►
From YouTube: Meshery Development Meeting (March 23rd, 2022)
Description
Meshery Development Meeting - March 23rd, 2022
Agenda -
- RJSF custom components [#5225]
- Meshery extension points (user_prefs)
- Update on implementation of Meshery Adapter for Cilium
Promote to beta.
- `mesheryctl system config eks`
Currently, using system calls to `aws eks`. Desire to migrate to aws go SDK and packages.
- Update on CodeCov CI coverage of unit tests (result and further works)
- Progress update on meshery-perf.
Volunteers (golang): on the ground floor of an adapter.
- Meshery event streamer implementation [#5220]
- Discussing UI bugs
- Progress update: Refactoring Meshery UI
- Discussion on Compatibility Matrix Summary
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
Oh,
if
you
don't
have
a
link
to
the
meeting
minutes,
ashish
tiwari
will
post
a
link
in
the
chat.
Let
me
share
them
here,
if
you're
on
the
call,
welcome
and
flip
on
your
webcams.
If
you
have
them,
and
you
at
least
have
one
article
of
clothing
on
and
it'll
be
nice
to
connect
with.
You
also
drop
your
name
into
the
meeting
minutes.
If
you
would
mesherie
is
a
cncf
project.
A
If
you
have
topics
today,
please
drop
them
in
the
format.
Goes
a
little
bit
like
this
that
you
would
put
your
name
here
as
the
individual
that
wants
to
talk
about
it
and
then
the
thing
you
want
to
talk
about,
and
so
we've
got
a
number
of
them
lined
up.
But
if
you
have
one,
please
drop
it
in.
If
we
don't
happen
to
get
to
it
this
week.
A
That
means
that
we'll
it'll
be
queued
up
for
the
next
time
that
we
meet
and
we
meet
a
few
different
times
a
week
for
meshary
and
it's
sibling
projects
so
there's
a
calendar
out
there.
I
think
some
of
this
information
that
I'm
giving
to
you
now
is
old
news
for
the
vast
majority
of
you.
There
might
be
a
lucky
winner
today
who
might
be
the
recipient
of
our
well,
our
our
what
geez,
I
can't
think
of
something
anyway.
Harshit.
A
B
A
A
A
C
A
So
there's
probably
some
background
noise
for
abby,
but
so,
as
he
goes
to
speak
to
this
I'll
fill
in
just
a
little
bit
of
background
for
some
of
you
that
might
not
be
overly
aware
of
well
of
the
the
concept
of
extensibility
within
measuring
mastery
is
a
management
plan.
It's
piece
of
management
software
that
manages
kubernetes
clusters
service
meshes
a
variety
of
things
and
in
trying
to
cover
all
the
bases.
A
There's
explicit
consideration
made
to
ensure
that
people
can
plug
in
new
capabilities
into
meshes,
so
it
has
a
plug-in,
an
extensibility
model
that
goes
not
only
for
its
back-end
capabilities,
but
also
mesherie's
user
interface,
and
so
there
is
a
set
of
document,
a
whole
area
about
extensibility
that
you
can
peruse,
and
so
abby
was
just
expanding.
One
of
the
extension
points
so
bobby
if
you're
back.
C
C
If
I
have
I
should
have
running,
I
I'm
running
kubernetes,
so
I
would
get
like
a
lot
of
other
sub
components
and
also
this
is
we
are
using
others
forms
for
dynamically,
generating
forms
that
is
to
be
rendered
on
the
canvas
that
a
user
can
fill
in.
You
know
in
place
of
the
yammers
to
get
to
the
configuration
they
want
so
because
we
are
using.
C
These
are
just
forms
inside
our
extension,
so
we
have
to
provide
the
support
for
the
custom
components
inside
extension,
because
the
two
forms
may
or
may
not
look
very
similar
because
it
may
be
the
need
actually
right.
So
this
form
in
the
ui
here
may
not
look
the
same
as
we
are
using
in
the
matchup.
A
No
one
has
heard
of
that
before,
and
I
don't
know
that
you
should
it's
just
another
of
the
1
billion,
what
npm
packages
available
sure,
but
it
becomes
very
it
or
something
like
it
is
really
helpful
for
measuring,
as
as
mescheri
goes
to
manage,
like
all
kinds
of
systems
I
mean
managing
kubernetes
would
be
just
you
know
enough.
Actually,
if
you,
if
you
were
to
go,
look
at
it,
managing
istio
would
be
just
enough.
Istio
is
a
a
bigger
beast
than
kubernetes.
A
A
To
do
it,
programmatically,
and
so
this
particular
library
assists
with
that.
It's
not
the
only
one
of
its
kind
out
there,
but
it
allows
all
of
you
and
you
know
meshary
itself,
to
describe
you,
know,
attributes
of
the
infrastructure
and
then
to
do
so
in
json
and
have
that
json
converted
into
react
components
so
from
the
left-hand
side
to
the
right-hand
side,
and
so.
A
A
A
D
D
So
basically,
this
extension
point
gets
plugged
in
dynamically
when
you're
running,
measuring
and
depend,
and
it
depends
upon
the
provider
which
you
use
while
signing
in
to
the
machine
ui.
So
right
now
I
will
be
talking
about
the
user
preferences,
which
is
basically
the
preferences
which
you
use
for
your
particular
measure
deployment.
So
I
will
be
sharing
my
screen
and
like
explaining
you
that
and
also
what
my
that
I
also
share.
What
I
did
in
my
pr.
D
So
this
is
how
your
machine
ui
would
look
like
when
you
log
in
after
logging
in
and
selecting
measurey
as
you
provide
remote
provider.
This
is
the
preferences
tab
and
in
there
this
is
one
of
this
remote
provider.
Tab
is
one
of
the
remote
pluggable
extension
point
which
plugs
in
when
you
run
your
measure
deployment.
So
what
I
initially
we
just
had
one
settings
that
was
start
mesh
map
on
zoom.
So
what
I
did
was
I
expanded
it
and
added
two
more
setting
that
is
snap
to
grid
and
hydrate.
D
Now
what
these
settings
basically
are
are
related
to
mesh
mark.
So
when
you
get
access
to
measurement,
you
understand
that
in
the
canvas
we
have
when
we
are
deploying
our
designs,
we
we
might
want
the
canvas
to
look
like
as
we
want.
So
if
we
go
over
here
here
are
those
settings
and
basically
I
replicated
the
setting
into
the
preferences
so
that
user
can
have
two
ways
of
turning
on
their
switches.
D
D
E
A
A
There's
it's
time
to
retire,
another
adapter
that
that
one
is
the
octarine
adapter
which
most
of
you
probably
never
heard
of,
but
that
service
mesh
was
acquired
by
vmware
and
they're.
Changing
around
they've
got
a
whole
10
zoo
platform,
and
so
anyway,
point
is:
there's
a
new
service
mesh
coming
forth,
psyllium,
surface
mesh,
and
so
naturally
mastery
should
interact
with
it
and
help
support
it.
A
D
Yeah,
so
clm
adapter
is
basically
a
new
adapter,
which
we
will
be
using
in
actually
so
right
now
there
is
a
lot
of
work
on
it,
so
that
it
remains
consistent
with
other
adapters
which
we
are
already
using.
So
there
are
many
places
of
work
like
if
anyone
is
interested
in
contributing
to
a
serium.
Adapter
then
feel
free
to
check
out
the
open
issues
and
maybe
try
to
adapt,
and
let
us
know
that
way,
it's
failing
so
that
we
can
improve
it
further.
D
So
recently
I
also
made
a
pr,
so
I
would
be
sharing
my
screen
to
show
what
I
did.
I
think
I
mean
I'm
descending
only
so
what
bi
is
basically
about.
First
of
all,
it
writes
a
workflow
which
that
workflow
is
about
statistically
updating
our
statistical,
statistically
updating
our
mesh
symmetry
serium
adapters,
repo
with
the
custom
resources
cm,
has
on
its
own
repository.
D
So
that
is
one
of
the
change.
Second
changes
about
refactoring
the
code
for
dynamically
and
statistically
generation
of
custom
resources
like
this
I
did.
This
was
required
just
because
so
so
the
code
for
each
adapter
is
to
remain
consistent
and
it
is
easier
to
maintain,
and
the
third
was
basically
I'm
updating
the
latest
latest
resources
of
serium
and
serium
service
mesh
into
our
machine
repository.
So
I
did
the
three
changes
and
I've
asked
for
the
groups
someone
can
distribute.
D
D
A
A
It's
not
actually
clear
to
me
where,
if
it's
the
ui
or
the
adapter,
that
needs
to
be
where
code
needs
to
be
changed,
but
so
to
demonstrate
this,
I'm
running
mesherie
and
it's
got
a
couple
of
adapters
connected.
So
one
of
them
is
the
psyllium
service
mesh
adapter.
That
pronov
was
just
showing
so
I've
got
it
running
locally
and
it's
connected
when
it
connects.
It
will
register
the
components
that
pranav
was
just
showing
so
that
that
pull
request
means
that
the
adapter
will
generate
components
for
meshri
to
manage
psyllium.
A
So
if
I
go
over
to
this
adapter
and
navigate
to
it
in
the
ui,
what
we
would
be
expecting
is
for
the
ui
to
look
like
this
like
it,
it
would
say:
okay!
Well,
if
you
want
to
manage
a
psyllium
service
mesh,
here's
the
various
operations
that
you
can
just
invoke
directly,
and
so,
when
you
navigate
to
psyllium,
you
should
see
a
very
similar
thing.
There
are
operations
written
into
the
adapter,
and
so
utzer
is
here
now
as
well,
and
do
you
guys
know
where
the
issue?
The
final
issue
remains?
D
I
tried
debugging.
What's
the
issue
I
was
doing.
I
was
doing
the
same
thing
just
right
now,
so
I
encountered
that
it
is
in
the.
I
think
it
is
in
the
ui,
because
it
is
complaining
about
add-ons
when
we
try
to
filter
the
add-ons
in
the
ui,
so
the
add-ons
field
is
empty,
so
the
ui
is
complaining
that
it
cannot
iterate
on
the
empty
array
kind
of
thing.
So
this
error
is
coming
till
now
what
I've
understood
I
will
update
more
regarding
it.
A
A
Okay,
let's
do
that.
Well,
no
or
I
mean
it
could.
Rather
it's
usually
the
combination
of
the
two
and
usually,
usually
you
can
address
it
on
one
side
or
the
next,
or
both
it's
generally
good,
to
address
it
on
both
because
in
concept
the
ui,
if
there's
no
add-on
present
or
if
there
was
a
a
bad
response
in
the
add-ons
section
and
by
the
way
I'll
try
to
go
to
the
add-on
section
for
istio
to
show
you,
I
think
I
I
hurt
that
adapter.
A
The
ui
should
be
resilient
as
well
like,
even
if
there
are
changes
to
go
into
the
adapter,
the
ui
should
still
be
resilient,
so
yeah
but
good.
So
that's
it.
That's
a
fresh
adapter!
I
think
after
that
portion
is
fixed.
The
adapter
will
be
ready
for
beta
state.
So
once
that's
fixed
it's
time
to
promote
to
beta.
F
F
F
Yeah
this
solution
using
the
the
aws
golem
sd,
so
this
describe
cluster
it
kind
of
gives
us
information
about
the
cluster.
The
information
in
three
times
looks
like
this
is
what
the
command
returns.
The
aws
eks
update
config.
If
this
is
the
structure
it
returns,
and
it
writes
it
to
the
config
file,
provided
here.
F
F
Yeah
this
one,
we
can
populate
this
with
struct
you're,
using
the
output
of
the
code
snippet
or
we
can
use
this
one,
the
rest
api
that
config,
so
we
are
still
kind
of
debating
on
which
one
to
use
we
haven't
made
any
decisions
here
and
then
that
is
what
we
shall
put
here,
because
the
original
implementation,
after
making
this
after
making
this
cli
call
it
writes
the
conf.
It
adds
the
configuration
to
the
utils.config
part,
and
then
it
calls
this
set
token
function.
F
F
F
E
E
So
if
we,
if
you
can
get
this
content
from
eks
cluster
in
rest.config
format-
and
there
is
a
way
to
actually
write
it
into
a
file,
so
let
me
share
this.
I'm
actually
trying
to
do
this
part
of
the
coding
now
because
from
the
code
snippet
what
he
has
shared
now
right,
so
we
can
return
the
token,
as
well
as
the
ca
data
in
a
rest.config
format
as
well.
E
So
from
the
region
as
well
as
the
cluster
name,
we
can
get
the
from
the
cluster.
You
can
get
the
rest
conflict,
so
we
can
get
the
tokens
as
well
as
this
is
exactly
the
stack
workflow
for
snippet,
so
we
can
build
this
rest
config
and
give
it
so
here
we
can
specify
what
all
things
are
important
and
we
can
return
that
and
then,
by
using.
E
A
Yeah
hussain
refresh
my
memory,
if
you
would,
when
we-
and
this
is
like
it's
specific
to
this,
question-
is
specific
to
eks
and
what
we're
doing,
but
it's
also
a
general
question
for
what
we're
doing
when
we
like
manipulate
cube,
config
for
mini
cube
or
for
for
any
kubernetes
cluster
that
doesn't
I'm
sorry
for
any
user
whose
cube
config
doesn't
contain
the
certificate
data
refresh.
My
memory
are
we,
as
we
go
to
extract
that
information?
A
Good,
thank
you
for
that
yeah.
We
just
I'm
making
a
note
that
we
need
a
warning,
slash,
informational
note
in
our
docs
that
that
lets
people
know
that
that
behavior
goes
on
just
so
that
they're
aware
like
I.
There
are
other
programs
like
docker
and
probably
I'm
rehashing
something
in
the
past,
but
I
had
wondered
about
this
last
week
and
just
and
so
okay,
so
we
are
rewriting
the
file
all
right.
A
A
You
know
it's
co-managing
that
file
along
with
other
systems,
and
so
I
think,
just
I'm
just
kind
of
talking
out
loud
that,
like
that,
makes
me
feel
better
with
mesherie
as
a
management
plane
as
a
kubernetes
manager
as
a
service
mesh
manager,
that
it
too
would
rewrite
portions
of
the
file
like
okay
but
yeah.
Having
a
note,
it's
actually
a
request
for
volunteer
to
add
a
note
to
the
meschery
docs
I'll
go
make
an
issue.
A
Oh
and
then
yeah,
so
just
some
commentary
on
what
hussein
and
secretary
we're
showing
is.
What's
great,
I
mean
like
I,
I
think,
you're
gonna
get
it
like
you've.
You
know
some
of
it
is
deciphering
the
data
that
you
get
back
from
aws.
You
know
the
snippet
that
you
were
sharing
some
of
the
husseina
and
some
others
that
had
worked
on
this
system.
A
Config
area
had
already
given
prior
thought
to
the
user
experience
that
should
occur
as
they're
going
through
as
they're
being
asked
a
few
different
questions
about
their
eks
connection,
and
so
it's
nice
because
there's
an
existing
foundation
to
so.
This
effort
is
just
a
bit
of
refactoring
to
really
to
really
make
it
a
lot
smoother
a
lot
well
to
eliminate
a
dependency.
This
is.
Hussein
has
been
on
this
been
carrying
this
torch
for
a
while,
like
which
is
to
every
time
that
a
dependency
is
eliminated.
F
A
F
F
B
E
Up
one
of
the
rest
api
calls
that
we
can
use
like
set.
There
are
some
handlers
where
we
can
directly
set
this
context
using
those
rest
apis.
So
if
you
don't
want
to
go
to
the
exact
cube,
config
file
path,
we
can
remove
this
set
context
set
to
ken
part
and
directly,
maybe
call
those
rest
apis
with
whatever
config
we
get
from
the
tks
cluster.
So
that
is
another
part.
A
Yeah
yeah,
I
am
from
yeah.
I
would,
I
think,
I'm
gonna
point
to
what
hussain
has
said,
which
is,
which
is
that
we
probably
are
doing
really
well
to
incorporate
what
you
guys
are
doing
and
improving
upon
the
existing
experience,
while
leaving
aside
the
it
like
that
that
question
that
I
was
asking
or
what
I
had
suggested
earlier,
the
problem
with
what
I'd
suggest
the
good
thing
about
what
I'm
suggesting
I
had.
The
question
I
just
asked
was
that
hey
look
like
we
don't
step
on
anyone's
file
and
they
don't
get
upset
that,
like.
A
A
A
Well
then,
you're
going
to
have
to
ask
them
all
the
same
questions
again,
because
you
need
to
re-establish
the
you
need
to
get
that
info
again.
Unless
you
take
that
info
the
first
time
you
got
it
and
persist
it
somewhere
separately
and
then
on
the
fly.
You
merge,
those
you
know
their
cube
config
with
the
certificate
data
that
you
got
and
it's
like
like
that's
just
I
I
want
to
retract
anything
that
I've
said
there.
A
A
A
A
G
And
it's
yeah,
so
for
a
beef,
my
task
was
to
increase
the
coverage
of
the
unit
test
on
cocoa
for
measuring
earlier.
It
was
like
the
unit
test
worked
on
the
measuring
ctl
and
some
of
the
internal
folders,
but
not
for
all
like
this
one
like
it
got.
It
worked
by
the
partners
being
inserted,
but
only
these
folders,
like
the
internal
and
measuring
sql,
are
what
you
can
say.
The
results
are
being
available
and
not
available
for
these
folders.
G
So
my
task
was
to
increase
to
make
the
code
curve
know
that
these
files,
these
folders
all
also
be
need
to
be
included,
and
what
I
did
is
that
I
also
had
to
write
some
of
the
unique
text
for
those
folders
and
the
result
this
year.
So
all
of
the
internal
folders
are
being
included
now
and
it's
now
live.
G
So
the
my
future
works
for
to
increase
the
coverage
of
this
test
and
to
write
more
unit
tests
for
these
folders,
because
some
of
the
tests
are
written
for
dummy,
because
only
then
the
code
curve
will
know
that
this
folder
is
there
just
by
writing
one
unit
test
alone,
and
it
will
be
done
in
the
future
and
yeah.
I
also
need
to
say
this:
if
there
are
any
volunteers
who
want
to
write
unit
test
for
measuring
project,
you
are
welcome
to
do
so
so
yeah.
This
topic
is
just
short,
though,.
G
A
A
A
So
we're
on
the
very
tail
end
of
dot,
six
zero
and
getting
it
out,
but
as
we
look
toward
capabilities
of
the
forthcoming,
the
next
release,
the
next
version
we're
going
to
take
and
split
nighthawk
out
of
mesherie
proper
out
of
the
mesherie
server
container
image
into
its
own
image
and
making
it
basically
its
own
adapter,
and
so
there's
groundwork.
That's
been
laid
to
do
so
I'll,
say
so
similar
to
aditya's
call
to
call
for
volunteers.
A
One
of
the
primary
points
excuse
me
is
well
to
facilitate
what
horizontal
performance,
testing
horizontally,
distributed,
performance,
testing
or
or
to
create
a
barrage
of
load
generators
that
can
generate
load
against
different
endpoints
all
at
the
same
time
and
then
sync
all
of
those
results
back
to
mastery
server.
So
that's
what
that's
about
there's
an
open
repo
is
quite
young.
I
think
ashish
is
going
to
make
sure
that
we
have
a
functional,
build
because
ashish,
I
think,
as
of
pr
six
and
five,
I
don't
think
that
the
adapter
is
building
successfully.
C
A
H
H
Yeah,
so
the
the
pr
itself
is
about
having
a
watch
flag
for
the
mesh
remove
command
and
it's
pretty
much
the
same
pr
as
aditya
had
made,
but
I'm
addressing
an
issue
which
he
had
also
faced
back
here,
and
there
was
a
discussion
that
all
the
commands
that
use
event
streamers
are
getting
stuck
at
verifying
operation.
So
what
I
figured
out
the
problem
was:
is
that
the
adapters
so
just
to
give
a
little
context.
The
way
this
works
is
an
adapter
sends
out
the
adapter
sends
out
like
the
adapter.
H
The
machine
server
sends
out
a
request
to
the
adapter,
to
let's
say,
install
a
service
mesh
or
remove
the
service
mesh,
and
then
it
the
adapter
then
sends
back
like
an
a
server
side
event
to
for
the
server
to
broadcast
to
the
clients
who
want
who
are
waiting
for
events.
So
the
problem
here
is
that
so,
for
example,
right
now
here
I've
you
know
just
have
the
these
two
requests
for
events,
but
if
I,
if
I
do
any
operations
here,
let's
say
I
think
I
have
yes,
I
just
removed
it.
H
Yeah,
I
think
this
is
so.
The
point
is
that
yeah,
so
the
we
can
see
that
the
ui
got
the
callback,
but
the
other
two
didn't
so.
The
problem
here
is
that
it's
the
the
problem
is
that
the
measure
the
clients
are
list
directly
listening
to
other
machine
server
code
that
handles
this
is
directly
listening
to
the
the
events
from
the
adapter
for
each
client.
H
The
problem
with
that
is
that
this
is
resulting
in
us
calling
the
receive
function
from
grpc
in
multiple
different
go
routines,
so
the
code
for
it
is
here.
So
this
is
the
code
which
is
being
run
for
every
single
client
that
wants
to
listen
for
events,
and
that,
like
eventually
goes
and
calls,
I
think
here
receive
message
yeah.
H
So
this
function,
the
primestream.c
message
function
and
the
documentation
actually
also
says
that
this
it's
not
safe
to
call
receive
message
for
the
same
stream
in
different
go
routines,
so
that
is
what's
resulting
in
only
one
of
the
clients
getting
the
event.
So
that
is
why
a
lot
of
the
the
commands
from
messy
ctl
or
otherwise,
which
are
listening
for
server
side
events
just
don't
receive
the
events.
H
What
solution
I
was
proposing
is
we
have
some
sort
of
code,
some
common,
let's
say
a
go
routine,
which
just
listens
to
events
from
the
adapters
and
just
pushes
those
events
to
each
client.
Let's
say
through
a
channel
to
each
each
client,
instead
of
the
clients
themselves
directly
listening
to
the
adapters.
A
A
A
A
It's
called
gnats
and
so
like
this
distributed
systems,
concern
about
messages
that
who's
listening
to
the
messages
who
gets
sent
them.
How
many
times
they
can
get
sent
received
whether
or
not
they're
acknowledged
and
like
all
of
that
like
it
starts
to
one
thread,
turns
into
many,
and
it
starts
to
get
hairy
quickly
and
so.
A
All
of
what
I
just
said
doesn't
really
help
you
necessarily
it's.
I
said:
okay,
go
think
about
it.
From
these
perspectives.
Fine,
hey,
go,
look
up
an
old
pr
with
the
term
go
search
for
broadcast
and
you'll,
find
some
examples
and
then
even
then
you'll
come
back
and
then
go
study
gnats.
Those
are
kind
of
the
three
things
that
I
said
and
it's
like.
Okay,
three
weeks
later,
you
come
back
with
like
having
thought
through
all
that
and
learned
that
stuff
and.
A
If
gnats
isn't
too
heavy
of
a
lift,
it's
the
one
to
go
use
or
like
it's
the
tech
to
lead
into
and
for
us
to
really
consider
like
the
the
its
potential
use
or
one
of
its
projects,
potential
use
like
like
nat
streaming
or
jet
stream.
It's
just
you
know
like.
Maybe
it's
not
the
full
broker.
Yeah
do
other
people
have
thoughts
on
this
suggestions
on
this
because
it's
it's
a
it's
an
issue,
we're
using
that's
primarily
in
between
mesh
sync
and
mesh
reservoir.
A
A
The
issue
that
you're
talking
about
is
ultimately
quite
important
because,
what's
going
on,
is
that
a
user
will
say
either
through
measuring
ctl
or
mesh
ui
they'll
say
I
would
like
this
operation
to
be
performed.
I
would
like
to
provision
a
service
mesh
or
something
like
that,
and
that
might
take
a
while,
and
it
might
take
many
steps
so
for
a
user
to
sit
there
twiddling
their
thumbs,
hoping
that
something
didn't
crash,
just
sort
of
waiting
to
see.
If
something
worked,
it's
a
lot
of
it's
a
lot
to
ask
of
them.
A
They've
got
to
give
a
lot
of.
They
have
a
lot
of
faith
and
measuring
to
sit
there
waiting.
Rather,
what
you
want
to
see
back
is
just
incremental
progress
yeah.
So
another
approach
that
I
would
also
wince
at
and
not
it
would
not
be
suggested-
is
to
instead
have
the
client
that
invoked
the
operation
poll
for
status
like
anyway,
what
the
user
needs
is
a
is
not
only
a
weight
indicator
like
a
spinner
sitting.
There
saying
wait,
because
eventually
the
spinner
isn't
enough.
A
Eventually,
most
of
us
have
looked
at
spinners
for
like
hours
when
actually
the
spinner,
the
operation
behind
the
spinner
died
like
30
seconds
into
the
thing,
and
they
don't
know
that
because
it's
just
spinning
forever.
So
that's
not
even
enough
like
they
need
some
sort
of
actual
progress
indicator,
whether
it's
the
actual
messages
coming
back
or
just
so.
A
A
Yeah,
it's
kind
of
it's
one
of
those,
it's
one
of
those
things
that
we
eventually
need
to
stop
talking
about
it
and
actually
go
through
the
design
spec
and
get
it
done.
That
is
teed
up
for
mesri,
zero,
seven,
zero,
a
focus
on
that
so
and
abi
and
pranav
have
given
some
thought
to
it
recently,
and
so
abi
has
linked
to
it
in
the
chat.
A
B
A
A
B
A
B
B
A
B
B
A
A
B
A
B
B
Okay
and
while
setting
up
our
docker
docs,
I
actually
use
this
docker
docks.
First,
I
mean
issue
of
a
lower
case
and
after
that
I
think
it's
a
operating
system
issue
from
there.
So
windows
operating
system
is
working
for
docker
docks.
You
can
see
that,
but
while
setting
also
rbm,
I
said
we
tried
to
set
up
the
rpm
and
this
command
actually
in
the
rvm.