A task-centric desktop...

[Liam Proven]

On 21 November 2011 18:35, Justin Gruenberg <justin.gruenberg at gmail.com> wrote:
> On Mon, Nov 21, 2011 at 9:28 AM, Liam Proven <lproven at gmail.com> wrote:
>> I can give you a detailed, point-by-point comparison of the
>> GNOME/KDE/XFCE/LXDE desktop and the ways in which it uses methods,
>> designs and techniques that were original Microsoft innovations from
>> the early 1990s, as opposed to and compared with existing prior art in
>> those fields.
>>
>> If you want details of which MS patents, the company is not disclosing
>> this, intentionally. So no, there, I can't help you.
>
> So we really do not have any idea on what their patent claims are...

MS has described the areas covered.

If you know your GUIs and their history and development /well/ - and
that means knowing them in the period /before/ Windows, not Windows 95
but before Windows 3 and indeed Windows 2 - then the descriptions of
what is infringing are really quite clear pointers.

The trouble is that we're talking early to mid 1980s for the history
of GUIs and WIMPs (which are /not/ the same thing) and then the period
from the mid-1980s to the mid-1990s for the stuff that MS came up with
which was new and different.

Very few people still around in this business actually remember back
then. Many just take it all as read, and maybe rely on sayings for the
history, e.g. "Apple stole the GUI from Xerox", and don't actually
know who took what from where.

Trying to reconstruct it *now* if you were *not* there is hard,
because it's all pre-Web and is very poorly documented in print.


> so how does taking shots in the dark at what they _might_ have patents
> over help?

Actually, yes, I think it does. (Although Rice-Davies applies here, of course.)

This is an important threat and that is not well-enough appreciated.

Linux is a Unix. No, it's not BSD code (or, well, /very/ little of
it), and it's not AT&T code, but it is indubitably *a* Unix. If
someone were willing to pay it could pass the X.Open conformance
tests, no bother.

That means it's a FOSS copy of a commercial system.

X.org is now /the/ official reference implementation of X.11 so it has
official status. But almost all the bits of Linux are, frankly, ripped
off. They are clean-room (mostly, kinda, ish, we hope)
reimplementations of code originated at AT&T or UCSD or UCB or who
knows what else. Almost all of it, from "ls" to vi to EMACS.

Now most of that seems to be fine now. The vendors don't care to chase
it. The only one that did was SCO and it failed, because it claimed
the /code/ was copied.

But as we progress up the layers, there is less and less diversity.

"ls" is "ls" is "ls". The differences between Solaris, AIX or Linux
are scant and don't massively matter.

But up at the desktop layer, it's different. There /are/
window-managers and desktops that are nothing like those on anything
else. Back in the 1980s, there was wild diversity: OpenLook was
nothing like the Mac System which was nothing like Windows... and GEM
was nothing like any of them, at least after Apple sued it into a
major change.

But now, the weird, the different FOSS WMs are obscurities. There are
things like ratpoison or twm or wm2 which look like nothing else. But
only a few techies use them.

Most people use Kwm or Metacity or Clutter - and they work just like Windows™.

File managers: the mainstream ones - Konqueror and Dolphin and
Nautilus and even Thunar - are significantly like Windows.

Ditto panels. Ditto application-launch menus. Ditto system trays.
Ditto desktop managers. (Think seemingly-small stuff, like how you
change your wallpaper, or what icons are present and how they look and
what they do.)

The main choices are:
KDE - a direct copy of the Windows 95 + IE4 "Active Desktop™", right
down to the way that filer-window contents are generated in HTML and
then rendered by the embedded browser.
GNOME 2 - highly Windows-like panels, menus, filer, desktop, etc.
Very much in 3rd place, Xfce - can be configured to be quite different
but contains distinct elements of the Windows implementations of
stuff.
LXDE - quite new, but the panel/menu/icons thing is straight out of Win95/NT4.

Also-runs that embody the same look and feel are IceWM and Fvwm95.

The Mac System, later MacOS - not Mac OS X, but classic MacOS -
evolved quite incrementally over 15y. It has some PC influences, but
they are quite scant. (e.g. the overlaid shortcut arrows on the icons
for aliases.)

But Windows 4, AKA Windows 95, AKA Windows Chicago with its GUI taken
from the Cairo project, was a single massive cohesive design effort to
produce a new friendly object-oriented (kinda ish) desktop GUI that
built on Windows 3 and thus on OS/2 and Windows 2, but which was
modern, new, and most of all, /different/ from anyone else.

It contains lots of new ideas that nobody had done before.

It's hard to believe now, looking back 16y, but it's true.

And the snag is that almost every new GUI since then until post-2007
uses some of those ideas. The /one/ notable exception is Mac OS X,
which is a weird fusion of bits of NeXTstep and bits of classic MacOS
and some new stuff.

Aside from it, they are /all/ Windows ripoffs. QNX is. BeOS was. OS/2
Warp 4 was, eComStation still is. GoMac on MacOS 9 was.

It is all stuff that anyone that started out with computers since 1995
will, entirely reasonably, just assume "that is how GUIs work."

But it isn't. It was all MS innovation. Yes, MS does innovate
sometimes, and some of the innovations are /really good./

> They have a number of patents covering everything from
> algorithms to user interfaces, and there is no certainty that those
> patents would withstand a serious challenge.  For all we know, the
> patents they "claim" could be the optimal snack to drink combination
> that results in the maximum productivity of their programmers.

We don't *know* but they've said what infringes, and from that, we can
work out /how/ given sufficient historical knowledge.

Before Win95, there was no such thing as a taskbar. Nobody did
anything /remotely/ like it.

Before Win95, there was no such thing as a Start menu. No system tray.
No Explorer window with a tree down the left hand side. This stuff was
all new and it is all MS intellectual property.

And much as I dislike MS and its approach to business, I think its
management would have been irresponsible and culpable if it did not
protect all this stuff.

And MS is not careless like that. It's /very/ litigious. Partly
because it is partly built and founded on stolen IP: MS-DOS 1 was a
direct ripoff of CP/M, and Windows 3 contains a lot of OS/2 ideas and
concepts, from IBM.

So I am sure it did, as, to be fair, it should.

Why it didn't enforce such things for over a decade, and then just
threaten, I don't know. I presume it was a tactic to "encourage" some
Linux vendors to sign up to IP-sharing deals.

> I think, as others have pointed out, claiming that the Unity and Gnome
> Shell interfaces came about because of Microsoft IP claims is without
> evidence.

[1] Main 2 Linux interfaces both use a MS look & feel
[2] MS rattles its sabre and indicates that the look & feel is protected
[3] Most popular FOSS Unix desktop (GNOME) junks its entire look &
feel and moves to a new one, despite much pain, a long, overrunning
development process & a very controversial new look
[4] 2 main Linux vendors /don't/ sign an IP-sharing agreement & shift
to 2 new different ones (Fedora -> GNOME 3, Ubuntu -> Unity), despite
upsetting thousands, maybe millions of users
[5] main vendor that /did/ sign (SUSE) announces that it is
re-emphasizing on KDE (an infringing desktop).

Does this /really/ look like a coincidence to you?

-- 
Liam Proven • Info & profile: http://www.google.com/profiles/lproven
Email: lproven at cix.co.uk • GMail/GoogleTalk/Orkut: lproven at gmail.com
Tel: +44 20-8685-0498 • Cell: +44 7939-087884 • Fax: + 44 870-9151419
AIM/Yahoo/Skype: liamproven • MSN: lproven at hotmail.com • ICQ: 73187508