[clue-talk] Vmware ESX / VI, WTF?

[Nate Duehr]

David L. Willson wrote:

> I think we're closer to a productive statement of points on which we agree and disagree,
> so I'll say this, on this point.  The reason why I always choose platform neutral
> products when such are available, is because I believe in the power of competition.  If
> I let someone lever me into Exchange, or ESX, or MAGIC HDIQ, or Commtouch, or some other
> choice-limiting product, I remove from myself, at least to a small extent, the benefit
> of competition for my business.  A Linux-based product is ideally positioned to
> encourage competition, because Linuxes are so very competitive with one another,
> providing unprecedented lateral mobility for users and software makers.  If it were
> "just me", then it would be only $100, as you say, but I think of myself as a leader, a
> trail-blazer, a hero, an icon in the place where God put me, so I think of every move I
> make as a "leading" move, something that others might follow, and that carries a
> terrible burden of responsibility.  :-)

Hmm, I can't see how realistically running VMWare's app on Windows is 
suddenly going to "leverage" you into Outlook/Exchange.  That seems like 
some serious paranoia you've got going there!  (GRIN)

Linux products on the Desktop are currently an also-ran, as far as REAL 
competition goes -- so far.  Nothing on the Linux desktop is compelling 
enough for the masses to switch to it from [insert whatever they're 
using here, probably Windows].  It's been promised for a long time, but 
Linux devs are too fragmented and PO'ed at each other to really produce 
anything useful.  I see Enlightenment is making a comeback in the latest 
Linux Journal... LOL... it died when the main developer couldn't get 
along with other folks, and the GNOME/KDE "competition" era started. 
Neither has produced a compelling desktop in almost a decade of 
development, and E is coming back?  Sad.  But not too surprising really. 
  E was a nice clean Desktop with a little eye candy to help you feel 
good about the choice, I suppose.  Certainly more lightweight and 
well-written than the mess GNOME/KDE are offering today... or back then.

As a server platform, the kernel and GNU tools are great.  That's 
"Linux" as most of us perceive it that wanted a way to set up servers 
cheaply.  Without Apache, it would be nothing.  And Apache grew up and 
runs on any OS now.  So Linux/Apache isn't the only game in town in that 
market anymore.

As far as the Desktop goes... bah.  It doesn't exactly "suck" if you 
load it ONCE and enjoy it, but the long-term upgrade paths and time 
wasted fixing things distros do that cause collateral damage via their 
packaging systems is pretty annoying after years of doing it.

I can get that from MS and Apple too, so "competition"-wise, Linux is no 
better than the alternatives there, from a quality perspective.

"Competitors" that always lose are pretty boring, in the long term.

Not trying to be a troll, just stating where we're at today -- something 
fans are loathe to do to ANY OS.

I'm not so worried about saying ALL this stuff (Linux included, but also 
Windows and Mac to different extents) stinks to high heaven anymore.

I figure if enough people say it, maybe the crazy folks that have the 
time to write code for free might have some motivation to fix it.  I 
doubt it though, really... deep down.   Even the paid folks don't have 
time to fix all that stuff.

Linux has and always will have the responsibility handed out with it 
that says, "Don't like something... YOU fix it."  Frankly, I'd rather 
pay someone else to do it.  I have other things to get done these days, 
and I never was a programmer.  Waiting around for some dev to fix 
something they broke that I was USING after an upgrade, really gets old. 
    At least with commercial desktops there's a fairly well-defined set 
of things they're NOT going to break on a whim -- in Linux, it could 
literally be anything.  Risky for the end-user who can't and/or won't 
program.

The performance of the commercial "leader" in Linux is pretty poor on 
the Desktop, and they are OBVIOUSLY not interested.  They're a server 
company.  [RedHat, that is.])

Since I can run servers without RHEL for free, and just fine... my 
"needs" aren't met too well by "commercial Linux".  All I really want 
out of commercial Linux distros is to stop changing things to 
differentiate themselves from "regular" Linux, so I can run the free 
stuff and the commercial stuff interchangeably, and not have to learn a 
completely different methodology for running a Unix-like system.

Unfortunately, that's not how the software biz works... commercial 
distros and commercial Unix (non-Linux) all do things a little 
differently on purpose... trying to lock you into their systems.  Those 
that tell me RH/Fedora is great, are usually completely LOCKED IN to 
that methodology.  And often don't even realize it.

So if this is all about "competition"... Linux isn't competing 
effectively yet... when it has 10% of the Desktop market share, that 
argument will start to get more compelling.

>>From a pure cost/benefit perspective Linux is not always the best choice, but immediate
> cost-benefit cannot be the only consideration, when a little short-term cost buys a
> long-term benefit.
> 
> Let's take a couple example cases.  You don't have to agree with me on any or all of
> them, but they are examples of when ~I~ choose to pay a little more to "do the right thing":
> 
> A Chipotle chicken burrito is $6, and they treat they animals nicely before they kill
> and cook them for me.
> A KFC chicken meal is less bucks, but the animals may not be treated humanely.

This one's a non-starter for me.  Animals raised as food are already 
treated as food, not as animals.  Cut 'em all loose and start hunting 
them instead of having food factories, if this is a goal to treat them 
better... I figure.

Seriously -- "humane" treatment of animals... that are already herded 
into a pen as food, waiting to die to become a meal?  Not exactly high 
on my "moral priorities" list.  There's a reason the word "human" is 
part of the word "humane"... it's intended as a concept for how you 
treat PEOPLE, not animals intended as food.

Animals domesticated as pets, that's a different decision -- the humans 
taking care of those chose to take care of animals and breed them to not 
be able to survive in the wild -- so they had better not mistreat them.

(I'm not a total "ogre", but humane treatment of food, isn't important 
to me.  "Humane" treatment of pets, is.)

> Oil is $x per unit of energy produced.
> Solar, wood, wind, alcohol, all cost more today, but as people choose to pay a little
> more for clean reproducible energy, the benefits of commoditization and competition
> become realized and the price difference shrinks.

Solar: I did my social responsibility and ran the payback numbers on my 
house, and found it to be 17 years, even with non-free-market tax 
incentives.  More without.  I live here in COLORADO where the weather 
will DESTROY such a system long before 17 years is up.  Too expensive. 
Glad folks are working on making it better though.  Get it down to a 10 
year payback in efficiency, and I'm definitely interested.

Wood: There's already limitations in place to not allow woodburning as 
primary heating unless you have nothing else due to pollution issues.  A 
non-starter in urban cities.  A family member already heats their home 
with wood here in the 'burbs of Denver, and the only way you can make it 
cost-effective is to have enough free time (retired in his case) to hunt 
down very large supplies of wood throughout the year, and many many days 
of labor to prepare it for use in a very high-efficience wood stove.  He 
ran his natural gas system about five days out of the last year, but 
calculating the amount of time he had to put into it to get there, I'd 
have about 10 weekends left a year to myself if I put in similar time to 
accomplish it.  Not to mention the serious pollution problem we'd have 
if even a significant percentage of homes in a large metropolitan area 
burnt wood as a fuel source.  I think he's technically breaking the law, 
unless he were to disable his natural gas system altogether.  Don't 
know.  Don't have time to do it, so it'll never matter to me until 
retirement age.  If it makes economic sense (he gets much of his wood 
for free, downed trees, neighbors cutting them down, etc... and word 
gets around that you're a "wood" guy, and that you have chainsaws... and 
will travel...) -- I'll do it.  Not right now.

Alcohol: Same problem with power produced versus amount of fuel, like 
solar.  Serious problems with affecting the cost of the food supply and 
additional environmental problems that would likely result from the 
additional farming activity if "everyone" were using it, too.  Good in 
some circumstances, not so good here locally.

So I'd like to believe you that paying the extra for the above 
"solutions" would work, but I don't think they're viable, long-term. 
They're "hedges" against high petroleum costs, but like Linux on the 
Desktop -- they're "also rans".

Natural gas, on the other hand... is here, not imported, and we've 
learned how to get at a whole bunch of it with horizontal drilling and 
other technology lately.  Burning it in vehicles is also technology 
that's already here... I'm amazed it hasn't taken off economically more 
than it has.

It'll start with large fleets, and others who can fuel up at the 
specialized stations for it... but it's ready to go, today... burns 
cleaner than petroleum products, and can sustain the country's needs for 
a number of generations as the other technologies come up to speed.

> Throwing away plastics costs me almost nothing, but has a long-lasting and negative
> effect on the environment.
> Recycling is much more expensive, but minimizes the negative impact on the environement.

Change that to "throwing away anything" and you'll feel like you're from 
the generation that grew up during the Great Depression.  Plastic 
baggies can be washed (people freak out at that one), plastic water 
bottles and $2 for less than a gallon of water was ALWAYS retarded (What 
made it so popular?  Haven't people ever heard of tap water?  I guess we 
did all that work to have indoor plumbing for nothing...), and the list 
of things we throw out goes on and on.

How hard is it to recycle a COMPUTER, trying to stay to things THIS list 
cares about...?  (Anyone want a large out of focus monitor?)

Why do we need new computers ever three-five years?  Could the software 
folks work on making the software for the one I already have MORE 
EFFICIENT instead of leading a race to the CPU speed frenzy we're 
already at?  Linux has some niche people doing this for older machines, 
but the majority want "gee wiz" stuff like Compvis?  Why?

On this one, I think we definitely agree -- recycling is the right thing 
to do, and always has been.  What's it got to do with VMWare versus 
[insert virtualization system that doesn't require Windows here].

> Windows is $x, but every person that only uses and only knows Windows is one more
> prisoner to MS's interoperability FUD.

This is the crux of it, really -- they're not "prisoners", and they 
don't feel like such.  They either get done with their computer what 
they want to, or they go do something else.

There's NOTHING that Linux offers them as far as FEATURES that 
Windows/Mac/Whatever don't also already do.  If Linux lived up to its 
HYPE about how being "open" creates BETTER products... they'd be beating 
down the door to install it.  But it really doesn't, or they would.

What does Linux *DO* that Windows doesn't that any regular person cares 
about?  Sure some of us geeks can name some esoteric things, but none of 
them are anything the average person cares about.

Show them something BETTER than Windows in Linux and clean up Linux so 
it's virtually maintenance-free, they'll be BEGGING to install Linux.  I 
promise.

(It worked for cars.  Early Honda's weren't good performers, but they 
built a reliability rating for themselves over decades, and also worked 
on OUTPERFORMING the competition while staying reliable as all get out. 
  Linux needs to start outperforming Windows/Mac.  Problem is... that's 
hard to do with volunteer coders...)

> Linux is $y and/or z hours of downloading and burning, and any person that knows any
> Linux, can change OS vendors with an ease unprecedented in this industry.

The average user doesn't want to KNOW Linux.  They want to know their 
APPLICATIONS and how to get around the desktop.  They're not IT folks, 
not interested in the slightest about how the "car" they bought works... 
they take it to a pro to get it fixed when it breaks, and they follow 
the manufacturer's instructions on maintenance.  They also rate "cars" 
on how much maintenance they require.

Linux needs to be low-maintenance, cheaper, and better for the average 
Joe to want it.

The OS as a battleground is only a luxury geeks with spare time can 
afford.

Everyone else has things to do (gotta cut up that firewood, for example) 
and isn't going to invest huge amounts of time to learn how to care and 
feed Linux.  Linux needs to care and feed ITSELF and also offer 
something better than the commercial offerings, for it to survive the 
Desktop market.  It doesn't.

In the server space, Linux holds its own by virtual of being stable and 
more security conscious than other systems.  Hiring syadmins hasn't gone 
away yet (the ultimate goal of the "best" server software ever), so 
there are people around to handle the care and feeding of Linux in the 
server farm -- it's just less care and feeding than Windows.

That's great.  But this is where the VMWare decision hits the road... 
they just don't care if they lose you because you won't run a single 
Windows desktop.  Given enough pressure, they might fix the problem, but 
it takes away time they would rather spend on their core technology -- 
virtualization.  Being competitive in the virtualization world takes all 
their resources... writing and testing cross-platform control software 
is probably one of the LAST things they want to spend time, money and 
resource on... for a Desktop that has less than 10% market-share.  They 
know deep down that Linux on the Desktop isn't there yet.  They'll make 
a Linux version when it is... or if it'll cause a SIGNIFICANT hit in 
their bottom line.

But demanding it of them will also indirectly make them take resources 
and time away from their core product.

Nate