[gentoo-amd64] Re: amd64 list, still useful? Was: btrfs

[Duncan]

On Thu, 05 Jun 2014 22:48:07 +0100
Martin <m_btrfs@××××××.uk> wrote:

> Resend (gmane appears to be losing my email for this list... :-( )

OK, forwarding to the list too (with a bit less snippage than normal,
to keep your message intact as I'm relaying) and replying below.

>
> On 05/06/14 16:35, Martin wrote:
> > On 05/06/14 03:00, Duncan wrote:
> >> So things should really be simmering back down pretty shortly.
> >> =:^)

> > Thanks for the good summary.
> >
> > Yep, I hit all the red "B" blockers... Quickly saw it was upower and
> > some confusion with systemd even though I've not selected systemd
> > anywhere and...
> >
> > I was too rushed to investigate much further and so added into my
> > /etc/portage/package.mask:
> >
> > # Avoid pulling in systemd!
> > =sys-power/upower-0.9.23-r3
> >
> >
> > Thanks for letting me know to await the news item and for the bits
> > to settle...

[Just forwarding that part and would delete it as I'm not replying to
it, were I not forwarding it for you too.  But I'm replying to the
below.]

> > As for systemd... I'm just wondering if the various heated air being
> > generated/wasted is as much rushed arrogance on the part of the
> > implementation as due to the grand ripples of change.
> >
> > The recent kernel DoS debacle due to misusing the long used kernel
> > debug showed a certain 'cavalier' attitude to taking over
> > functionality without a wider concern or caution to keep projects
> > outside of systemd undisturbed... Or at least conscientiously
> > minimise disturbance...

Agreed, and for quite some time I that attitude was why I was delaying
my own switch, tho I expected I'd eventually make it.

But backing up a bit to reveal the larger picture...

Developers in general aren't always exactly known for their ability to
get along with each other or with others or necessarily the wider
community.  Certainly there's many examples of this in the FLOSS
community, and from what I've read of the proprietary development
community it's no different, save much of it happens behind closed
doors, with public appearances moderated by the PR folks.

Actually, I've a personal experience that rather changed my own outlook
on things, that I believe explains a lot of the problem here.  The
following gets a bit franker and more personally honest than most
discussions and I'm not really comfortable saying it, but it's
important enough not to skip as it illustrates a point I want to make.

I don't ordinarily talk about myself in this way, but the fact is, on
most tests I score well above 90 percentile IQ.  Typically, depending
on the test and whether I'm hitting my groove that day or not, I run
95-97 percentile on average in most areas (tho in composition I'm
relatively low for me, 70s). (FWIW, I've always been slightly
frustrated.  The MENSA cutoff is supposed to be 98 percentile and I
typically score tantalizingly close, but not quite!  It'd be nice...
=:^( )  In technology and computer areas I'd guess I'm a bit higher,
perhaps 98 percentile or so.  95 percentile means about 19 out of 20
people score lower, 98 percentile is 49 out of 50. 

But, this level of attainment presents its own set of difficulties,
difficulties I'm intimately familiar with, but obviously not to the
level these /real/ geniuses, the big hero coders of our community, are.

I still remember the day I actually realized what dealing with a
mentally challenged individual actually was, back in about 8th grade or
so.  He had come to visit a next door neighbor and we set out to climb
a local butte, me not yet understanding his difficulty -- I knew
there was /something/ different about him, but I didn't know what, I
just accepted it, and him, as basically my equal, as I had been taught
to accept and treat everyone.  But climbing this butte didn't simply
involve a hike, as is the case with many hills/buttes.  It involved a
bit of relatively minor technical climbing, "chimneying", etc. I
had done it with a group previously, but wanted to try it again, for
the exercise and challenge.  But I didn't want to do it alone, and this
guy was agreeable to trying it, so we set out.

Everything went well, considering, but it did take somewhat longer than
I had planned and our ride back got a bit worried and alerted the
authorities.  Fortunately, they didn't have to pull us off the mountain
(or scrape us from the bottom of the chimney), but we got in a bit of
trouble.

When I got home, Mom asked me why on earth I'd take a r* guy up a
mountain like that.  I was flabbergasted!  I didn't know!  And to think
I took him on that climb that was slightly challenging for me
(something I'm not sure my Mom knew, and that I didn't tell her!), what
must it have been for him? I was perhaps rather fortunate
something /didn't/ happen, altho now I realize that despite (or even
perhaps because of) his challenge he was remarkably resilient, and may
well have picked himself up and continued better than I would have if
something had gone wrong and either one of us was hurt

That night or perhaps the next day, as I thought about it, I realized
what had happened.  I was so used to, as a matter of course, dropping
to whatever level was required to meet people at their own level and
treat them as equal, that I didn't realize I was even doing it.  To me
it was just the way one interacted with others. What I had originally
noticed different about him, that I couldn't put into words before as I
simply didn't have the experience or concept, was that I had to drop a
bit more than normal, but I was so used to doing it for pretty much
everyone, that I didn't even realize I was doing it, or know what it
was... until I was forcibly confronted with the fact that this guy was
(to others) noticeably below average. But to me he was simply a bit
more of the normal that I always did, and that I thought was just the
way it was to interact with /anyone/.

Since then I've obviously become a bit wiser in the ways of the world,
but realistically, I really do seldom meet people /really/ my equal in
the real world, and that has really distorted my experience, and to
some extent my attitude and picture of the world.

But that was only experiencing the one side.  I consider myself
fortunate to have actually had the opposite experience as well.  A bit
over a decade ago I was with a Linux and Unix friendly ISP that had a
lot of real knowledgeable folks as customers, including one guy that was
one of only about a dozen with direct commit privs to one of the BSDs,
and several others that were in the same knowledge class.  While I
may well be at the 95-97 percentile range, for the first time in my
life I was well outranked, as several of these guys were at the 99th
percentile or better I'm sure, plus they had likely decades of
experience I didn't (as a newbie fresh from the MS side of the track)
have!

That was a humbling experience indeed!  To that point, I had been used
to being at least /equal/ to pretty much anyone I met, and enough above
most that even if I happened to be wrong I knew more about the
situation than pretty much anyone else, that I could out-argue them
even in my wrongness.

Here the situation was exactly reversed, *I* was the know-nothing, the
slow guy that everyone else had to wait for while someone patiently
explained what was going on so I could follow along!

I **VERY** **QUICKLY** learned how to shut up and simply read the
discussion as it happened around me, learning from the masters and
occasionally asking a question or two, and to be *VERY* sure I could
backup any claims I DID make, because if I was wrong, for the first
time in my life I was pretty much guaranteed to be called on it, and
there was no bluffing my way out of that fix with THESE guys!

That had roughly the same level of effect on me as the earlier
experience, but at the opposite end, something I rather badly needed as
I NEEDED a bit of humbling at that point!

Now here's the critical point that I've been so brutally honest to try
to present:  What happens to the *REAL* 99 percentilers, the guys who
*NEVER* have that sort of humbling "OOPS, I screwed up and better
shutup!  These guys know more than me and if I'm wrong they're not
afraid to demonstrate exactly why and how!" ... experience?

Unfortunately, a lot of them are a**h***s!  Why?  Because they're at
the top of their class and they know it.  Nobody can prove them wrong,
and if somehow someone does, they simply don't know how to react, as
it's an experience they very rarely have.  Even on things they know is
simply opinion, they're so used to having absolutely zero peers around
that can actually challenge them on it, that they simply don't
know /how/ to take a real, honest challenge when it comes.

Which BTW is one of the things I find so amazing about Linus Torvalds.
I doubt many would argue that he's at the 99 percentile point, yet
somehow he's a real person, still approachable, and unlike most folks
at his level, actually able to deal with people!

At the other end are people like Hans Reiser.  He was and is a
filesystem genius, and reiser4 was years before its time, yet never got
into the kernel despite years of trying, because he was absolutely
horrible at interpersonal relations and nobody anywhere near his
level could work with him, because he simply didn't know how to be
wrong.  Unfortunately learning that was literally a fatal experience
for his wife. =:^(

Take it from someone who is in many areas 90 percentile plus, but who
counts that experience sitting at the feet of /real/ masters as perhaps
the single most fortunate and critical experience in his live, because
he learned how to be wrong, that's NOT an easy lesson to learn, but
it's an *EXTREMELY* critical lesson to learn!

Think about that the next time you see something like that kernel
command-line debug thing go down.  Poettering and Sievers are extremely
bright men, genius, top of their class.  And Poettering in particular
is a gifted speaker as well (researching systemd I watched a number of
videos of presentations he has done on the subject, he really IS an
impressive and gifted speaker!).

But, they don't take being wrong well at all, and they have a sense of
entitlement due to their sense of ultimate rightness.

Never-the-less, however one might dislike and distrust the personality
behind them, both systemd and reiserfs (and later reiser4) were/are top
of their class for their time, unique and ahead of their time in many
ways.  There's no arguing that.

I didn't and don't like Hans Reiser, but I used his filesystem
(reiser3), and still use it on my spinning rust drives altho I've
switched to the still not fully mature btrfs on my newer ssds.

Unlike Reiser, I don't know so much about Poettering and Sievers
personal lives and I surely hope they don't end up where Reiser did for
similar reasons.  But similar to Reiser, I use their software, systemd,
now.  And there's no arguing the fact, it's /good/, even if not exactly
stable, because they continue to "gray goo" anything in their path, and
haven't yet taken the time necessary to truly stabilize the thing.
While I never used it, from what I have read, PulseAudio was much the
same way as long as Poettering was involved -- it never truly
stabilized until he lost interest and left.

Unfortunately I think that's likely to be the case with systemd as
well; it won't really stabilize until Poettering loses interest and
moves on to something else.  And for people who depend on stable, I
really doubt I'll be able to recommend it (if you can avoid the
gray goo, I really don't know if that will remain possible if he
doesn't lose interest in another couple years) until then. But it /is/
good, good enough it's taking the Linux world by storm, gray goo and
all.  If systemd could just be left alone to stabilize for a year or
so, I think it'd be good, /incredibly/ good, and a lot of hold outs,
like I was until recently, would find little reason not to switch, once
it was allowed to stabilize.  But when that's likely to happen
(presumably after Poettering moves on), I really haven't the foggiest.

Meanwhile I'm leading edge enough (I'm running git kde4 and kernel,
after all), and (fortunately) good enough at troubleshooting Linux boot
issues when I have to, that I decided it was time, for me anyway.

So as you can see, while I've succumbed now, I really do still have
mixed feelings on it all.

But meanwhile, try applying the "do they actually know how to be wrong"
theory the next time you see something happening elsewhere, too.  It's
surprising just how much of the FLOSS-world feuding it explains!...

Tho this is one area I'd be I'd be /very/ happy if I /was/ wrong about,
and suddenly all these definite top-of-their-field coders started
getting along with each other!  Well, we can hope, anyway (and while
we're at hoping, hope the lesson in being wrong isn't data eating code
teaching them how to be wrong, or security code either, as seems to
have been the recent case with openssl!).

-- 
Duncan - No HTML messages please, as they are filtered as spam.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master."  Richard Stallman