Dr. Jerry Pournelle

Email Me


Why not subscribe now?

Chaos Manor Subscribe Now



Useful Link(s)...

JerryPournelle.com


Hosting by
  Bluehost


Powered by Apache

Computing At Chaos Manor:
May 8, 2008

The User's Column, May, 2008
Column 334, Part 1
Jerry Pournelle jerryp@jerrypournelle.com
www.jerrypournelle.com
Copyright 2008 Jerry E. Pournelle, Ph.D.


The big news in the financial papers, as well as the computer press, is the complicated dance between Microsoft and Yahoo, with Google clandestinely pulling strings from the side. Of course Microsoft's real perceived enemy is Google, and its fear of what Google will do next that drives Microsoft CEO Steve Ballmer's obsession. Microsoft thinks it has to take an aggressive stand against Google, lest Google carry the war to Microsoft's essential core. Nevertheless, last weekend Microsoft withdrew its final offer of 32 dollars a share in response to one of the Google overtures.

The result was predictable, and any day trader who had a Yahoo buy order at 20 and a sell order at 24 could have made a fair amount of money on Monday. It appears that a fair number did: when Yahoo hit 20 the sales volume went through the roof, and there was another volume spike at 24-25. Of course day trading is a very risky business, and I don't advise it; but a sharp decline in Yahoo after Microsoft's offer expired was easy enough to predict, while Yahoo's insistence that their stock is worth more than the 32 or so that Microsoft offered helped shore up the price.

As to Google's role, Steve Ballmer's letter to Yahoo spent a lot of time on that: Google has offered Yahoo a deal in which Yahoo will carry search advertising from Google. From Yahoo's view this may be a good deal, but if any substantial part of Yahoo's income is from that source, it makes Yahoo far less attractive as a Microsoft acquisition.

The dance will continue, and this is probably not the last act.

It shouldn't continue. One reason Microsoft walked away from the deal was that Ballmer found dissension in the ranks: a lot of Microsoft's genius brigade didn't want Yahoo, and isn't all that enthusiastic about Internet advertising revenue to begin with. They contend that Microsoft is a software company that sells code, and they ought to put their best talents into that, not hare off after Internet advertising revenue just because there's so much growth money in it. After all, there's a lot of growth money in custom leather goods — Coach stock went from 10 to 50 in four years — so why not try a line of purses? Given past performance, Microsoft is as likely to know as much about high fashion and leather goods as it does about making high growth money from Internet advertising.

The problem here is the endless search for growth. It's no longer good enough to make a decent and sustained profit. In order to sustain the large price to earnings ratios that everyone expects, you have to keep growing, even if that growth gives you lower profit margins (see below). In today's financial environment it's no longer good enough to do one thing and do it well; now you have to do a lot of things and keep doing more of them, and how well you do them isn't quite so important.

Now, one can argue that it has been a while since Microsoft did one thing — sell code — and did it well, but you can't argue that what they did under Gates's leadership wasn't a successful strategy. Few remember that Microsoft wasn't originally very forcefully in the applications business. Gates used to tell the story in the 1980's: "I went to all the software developers and asked them to write applications for Microsoft Windows. They wouldn't do it. So I went to the Microsoft Applications Group, and they didn't have that option."

Microsoft grew like crazy by selling code. The company didn't start with a monopoly: that had to be built. One strategy was to bet on Moore's law: if a product would work at all, ship it. It might not be very fast, but new hardware was inevitable, and the hardware would bail Microsoft out. One example was graphics processors: in the early days of Windows a page down operation in a word processor could take as much as half a minute. In those days most professional writers used DOS-based text creation and editing programs for just that reason. Then along came ATI with their "Windows Graphics Accelerator" board (followed quickly by competitors) and everything changed. Windows became usable. WYSIWYG editors really are preferable to line editors, and Windows became the standard office OS. Microsoft followed by combining Word with Excel into the Microsoft Office suite, and that crowded out WordStar, Word Perfect, Lotus 1-2-3, Quattro, and other then-beloved rivals.

I agree this is a simplification, but not by all that much. My point is that Microsoft grew to be dominant by selling code, not by haring off to invest in movies or fish farms or windmills.

Bill Gates built Microsoft in a competitive environment. It's true enough that Gates had a head start, largely because Dottie Killdall refused to sign an IBM non-disclosure agreement, and the IBM representatives went to a then rather obscure company called Microsoft for a PC operating system, thus giving Microsoft a virtually guaranteed income stream. On the other hand, many companies get a big break; as von Moltke the elder famously said, in the last analysis luck benefits only the well prepared.

Microsoft took advantage of its luck. The company grew into a virtual monopoly, so that Microsoft was its own competition, but Gates never forgot that everything changes every few years. Today's market giants are forgotten tomorrow (who uses WordStar now?). Moore's Law ensures that: as the hardware base changes, software that used to look pretty cool becomes slow and obsolete. In a volatile market like that, a company can fall from grace rather quickly, and many have.

This appears to be one of those times. There are alternatives to Microsoft now. Apple is one of them. Fueled by iPod and iPhone revenues, Apple has built an integrated system of hardware, operating system, and applications, bundling many of the applications into the basic product; and they keep it all up to date. If that reminds you of the early days of Microsoft Applications, perhaps it should.

Apple is succeeding. Many IBM executives carry Apple laptops to presentation meetings. There are other alternatives to Microsoft. Whether Linux is ready for Aunt Minnie yet, it's certainly very nearly ready for Sidney Cubicleworker.

Robert Bruce Thompson observes

I was stunned at Maker Faire to see how dominant Apple notebooks have become, at least among that group. I'd guess that literally 90% of the notebooks I saw were Apples. In fact, at one point I was sitting there in the break room looking at the back of someone's notebook screen trying to figure out why it didn't have an Apple on it before I finally realized that it was a Dell.

Alas, due to my radiation therapy, I wasn't able to get to Maker Faire, the annual O'Reilly convention of gadget and gizmo tinkerers. I hope to next year. Note that Maker Faire is attended almost entirely by early adopters. I know a lot of my readers and subscribers go there. They are trend setters.

The upshot is that Microsoft has competition in its area of core competence. There was a time when Gates would have understood that. Possibly he still does. It's not so clear that the rest of his company does.

Some Future Facts

In my judgment, Yahoo CEO Jerry Wang's stubbornness — from whatever motive — allowed Microsoft to dodge a bullet. If Microsoft wants in the Internet advertising revenue business, there are better ways using Microsoft's own core competence in applications code writing to exploit that. Redmond need not engage Google on its own territory. Frontal assaults are seldom a good idea in war, and for good reasons.

What is certain in the future is that the whole hardware basis of the computer industry is changing. Multiple core chips are cheap and getting cheaper. The hardware already has potentials that the best software can't use, and that trend will continue. Moreover, as specialty applications exploit the hardware capabilities, users will see what their computers can do, and will get increasingly unhappy with the performance of their operating systems and standard applications. If the computer can do THIS, why does Outlook hiccough when downloading my mail?

This is the crucial ground for future competition. Whoever wins that will be able to move out into the Internet advertising revenue competition. The reverse is not necessarily true: having a good revenue stream from Internet advertising does not necessarily generate a capability in providing better applications and operating systems. On the other hand, one may invest that revenue into building better operating systems and applications. Apple seems to have done so.

Not only is Moore's Law inexorable, but we are well up on the rising side of that exponential. Doubling the capability of systems with 64 Kilobytes of memory seemed important at the time, but there were still serious memory limits. Doubling the capacity of 128 Kilobyte 8" floppy disks removed some inconveniences, but we are now at a stage when disk storage is nearly infinite, and the bandwidth of read/write operations has become a limiting factor. I could list more but surely the point is made?

Today there are no applications or operating systems that are "good enough" for the hardware we can all afford. Within five years the programs we run today will seem ridiculously slow and limited. The company that realizes that will win big.

The Fiduciary Racket

As I said, there was a time when a company that steadily made a 10% return on investment was considered a solid company. No longer. Now the financial analysts want growth and lots of it.

Worse, the legal eagles have adopted the theory that corporate directors have a fiduciary obligation to accept any offer that greatly benefits stockholders, regardless of the strategic consequences, and for that matter regardless of the ethical implications. If they don't, they can be sued personally. The notion of a limited liability corporation has been undermined.

It's worse now in the days of Sarbanes-Oxley, but in fact it was always pretty bad. We all know some examples. In my case it's wash and wear silver-tan shirts with epaulettes. I have worn those shirts for forty years. I bought them from specialty markets. I am not sure what the annual sales volume was, but it was enough to sustain the manufacturers even if there was no annual growth. Then suddenly, those shirts and bush jackets caught on. The small companies that made my favorite brands were bought out by the big boys. Everyone was selling silver-tan shirts with epaulettes. Then the fad wore off. Sales fell to levels considered insignificant by the big outfits, and they dropped the line entirely. As a result, my old sources are gone. I am sure everyone has a similar story: a small steady brand is caught in a fad, bought out by a big outfit, then killed off when the fad dies. The original stockholders would probably have been better off with a steady profit year after year, but the demand for "growth" added to the fiduciary obligations of the directors prevented that.

The worse case I know of came when Tom Clancy published The Hunt for Red October with the US Naval Institute (USNI), a non-profit association that had never published fiction before. Clancy had the great good fortune to have written a fascinating novel — and to have got a copy into the hands of Ronald Reagan, a president who unlike his successors actually read books. A Time reporter noted what the President was reading and got that into a Time squib, and the predictable happened: Clancy's book became a runaway bestseller.

USNI was unable to handle the logistics of a runaway bestseller and sold the book to a commercial publisher for a handsome profit. At this point the lawyers appeared, and told the USNI board — composed mostly of high ranking Naval officers both serving and retired — that they had a fiduciary obligation to USNI as an institution to retain financial control of Clancy's characters and extract ransom for them. Several board members found this dishonorable and resigned. The resulting settlement is sealed, but the stench from the legal theory of "fiduciary obligation" has soured the atmosphere of corporate board members ever since.

We now note that Yahoo's Jerry Wang is under pressure from his stockholders and board for terminating the Microsoft negotiations. We can expect the usual result: a bunch of lawyers will get rich in this game. Meanwhile my opinion remains the same: Microsoft dodged a bullet. Now Redmond has an opportunity to reevaluate its future. We can all hope they make use of it.

iPhone Update

I have carried my iPhone for a couple of months now. For various reasons I haven't fully used all of its features, and in fact many of them don't work — attempts to use the Map to "find" the local Trader Joe's, or the stock feature to find out the current price of Yahoo, invariably are met with the notation "update failed." I'll have to take this to the Apple Store in Fashion Square — fortunately I know where it is since my iPhone can't update to find it — and find out why nothing works any more.

It does work as a telephone, and it sends text messages. Attempts to connect to Safari get the message that "Safari can't open the page because it can't find the server." It then tells me I must click OK to this discouraging message before it will do anything else. Safari used to work just fine, so I don't understand what's going on.

I have tried turning the phone off and letting it come back on, but that doesn't help. I'm not sure what else to do, except go to the Apple store. With luck I'll have a happy ending in a week or so. Meanwhile, the iPhone is cool and neat, but if something doesn't work, it's pretty hard to find out what to do.

Griffin Technology ClearBoost iPhone Case

The Griffin ClearBoost solves one iPhone problem: it gives me an extra bar of AT&T telephone connectivity. It's also a pretty good case for protecting the iPhone.

The ClearBoost has an antenna booster built into the case. It seems to work in low signal areas: I pretty consistently get two bars with the case, and only one without it. To do that I have to enclose the iPhone in the case, shut the phone down, and restart it. That works. And yes, I've tried just restarting the iPhone without installing it in the case. That will sometimes help, but not regularly, while adding the case and restarting nearly always works.

The ClearBoost has a small antenna extension that serves as a convenient handle for pulling the iPhone out of my shirt pocket where I usually carry it. It is custom made to allow access to all the iPhone controls and buttons. The case is fairly easy to install; taking the iPhone out of the case requires a dime or some other means for prying the front and back apart, but it's easy enough to do.

The only drawback, and it's minor, is that with the ClearBoost case installed, the iPhone won't fit into its docking cradle. However, my iPhone came with a simple cable connector that does the same thing as the cradle, so that's no problem.

I have more than a dozen iPhone cases, but the one I use is the Griffin ClearBoost. It's good as a case, and my experience has been that the signal booster feature works. Highly recommended.

Reports

The motto of this column is that I do lots of silly things so you don't have to. Most times that's true; but lately, for reasons explained in a previous column I haven't been able to keep up. Fortunately, a number of very competent readers have leaped into the breach. Here are several Chaos Manor reader reports:

Subject: Lenovo trials and tribulations

Jerry,

While I have been very happy with my Mac Book Pro, I had the need to run some hardware that works only with XP and needs a PC Card slot. Apple has gone to express card slots only, which forced me to look at PC hardware. I have had great success with ThinkPads in the past, so I purchased a full up T61 which has similar hardware to the Mac Book Pro. 2.5 GHZ core 2 duo with 6MB of cache, 4GB or RAM, 200 GB 7200 RPM drive etc. In short, a very nice system. Also, the price was only $1500.00 - a far cry from a few years ago where it would have been over 3K. I was a little excited. Note, no surprise to you, that even though the system came loaded with Vista business, it included restore CD's for XP - just in case. Since I am running an older protocol analyzer, XP seemed the right choice. Let the games begin:

1. Restoring XP from the recovery CD's took over an hour and a half. XP was up within 30 minutes, the rest of the time was customization of XP by the Lenovo software - not a good sign in my opinion. But, 1 hour 30 minutes or so after I started, I had an XP system with all things Lenovo installed.

2. I noticed that included with the Lenovo software was Norton Internet Security, a package I detest for various reasons, not the least of which is that it eats gobs of CPU cycles getting between me and everything on the network. While I may be wrong, I would guess it will expire after 30 days or a year and start bugging me to pay more money.

I practice Pournelle safe computing - there is a strong firewall and router between my internal network and the Internet, I install all of Microsoft patches except for IE7 and Windows Genuine Disadvantage etc.

I uninstalled Norton, and installed McAfee Virus Scan Enterprise version 7, the last reasonable virus scan product they produced. It does not expire, and does not try to get in the way of all network access. A reasonable compromise given the rest of my setup.

The troubles began when I tried to install the protocol analyzer software, which is a 16 bit application - or at least the setup program is 16 bit. I got the error "Unable to load virtualdevicedriver c:\program~1\Symantec\ and so on. I ran regedit and found that indeed, there was a virtual device driver entry in the current control set /control/VirtualDeviceDriver. Deleting the entry and rebooting did not fix the problem. I googled the problem and others have had it. Apparently there are enough problems un-installing Norton and Symantec products these days, that Symantec has written a removal tool for their software.

I found, downloaded, and ran the tool. It trundled for about 5 minutes, announced that it had fixed everything and I rebooted. Tried to run setup again, and the problem is still there.

I have nuked from high orbit and am installing XP form my own VLK disk.

It is interesting to contrast Lenovo's latest and greatest with the similar Mac Book Pro.

The Mac Book Pro is an integrated design. Everything fits and pretty much works and works together. Lenovo's product has hardware as good as the MAC, but in software things fall apart. XP is from Microsoft not Lenovo. Lenovo has the need to add "Value" to the O/S and has to do so by adding add-ons - a lot of add-ons. A lot of these add-on's do the same things that standard XP components do. Perhaps better, perhaps not. It is somewhat disconcerting to see over 10 different programs starting when the system boots with Lenovo name on it.

In addition they install third party software that probably expires and is not necessary optimal. And the worst is that the third party software will not cleanly un-install.

Phil Tharp

This is one more example of Microsoft neglecting its core competency. Either make Vista work properly, or work with your major customers — heaven knows Lenovo has to be a major customer — to smooth things out, write the proper drivers, etc.

I don't know whether Vista can be salvaged. It's certainly possible, and if Microsoft put the effort into it that has been wasted on the Yahoo pursuit it might be. Another possibility is to make a workstation version of Windows Server 2008. From all my early reports that might well be possible, but it would require writing drivers. A lot of drivers. Microsoft used to be good with drivers, and even better at getting out Driver Developer Kits. Maybe they need to relearn that.

Managing Outlook

In my previous column I spent considerable time on Outlook problems. One problem was the location of the Outlook pst files. This prompted a number of suggestions from readers. I found this one from Lewis C. Brevard worth your time:

JERRY: I keep both My Documents and Outlook PST files in locations of my own choosing.

(1) My Documents has been redefined to C:\LCB (and called C LCB on my desktop and in most applications)

(2) PST files are C:\LCB\OUTLOOK.PST\

This makes backup easy and avoids the treacherous \Documents and Settings mush of configuration and temporary files.

DETAILS

(1) is done by right click and selecting Properties of My Documents and setting the Target Folder Location.

I also right click and rename it from My Documents to C LCB to remind myself. Some dialogs still show My Documents but a number of them, including in Word show the new name.

(2) is a bit trickier. I set up Outlook on a new machine then close it. Go to the normal location in C:\Documents and Settings\user\Local Settings\Application Data\Microsoft\Outlook\Outlook.pst and RENAME that to created-outlook.pst-hide

When next starting Outlook it freaks because the Outlook.PST is missing and lets you go find it.

At which point I go to my C:\LCB\OUTLOOK.PST\MainOutlook.PST

LAURENCE C. BREVARD

That's a bit like what I used to do, but with considerably more control. I'll give this a try. For the moment I am still using Office 2003; I have mixed reports on what happens when you convert to Outlook 2007. This story is far from finished.

Linux in England

Finally, we have a reader report from England:

Leopard and Ubuntu in England

Hi Jerry I have been a long term fan of your Byte columns and was delighted to find you are still keeping it up at Chaos Manor. My brother went through some of the treatment you are having, and my thoughts and sympathies are with you - good luck with the X-rays. (He recovered from his tumour quite well and that was twenty years ago)

Your experiment with Apple Macs inspired me to have a go, after many decades of using Windows. I used to think Macs looked great, but they were always a bit too expensive for me. When the Lisa came out I confess to lusting after one, but the price tag didn't fit a student budget at the time.

I have been playing around with Linux ever since it first appeared, and the difference between the early releases and the current distributions is dramatic - no more trying to guess what the dot clock frequency of your VGA card might be, and reading the warnings about how you could fry your system if you got it wrong! You just put a CD in the drive, boot up and away you go - I found that Ubuntu installed quickly and cleanly (far quicker than Vista, which is a nightmare). It found every device I need to use - even found my HP All-in-One and was able to scan pictures more easily and to a better resolution than I could with Windows XP.

I decided to compare the Mac, and after wasting some time ordering one on the Apple Store (only to be told it would take about ten days to arrive), I cancelled the online order and went to an Apple shop. It took them two and a half hours to fit some extra RAM in my new Mac Mini, but I finally got it home and hooked it up to my network. It really was easy to get it running - or so I thought ! Silly me, I had forgotten to become an American Citizen, and my nasty old UK keyboard didn't quite work, even though I had told the Mac I was in Britain, and had a UK keyboard.

Google to the rescue, as Apple help didn't help much. Aunt Beatrice (your aunt Minnie's British cousin) wouldn't have been impressed at the lengths I had to go to to get the keyboard working. Anyway, I persevered, and finally I can type £, @ and " using the keys labelled as such. Thanks go out to those brave Brits who found this out before me and published their fix on the web.

Next - I needed to print something and my wife wanted a photo scanned. No problem, I thought - if Ubuntu could find the HP All-in-One, then the Mac should have no problem - how wrong I was! Two more hours and a 100Mb download later we were finally in business, printing and scanning beautifully.

I don't really think there is an OS that Aunt Minnie (or Beatrice) could manage entirely on their own, but I would like to say that its a close call between Leopard and Ubuntu. The Mac OS is beautiful and simple, and I am increasingly becoming enamoured with it. Ubuntu is free, with all the office software you need right there from day one, and is pretty easy to use. Vista is just awful, so no point in even going there !

I will keep using Leopard, and I am very impressed with the Mac Mini, which is a joy to use and so compact you would think it was just a cd drive sitting on the top of the PC tower. I will also use Ubuntu, as it Just Works - you should try it.

All the best
Keep up the good work Regards, Andy

I haven't encountered quite so many difficulties, but I understand you perfectly: Mac enthusiasts leave out many of their problems in their reports. So do Linux enthusiasts, but I am pleased to learn that you had few problems. I'm inclined to agree that Aunts Minnie and Beatrice need nerdy nephews if they are going to get very far. On the other hand, I am pretty sure they'll get further with a Mac than with Windows.

I'm still learning about the Mac, and we have a lot of new Mac equipment I haven't reported on. Alas, there's just me: and while I intend to give Ubuntu Linux a thorough workout, it will have to wait until I'm done with the Mac. I'm dancing as fast as I can...