[Tfug] 2 weeks of Hackintosh fun..

[Bexley Hall]

--- On Sun, 11/9/08, Shawn Nock <nock at fastmail.fm> wrote:

> Bexley Hall wrote:
> > 
> > Take away her IT department ( <grin> ) and give her a live CD.
> > Give her a brand-spanking new machine to sit in front of.
> > Supposedly, "an hour or so" (I believe that is the quote?)
> > later she should be back where she started, right? 
> >
> > <frown>
> 
> Point taken, but I'd spin it a bit differently. Some
> people will have trouble with a procedure and require 
> assistance no matter how simple it may be.

Yes.  But, the FOSS community does very little to help
ease those problems, IMO.

First, there is no way to know a priori if this distro
will run on this particular piece of hardware -- unless
you get someone you trust to give you that assurance.

So, if things don't go right, the user wonders if "he"
did something "wrong" -- or if his machine is "broken".
OTOH, if he is trying to install something on a Windows
platform (I assume the same to be true for Mac), he's
got a lot more confidence in the fact that it *should*
work (even if it doesn't, currently).  I.e., there's
very little chance that "this camera won't work with
my PC (and, by PC, he means hardware+OS)"

The FOSS community also has done very little (IMO) to
make things friendlier for the user.  Yeah, sure, you
get a pretty dialog box and glitzy garphics so you can
type in your email address instead of having to enter it
in a config file...  But, when something doesn't work,
you get some cryptic error:
   "assetion failed:  line 23 foobar.c"
Of course, MS is no better (no comment wrt Mac).  But, here
is a place where -- if your *GOAL* is to come up with a
better UBIQUITOUS computing platform -- you can easily
add LOTS of value!  E.g., in the product I am working on
currently, "errors" are pointers into the online documentation.
So, instead of a stupid error code, you get a section of the
manual that tells you what/why -- assuming the code can't
work around it for you.  :-(

Likewise, the FOSS community has done very little towards
improving the *quality* of the software it releases.  I am
*so* tired of the "well, you EXPECT some bugs" rationale.
Do you expect your car to spontaneously shift into reverse
and explode in a ball of flames?

Instead of improving the existing functionality, there seems to
always be a drive to add new bells and whistles.  As if folks
are easily bored with the prospect of "making it WORK (right)".

The documentation follows the same set of issues.

It just makes the whole system reek of "not finished".

> The support options are, to my mind, identical for Linux and Windows,
> Either marry savvy guy/girl or pay for support from a vendor.

I suspect the practical support option for most Windows users is
"ask someone -- ANYONE -- at work/play/home" as, chances are, they
will at least be a *user* of Windows.  Not true of Linux or any
of those other OS's.

> RedHat, Novell (and etc) will give you a number to call that is
> probably just as
> useful as GeekSquad. Some vendors support the hardware and
> software on their Linux based machines. GeekSquad and their ilk 
> are incredibly popular because the 'mainstream' doesn't
> understand computers no matter what the OS.
> 
> Support options aside, I give my mother even odds with a
> Ubuntu CD and a Windows CD (which is to say, not good either way).

I'd suspect she would have a working Wintel machine in "about an
hour" -- though adding to that would require work.

It is *VERY* hard to deal with users.  Especially for ubiquitous
devices.  You can't impose some prerequisite:  "You must be THIS
TALL to go on this ride".  And, you don't *want* to as that means
lost sales opportunities!

And, it is painfully expensive to support users.  The current
support model shifts the cost of support entirely onto the user.
As if the PC had two pricetags on it:  "Price for experts:"
and "Price for dummies:"  <shrug>

With this model, there is no incentive for the manufacturer to
improve the quality (in this case, usability) of their product.
Until the market shifts away from them (not likely for MS).

If, OTOH, "support" came out of "development's" cost center, then
the folks in Development might be more motivated to produce a
friendlier device!

Now, think of this in the FOSS context.  A developer has no excuse
for releasing a buggy piece of code, an undocumented application,
an untested application, etc.  You can't whine that your *boss*
didn't give you time to finish it "the way you wanted to" (which
implies that you really would have done a bang-up job -- had you
only been given the opportunity  ;-).  And, you can't complain
that you were trying to meet the "christmas rush" to boost your
sales for the year.  Or, that some piece of code that you were
relying on (a library? another application??) didn't behave the
way you *expected* it would ("Use the Source, Luke").

So, why WOULDN'T the FOSS products have *impeccable* quality/support?
At least MS has the excuse of "The Marketplace" and "stockholders"...

Apple is obviously trying to take this approach.  I can't attest
to the quality of their products... every time I get a Mac to play
with, I end up giving it away to someone who needs/wants it more
than I.  <shrug>

OpenBSD tried to adopt the positioning of "security".  NetBSD opts
for "universality" (?).  No idea where FreeBSD has gone in the years
since I last played with it.

Instead of NetBSD's "Of course it runs NetBSD" slogan, there needs
(IMO) to be a push for an "Of course it works" FOSS OS.  :<