Re: [NTLK] [OT] Apple, Macintosh, 3rd Parties, and Market Share

From: David Ensteness (denstene_at_mac.com)
Date: Sun Apr 25 2004 - 09:48:23 PDT


> I'm trying to spare your blushes here because if your best example of
> innovation on a platform is a comms port then it's getting scary.

No actually it was included in a list of many .... Short memory you
have there Joel. And since when is I/O not innovation? Was ADB not a
big deal? A daisy chainable, powered, hot swapable [for the most part],
peripheral bus? Huh, sounds like USB ...

Want me to list more?

*AirPort - first and best.
*Altivec - argue as you might Apple does do R&D and fund AIM alliance
projects, they also build the supporting chipsets.
*The G5 logic board architecture - in a world where system throughput
is making faster processors worthless, 6.4GBps throughput to the
processors is sorta a big deal, and I doubt it was just a "couple hours
work" to put it together.
*The iPod - funny, its all "prepackaged" as you said, but no one else
can touch it ...
*iTunes - bought/adopted from SoundJam yes but very changed since in
its inner workings and functionality.
*The iTunes Music Store - Hmm, 2.5 million downloads a week, 70% of the
market, over 50 million downloads at last count two months back. Not
bad when online music services had been around since 2000 and done
nothing ...
*Motion - Talk to a video person if you don't think its cool ...
*Soundtrack - available for Linux as well ...
*GarageBand - DTP for upstart musicians?
*Xcode - huge hit with programmers, read some reviews.
*Expose' - Some people prefer virtual desktops, some do not, but
Expose' is a nice step forward in window management.
*Quartz and Quartz Extreme
*The use of the QuickTime API in Mac OS X
*Xserve - If you don't see this as a big deal, you haven't much
knowledge of it.
*Xserve RAID - PLEASE, PLEASE find me another, cheaper 3.5TB hardware
RAID 5 system that is certified for OS X, RedHat and YDL, Windows, and
Solaris that is only 3U in size ....
*Rendezvous - OpenSource is our friend, a friendlier Zero-Config IP
networking standard, nothing bad about that. Self discovery and
configuration over IP networks, man, its like AppleTalk but fast and
cross platform.
*Webcore - Can't wait until Keynote is able to render HTML for me
within a presentation ...
*Keynote - heck, 4 hours to do a PowerPoint, 45 minutes to build a
likewise Keynote presentation, lots of refinement I would like but a
very excellent start.

> The innovation was in realising the need for better connectivity and
> creating FireWire which, in every possible way is superior to USB, but
> may
> wither because USB is so ubiquitous. Damnable market share again.
> Peripheral
> manufacturers have a very easy choice now that USB 2.0 High Speed is
> "good
> enough" - 1 fast port for all users - simple.

How so? FW is now ubiquitous on DV cameras and video equipment, has
been for years thanks to Sony pushing the 4-pin version. I have yet to
hear of a DV camera that has USB anything on it ... I work at a TV
station ...

> This is just small beer - lets turn this conversation around and see
> what
> would make Mac OS a no-brainer choice for everything from our Mums to
> the
> most hard-bitten pro.

The everything to everyone approach - I remember that, it let to 72
different models of Performa, it led to continual product shortages in
the product lines that were selling and over abundance in those lines
that were not all through the 1990s. Worked great then ...

> Some of the big issues are complexity, data security and interface,
> none of
> which Mac OS X makes any attempt to redress or has any ideas about.

Do you honestly think that Mac OS X does not make UNIX simple? Have you
used Mac OS X? Have you used BSD?

> There is all this computer performance but little of it is being used
> to make good
> guesses about what I want.

MS is very big on trying to have software guess what the user wants
next, they set up systems to collect user actions and then react to
them. I do not want this. I do not want the company's software deciding
"where I want to go today." Those are choices I want to make myself. A
computer is a means, not an end. The Newton, which was your example for
all of these, does not guess what I want unless I tell it to use the
Assistant. That is important because system such as WinXP that try to
interpret a user's needs generally get in the way of the user. I have
no misconception about the computer doing the work for me, the computer
is a tool, I use it to do the work.

> 2/. Dump the file system. This archaic piece of rubbish should give
> way to a
> soup-like database system. MS are going probably to do it anyway and
> make us
> look stupid, so get in there and do it properly. The full monty with
> roll-back, field-level locking - the works. Guarantee that users will
> not
> lose hours of work and hearts and minds will be captured.

Get up on your reading, both Apple and MS are developing database
attribute based file systems. MS says they may be including it in
Longhorn which is now due sometime in mid to late 2007, Apple does not
say when they will include it but it seems likely it will not be until
at least 10.5 which means it will be around 2005 or 2006. Roughly the
same timeline that MS has for an OS that runs on a database.

> Make the Finder spatial again.

You are not comparing your own notes. If you use a database file
system, you do not want a spatial interface because a database is not
spacial. You want a level of abstraction so that database attributes
can be edited from anything that can access the file so that users have
universal tools. That is why Column view is a big philosophical leap
forward, the old Finder was mostly spatial, that was terrific, computer
interface was based on the office/desktop metaphor. Except as computers
continue to evolve that metaphor makes less sense. Less users have
worked in the office without computers on which that metaphor was based
so much of the familiarity it brought has been lost. Eventually
computers will need an interface that is native to how they really work
and not native to a model few have used in the real world for ten to
fifteen years. As this happens the interface should become less spatial
because computers are not spatial. This should also allow a greater
number of interface points for any given piece of data.

Oh, a good counter example: Your Newton is not spatial.

> Be consistent.

Mac OS X's user interface is actually quite consistent, care to point
out where it is not? I am sure there are some areas, but I would be
curious how pervasive you think this inconsistency is and where you
think it shows up.

> Don't hide functionality behind glitzy or invisible controls.

Here I am going to quote you:

> Too much choice is bad for mental health so hide a lot
> of the complexity with more intelligent code and better UI design.

OK, less choice, hide complexity, but don't hide it and don't put a
pretty face on it. I am confused here, want me to go make you some cake
so you can eat it too?

Its a very thin line to walk, just because you do not like the method
does not make it bad, they have made different design choices than you
might have, but I am still interested in how much experience you have
with Mac OS X as you stated you do not yet use it.

I know in a perfect world and all.

> Seriously improve responsiveness.

Have you used Mac OS X v.10.2 or 10.3? Its a four year old distribution
that we are working with, sure its roots are older but X has been
around in this form since 2000 sometime and it has had three revisions
in that time. If you have not spent much time using it in the last two
years then you may not know that criticizing it for speed went out of
style when it continued to increase in speed year after year.

> Actually design the software before building it. Hire the right people
> and
> do the R&D to move things forward. Stop tinkering and start innovating.

Have you any knowledge of who they have hired or what their software
development model is or are you just talking?

> 4/. Stop using metal for laptops. This has to be the dumbest materials
> choice in the history of the company.

Based on what? People absolutely love the look, it is the envy of the
industry since the release of the Titanium, yes it can scratch, I have
one that is scratched to hell at work, it was my predecessor's. I also
know several people who use them in the field for their main computer
and the things look mint ... they tend to have scratches more than
plastic cases [visiable anyway] but that alone does not make them a
horrible choice. I have heard no reports of the Aluminum ones
scratching up badly. There have not been any bad case cracking issues
like with the 5300/190 series, I would say that was a worse materials
choice, two entire lines that had their cases become brittle and crack?
Sorta a big deal if I recall.

> WiFi might actually work properly and users would not feel like
> they were being electrocuted/frozen/cooked.

Works well on the Aluminum PowerBooks [the current 12", 15", 17"] from
the reports I read. I will have one next week so I can tell you from
personal experience at that time.

Oh and who thinks that they will be electrocuted?

> And this is just for starters.

I would have to ask, how much use have you had of Mac OS X and what
versions?

> What would you add?

Oh I would certainly do many things differently, I think a database,
attribute based file system where different applications had access to
the same attributes and shared functionality, similar to the OpenDoc
concept, would be excellent. Something where work focuses more on the
data and less on the applications that manipulate it. Examples would be
e-mail on the Newton or Be OS.

I think that I should be able to have complete metadata, both
file/application/system and user defined. I should be able to edit it
in my file viewer/Finder. I should be able to organize it there and
that organization should be respected by other applications by iTunes.
I also feel that it is distinctly possible Apple will go this
direction, I certainly hope so. As for hiring the right people, Be OS
had the most advanced file system to date in terms of database
integration, and many of the lead programmers from Be are now working
at Apple on the Finder and file system.

David

-- 
This is the NewtonTalk list - http://www.newtontalk.net/ for all inquiries
Official Newton FAQ: http://www.chuma.org/newton/faq/
WikiWikiNewt for all kinds of articles: http://tools.unna.org/wikiwikinewt/


This archive was generated by hypermail 2.1.5 : Sun Apr 25 2004 - 12:00:02 PDT