So
each
profile
might
they
might
have
a
bunch
of
profiles.
Each
profile
might
have
a
bunch
of
results
and
there's
been
gosh
a
lot
of
thought
that
was
put
into
the
ux
around
what
if
people
you
know
like
use
cases
like
what,
if
people
have
the
same
profile,
name
performance
profile,
name
twice,
or
how
do
we
display
all
these
test
results
but
not
overwhelm
them?
And
if
you,
if
you
play
with
that
command
for
a
little
while
it
might,
you
might
have
some
Reflections
on
your
thinking
around
the.
A
D
It
might
be
boring
because
you've
been
there
done
that,
but
for
others
of
you,
if
it
is
boring
and
you've
been
there
done
that,
then
you
might
want
to
come
and
answer
quite
field.
Some
questions
from
from
other
learners,
so
we'll
use
this
same
time.
I
expect
it'll,
probably
be
about
a
two
hour.
Workshop
it'll
be
live
streamed,
but
recorded
in
case
you
miss
it
and
you
want
to
go
back
and
board
yourself
again.
A
Nice
yeah,
so
so
the
next
yeah
next
week
there
would
be
at
the
workshop
again
in
the
timing
of
this
meeting,
so
yeah
good
good.
It's
a
wrap
right.
Okay,
so
yeah
the
future
like
better
than
like
next
time.
Yes,
so
yeah,
we
are
out
of
time.
So,
okay,
we
will
discuss
these
all
in
the
next
next
development
call
meeting.
So
since
then
till
then
you
can,
you
know,
join
the
newcomers
meeting
in
the
next
machine,
build
and
release
meeting.