►
From YouTube: 2023 02 03 Docs Office Hours
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
B
Yes,
a
very
small
discussion
on
how
should
we
interact
with
the
contributors,
because
I
am
a
mentor
of
a
project.
So
I
was
just
wondering:
okay,.
A
A
A
April
or
May
date
is
not
yet
set
the
freeze
date
for
their
their
code.
The
next
level
freeze,
if
you
will
I,
think
it's
end
of
March
and
assuming
one
month
to
two
months
to
for
between
deep
freeze
and
final
release,
April
or
May
they'll
release
they'll
ship
or
release,
and
they
will
ship
without
open,
jdk
11.
A
they're
preparing
for
the
day
when
jdk
11
is
no
longer
interesting
to
their
operating
system
community.
So
what
the
the
doc
Sig
in
general
for
your
info
has
decided
is
that,
with
the
release
of
Debian
12
we're
going
to
change
the
documentation
for
Linux
installs
so
that
they
guide
people
to
use
Java
17.
job
11
will
be
continued
to
be
supported,
but
we're
going
to
tell
them.
This
is
how
you
install,
and
you
should
use
Java
17.
that
way
we
can
use
the
same
set
of
installation
instructions
on
all
platforms
without
having
to
explain
that.
A
Oh,
if
you're
on
Debian
12,
you
can't
get
Java
11
without
special
extra
work,
we'll
make
the
same
change
with
windows
installs.
That
way,
we
start
the
process
of
persuading
people
that
they
should
be
using
Java
17,
more
and
more
Java
17
Works
quite
well
been
supported
now
for
two
full
LTS
cycles
and
by
the
time
we
get
to
May
or
June,
it
will
be
yet
another
full
LTS
cycle.
So
by
then
we
will.
We
should
have
plenty
of
experience
with
Java
17,
very
comfortable,
supporting
it
any
questions
there.
C
Unmute,
how
long
do
we
maintain
an
LTS
release
of
Jenkins.
C
A
No
so,
and
and
when
we
use
the
word
when
we
use
the
word
supported
here,
let's
be
very
precise.
What
it
really
means
is
with
security
fixes
right
so
right.
A
If
there's
a
security
issue
in
Jenkins,
core
security
issues
in
Jenkins
core
are
fixed
only
on
most
recent
weekly,
so
weekly.
If
you're
not
upgrading
every
week,
you're
making
a
mistake
if
you're
choosing
to
run
weekly,
you
should
upgrade
every
week
and
then
most
recent
LTS
and
then
in
very
rare
occasions,.
A
And
I
think
it
happened
once
in
the
last
two
years.
Lts
minus
one
also
gets
a
security
fix.
A
So
that's
that
it
truly
is
quite
rare
and
I
think
maybe
once
in
two
years,
maybe
once
in
three
or
more
years,
this
has
actually
happened.
A
All
right,
yeah
so
now
same
thing,
same
storytelled
for
platform.
Support
is
dropped
for
an
operating
system
when
the
operating
system
provider.
C
A
C
So
there's
isn't
there
an
Ubuntu
release,
that's
expiring!
This
year,
too,.
C
A
A
B
Yes,
I
understand
it's
a
very
generic
question
so,
like
this
question
came
to
my
mind
when
I
was
observing
two
kinds
of
conversation
on
two
channels,
so
first
channel
is
G-Shock,
sick,
Channel
and
second
one
is
the
plug-in
Health
scoring
channel
so
which
one
should
we
use
for
debugging
or
helping
them
to
set
up
the
local
environment.
Just
a
small
question.
A
So
and
I
think
the
answer
is
either
wherever
they
are
you're
welcome
to.
If
you
feel
like
it's,
the
noise
level
is
becoming
too
great
in
the
gsoc
Sig
Channel
you're
welcome
to
redirect
them
to
the
other
channel.
It's
whatever
works
for
you.
A
So
you're
saying
the
gstock
Sig
channel
has
conversations
and
so
does
the
so
does
these
plug-in
Health
scoring
Channel.
A
B
We
have
me
looking
out
at
the
editor
messages,
so
I
could
not
find
the
reason
behind
one
of
the
error
messages.
So
I
thought:
let's
redirect
it
redirect
that
person
to
the
channel
so
that
Adrian
can
have
a
look
and.
A
A
I
believe
I've
seen
some
conversations
on
the
on
the
plug-in
installation
manager,
tool
Channel
as
well,
so
it
fits
with
what
you're
describing
that
sometimes
now
that's
the
conversations
come
and
go
right.
They
arrive,
they
happen
and
they
then
they
and
they
may
end
or
they
may
die
without
an
answer
and
those
are
all
very
real
risks.
If
you
choose
a
a
narrowly
focused
Channel
and
ask
a
general
question,
you
may
not
get
an
answer,
so
it's
it's
a
risk
of
having
multiple
channels
and
I
think
it's
an
okay
risk.
B
Sure
that
answers
my
question
and
I
have
one
more
question
on
slightly
related
topic.
So
few
weeks
ago,
I
saw
your
email
asking
for
a
poll
on
discussing
the
discussing
more
on
and
refining
the
topics
that
we
already
have.
So
after
that
I
actually
replied
here
that
I'm
interested
for
the
plug-in
Health
scoring
idea
and.
B
I
went
some
on
vacation,
so
I
could
not
be
in
loop
at
that
time.
So
did
I
miss
anything.
Can
I
help
in
on
that
regarding
the
plug-in
Health
scoring
project.
A
Good
question
so
I
didn't
schedule
a
session
for
plug-in
Health
scoring
because
I
believe
the
two
people
who
were
willing
to
be
mentors
are
you
and
and
Adrian,
and
with
the
with
Adrian
and
you
as
the
two
mentors
I
figured
you'd
just
chat
with
each
other.
So
I
did
not
schedule
a
poll
with
for
with
Adrian.
For
that
one.
The.
A
We
we
did
a
poll
for
I,
did
a
poll
for
plug-in
installation
manager
tool
and
about
five
of
us
met
about
12
hours
ago
and
discussed
it
so
Freya
Freya,
meta,
Bruno,
varached
and
John
Mark,
Mason,
Chris,
Stern
and
I
all
met,
and
we
talked
through
some
ideas.
I
can
certainly
link
you
to
the
notes
from
our
meeting
and
I'll
post,
the
recording
from
the.
A
A
So
you're
welcome
to
those
notes
and
I'll
post
a
recording
of
it
later,
probably
within
the
next
24
hours.
B
B
Okay,
so
if
I
wanted
to
discuss
more
on
the
plugin
Health
scoring,
then
I
would
reach
out
to
Adrian
right
right
exactly.
A
B
A
Yeah,
one
of
the
things
that
came
out
of
the
discussion
with
the
five
of
us
was
ways
that
the
mentors
can
prepare
themselves.
So
it
wasn't
as
much
about
what
the
candidates
can
do.
This
one
was
about
how
we
can
do
and
one
was
review.
The
existing
code
become
familiar
with
it,
improve
the
test
coverage
by
writing
some
tests.
Well,
one
of
the
things
that
came
out
of
that
rather
embarrassing
was
that
hey
that
particular
thing
was
not
actually
reporting
test
coverage,
and
so
so
we
opened
up
the
job
and
saw
oh
wait.
A
A
second
look
at
this
there's
no
coverage
report
over
here
on
the
left
hand
side.
So
how
would
a
new
contributor
even
know
which
things
are
not
covered
and
and
then
we
said,
should
we
call
that
a
good
first
issue
and
let
somebody
else
enable
coverage
and
I
said
no,
because
I
had
just
had
some
interactions
with
coverage,
reporting
and
I
knew
how
to
implement
it
very
quickly.
So,
rather
than
have,
somebody
else
go
learn
the
things
that
I
had
recently
learned.
A
B
A
A
A
Okay,
sorry!
This
is
a
bad
choice
for
the
example.
Let
me
get
a
better
this
one.
This
one
has
a
complicated
build
plug-in
because
I
need
it
complicated.
Let's
choose
a
simpler
one,
just
a
minute:
diraj,
that's
a
little
awkward
that
I
choose
one
that
looks
so
terrible
as
that
just
a
minute
here
we
go
okay,
so
here
is
an
example
of
build
plug-in
that
shows
us
replay.
A
Here
we
go
what
you'll
see
here
is:
that's
all
it
takes,
and
what
this
says
is
use
containers
don't
stop
if
one
of
the
tests
fails,
I
want
all
the
test
results,
and
this
will
go
away
soon.
Artifact,
caching,
proxy-
and
here
it
is
this
little
bit
right
here-
run
a
configuration
of
Linux
running
on
either
Java
11
or
17
as
the
first
choice,
and
you
will
automatically
get
a
coverage
report.
A
A
B
A
B
Yeah,
so
this
is
helpful.
I
can
look
at
the
repository
and
see
whether
we
have
already
have
this
coverage
report
or
not
because
I
trust
Adrian
sets
up
everything
that
needs
to
be
there
and,
if
not
I'll,
you
know
discuss
with
him
that
hey.
We
should
have
this
and
then
go.
A
A
They're
nine
statements
are
covered,
33
are
not
covered,
so
here's
a
an
opportunity
for
now
now
some
of
these
jagruti
proved
to
us
that
some
of
these
are
very
difficult
right.
She
tried
to
add
tests
to
this
one
to
node
label
parameter
and
its
coverage
report
looks
like
it
should
be
nice
and
easy
to
do,
but
when
she
attempted
it
she
found
there
were
all
sorts
of
complex
things
hiding
here
that
she
just
couldn't
couldn't
quite
see.
A
A
B
A
No,
usually
what
I'd
just
do
is
open.
My
IDE
open
this
plugin
in
the
project
find
that
class
and
click
the
button
that
says
generate
tests
for
that
class
and
what
what
the
IDE
does,
in
my
case,
netbeans,
but
also
IntelliJ
and
visual
studio
code,
have
this.
It
will
generate
stubs
for
each
of
the
methods
that
are
public
and
then
I.
Take
those
stubs
and
I
start
filling
in
actual
assertions
for
them.
A
That,
oh,
oh,
okay,
that's
a
that's
a
a
very
fair
question
and
the
way
they
know
that
is
usually
because
someone
has
created
an
issue
in
the
Jenkins
issue.
Tracker
that
says
just
a
minute:
I'll
bring
up
some
issues
to
show
someone
will
have
created
an
issue
in
the
Jenkins
issue.
Tracker
that
says
something
like
improve:
embeddable
build
status,
plug-in
test
coverage.
A
So
does
so
what
this
is
a
more
general
purpose
statement
it
says:
look
your
goal
is
to
improve
the
test
coverage
and
I,
don't
care
which
thing
you
improve
test
coverage
on
pick
one,
and
it
then
says
here's
how
you
review
the
current
coverage
and
in
this
case
it
opens
up
this
page.
That
shows
you
an
awful
lot
of
red
right,
there's
a
lot
of
red
there
so
that
that
should
be
doable.
A
B
A
Well,
and-
and
the
reason
is
exactly
things
like
this
right-
okay,
so
vandit
Vandy
suggests
said:
hey
I'd
like
to
work
on
this
one
and
that's
that's
a
common
common
thing.
Someone
will
say
I'd
like
to
work
on
this
one
that
was
three
days
ago,
and
we
don't
have
a
pull
request
yet
and-
and
the
reason
is
sometimes
a
new
contributor
will
look
at
the
code
and
say
I'm.
Sorry,
I,
don't
understand
how
to
write
a
test
for
this.
That's
not
shocking
right,
we're
not
not
at
all
dismayed
at
that
it
happens.
A
Sometimes
these
things
are
really
complicated
or
obscure
or
or
the
person
who's
saying.
I
want
to
write.
A
test
has
never
written
a
junit
test
before,
and
this
is
probably
a
really
unfriendly
way
to
put
them
into
junit
testing,
because
Java
based
junit
testing
in
Jenkins
has
some
unique
quirks
right.
It's
got
some
things
that
are
odd.
C
B
Yes,
it
did
I
think
this
is
a
good
place
of
Jenkins
cheetah,
where
we
can
create
a
generic
issue
to
cover
increase
the
test
coverage
so.
B
A
Yeah,
okay,
well
and
and
now
be
be
forewarned.
There
are
I
think
Adrian
may
actually
be
tracking
his
his
the
issues
for
plug-in
Health
score
in
the
the
plug-in
Health
scoring
Repository.
A
A
In
fact,
if
you
want,
we
can
do
it
right
now,
yep,
no,
no,
probably
best
you
work
with
you
work
with
with
with
Adrian
on
it,
because
if
I
create
it
it's
written
by
me
and
my
framing
and
with
my
instructions,
it's
much
better
that
that
he
can
tell
hey
here's
how
you
could
do
this,
and
he
may
also
say
no
improving
test
coverage
is
not
helpful
at
this
point
because
he's
already
pretty
well
covered.
B
Right,
yeah
because
I
think
he
thinks
think
of
the
things
way
before
so
I
think
he
might
have
some
idea
or
some
things
that
he
wanted
to
share
about
this.
So
it
will
be
better
if
we
discuss
first
right.
A
And-
and
that
makes
that
makes
perfect
sense
if
we
think
about
it.
Well,
if
we
look,
if
we
just
look
at
his
at
the
current
coverage
report,
it's
actually
pretty
good
right.
So
this
this
says
there's
an
awful.
There
are
an
awful
lot
of
things
that
are
very,
very
well
covered.
The
green
and
even
the
the
whatever
that
color
is.
Maybe
yellowish
is
over
70
percent
in
most
cases,
so
that
this
is
actually
quite
well
tested
already
and
so
telling
a
new
contributor
hey,
go
write
tests
for
this
may
not
be
that
helpful.
B
Right
I,
we
are
seeing
these
green
and
yellow
more
because,
when
I
was
in
turning
on
this
project
on
every
PR
Adrian
suggested
me
to
add
some
tests.
So
that's
why
I
know
this.
This
is
not
shocking
to
me
so
that
he
he
follows
best
practices.
So
that's
really
nice.
A
A
C
A
It
just
some
people-
kosuke
kawaguche,
for
instance,
doesn't
write
a
lot
of
tests,
but
he
writes
great
code.
Mark
Wade
doesn't
write
great
code,
but
he
writes
a
lot
of
tests
and
we're
both
useful.
A
Yeah,
actually
it's
it's
a
much
scarier
graph
for
the
git
plug-in,
because
the
the
reason
it's
scarier
is
because
the
git
plug-in
started
well.
The
git
plug-in
is
a
source
of
an
entire
talk
that
I've
presented
about
this.
This
pattern
for
the
first
18
months
of
the
life
of
the
git
plug-in
there
were
zero.
Automated
tests,
zero,
none,
wow
and-
and
that
was
the
right
thing
to
do-
for
the
git
plug-in,
because
in
those
early
days,
get
looked
an
awful
lot
like
Linus's
experiment.
A
That
was
never
going
to
go
anywhere
now:
Looking
Back,
Now,
15
years
later
or
10
10
12
years
later.
We
laugh
at
that
because
of
course,
git
runs
the
world
now,
but
all
those
years
ago,
when
they
first
started,
it
looked
like
a
really
strange
experiment
from
Linus
Torvalds,
and
so
the
fact
that
they
wrote
notice
was
not
a
big
deal
right,
but
then
people
like
me
got
dependent
on
it
in
production
and
got
very
grumpy
when
it
broke,
and
so
we
started
writing
tests.
A
A
Yeah,
so
no
and
I
wasn't
the
only
one
who
did
that
there
were
lots
of
people
involved
before
me,
but
but
yes,
the
the
story
and
and
really
it's
a
it's
a
fun
topic
for
a
there's,
a
reason
why
I
give
the
talk
about
it,
because
it
was
the
right
decision
at
the
time
to
not
write
tests
and
when
we
started
implementing
tests
it
was
because
we
absolutely
needed
the
tests
now.
That
means
there
are
some
things
in
this.
A
B
B
Nothing
as
that
was
it
thanks.
A
Okay,
great
all
right,
so
let's
then
I
propose,
let's
spend
the
last
15
or
20
minutes,
looking
at
pull
requests
in
particular,
because
several
of
these
poll
requests
are
things
of
beauty.
It's
really
impressive.
So
let's
look
at
this
top
one
first,
because
I
haven't
looked
at
it
in
a
week
or
two
Mandeep
Singh,
one
of
the
candidates
for
I,
think
candidate
for
Google
summer
of
code
and
trying
to
contribute
decided
to
take
on
a
an
issue
that
we
never
said
was
a
good
first
issue.
A
It
is
not
definitely
most
definitely
not
a
good
first
issue,
because
what
it
says
is
the
Linux
install
guide
has
no
upgrade
instructions,
it
just
tells
you
how
to
install
not
upgrade
and
we've
had
lots
of
people
ask
well.
How
do
I
do
an
upgrade
and
we
point
them
to
the
change
log
and
we
tell
them
read
the
upgrade
guide,
but
we
have
no
instructions
that
provide
General
guidance.
So
vandit
is
not
10
years.
Working
on
Jenkins,
like
some
of
the
rest
of
us,
are
brand
new
contributor,
but
decided
wants
to
write
this.
A
So,
let's
take
a
look
at
this
upgrade
guide
and
and
I'm
thoroughly
impressed.
It's
not
done
yet
because
the
the
problem
is
really
hard.
That
bundit
is
trying
to
solve
and
you'll
see
the
long
list
of
evolution
that
it's
been
through
as
people
have
given
comments
have
looked
at
it
have
said:
no,
we
got
to
correct
this.
It's
really
really
a
nice
piece
of
work
that
vandit
has
done
now.
A
C
A
C
A
C
A
I
thought
that
in
general
lists
did
not
end
with
a
hard
stop.
Meg
help
me
on
that.
What's
what's
the
rule
for
list
items.
A
B
A
A
A
C
C
A
So
so
there,
if
you
can
do
if
you've
got
a
snapshotting
file
system.
You
should
absolutely
use
it
for
your
backups
and
not
bother
with
any
of
the
other
techniques,
and
then
it
does
a
backup
of
everything
it
backs
up,
Secrets
everything,
because
it's
doing
the
file
system
level,
backup
it's
actually
not
even
a
backup
a
snapshot
is,
is
really
just
a
writing
of
some
pointers.
C
A
A
All
right,
so
we
got
that
now.
Let's
continue
our
reading
oops
and
where
were
we,
we
were
documentation
installing
updating,
Jenkins,
okay,
so
Linux,
oh
wait
a
sec!
Let's
read
here!
All
right
Jenkins
can
be
installed
with
operating
system.
Specific
installers
when
upgrading
the
process
used
must
match
the
original
installation
method.
A
A
A
C
B
C
Getting
ready
to
up
the
the
thing
that
none
of
us
you
know:
I
am
bad
about
backing
stuff
up
because
I
know
my
step.
My
system's
never
going
to
fail,
I'm,
never
going
to
lose
stuff
and
really
need
that
back
up.
But
but
when
you're
doing
an
upgrade,
you
need
to
be
prepared
for
a
real
possibility
that
you
are
going
to
need
that
backup.
C
So
you
want
it
to
be
as
clean
as
possible,
so
it
isn't
something
and
then
actually
yeah,
maybe
that
before
you
upgrade
just
these
and
the
third
one
is
a
statement
about
Java.
But
what
am
I
supposed
to
do
with
that.
A
C
A
It
needs
a
verb
right.
It
should
be
assure
that
Java,
Java,
11
or
Java
17
are
installed
on
the
controller
and
on
all
the
Jenkins
agents
right,
and
this
is
why
you
have
to
assure
that,
but
really
it
needs
just
like
there's
an
Ensure
you've
created
a
backup
and
verify
this
should
be
ensure
that
Java,
11
or
Java
17
are
installed
right.
C
A
Right
right,
good
point:
yeah
yeah!
So
well,
if
isn't
it
almost
that's
part
of
the
backup,
because
the
backup
instructions
tell
you
that
you
can't
be
considered
that
you
need
a
consistent
backup
and
you
can
get
a
consistent
backup
either
by
using
a
file
system
snapshot
or
by
quieting
down
and
getting
all
the
jobs
clear.
C
A
A
C
A
C
A
It
just
it
just
requires
more
work.
It
actually
can
be
done.
I'm,
just
not
sure
that
we
want
this
actually
to
be
numbered,
because
I'm
I'm
not
entirely
sure
that
it's
helpful
to
the
user
to
see
the
number
as
much
it
is
just
the
heading
level.
They
see
hey,
stop
your
Jenkins
instance.
Okay,
that's
a
step!
I
need
to
do,
perform
a
backup
yup!
That's
that's
a
good
one.
A
A
C
A
A
Interesting
that
it
doesn't
say
anything
about
steps
to
perform
on
agents.
Maybe
that
comes
as
a
separate
thing,
because
inbound
agents
need
to
be
updated
as
well.
Oh
yeah,
outbound
agents
take
care
of
themselves,
but
inbound
agents
need
need
some
well
okay.
So,
let's
all
right!
Here's
Darren's
tutorial
and
I
like
that,
and
now
the
post
upgrade
Setup
Wizard
verified
the.
C
C
A
C
A
A
See-
and
this
is
missing-
a
concept
that
we've
been
recommending
multiple
times
in
recent,
which
is
that
before
performing
the
upgrade
so
before,
starting
the
upgrade
before
executing
this
command,
for
instance,
or
this
command,
oh
dear,
that
needs
to
be
upgraded.
Okay,
just
a
minute
update.
A
Okay,
so
let
me
fix
that
as
well,
then
moment
app
get
apps
get
installed.
Okay,
this
one
I
believe
is.
A
A
A
A
A
A
I'll
I'll
have
to
check
it
myself,
that's
one
that
that
I
need
to
check.
Let
me
make
myself
a
note
here.
A
Check
that
upgrade
can
operate.