jp.jpg (13389 bytes)Picture of Jerry Pournelle


read book now



Current View
BOOK Reviews

This week:
Next week



March 1 - 7, 1999

emailblimp.gif (23130 bytes)


The current page will always have the name currentmail.html and may be bookmarked. For previous weeks, go to the MAIL HOME PAGE.


Fair warning: some of those previous weeks can take a minute plus to download. After Mail 10, though, they're tamed down a bit.

IF YOU SEND MAIL it may be published; if you want it private SAY SO AT THE TOP of the mail. I try to respect confidences, but there is only me, and this is Chaos Manor.

PLEASE DO NOT USE DEEP INDENTATION INCLUDING LAYERS OF BLOCK QUOTES IN MAIL. TABS in mail will also do deep indentations. Use with care or not at all.

I try to answer mail, but mostly I can't get to all of it. I read it all, although not always the instant it comes in. I do have books to write too...  I am reminded of H. P. Lovecraft who slowly starved to death while answering fan mail. 

If you want to send mail that will be published, you don't have to use the formatting instructions you will find when you click here but it will make my life simpler, and your chances of being published better..

This week:
Monday -- Tuesday -- Wednesday -- Thursday -- Friday -- Saturday -- Sunday



Monday, March 1, 1999

Dr. Pournelle,

It is possible that Norton Utilities is causing your shutdown problems. I ran Norton Live Update for Norton Utilities 3.0 a week or so ago and there was a note that it had fixed a problem which could prevent a system from shutting down properly. I hadn't noticed the problem so I didn't pay too much attention, but then I normally run System Doctor at startup and shut it off once it has finished scanning. It's probably worth running Live Update if you haven't done it recently.

Illegitimi non carborundum.

John Guppy

I will try that, but in fact with the Windows 95b system SOMETHING I did finished off the problem. I tried closing all the tray functions manually, and everything else. I also did the USE DOS TO EXPORT THE REGISTRY TRICK (see Mail30.html) and reimported; and some evil and potent words of magic. Something did it: the system shuts down properly now. Still, I will do Live Update shortly on it. It's good advice. For all the curses I have expended on Norton System Doctor, it and Norton UNINSTALL are quite worth while, and I recommend them.



I write to complain. Not about you, of course, or about the monopoly at Microsoft, or President Clinton. No, I am currently fighting a war of attrition on my computer over which utilities to use. I have tried both Norton’s Utilities and Nuts and Bolts 98, and have some things to report for comment.

First, my system is new, with good hardware. Win98 is not beta, but retail version. Updated drivers on all hardware, and no regular system crashes. System shuts down well. No recurring problems of any kind.

Nuts and Bolts 98: First, the "98" in the name is a reference to the year of release, not to compatabiltiy to Win98. While the "System Requirements" shows Win95/98, don’t be deceived - this package does not work well with Win98. I installed the program, and used its system check tools first. No problems reported. I then tried the optimization tools. Bomb shelter, the "crash guard" feature, alerted me to a potential software crash...a component of Nuts and Bolts was rescued. I tried optimization tools. Crash. Again, this time got to speed tune my harddrives. All went well.

Moved to registry tuning. (I manually saved first, I am only a small fool.)

Reduced the size of the registry by removing unneeded entries. Reboot. Win98 says, "Corrupt Registry, restoring backup". Boot, Nuts and Bolts has "invalid license", has to be reinstalled. I uninstall instead.

Install Norton Utilities. System doctor comes on line, informs me, "virus files are out of date, click here to update." I click, and am informed that this version doesn’t support updating (using a "full" demo). OK. SysDoc says, "Rescue file out of date, click here." I click. "This version doesn’t support rescue disk." I run Tune-up. Crash guard warns that SysDoc is going to crash, and rescues it. SysDoc reports a virus, and click here to clean. I click. Can’t clean these files, click here to delete them. The files are .exe files that I routinely use in Quicken. To be sure of the virus, I run McAfee virus scan. No viruses. Run McAfee virus update, get latest .dat files. Scan. No viruses. Crash guard rescues SysDoc. I run Registry tune up, which will automatically reboot system upon completion. System fails to shut down. Manually three finger it down, it shuts down. Rerun Registry Tune, with all other programs closed from system tray. System fails to shut down. On reboot, SysDoc reports, in order, Virus, Virus data out of date, rescue disc out of date. I close SysDoc, it crashes, is rescued, then shuts down properly. I uninstall Norton’s.

Both packages claim they are Win98 compatable. Neither runs well, but both handle Win95 wonderfully. MaxPC gave Nuts and Bolts a 9/10 and a KickAss award, but acknowledge, only in response to a letter, not in the article, that they did not test on Win98.

My question to your readers, and yourself is, what, if any, set of utilities works for Windows 98?

Bryan Broyles

Across the pale parabola of joy... Ralston McTodd

Good questions. First, Norton Utilities, installed on Windows 98, works "just barely" until you let it do the Live Update. Once it updates itself, it's considerably better. Having said that, I have to say also that I'm in the middle of finding out the rest of the story. I'll tell you more when I get the chance. Of the lot, I have more confidence in a recently updated Norton than anything else. But I haven't got one installed just now: recall that I have long told people, if you have Windows 95b running, and NT, you DO NOT NEED WINDOWS 98. I've taken that advice myself, so that I don't have a lot of machines running 98. I do have one, and I'll shortly put Norton on it, and let it update and tell you the results.

I'm reluctant to allow this to happen on BIGSYS, which is a "main machine" and just at the moment my other 98 systems are dismantled, but I'll get to this real soon now…




Tuesday, March 2, 1999


For reasons you will see below, I was getting somewhat different layout results in VIEW and MAIL (how they got to be different I do not know). I mentioned it and got this reply:

Why there is more space in lists in currentview than currentmail?

Bruce McFarling, Callaghan



This is simple. In Current Mail, your lists have been set up like

<li>List item 1</li>

<li>List item 2</li>


etc, while in Current View, the lists with wider spacing have been set out as

<li><p>List item 1</p></li>

<li><p>List item 2</p</li>

Why? I don’t use home page, so I can only guess. However, the reason that


you have <p> tags is to hold a left alignment on these items: that is, the

full tag is

<li><p align="left">List item 1</p></li>

<li><p align="left">List item 2</p</li>


As a side effect of adding that alignment property, you end up with a list of paragraphs, which a lot of browsers will render with extra linespacing.


Bruce McFarling, Newcastle, NSW

Thanks. I was pretty sure it was something like that, and if I looked close enough at the code I would see the difference, but I wouldn't necessarily understand why it was doing it that way; so asking would get a more coherent explanation. Which it did, and thanks again. I'll fix it.

I also got this:

Dr. Pournelle,

You’re adding </P>, a paragraph marker, in your line items, before moving on th the next line item:



<li><a href="#shutdown">Shutdown yet one more time</a></li>

<li><a href="#Norton">Utilities Complaint and query</a></li>






<p><big>Highlights this week:</big> <ul>

<li><p align="left"><a href="">BYTE IS BACK</a></p>


<li><p align="left"><a href="">Rotary




<li><p align="left">&;nbsp;</p>




&;nbsp also adds a blank line

I don’t know much about Front Page, but perhaps you hit the enter key, which the program assumed you wanted as a new paragraph inside the line item. You can see this in the HTML source as </p></li>.

George Laiacona III

which explains in even more detail. This is of interest if it happens to you, and it's well to get it on record. Thanks to all who answered.


Dear Dr. Pournelle,

I just finished reading your new column over on the resurrected Byte site.

I found your comments about mice interesting.

I bought a new Logitech MouseMan Wheel last week after having ogled it at the local Office Max for months. I liked the idea of the more sculpted shape and the extra thumb button. I assigned it to the Ctrl key so that I can make multiple selections in Windows Explorer with my mouse while holding my newborn with my other arm.

But if wishes were fishes I’d buy a mouse made by a Motie in a heartbeat. The first time I grabbed that Logitech mouse I thought, "Well, gee, it’s nice and everything but it isn’t nearly as contoured as I thought it would be." The thought of having a Motie measure my hand and mold that MouseMan with her (or is it his?) tube of magic plastic gave me pleasurable moments of fancy.

Anyway, enough woolgathering. The MouseMan Wheel on a 3M Precise Mousing Surface is a joy to use. And now I see that Microsoft has copied the design.


Steve Erbach

Welcome aboard. I suspect we'll be getting a lot of people here through the new BYTE (at least we can certainly hope so…)

I have always wished I had a couple of Watchmakers around, too. For those who don't know what this conversation is about, have a look at THE MOTE IN GOD'S EYE by Larry Niven and Jerry Pournelle.



Interesting mail from Jim Ransom about the Rotary Rocket Rollout. Jim has been my Executive Officer in the Citizen's Advisory Council on National Space Policy for years. When we began he was at Aerospace Corp. Now he does consulting work including with Rotary Rocket. His observations are on target:


Dear Jerry,

As you observed as we talked in the Rotary Rocket Company high bay next to the Roton Atmospheric Test Vehicle (ATV), the ATV is roughly comparable to the DC-X in that it is a landing system technology demonstrator. Where it differs from the DC-X is that once the ATV test flights are complete, the Roton vehicle low speed aerodynamics and control laws will have been measured in full scale, rather than the subscale of the DC-X and the X-33. The ATV also tested the manufacturing methods for the liquid oxygen and kerosene tanks for the orbital Roton vehicle. The Roton enters base first, so the ATV does not need to perform comparable tests to the rotation maneuver which was necessary for the DC-X to demonstrate for the operational Delta Clipper side entry profile. The ATV is only a big, hydrogen-peroxide powered helicopter, but it demonstrates a level of commitment and follow-through toward an operational SSTO vehicle in its structures and configuration that the DC-X lacked.

Jim Ransom


For those to whom this makes no sense you can either ignore it, or go to the space sections of this web site and look for more information. DC/X was a scale model VTOL ship that was built through the efforts of the Council. X-33 is what was supposed to be an X project that became Venture Star, which, whatever it is, is not an X project. X projects are the most successful government program in R&;D history in my opinion. Now back to your regularly scheduled stuff.



I have greatly enjoyed both your previous "BYTE" lifestyle as well as your *new* independent lifestyle (and actually more so). Upon visiting Chaos Manor this morning I saw your reference to the new BYTE syte and went to read your latest column. But I must report that your paragraph in the lower portion of the page with the title "COMPUTING AT CHAOS MANOR" is now obviously out of date since you indicate that you have not been contacted and know nothing of the new BYTE web site...but you must!

best regards and thanks *alot* for all of the useful info, insight, and discussion.

Tony O'Brien []


Right. I will start to fix all those things as time goes on, and thanks to all who are pointing them out. KEEP IT UP. This disorganized chaotic place is going to be in transition again for a while.

I am getting along famously with the new editorial people, they are working to get back BYTE staffers, and it all looks good, but it's also making for a very busy time…


My name is Charles Lowry and I am a supervisor for the Iomega Return Authorizations Department.

I have enjoyed your column in Byte for many years and while reading your web site I discovered you were having some problems with your Zip drive.

First, let me state that although I work for Iomega, this email is my own, and is entirely independent from Iomega Corporation -Really!

As I understand it, you have two internal Zip drives in question, one on which you can write data which can be read by other Zip drives, and one which only reads data on disks which it has written to. This second drive has a slightly misaligned read/write head. That’s all it is. This type of problem is very rare and is remedied by replacing the drive, which will be for free given that the drive is still under warranty.

Let me take a moment to address the issue of "Clicking" drives in the hopes that I can help solve some Zip drive user problems.

Iomega states that only .5% of all Zip drives exhibit the "Clicking" problem. I was instrumental in discovering this figure, and it is accurate. What this figure does not determine is how many drives click due to mechanical malfunction from Iomega manufacturing. In my position as Call Center Supervisor, I have opportunities every day to talk to customers to determine the source of their clicking problems and develop resolution strategies. I have discovered that the majority of clicking drive problems are not due to manufacturing but are the result of other factors. My research and experience shows that less than one out of four drives malfunction due to manufacturing. Given the industry standard of 3.5% failure rate in 1 year, a less than .125% failure rate is pretty good. Unfortunately that still leaves 25,000 users with problems, which Iomega is trying to make right.

I would like to give to you and your readers a number of reasons why Zip drives click, and possible resolutions.

All the below mentioned problems may result in a clicking drive.

Windows 95 Rev A. apparently has difficulties properly registering removable media.

Solution: Uninstall Iomega Tools, delete the Tools_95 Directory, Upgrade to Windows 95 Rev. B or better and reinstall the Iomega Tools. If upgrading to latest version of Windows 98, it may not be necessary to reinstall the Iomega Tools.

Upgrading to Windows 98 from Windows 95 which has had the Iomega tools installed is problematic. Windows 98 includes native support for all current Iomega products and normally does not require the installation of the tools. Upgrading from Windows 95 which already has the tools installed often results in a driver conflict which results in the drive making a clicking noise as Windows tries to resolve which driver it is supposed to use.

Solution: Best method is to completely uninstall Windows 98 to remove all traces of driver conflict and then reinstall Windows 98. Otherwise, completely remove all Iomega products from Windows 98 and reboot the system. This often removes all the Iomega drivers from Windows. Shutdown the system again, and while shutdown, reattach and power on all Iomega products, Restart Windows. Windows will usually find the "new" devices and setup the drivers properly. If Windows 98 does not recognize the Iomega products upon reinstallation, The Iomega tools can now be installed without driver conflicts.

Move the drives away from magnetic fields. It is very difficult to provide complete magnetic shielding for any device, and specially for removable media. Remember that these products use very fine read/write heads which are designed to be very sensitive to small fluctuations in magnetic fields. Placing a Zip or Jaz drive on top of a monitor, in direct sunlight, dropping it, or placing it close to a power supply is a formula for disaster. (As a note, external power supplies are much worse EM wise than internal supplies due to internal supplies being encased in a metal shield. Still, I would place any internal Iomega drive as far away from the internal power supply as possible).

Dropping the drive even from a short distance could displace the delicate heads. We would never think of dropping our computers or VCR’s 3 feet or more; Zip drives need to be treated with the same respect.

Solution: If the drive starts misbehaving, replace the drive. Note: This is probably the source of your Zip drive problem. The drive or system in which it was installed may have been mishandled at some time, causing the heads to become slightly misaligned as compared to other Zip drives.

Lastly, Keep your drive clean! I drape a towel over my external zip drive when not in use to insure dust and dirt do not accumulate on the door. Dirt is probably the worse cause of clicking. These drives run at close tolerances and can easily be destroyed by errant hair and dust. Again-keep it clean!

Thank you,

Charles Lowry

ERC Supervisor, Iomega Corp


Thanks. I'll have some comments another time, but this ought to tell readers what to do about ZIP.


I was going down the list of books you recommend, and I noticed the absence of one I have found very useful. I recommend it to all would-be web page designers.

It’s called TYPE AND LAYOUT by Colin Wheildon.

While your own web pages are reasonably restrained, I’m sure

you’ve seen pages in garish colours, funny fonts and/or with twirling icons that make you reach for the "stop animations" button. The errors on such designs are easy to spot, but what about the less obvious ones?

This book is not specifically geared to web-page design, but the overall format of any printed page, with a lot of references to advertising (which is how I originally found it). For any page (web or print) it can teach you what not to do, why not to do it, and, if you persist, what will happen to reader comprehension and interest.

It’s only fault is its cost (Cdn$40 in paperback), but I checked and does have it. Oh yes, it’s also easy to read. (surprise!)

Trefor Thomas []

To be civilized is to restrain the ability to commit mayhem. To be incapable of committing mayhem is not the mark of the civilized, merely the domesticated.

Thanks. I'll have a look at it.






Wednesday March 3, 1999

Subject: Ghosts in the machine

Dear Jerry:

Great to have you back on the "new" Byte...I’ve had too much fun with CMP regarding my Byte subsciption to feel real happy about the transition to the web site, but that story will have to wait for another note...

In the meantime, your comments about modem issues inspired me to share my experiences from my recent stint at tech support for one of the "Big 5" of the PC bidness...I’ve got many horror stories, but I’ll concentrate on purely technical stuff for now.

You wouldn’t believe how many calls I’ve taken from folks whose modems wouldn’t respond to DUN, or other dialers; the problem is usually a "Ghost" modem - thanks to PnP, Windows has the marginal propensity, especially after new hardware, or occaisionally new software, has been instaled to create, on its own, additional hardware settings in the registry; and this applies to Win 98, as well as 95.

(I’m sure the folks at Microsoft would challenge this statement, but I’ve done this fix too many times to care about what the "experts" at MS think...)

What you have is a duplicate, or "Ghost", modem (or other device - this is also a common problem with sound cards) in the system.

You often won’t see these 2nd, 3rd, and so on, modems in the Device Mgr. unless you reboot and go into safe mode - this is step #1.

(Even if you do see them in regular Win, I usually had the customer go into safe mode, just to make sure we had an accurate picture of device status.)

>From here, the procedure will sound familiar - except save yourself a


keystoke, or mouise click, by holding down the Win key, and press the Break key - this will jump directly to the System Properties dialog, so you can avoid the Ctrl Panel/System icon route...

Next, click on the Device Mgr. tab, open the modem folder, and remove each device you find.

Then, if you have an "Other Devices" folder open it and check for modems in there - you may only see a line thats called "unknown device", or you may see another modem - remove either of those, also. On some systems you’ll also find dupes of sound cards, USB ports(Win 95), and other junk - you can remove this stuff also, but if you don’t have the correct drivers, this junk will just end up in "other devices" after you reboot; if the stuff isn’t causing any problems, just leave it there and fix it later.

Next, open the Ports folder. If your using an internal modem, which is probably assigned to Com2, remove Com2 if its listed - and here’s the trick: now reboot, go into the BIOS setup, and check the BIOS configuration of your Com ports - if your modem is supposed to be on Com2, make sure that Com port is disabled in the BIOS so the internal modem use the system resources that would normally be assigned to that Com port (sub Com1 in the above if your modem is on that port because your pointing device is on the P/2 port); On external modems, the ports will probably be OK, but it won’t hurt to check.

Reboot. Win should find the modem, and ask for the driver disk if needed.

(Sometimes to speed up calls we would have the customer also remove any Com ports from the Device Mgr, but this often isn’t needed. If removed, Win will reset the ports on reboot without user intervention.)

As a finishing touch, after Win restarts, open the Control Panel, dbl-click "Modem", click on on the "Diagnostics" tab, and you should see your modem on the correct Com port - click on the port to highlight it, then click "More Info" - Win should come back after a few seconds with a dialog that shows your port info, and a "Command/Respose" section that "verifies" the modem is working properly (we would use this to prove to prove to customers that the reason for their problem connecting to the internet WAS NOT the modem.)

If you find your modem isn’t on the correct port, you can go back to the Dev Mgr and "force" it onto the correct one by opening its properties, removing the check mark from "use autonmatic settings", and selecting another "basic configuration" value from the list - trial and error will usually find the right one as indicated by the change in Com port and IRQ.

This applies to "Standard" modems - if you’ve got one of the new PCI Winmodems all bets are off - those damn things are IMPOSSIBLE! They often refuse to use the available Com 1 or 2, instead trying to setup on the same IRQ as your video card...I’ve used USR (now 3Com) modems for years, and recommend them to anyone who asks, but I sure can’t recommend their Winmodems, or any others of this new, cheap genera - to make them work would often require several re-installs, and then the reliability was questionable - OEMs love ‘em cause they’re dirt cheap, but as the old Mechanic once said: "the bitternes of poor quality remains long after the sweetness of low price is gone"...

Well, the hour is late...I hope this info will help with any future modem issues. The sound card "Ghost" fix requires a registry hack, so I’ll pass that one along later. Modem and sound card "Ghosts" were the most common tech support calls, followed by totally trashed systems requiring a Win deltree/reinstall.

I’ve been a Byte subscriber for over ten years - the main reason I kept renewing was I didn’t ever want to miss a month of "Computing at Chaos Manor"!

Best Wishes,

Terry Farnand

Terry &; Nancy []

Portland, OR

Thanks for the kind words, but even more for the fix. I'll apply that next time.


Dear Dr. Pournelle,

I just finished the "AAAS Hears Mars or Bust" section of March Chaos Manor and though I would drop you a note on it.

While at the founding conference for the Mars Society last August (you should catch the next one... very good gathering) I approached Robert Zubrin with the question of adopting Dr. Walford’s diet for his Mars Direct missions. I pointed out the not inconsiderable savings in mass as well as the heath benefits of the calorie restricted diet. I was rather disappointed to have him dismiss it out of hand. He said that he hadn’t heard of Dr. Walford or read much on calorie restricted diets but insisted that water mass was the larger issue and then walked off.

I would like to think that even with his dismissal of the idea, a seed may have stayed with him through to this AAAS meeting where he would have had a chance to hear Dr. Walford himself speak. I am hoping that he may have taken a deeper look at the proposal after a more... esteemed personage spoke.

I haven’t heard anyone other than myself talk of adopting this diet to long duration space missions but it strikes me as the only reasonable way to go about it. Over a multiyear mission, the dry mass of food alone would amount to a savings of hundreds of kilograms. As for the water, one of the primary needs is digestion... the less digesting you do the less water you need (although the saving here would be considerably less).

The astronauts would have to work their way down to 1800 calories a day... which in turn would leave you with the slimmest, healthiest crew. You would be leaving behind a few tens of kilograms of nonessential fat. Okay, this is getting down to picking nits... but every kilogram counts when you are tossing it from the surface of the Earth to the surface of Mars... and back again!

Almost the entirety of the literature that I have found of diet in space has been on preservation and palatability. While preservation is essential, palatability is not. Our planet was explored, mapped and conquered by sailors eating food that would get a restaurant condemned today.

Healthy and nutritious... and I would add as *little* as possible.

Healthier astronauts and large weight savings, if I’m wrong maybe you could help me see where. If on the other hand, you see a possible benefit to this idea, maybe you could broach it to others who can do something with the idea.

Clint Johnson

I never thought on that one. I have to say that given my druthers we would build a Lunar Colony before we do Mars. The Moon is only 1.5 seconds away for communications, and a few days for sending a hard lander with what we forgot. With Mars you have to have thought of it all. But Zubrin makes a good pitch, worth seeing.

I'll have to think on 2000 calorie a day diets for space exploration. Sounds interesting at first shot.

I am sure that Zubrin did not meet Walford. Different days.


Anyone have what is needed here?

Dear Dr Pournelle,

Great news about the new Byte - I hope it works out!

At work here we have some software for dynamic modelling of machinery (mostly in Fortran) to which we keep adding bits as needed. We plan to add a graphic front end with little models of motors, gearboxes and so on, and the idea is that you click on a motor icon, pull it to the workspace, then drag up a gearbox and connect it, and so on.

Is anyone out there producing a general development kit for this sort of thing? I know there are modelling packages available, but they tend to be not very efficient. Our optimised Fortran models tend to run for days on some problems, so we need to stick with our lean core of code.

We have a front end written in Visual Basic, but it is based on forms, and you have to remember that motor 3 is connected to pulley 27 and the suchlike or it all falls apart.

We used to be doing all this on Unix workstations, but graphics in Windows seems more complicated somehow.

Thanks for all the information; I’m having a great time working my way through past Mail and View pages.


Peter Smith

WBM Pty Ltd

Brisbane, Australia

I'm heading out the door, but perhaps a reader will have the information you need. Best regards.




In Seattle




At Microsoft




Saturday March 6, 1999

After all the mess about web sites that suck:


Subject: sites that suck


Do not waste your time worrying about how your site looks, or what the proprietor of thinks about it. You provide the content which those of us that can read enjoy. We know why your site is around, and we appreciate it.

Even the man’s domain name is offensive - it clearly shows that he doesn’t have the intelligence or the creativity to provide any meaningful content of his own, so he has to try and drag down those who do.

Brandon Stenger []

That is of course my view of the matter, but authors are all alike: if this morning I get 8 favorable reviews including a starred Kirkus and a fanzine with a circulation of 23 and a readership of 9 with a badly written slam of my latest book, I'll brood all day about the fanzine, wondering if this guy knows something I should and don't. It's one of the hazards of the profession and there's little to be done about it. But thank you, and all the other readers who have responded in much the same way.


Subject: sites that suck


> or why he is so well qualified


Why, Jerry, he’s the self-proclaimed arbiter of good web page design. If you don’t believe me, ask him. He’ll tell you himself. And there’s more than a little envy there. The vitriolic diatribe he posted as the first message in that thread drips with it. I suppose good web design practice suggests that he should have used a green font for that.

I think the main problem here is that different things are important to different people. Although that site gives lip-service to content, what it and its readers are really concerned about is appearance. From their point of view, a visually attractive site is the sine qua non. Content is purely secondary. What I find ironic is that this site isn’t anything special visually. As a matter of fact, it’s not anything special, full stop.

They’ve confused the concept of "web pages that suck" with "web pages that are not visually attractive". Considering content and presentation, their site falls in the average category for both. Your site contains excellent content and below average presentation. So what? I’d rather you spend your limited time working on the steak than the sizzle.

Illegitimi non carborundum.



Robert Bruce Thompson

Oh I agree, and of course I was being rather heavy handed with the irony in my question; but I do wonder about some of his contributors. I took the trouble to look up some of their web sites (those that have them; many do not) and found for instance the young lady whose note I quoted seems a decent sort, with a nice family of which she is no doubt justly proud; why would she participate in this silly effort to get at me? Or for that matter the whole enterprise? She can hardly pretend that her letter was intended as constructive criticism given its vituperation, or that she was trying to be helpful to me or anyone else. There are a couple of others in there, people whose sites seem well designed for what they intend, and who seem like decent people: what impels them to fall in behind this particular tumbrel and cheer as we head for the guillotine? It is not as if they resent my politics, or my views on Microsoft or Linux or OS/2, because they carefully begin by saying they know nothing whatever about me. So why comment at all? Surely not all are the petty sort you describe?

I confess a somewhat morbid fascination. Novelists tend to that sort of thing, looking at the darker elements in otherwise sunny dispositions, especially novelists with training in abnormal psychology. Well, better that the lady take out her frustrations on me than on her children.


Dear Jerry,

I forget where I read this, probably in Personal Computer World. A couple of years ago Sun Microsystems achieved a substantial productivity increase. On the orders of tha CEO all copies of Powerpoint were wiped. People reverted to producing standard reports instead of spending hours preparing pretty presentations.

CONTENT is the name of the game. A few well written lines making a point will always hit home. if you need to dress it up with excessive presentation then then it’s basically smoke and mirrors.

Interestingly when I first started producing web pages (Intranet) I used Word and came up with something similar to your earlier layout - properly spaced text on a relaxing background. 2 years on I see no reason to change. I pride (kid?) myself that I deliver content, the kind my target audience wants.

Marketing types want to attract EVERYBODY to their site. Poor content providers need smoke and mirrors. People like you need to concentrate on delivering the message to people who WANT to read it.

Keep up the good work. I would guess that the readers who think the site sucks are the ones who would not gain anything by reading your stuff anyway.

Keep on sucking!

David Cefai

That does sum up my view fairly well. Thanks.

Sinus Problems.

I use a product called Potters Antifect. It is a hebal remedy that always clears any stuffiness and does not cause drowsiness. It is commonly available from UK chemists and can even be bought in Boots (largest UK retail chemists).


Please post some photos asap.


It would also be nice to have a gallery of the folks that you refer to in the column - family, friends and colleagues. I do understand how this could expose them to risk and why they might not want to.

Web Site Design.

Please ignore the detractors. I read CMM every day. It is the only site I read every day. Make detail enhancement changes as and when required. Wizz bang graphics, colours, peculiar fonts and such just slow down the reading experience.

Web Site Enhancement.

I would like to see todays view on a page I could bookmark. I could then pop straight in to see if today had been updated and then off again. Keep the week view though as it is most useful for catching up when I have been away for any reason.

Keep up the most excellent work,


Your wish has been granted:

is always the current view page, and

is always the current mail page. My wife doesn't want me to put up pictures of the current truck. I fear she has some legitimate security concerns, although we have invested in a lot of security equipment. You can see pictures of Roberta, Eric, Alex, and Larry Niven over in the pictures section: go to the home page here and look for pictures. And I am working on getting the search engine set up for this screwy place. Thanks for the kind words. I'll look for the herbal remedy, but I doubt it's available here; and I have to confess I get so many recommendations that if I tried them all I would be awash...


Steve Gibson (of SpinRite fame) has a free utility called TIP (Trouble in Paradise) that tests Zip &; Jaz drives. It’s at Might help.

Keep up the good work; I’ve long enjoyed it!

Thanks for the pointer. I need to visit Gibson's latest. Used to use SpinRite all the time.


Dr. Pournelle,

Content is indeed more important (for your site) than whizbangs and pretty graphics. But whizbangs and pretty graphics have their place, whether printed or on web. I bought my first issue of BYTE at the college bookstore and my purchase was influenced by the intriguing cover art of Robert Tinney.

The problem, or conflict, is that those who know of your printed work know what to expect, while members of a potential new audience have no previous experience with you or your work. But they have ample experience surfing the web, much more than this subscriber, and from that standpoint your web site probably sticks out like a sore thumb.

This new medium is still developing its own rules—evolving, and as it matures the most adept at it will be those who learned it as their native "tongue." Just as my generation doesn’t remember when there wasn’t television, succeeding generations don’t remember when variations on pinball were the only arcade games (I remember the first time I saw a Pong arcade game in November 1977). The next generation has grown up with home computers, or at least a Nintendo machine.

Nonlinear text (hypertext) is so new that all but the youngest surfers can remember the first time they encountered it. For me it was in about 1998. But I also remember one of its precursors, context-sensitive help.

"Old-school" surfers (loosely defined, the sake of discussion, as those who completed their formal education before the hypertext revolution) feel quite comfortable with your format. But for the designers who are defining the new medium—those who have experienced a "paradigm shift" and are immersed in the new model, it grates on the nerves, in approximately the same way that "Japanese English" grates on the nerves of native speakers of American English. When I spent two weeks in Japan I found my language creeping toward "Japanese English" in order to be understood, and when I returned to the US it took a couple days for me to come completely back to American English. You’ve been creeping toward web/hypertext in the same way.

To those of us who have done useful work in a 64K environment, who have hit the 192K limit on a single-sided floppy disk, the mostly-text content of your page is like going home again. But over time we and your page will evolve.

As a college student in the mid eighties, I remember reading research in an English or composition journal documenting the fact that student writing done in a graphical environment had more errors and was generally lower quality writing than student writing done in text-based word processors. Now that almost everyone uses a graphical environment, I doubt there would be much difference, because we have all become more accustomed to the GUI approach. But inexperienced users still occasionally use something like Paint for word processing, and see each word as an image to manipulate.

Maybe another generation division is whether your first word processor was text-based, or graphical. I suspect that the majority of those who truly think your web page sucks have little history with book-based dictionaries or command prompts. Or they have, but have been immersed in web/hypertext so deeply that they have made the paradigm shift to the new medium.


Charles Duell


Interesting observations, and probably on target. This leaves the question of what to do about it: how can I manage to convey, on my front page, just what is available here? I'm not at all sure I can, meaning that those who surf by without reading the opening paragraph may simply have to be lost; a few may come back from hearing about this from others. And of course the new BYTE site ought to be attractive enough and that may send more this way.

But I think I will have a contest on redesigning my home page…




Sunday March 7, 1999


Although most people on the web are new to it, I remember Al Perlis talking about it as the coming thing (in a web context) at Xerox PARC in 1988. (I had been invited there to talk about a practical example of computer chaos that had bit us.) The first Macs had HyperCard, too, based on a PARC idea.


Harry Erwin []

Well, of course Ted Nelson talked about hypetext in the very early 80's and at the very first Hacker's meeting had been working on it for a long time. I believe I wrote that up in about 1982 or so (I need to find a way to get all my early columns up on line). I am not at all sure that Xerox PARC was the originator of Hypertext; I think Ted Nelson deserves the credit on that one.


Internet Ads are like flea bites: each one’s just a little itch, but it doesn’t take many before you have a major rash. It’s not the actual ads that make me itch so, it’s the time it takes to load them. As a result, I use filtering software JunkBuster (from, with a block list from (see for the maintainer) to block the graphics associated with most ads. I now get something for nothing: I get content that is paid for by selling my mindshare, without actually delivering it.

Most ads are fairly innocuous window dressing around the content of some of my favorite sites. A few, like the bait-and-switch images that look like search forms, actually are really annoying. As my kids grow up, I suspect I’ll find some ads that are just plain unacceptable. However, ads on web pages, in general, are a pretty good deal. For a few pixels, CPU cycles, and eyeblinks, I get content that would normally cost money directly. Occasionally I even see an add that amuses or interests me.

Without ads, sites like Wired News and the resurrected online would have to follow the lead of sites like the recently subscription-only Slate, and probably with about as much success.

In order to work well, ads seem to require bandwidth hungry graphics that need to be loaded from central servers. This means that browsing is often interrupted by messages such as "Connect: Host contacted, waiting for reply...". This happens whenever the ad servers are busy. On slow links, the time it takes to download the image adds insult to injury.

The net result is that even otherwise good ads are often self-defeating: the delays annoy me badly enough to neuter the ad. So much so that I downloaded free filtering software from one site and a comprehensive set of ad filters from another and installed them on my gateway machine. Voila! no more ads. This took all of an hour, even given that I was installing a proxy server under Linux, not just a browser add-on.

And, as Jerry Pournelle pointed out: Voila! no more incentive for to give away content. That graphic banner pointing to "SUPER.HUMAN.SOFTWARE - Use this power wisely. Click here to demo it now?" I don’t even see it. All I see is an empty white rectangle.

This means that I probably will never know what SUPER.HUMAN.SOFTWARE is, and’s advertising revenues will suffer just that little more due to my lack of clicking through to their sponsor. And they would know if I followed it: the URL is fashioned to allow them to record the site and often which ad on the page I clicked on.

As a result, Jerry and other contributors are that much less likely to find a market for the content we go there for. So I’ve achieved a pyrrhic victory: I get faster serving to less and less valuable content as the ad dollars cease to flow.

This is a classic tragedy of the commons: as I persue my private gain, I undermine the very resource I value. Well, maybe my absence from the ad clicking crowd won’t drive under, but extend the effects of my personal gain to a large percentage of the browsing community, and it would.

I don’t see any easy answers, and I enjoy fast browsing enough that I’ll keep filtering, even though I do occasionally have to revert back to unfiltered access to get some quirky sites to work. It’s just a few clicks here and there to enable or disable the proxy, after all.

I can honestly say, though, if it weren’t for the delays created by the finding and transferring the ad images, I wouldn’t bother filtering. Filtering just wouldn’t be worth the time and inconvenience if I didn’t get that fleabitten feeling without it. I’d even gladly pay for ad-free content, just as I pay for public radio and pay-per-view movies.

Unfortunately, I know of few sites that offer this. Slate offers subscriptions with benefits. I might even subscribe if I end up there very often, but I suspect I’d still have to wait for the ads since they don’t mention ad-less browsing as a benefit of subscribing. To be fair, Slate seems to load pretty quickly.

I’d like to hope that faster connections for all will solve this for me, but then the ads will be full motion video and sound, always expanding to fill the available volume...

Barrie Slaymaker []

This continues a discussion begun when Barrie recommended junkbuster; I pointed out that I have mixed emotions on Internet ads, since without them there would not be enough revenue to continue the resource; it's a bit like television that way.

I am not sure what the conclusions are, but there is much to be said here.


Dr. Pournelle,

As to the origin of the WWW, I think you’ll have to go back to Vannevar Bush, the electrical engineer who helped develop the analog computer (and the atomic bomb). In 1945 he wrote a groundbreaking article entitled "As We May Think", in which he described a device called the "memex", designed to extend human memory by providing the means to organize information associatively. The "memex" was never built, but others picked up on the idea.

Here’s a bio link on Bush, a great American scientist:

Donald W. McArthur

Of course Bush is also the chap who told Congress not to worry about Intercontinental missiles, and moon rockets were impossible, and space travel wouldn't happen for a century, but yes, he had his moments of greatness too.

Eric Pobirs also points out that Bush probably invented hypertext and was the inspiration for Ted Nelson. I expect he's right.


Subject: Eagle One and page faults

Good evening,

Before laying waste to the Registry, may I suggest two options:

1: Reseat the memory modules. Of the three machines I’ve had this happen to, all of them have had something amiss with main or cache memory. In the case of PC100 DIMMs, make note of where the "illegal op" message occurs, then swap the sticks in their respective slots (if there are two or more) and see if the address moves. If it does, remove one and see if the error repeats. If not, the offending piece is lying on the table.

2: Use the old standby of Win95; laying waste to everything under the "System Devices" category in Device Manager, then allowing the machine to regenerate this portion of the Registry during the course of inevitable multiple reboots.

I’ve had success in the Registry with scalpel and the occasional small caliber weapon; fire and sword are not considered precision instruments in such a closed environment.


Tim Bowser, MSgt, USAF (ret.)

Tim Bowser []

Good advice, of course. I am actually going to remove most of the boards and reseat everything, slowly: this machine was very stable, then I moved some boards around, and whammo! I'll start over, and usually my first move is to lay waste to all the devices and let it regenerate. Usually by adding the cards one at a time: that seems to work better than letting it find all of them at once. Why I don't know.

Thanks, and best wishes.





Chaos Manor home

Entire contents copyright 1999 by Jerry E. Pournelle. All rights reserved.
Comments and discussion welcome.

birdline.gif (1428 bytes)