Strategy for fixing Bug #1

Wed Dec 27 21:38:29 UTC 2006

Below is the response from Rob Landley on the Vista 64 issue:

I'll format it in interview fashion so it's clear who's saying what.

Wes: In your essay, you assert that Windows-64 doesn't exist and that MS
still needs to provide an OS to fill the 64-bit desktop demand by 2008.
However, you don't really backup your claim that Vista is still a 32-bit OS.

Rob: We wrote that over a period of about 6 months during which Microsoft's
release
plans changed something like 3 times.  That part was written well before
Longhorn's plans firmed up enought to know what they were actually shipping,
and in a period when every major new feature had been cut.

Microsoft has had a number of 64 bit platforms over the years.  The Alpha
version of NT was 64 bit, and they had an itanium version of windows years
ago (which has been discontinued and resurrected at least twice; they insist
an Itanium version of longhorn is coming soon).  An x86-64 version of server
2003 shipped last year:
http://www.informationweek.com/showArticle.jhtml?articleID=161502192
And although everybody I know who evaluated it declared it essentially
unusable, this is nothing new for the first version of a microsoft product.

Recently they've realized the importance of x86-64 and have been moving to
catch up:
http://www.informationweek.com/showArticle.jhtml?articleID=170700287

Although a year earlier they were still claiming that the Itanium version
was
the important one:
http://news.com.com/2008-1016_3-6137579.html

*shrug*  They change direction a lot, and then retroactively claim that's
what
they were always doing.

Wes: Vista 64 comes in the same box with every edition besides "Starter,"
and it can address 8+ GB of RAM. It can also run basically all Win32 legacy
apps as well.

Rob: You've tried these legacy apps under it then?  My friend Stu in Austin
has and he has a longish list of applications he's already seen it can't
run, including the database underlying Exchange.  (I believe their current
approach to fixing that particular problem is to announce a new version of
Exchange they want everybody to migrate to.)

Wes' response: No, I haven't. I was playing devil's advocate, but didn't
make that clear, sorry. I have no doubt there are issues w/ it today, but I
also have faith in Microsoft's ability to preserve its monopoly by throwing
devs at these issues until they're basically solved. The theory, which could
very well be the reality by 2008, is that Vista 64 supports Win32 apps (but
not Win16, DOS, etc.) I just think we can't bank on MS not solving these
problems. It's not like they don't know they need good Win32 compatibility
in Vista 64. (I know you know that too, I'm just trying to clarify where I
was coming from, not debate your point.)

Wes: [Vista 64]'s a bit hamstrung by the driver situation, but that will
almost certainly improve quite a bit by 2008.

Rob: And improved == solved?  How do you know?  You could make exactly the
same
statement about Apple.

Wes' response: Improved != solved, but they have a monopoly, so they don't
need to be perfect, just good enough to remain the only game in town (not
that they're competing on quality, but they have to cause no more pain than
the perceived pain on the part of OEMs and end users to switch to something
else). Yes, you could make the same statement about Apple. Again, this is
just me having faith in MS's ability and desire to preserve their monopoly,
whatever the cost. I think we have to proceed assuming that to be the case.
Especially when we're talking about temporarily putting our ideals of
software and encoding format freedom on hold to take advantage of this
transition window. The opportunity had better be real and worth it. If MS
has already lined up all the pieces on the board to execute their strategy,
then there may not be as much of an opportunity here as we thought. (I don't
think that's the case, though.)

Wes: So, in what way is MS not already sitting pretty w/ regards to the
32-to-64-bit transition?

Rob: I said don't write them off.  The 32-bit version was the clear focus of
longhorn until a few months ago (they were struggling to ship _anything_),
and now they're moving to catch up.  Maybe they will, I don't know.

Wes' response: OK, so it sounds like Vista 64 is the Windows-64 referred to
in the essay, but it's existence is new enough and unproven enough that the
jury is still out and thus the window of opportunity is not necessarily
closed here. (Correct me if you disagree, Rob.)

Wes: A few of us are discussing your essay on the ubuntu-devel-discuss
mailing list, and we hit this point of confusion, so I thought I'd see if
you can clarify. Might be good to add some clarification to a new revision
of the essay as well.

Rob: That bit was written something like april.  When we started the essay
it was
1) unclear whether the most recent ship date for Longhorn meant anything
more
than the previous 5, 2) what features (if any) would make it into the final
version.  The 64-bit release was very much not a priority. :)

Now Microsoft's found 64-bit religion the way they found the internet: at
the
last minute, and in a big way.  There's a lot of hype about Longhorn (32 and
64 bit versions alike), and on the 64 bit side it's very unclear whether
Microsoft is in any better position with win-32 backwards compatability or
driver support than apple is.

I don't know.  I'm not a windows expert, and the game is still in play.  The
switch to Windows 3.0 happened more or less unopposed.  This time we're
pointing out when the transition point is and what the requirements are to
play in the new space.  (Apple already knows.)

Wes' response: Would you say, Rob, that is equally unclear where MS is at w/
Win32 compatibility compared to Linux/Wine as well? Are we not in a better
position than Apple there? Or about the same? Surely not worse...

So, in light of all this, it sounds to me like there is still a pretty huge
opportunity here. If Ubuntu can put together a formidable 64-bit desktop
offering (the major missing pieces, in my opinion, being tight Wine
integration and Codex support), then we can start working on an OEM
pre-install strategy. But I think we need something to sell before we go
knocking on doors.

Also, Rob or Eric, would you mind updating us (and perhaps the essay) on
where things are at with Linspire and the Codex? Is there a webpage where we
can go for updates or ways we can help out?

I'd love to see Ubuntu position itself as a contender for the 64-bit desktop
crown, and thank you Rob and Eric for writing this essay to bring our
attention to the opportunity.

Wes Morgan

-- 
"Small acts of humanity amid the chaos of inhumanity provide hope. But small
acts are insufficient."

- Paul Rusesabagina, Rwandan and former hotel manager whose actions inspired
the movie Hotel Rwanda
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.ubuntu.com/archives/ubuntu-devel-discuss/attachments/20061227/17a0d43c/attachment.html>