Dr. Jerry Pournelle

Email Me

Why not subscribe now?

Chaos Manor Subscribe Now

Useful Link(s)...


Hosting by

Powered by Apache

Computing At Chaos Manor:
May 22, 2007

The User's Column, May 2007
Column 322, part 4
Jerry Pournelle jerryp@jerrypournelle.com
Copyright 2007 Jerry E. Pournelle, Ph.D.

Continued from last week...

The great furor over the great Microsoft Patent Scare (see last week's column) has quieted, and many have concluded that it was but another instance of Microsoft sowing Fear, Uncertainty, and Doubt. While spreading FUD was undoubtedly one of the goals of the Microsoft press release, I would not be too hasty in concluding that it was all pure bluff. Microsoft has very good lawyers, and a lively appreciation of the current political situation. That latter was not always true: Gates hated politics and political infighting, and until the monopoly lawsuits the only Microsoft Washington presence was a sales office; now, however, Microsoft is a major player on the Washington scene, and its lobby staff are both very good and nearly omnipresent when there's a Microsoft interest at stake.

The player to watch in this game is IBM. IBM is the only major user of Linux and Open Office with the resources to stand up to Microsoft. It also has no real interest in doing so. It is extremely likely that Microsoft will make IBM a fairly generous offer that includes cross licensing and the agreement not to sue each others' customers - i.e. a deal very similar to the one Microsoft has with Novell - and that IBM will take the offer.

This will leave Red Hat and other Linux distributors on their own. What Microsoft does after the game board is cleared of the big players with large legal resources will be interesting.

Meanwhile, the open source software coalition is pinning its hopes on the tricky legalities involved in patent law and the cleverness of the new GPLv3 (GNU General Public License Version 3) to stop Microsoft cold. Many ingenious arguments are used; see for example this link at Groklaw. Some otherwise savvy people are enthusiastic over those arguments. I wouldn't be.

There is one major wild card in this deck: the US Supreme Court has never ruled that software is patentable, and some of the open software coalition plan to pursue this. The argument is that software is a set of algorithms, and algorithms have long been ruled unpatentable. Software can be copyrighted, which protects a particular implementation of an algorithm, but not the algorithm itself. This is not a frivolous argument, and a number of talented legal theorists believe the case can be made. On the other hand, the implications of ruling software unpatentable are huge, and the Supreme Court does read the newspapers including the financial journals. If I had to predict the outcome I'd bet that the SCOTUS will not overturn software patents; but I'd sure not be comfortable making a very big bet.

As to how Microsoft may profit from it's patents, Microsoft isn't likely to start suing its biggest customers. They are, however, perfectly capable of making life difficult for publishers and developers and distributors of Linux and other programs that violate Microsoft patents - indeed, it could be held that Microsoft has a fiduciary duty to its stockholders to exploit the value of those patents by all legal means.

I do not believe we are past this can of legal worms.


I went to WinHEC 2007 in the Los Angeles Convention Center last week (May 15 - 17).

It was the smallest WinHEC I have ever seen. Lots of sessions, but I estimate that attendance was about half what it has been.

There was also a lot less enthusiasm. Chairman Bill Gates gave his usual presentation, but he started an hour later than usual - for which I thank him, since I hate those dawn sessions - and he may as well have phoned in his presentation. I have not seen him less enthusiastic, despite trying to put the best face on things.

WinHEC is supposed to be about hardware and the future, and there was some of that, but not as much as I expected. Instead the big emphasis was on convincing the attendees that the Vista Launch was a great success, and that Vista is solid and here to stay. The first part, about the success of the launch, was explicit. There were charts and graphs, and the triumphant statement that after four months the number of copies of Vista sold exceeds the entire installed base of the major opposition. Microsoft commands market share and Vista is Microsoft's new product, and there's an end to it.

The insistence that Vista Marketing worked, and there was great enthusiasm about the launch - we saw scene after scene of people all across the world being enthusiastic at the launch - brought an unfortunate association to my mind: it reminded me of CBS and Katie Couric, insisting that the "launch" was successful long after the ratings fell. I hasten to add that there's not much to that mental association, and I probably ought to drop it. Still, showing launch scenes doesn't do a lot to convince me that the product is meeting everyone's expectations.

I had hoped that there would an announcement that Microsoft had abandoned the insane pricing structure of Vista, with its myriad versions, most of them resulting from crippling one or another feature in Vista Ultimate. I can't understand why they do this. If I had a product I was proud of I'd want everyone to use it. I don't allow my publishers to print up abridged copies of my novels to be sold at cut rate prices. When I put a scene into a book, it's because I think it belongs there. Alas, there was no discussion of Vista versions at all.

WinHEC press diehards
The diehards of the technical press, who stayed on after Chairman Bill Gates left. It's about as small a group as I have seen at any WinHEC. [View larger]

There was discussion of things to come, and demonstrations of the uses of Vista in the home to set up home networks and entertainment. It was impressive, of course. Such things usually are. And it all works in the demonstration. Things get connected and just automatically network properly. In practice, users are not quite so happy about it all. In my own case, I have real problems with Vista and networking. I am sure it's all my fault, but I still have the problems, and I am not alone. I am sure it will all be fixed, eventually, but for now Vista can drive me nuts.

What I got from the opening WinHEC sessions was that much is happening, things will be fun, and it will all work Real Soon Now. I also conclude that until then there's no reason to change my recommendation. If you can manage without Vista until the first Service Pack is out, do so. You will save yourself a fair amount of frustration. Stay with XP a while longer.

Roger Kay Presents

Microsoft and Waggener-Edstrom are not stupid. Far from it. They realized that while their reassurances to the technical audience - WinHEC is almost entirely filled with technical people, real geeks not marketing types - and to the general press would probably get their cheerful message about the Vista Launch across, the technical press probably wasn't buying it.

They arranged for Roger Kay, President of Endpoint Technologies Associates and a pretty thoroughly credible guy, to make a presentation just for the press. As I surmised, few - I think none -- of the general and market oriented press people came. The presentation was made in a small conference room and there were fewer than twenty sitting at the round table.

He began by repeating Gates's statement: in these early days of Vista, there are more Vista systems already in place than Apple's entire installed base. Microsoft has the numbers to prove it. Of course back in the heyday of General Motors I am sure there were more Chevrolet sedans sold each year than the cumulative total of all BMW automobiles ever sold, but I wouldn't infer from that most people were happier with Chevvies and preferred them to Beamers. It does mean that Microsoft can't be ignored, and software developers will see that big market share when making decisions about allocation of effort.

Roger Kay waxes eloquent
Roger Kay waxes eloquent to the technical press. [View larger]

Kay's presentation said many more of the same things Gates had said, but went into details. For example: technical support calls for Vista problems are down to under half the number received at this point after the XP launch. That sounds impressive until you think about it: when XP was launched there was far less on-line support available. People weren't so used to using the Internet to find problem solutions. They are now. The number of tech support calls may be down, but this doesn't necessarily imply there are fewer people with problems. It does mean that fewer result in calls for support.

Kay emphasized the increases in "telemetry" built into Vista. Microsoft is finding out about driver problems faster, and getting more information on what to do about them. Problems with major software are worked out, with Microsoft helping the publisher, and in some cases has inserted a small work-around into the Vista code itself. This has been very effective. One presumes these will trickle out with the now routine updates, but it's certain they'll be in Service Pack One...

Kay repeatedly brought up the horror stories from the early days of XP, showing that there seem to be fewer about Vista. This is not only correct, but it's good to remember. The Vista rollout has not been worse than the XP rollout, yet we are dealing with a much more complex mix of devices and software.

Kay called his talk "Managing a Complex Ecosystem," and I noted that Microsoft presenters in other sessions used the word "ecosystem" a lot. I suppose this is inevitable. I would also think it deplorable but it's probably too late to do anything about it.

When I was an undergraduate, the word "ecology" was fairly new in intellectual circles. It was used a lot, and meant, more or less, that "everything is connected to everything else," and "you can't do just one thing," which is to say that whatever you do, it will affect other parts of the environment, often in ways you didn't expect.

There were a number of essays and articles about ecology, and a number of science fiction stories illustrating the above principles, so in 1953 or so I thought I would take an upper division course in ecology. Imagine my astonishment when the professor told me I could take the course just after I completed differential equations. In those days ecology meant not merely claiming that everything affects everything else, but trying to find some quantitative description of those connections. That's all changed now, and I suppose I ought not wince at the thought of the computer-using community being "a massive ecosystem."

When I put it to Roger Kay that the usual advice to sophisticated users is to wait for Service Pack One before adopting a new Microsoft OS, and it seemed to me that is good advice regarding Vista, he wouldn't argue with me; which says all I needed to know.

I Have Seen the Future

I don't have time to do them justice: but the pocket computer first described by Niven and me back in the early 1970's in The Mote In God's Eye is coming onto the market. One emphasis in previous WinHEC conferences has been on ultra-portable devices. They're almost here. Small enough to carry, wireless communications, audio and video recording, very readable screen, the ability to hold thousands of books, keyboard large enough to be useful as well as a TabletPC touch screen, and a full fledged operating system complete with handwriting recognition: several were on display in the WinHEC exhibitions.

These machines have reasonable battery life and do nearly everything. At the moment they are priced above $2000; but we all know what happens to electronics prices. When these get below $500, and they will, they will become ubiquitous like cell phones. I expect clothing fashions to adapt to the necessity for carrying one, and it will replace cell phones, radios, PDA's, cameras, and essentially all personally carried consumer electronics. When that happens - and when it's as easy to read on this device as it is to read a paperback book - will anyone want to buy paperback books and carry yet one more object around?

Coincidentally but appropriately the Author's Guild sent out a warning about publisher practices in this changing age (link). The Science Fiction Writers of America is putting together a statement of support for the Guild, and there are signs of author solidarity sufficient to cause some publishers to stand down.

I believe we are seeing the beginning of the end for the paperback book business as we know it. I realize that may sound extreme; but I mean it. I'll expand on this another time.

Piggy Old Outlook

Outlook 2003 remains piggy but considerably less so since Microsoft sent a new "update" canceling the old one. I have, however, restricted Outlook 2003 on my XP main communications system to one processor. To do that I have to open Task Manager, Processes, choose OUTLOOK, right click, choose set affinity, and tell it which CPU it can have.

Since Outlook uses up all the CPU power it can find, but doesn't need both CPU's - most of the delays are in bringing in mail, and are a function of I/O problems rather than lack of processing speed - I do not want Outlook to take over all the power of my machine whenever it feels like it. It runs about as fast on one processor as on two.

If Outlook has both processors, it can interrupt my work when I am typing in Word or FrontPage, and that's irritating to say the least: I hate typing half a sentence without seeing any words appear on screen. I am, I confess, a very sloppy typist. By restricting Outlook to one CPU I don't slow piggy old Outlook by much, but I do allow most of my other programs to work without interruption.

The problem is that I must set the CPU affinity each time Outlook closes and is reopened. This is annoying. I was resigned to living with that annoyance when I got this letter:

Subject: Permanent affinity

I just read today's "Computing At Chaos Manor" and was happy to see that that affinity trick helped in making Outlook less piggy. I've used the set affinity trick more than a few times over the years with recalcitrant software with temporary bugs.

For what its worth and although there may be another way (and of course YMMV, etc.) - to set application affinity permanently so that it survives a reboot requires use of imagecfg - a good description and instructions are at this link.

Warm regards, Gary

I confess I haven't tried this yet: with a big trip starting this week, I haven't been in an experimental mood. I'll get at that when I get back. If anyone does try it, let me know the outcome.

Coming Up

I have a large collection of gadgets, some wonderful like the handsome Crosley unit that plays LP's, FM Radio, and converts whatever it plays to digital format, all rather painlessly. I have a new kind of electric toothbrush that works quite well. There are Otter boxes. Indeed, there's quite a list of gadgets in the queue.

There are also substantial items such as the MAXTOR SHARED STORAGE II (link) with a full terabyte of networked storage. Yum!

I also have the X60 Lenovo TabletPC, which I like quite a bit. It runs Vista and I am still having problems getting Vista to communicate properly with my ISP; that's not the X60's fault. Wireless networking is also a problem. It wasn't at WinHEC. I was able to connect to that network without problems.

When I got back home it was nearly impossible to connect to my home wireless network, but that was eventually solved by shutting down the Tablet and rebooting. Then it was automatic.

Alas, the convenient little Windows XP communications icons that let you search for wireless networks don't exist in Vista, or at least I can't find them. I admit I haven't had a lot of time to play with this. My point is that I shouldn't have to. This ought to work right out of the box. It doesn't. Another strike against Vista.

The good news is that Niven and I have finished Inferno II, so I should have more time to play around with all this and more when I get back from this trip to the DC area.