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:
The Mailbag

Jerry Pournelle jerryp@jerrypournelle.com
Copyright 2009 Jerry E. Pournelle, Ph.D.

November 29, 2009

This discussion began in my Advisors Conference.

Rich Heimlich asked

I'm still struggling to find a solution to allow me to get my RSS feeds into web-based e-mail programs that would result in one e-mail per story per feed.

Turns out this is much harder than it appears. I've even thought about writing my own program to do it but if you haven't touched Visual Studio in a decade or more like me you'd be SHOCKED at the differences today. Virtually all the code now is tied into endless pre-written routines that, if you don't know them, you're entirely lost as to their function.

I found an interesting site called FeedMyInbox that's interesting to look at just for the sake of how simple a website providing a service can be. However, it only retrieves feeds once every 24 hours and then only sends "digest" versions. That is you get one e-mail for every feed and that one e-mail might have 50 stories in it with no way to mark or remove the stories that you either read or don't interest you.

Thus it defeats the whole purpose. Plus there's no filtering so if you want that you have to join ANOTHER service like FeedRinse and then use the link it gives. So now you've got the feed, the feed filterer and the feed provider and your e-mail app. I hiccup anywhere and you're toast.

There's a general consensus that RSS isn't as successful as "it should be" and my take on that is because it was packaged and sold incorrectly from the start. The first mistake was the idea of needing a distinct app (a feed reader) to take advantage of it. Dumb. No one wants ANOTHER app that they have to go to. They just won't bother.

Second, the main way people access them outside of readers is in a browser and that's not so great either. Again, you can't filter out the noise. You can't delete the bulk of stories you don't want getting in the way of the good stuff and you can't easily mark those you've read as read or get rid of them so they just keep hanging around.

Imagine if your e-mail worked that way. Lastly RSS is, to me, just like an e-mail but automated. I want the info. Just let me work it like e-mail.

I've recently switched over to Gmail (and was surprised to find that it works so well integrating with my own POP3 server so I don't need to use a gmail e-mail account) because of my move to a Motorola Droid (Google Android) phone. It's really changed my life in a short time. Thunderbird just couldn't compete.

The problem is I lost contact with one of the key tools I use to stay on top of what's going on in the world. This is the perfect use for a tray-based local app where no service needs to be relied on. If I just could figure out how to do what needs to be done in VB it looks like a weekend of coding. Oh well.


Eric said

Have you examined www.feedmailer.net for your desired function? It seems to fit the bill but I'm not an RSS user. I've never seen the point unless I was to be severely bandwidth limited and wanted to avoid checking a site without knowing what new stuff was there.

I find it very annoying that the RSS reader in Live Mail is on by default using the MSN feed, and killing it is much more difficult than it should be. The main objection is that the indicator for unread items looks like unread mail to a casual glance. Unread mail I care about, the deluge of blather, not so much.

----- Eric Pobirs -

And Rich replied

Feedmailer is no longer functioning and is up for sale. Once you try RSS in e-mail and look at it just like any other form of information it takes on a much different light. At least it did for me. It was useless to me outside of a mail reader where I could apply any kind of filtering and segmenting by folders and such.


All of which was preliminary to the item I found interesting. Rich Heimlich said

Jerry, I know you used to play around with QB and VB back when it was called VB For MS-DOS (or VB-DOS for short). We used to talk a lot more by phone in those days.

You would be SHOCKED at what VB has become these days.

A typical section of code might look like this:

Private Sub GetChannelInfo()

  Dim rss As XmlNodeList = GetXMLDoc("rss/channel")

  Title = rss(0).SelectSingleNode("title").InnerText

  Link = rss(0).SelectSingleNode("link").InnerText

  Description = rss(0).SelectSingleNode("description").InnerText

End Sub

Those are SIMPLE examples. I've seen things that look like:

Title = xmlpath.xmlexplorer.xml.xmldocument.equals(doc)

What the hell? And you need to know all these inter-relationships and all the various types of what we would have called Libraries to know what each .segment is and does. The coding environment is so busy now.

You type just the letter X on screen and a pull-down pops up with literally hundreds (maybe into the thousands) of items you can type next. You can hit Tab when you highlight one and it enters it for you taking you to the next part. So if you type XML and the dot the pulldown fills with all the acceptable next options. for XML alone there are 69 entries and each of those has more below them.

For someone like me this looks like no one really "codes" anymore.

It's all just a series of IF/Then For/Next statements wrapped around convoluted calls to an endless array of outside routines. I had a friend come over who's a professional VB 6 coder and he couldn't even do anything with this stuff.

When I talked to one of my .net friends he made it sound like his whole life is staying on top of what the various arrays of options are.

Rich Heimlich

And that got me thinking. When the micro computer revolution began there was a parallel movement on code generation, proving programs, structured program design, and the general philosophy of programming. The goal was to build in program reliability so that programs would do what you expected them to.

There was also emphasis on program readability. Now this is supposed to be accomplished by extensive commenting, that that says nothing about program structure.

Back when there were debates about programming languages, a great deal of the debate was on top down structure. Many programs were actually fairly simple: the main program was a series of calls to sub-programs and procedures. The program structure was enforced by rigid rules: variables had to be declared at the level they would exist, type checking insured that subroutines didn't change an integer into something else (or that an overflow couldn't come back as a program instruction), there was range checking to prevent overflows, and such like.

There were several benefits to this. The first was that programs were fairly readable. You could get a pretty good idea of what the program was doing, and how it was doing it, just by reading the code. Second, you didn't have to be able to simulate the compiler in your head to be able to write code: there were rules, and if you followed them, the program would compile and very likely do just about what you expected it to do. If you had made a programming error the compiler would very likely catch it and reject it for you. Getting a complex program to compile was sometimes tricky, but there was a big payoff in reduced debugging times.

When the small computer revolution began, I speculated that the real revolution would come when it was more important to know what you wanted the computer to do than to be expert in teaching the computer to do things. Computer languages needed to reflect that.

I still believe that is true. For a while Visual Basic - insufficiently structured in my judgment, but still very useful - seemed to be headed that way, but Rich's example shows that trend has reversed.

Our hardware is now fast enough that we don't need hundreds of pre-defined program commands and structures in our languages. It is now time to develop programming languages that reflect that.

Apple's Mistake.


From Graham's site:

I don't think Apple realizes how badly the App Store approval process is broken. Or rather, I don't think they realize how much it matters that it's broken.

The way Apple runs the App Store has harmed their reputation with programmers more than anything else they've ever done. Their reputation with programmers used to be great. It used to be the most common complaint you heard about Apple was that their fans admired them too uncritically. The App Store has changed that. Now a lot of programmers have started to see Apple as evil.

How much of the goodwill Apple once had with programmers have they lost over the App Store? A third? Half? And that's just so far. The App Store is an ongoing karma leak.

Roland Dobbins

Well, whatever Apple has done, there are plenty of Apps being developed; something like 100,000 applications so far, and more every day. They can't have lost too much good will.

There are problems with buying Apps from the App Store. I had considerable difficulty last night because of their odd procedures. Apple has done well with the App Store advertisements - "There's an App for that," has become part of the American vocabulary - but the things you have to do to buy an App can be greatly improved.

I don't think of Apple as evil, but they do sometimes get priorities out of order.

Last month we heard from Dr. Hume, a lawyer and psychiatrist without much computer hardware experience, on building a new machine.

Here is his latest report:

Lessons Learned


More than a month after needing a new machine and hurriedly building one, I have learned a few things.

First of all, I still have a six-year-old Dell. Two other identical machines have died in the past fourteen months, so I need to be ready to replace this sole survivor when it inevitably dies. So I've been doing the homework I would have done before building my first box.

Quad core CPU's are the way to go. They can do several things at once far better than dual core CPU's. There is no comparison between my quad core and our dual cores. The current sweet spot in performance is the Intel Core i7 860, at $290. But the sweet spot in value is the Core i5 750, for $200. The main difference is that the latter is not multi-threaded. OTOH, few apps can make use of all eight threads. Then there are the AMD quads, which are now down under $200 - in one case under $150. Their performance is roughly comparable to the i5.

The motherboards for the Phenom II X4's are cheaper than the ones the i7/i5 use. Further, some come with decent graphics built in. And some of those come equipped with dedicated graphics memory. You save $100 on MB and graphics card alone, before you even consider the CPU's and the memory.

I could have saved a little money by re-using my "optical drive" - my DVD reader/writer, according to reviews. Not much has changed in the past few years.

I am expecting better-performing less-expensive solid state drives this spring, so I'm not going to push for faster HD's. SSD's apparently transform one's experience as much as going from dial-up to broadband.

For a machine that I will not overclock, I will get a Micro ATX motherboard and an mATX mini tower case. For a case, I will likely stick with Antec (probably the NSK3480), but Rosewill is making inexpensive cases that are getting good reviews. For the mATX motherboard, the Gigabyte P55M-UD2 is still considered the best.

As for memory, 8GB is too much. 4GB is more than enough for almost all purposes. Memory faster than DDR3-1333 is essentially wasted on most machines. The reviews say that in "real world" testing, faster memory makes essentially no difference in performance. I could have saved $130 by buying half as much memory ($100 then, $120 now), and DDR3-1333 instead of DDR3-2000 ($30 difference for 4GB).

My cooler performance has fallen off, so I opened the case. Lots of dust. Specifically, bird dust. We have two cockatiels that live ten feet from my computer. Cockatoo and cockatiel feathers break off at the ends to form a nice, light airborne dust. Even the cooler on my video board was caked with the stuff.

But blowing out the dust and putting in new fans (Noctua fans are very quiet) did not drop the temps much. Basically, the thermal intermediary material (TIM, AKA thermal grease) of my Freezer Pro 7 may have cooked off and is less effective. Soon I will pull the cooler off and reinstall it, but this time with the correct procedure. Before, I just slapped the cooler on top of the CPU. But this meant that the TIM was sandwiched between the pieces of metal, holding them apart. The preferred approach is to try to get the surface of the cooler to touch the surface of the CPU heat spreader (it looks like a cover) as much as possible. The purpose of the TIM is to inhabit the little spaces that would otherwise make air-filled voids. Since air is a thermal insulator, it needs to be avoided. At the moment, the highest rated TIM is GELID Solutions GC-Extreme Thermal Compound.

Overclocking memory and CPU is like hot-rodding. Back when I could tinker with my cars, I did so. It was fun. So is computer modding. But it's an end in itself. In fact, I have separate BIOS profiles for fun and for normal use. I want this machine to last.

But there is enough of a market for tweaking one's machine that both AMD and Intel cater to it, and motherboard makers sell multiple models aimed at OCing.

Anyway, I figure I can build a new machine with a great-performing quad core CPU for around $500.


First off I'd get to work if not replacing your six year old machines, then at least backing them up onto newer hard drives. The average useful life of a hard drive is about three years. They usually give some warning when they're about to fail - they slow down a lot due to what we used to call "soft errors" (retries), and often they get noisy; but they don't always give you warnings. My experience is that the drives are the first things to go with computers (second for me has been the on-board battery that keeps the clock working when the machine is off). Better would be to replace the machines. I give that advice but I note that I have several Year 2000 era machines running, and it's pretty well time for me to replace them...

I agree with your sweet spot analysis. It's a neat question whether it's worth the extra $90 to have the multi-thread capability. If this were to be the only machine, I'd say go for the multi-thread; there are increasing numbers of programs with multi-threading in mind, and that will really speed things up. And while AMD has come a long way in catching up, I'd still stick with Intel.

On cooling: dust filters get clogged easily. Bill Godbout used to have monitor programs that would bug you about cleaning your filters. Of course that was in S-100 days when cooling was even more important than now, but it's still vital to keep the air flowing through the system, and dust is ubiquitous in home environments even without cockatiels. I guarantee you that a Husky can clog the works on computers in rooms she never goes into...

Even more important is the way you spread the thermal paste. It's simply not true that if a little is good, more is better. As you note, the purpose of the stuff is not to be conductive but to fill in the tiny voids of metal to metal contact. I again caution you: the chip coolers that come with Intel CPU's are generally not adequate, and you're better off with any of several third party chip cooling systems.

Built in audio on most standard brand motherboards is now pretty well good enough for any system that isn't going to be part of a very fancy home theater or some such. It's probably true with video as well. At one time screen text was affected by the video card - they had character generators in them. Now that's all done by the operating system from fonts on your hard drive, but I still have the superstition that ATI video boards have the best character generators (as they did back when that mattered) because readable and attractive text is one of my major requirements for any system. Actually, I suspect that on-board video is good enough for most purposes. Whatever your video system, it's important to keep the video drivers up to date.

Solid state "hard drives" are coming; I am about to install one in my MacBook Pro. Long ago I said that silicon is cheaper than iron, and solid state drives would replace spinning metal. It wasn't my most timely prediction, but it does seem about to come true - but I would not let that stop you from getting a few new drives to back up any that are more than three years old.

Peter Glaskowsky reminds me:

I recently learned that some companies are offering SATA hard-disk adapters for the optical-drive slots in the new MacBook Pros, and I'm tempted to put a large solid-state drive in mine. I suspect I'd like the extra space and performance more than the built-in optical drive, which I can move to an external enclosure.

Link to PowerBook Medic page.

. png

There is continuing progress in hardware, with many exciting developments here or almost here. It's about time I started putting together at least one more system capable of taking over from any main machine at need. I'll almost certainly build a sweet spot system for that, and I expect it will be about as fast as the best system here.

Brian Bilbrey reminds me

You should still be recommending Building the Perfect PC. While the specific recommended hardware has changed, all the what-to-think-about advice and how-to-build advice is as fresh as the day it was picked. And of course, there's current hardware recommendations at http://forums.hardwareguys.com/, the supporting site for Bob's PC hardware books.

All of which is quite true.

You're not alone w/ the Firefox freeze problem.


Happens to me frequently, despite the recommendation to use the Flashblock extension plus my use of Adblock Plus as well.

So, I am increasing my use of SRWare's IRON browser (the source code starts with Google CHROME then all the 'phone home' tracking features are stripped off) as it seems to be the fastest tabbed brower.

73s & best regards de
John Bartley K7AAY

Let's do the math: CO2 is 3.42% of all greenhouse gases. Man-made CO2 is 3.2% of all CO2. 0.0342 * 0.032 = 0.1% So, if all the planet went 100% nuclear + solar power _tomorrow_, greenhousing drops by 1/10 of one percent. Should CO2 shedding stop 100%, another greenhouse gas would replace it; dihydrogen monoxide, which we *cannot* control (it absorbs 20X more IR in the same spectrum). Let's not cripple our economy w/ cap+trade before we understand what's really going on!

The only speed that really matters with a browser appears to be the server speed that feeds the browser. Perhaps not, but that's been my experience.

I continue to use Firefox and I continue to be angry when I am in the middle of ordering a product, there is an Internet call, and wham! I am told I need to update some Firefox addon - and then have to start over back where I was ordering vitamins or whatever. Ah well. The tabs do work, and I keep using it.

Firefox Problems


Some things just do not make sense. Take the Mozilla group for instance. I have used Firefox and Thunderbird since they came out but have become disillusioned since starting to develop a website. I prefer to not use a web form to solicit emails. I want to use the hmtl mailto tag to link to the default email program. I first configured the script

<a href="mailto:name@isp.net" <mailto:name@isp.net> >
name@isp-life.net</a <mailto:name@isp-life.net%3c/a> >

and tested it in my hmtl editor. Clicking on the link took me to Thunderbird. Next I copied the script into a web page and ran it with Firefox where it did not work. After chasing my tail for far too long I determined that when I used IE the link worked as it should. This seems strange that a link activated with IE would work but not when using one Mozilla product to another. Searching around the web forums I found out that there has been a problem for sometime in this area. I determined that it is not a Thunderbird problem by selecting gmail as the default email program where once again Firefox failed where IE functioned correctly. If the reports are correct part of the problem lies with multiple windows users. If one user uses a Firefox Thunderbird combo and another uses hotmail and another gmail then mailto does not work. Whatever the problem is, it is disappointing that software so highly regarded can not handle a simple task like this. >From a website development point I will have to abandon any attempt to use mailto and hope that I do not have anymore problems with Firefox. It is not a case of me switching browsers - I have to write for all browsers and having viewers having to fiddle Firefox into performing a simple link function is not an option.

Disappointed in Firefox,

Jim Potter

Ps - Just tried your Email Me link from Firefox and it failed. With IE it linked through to Thunderbird!

As I said above, Firefox continues to annoy me, but not to the point that I stop using it; and they do tend to get problems fixed as time goes on.

On the other hand, I can't really complain about Internet Explorer in its latest incarnations, and it's mostly out of habit that I don't use it now. I do caution you: if you're doing updates of Microsoft software, it's rather important that you shut down rival browsers and use IE for those operations. They will go much smoother.

The User's Column, November, 2009

"The premise of progressive education is that teachers have nothing to teach: they are supposed to induce their pupils to think for themselves."

And then when you go to work in the real world the last thing most companies want is for you to think for yourself. That is too threatening to management. Management exists for the sole purpose of doing the thinking and everyone else is just supposed to do what management says. So I guess college is worth it if you're going to be in management and a waste of time for everyone else.


That hasn't been my experience. I worked with Boeing, Aerospace Corporation, and North American, and I was pretty well given assignments and support, not "management" as you describe it. I've never been in one of those situations with the pointy-haired incompetent manager. I did have a rather imperious supervisor at North American once, but he was a famous WW II war hero, and he didn't really interfere with what I was doing; he just wanted to show he was necessary, when we all (including him) knew that he wasn't, and the company kept him on because he had The Medal. But even that was minor, and accommodating him mostly consisted of listening to rather interesting stories of the Pacific War.

On IP V6

I just came upon this info that I thought you might be interested in from the Microsoft Directory Services Team. I will give a snippet below and then include the link.

"Since IPv6 is the preferred protocol for Vista and later Operating Systems I was going to implement an IPv6 Site/Subnet association in Active Directory Sites and Services and see if this would improve DFS referral ordering of the in-site DFS target server. I checked the configuration of IPv6 on ContosoFS1 and found the protocol had been unchecked; which is not a good practice.

It's a common misconception that unchecking IPv6 disables the protocol when in fact all it does is introduce transient errors. Windows Vista and later operating systems heavily rely upon IPv6 for internal operation, which means the protocol cannot be disabled or uninstalled entirely. Unchecking IPv6 on the adapter settings only unbinds the protocol from the NIC and the OS can still attempt to send remote traffic to the NIC where it never hits the wire.

For those who have some aversion to having any IPv6 traffic hitting the wire there is a "supported" [not recommended by the Microsoft Product Group] way to disable outbound IPv6 using the DisabledComponents registry value as directed in KB929852. Contrary to what the article states, it is possible to simply use Group Policy Preferences (TechNet link) to disable IPv6 domain-wide - there is no need for a custom ADMX.

NOTE: It is beyond the scope of this blog to determine which components of IPv6 should be disabled in a given environment using DisabledComponents. To completely disable all External use of IPv6 configure DisabledComponents to ffffffff. Also, if you find IPv6 remains checked in the UI after configuring DisabledComponents in the registry rest assured the protocol is disabled for all remote traffic."

Another TechNet link


Peter Glaskowsky comments

IPv6 doesn't seem likely to be significant for consumers any time soon, though I gather it's of immediate importance to ISPs and such.

That may be okay. It occurred to me a while back that Network Address Translation, which IPv6 makes unnecessary, is probably the single most valuable form of protection for consumer PCs.

Does IPv6 even have an equivalent, or are all IPv6 systems effectively naked on the Internet?

. png

And Eric Pobirs notes

From what I've read, there is active resistance from the IETF but they once thought NAT on IPv4 was evil. People went ahead and did it anyway without an official standard because it was needed.

That basic level of protection afforded by NAT is going to make it an inevitable IPv6 implementation when it becomes real for consumers. The consumer sector is just too generally clueless about security to give up so simple a protection. Plus NAT-PT will help in the transition to IPv6 for consumers.


Captain Morse adds

I haven't been able to find anything definitive about Comcast's (my ISP) plans for IPV6 for residential customers. It doesn't seem to be very high on their "todo" list.

Ron Morse

The relevant Wikipedia page says NAT-PT is dead, and while it mentions a de-facto replacement (Transport Relay Translation, aka TRT), the RFC for that implies that it doesn't support IPsec, which seems to disqualify it for widespread consumer use.

There's a long paper describing various alternatives for bridging consumer IPv4 equipment to the IPv6 Internet, but too long; didn't read. :-)

. png

Eric sums up:

Scarcely any ISPs are forthcoming on the subject. (Network World link) I don't expect to hear much about IPv6 for consumers until well after they have other areas of their operations covered. Which is likely just as well. The need is far greater in items like cell phones, which need more direct contact with the world and preferably each their own IPv6 address.

Getting consumers rolled over is going to be a very long-term project. A decade from now I expect most homes to still run IPv4 internally, though by then at least nearly all of the machines in use will be running an IPv6 capable OS. If the switch were forced today you'd have a zillion people with twelve year old systems screaming bloody murder, no matter how inexpensive and better a new system would be. (Earlier this week I was at a Fry's location and met a guy trying to find a replacement 30 GB drive for his neighbor's laptop. He figured if the 320 GB drives {the only parallel model still stocked} were $60, then a 30 GB unit should be no more than $30. I had to explain to him that they simply don't make those anymore and that the cost of replacing the drive would be far better applied towards a new laptop. Even the $300 bargain basement units were going to be far better than any unit old enough to have shipped with so little drive capacity.)

The big problem, once the majority of computers are up to snuff, is replacing the tens of millions of routers out there that cannot be upgraded for IPv6. I expect the ISPs are hoping this happens by attrition, with IPv6 capable unit being used in new installations and purchase to replace failed units. But that could take a very long time. The store shelves are still mostly filled with gear that will never do IPv6.

Computing at Chaos Manor, November 9, 2009 - Windows 7 and the HP Tablet

Jerry -

Your recent column (11/9/2009) included the following comment:

"The other system still running XP is LisaBetta, the HP Compaq TabletPC. She's far too old and slow for Windows 7."

I just wanted to mention that at least under certain circumstances, the HP TabletPC will run Windows 7 just fine, although it doesn't support Aero. I have two of them (TC1100s), each of which has 2G of memory (the maximum the machine will recognize) and a 128G solid-state drive (also the maximum the machine will recognize). You have to load some of the tablet-specific drivers in XP-compatibility mode, but everything, even the Q menu, behaves as you would expect. There are several places on the Internet that discuss how to accomplish this (and yes, they are web sites for people who are fascinated with HP's tablets of that vintage).

Performance is not bad at all, particularly when you take into account the age of the equipment


I'm not at all certain if any of this would be desirable enough to chase just to get more life out of your tablet, but until somebody has something that I find as compelling as the TC1100, it's the machine I will use. I'm the director of I/S at a corporation with revenues of ~$2-3 billion, so I could get something else if I wanted to.

I've followed your column for the better part of my 25-year career. Thank you for the insights you have shared in that time! More than once they have helped make my life easier. Plus, I just plain enjoy reading it.

Jeffrey D. Elmer
Director, Information Systems
Dairylea Cooperative Inc.

Thank you. I am looking into the possibility of getting Windows 7 onto LisaBetta. She's a bit old and slow - not terribly so, but she is pretty old - and it would be interesting to see if Windows 7 would buck her up. A TabletPC and OneNote are still the best research tools I know of, and the old Compaq TabletPC is the most useful form factor tablet I know of: I can remove the keyboard or carry it along at need.

I expect the development of netbooks will lead to a new generation of tablet that contains Bluetooth telephone connection so that something you carry in a small messenger bag becomes your PC, email, portable memory, OneNote research system, and your telephone; as well as contains a decent still and video camera. Solid state "hard drive" of course. Lots of memory. They could build that now but the battery life would be a problem. I expect to own one some day. Until then perhaps I can keep LisaBetta going. Thanks.

Computing at Chaos Manor: Comment on network neutrality

Hi, Jerry...

I'd like to point you to David Isenberg's discussion on network neutrality at this link.

I think there is an inherent conflict between the commercial ISPs like Comcast who provide both "content" and connectivity. We seen this already where carriers restrict third-party applications (particularly VoIP traffic) on their own particular network. The result is a balkanization like the current cell-phone imbroglio being played out with the iPhone and ATT.

I appreciate your argument that this is a blip in the road, and that if carriers practice discrimination that the customer can just go to another carrier. In our location there isn't much choice, and in many parts of the country there may not be any other vendor.

Best wishes,

--- Larry
Lawrence Keyes
Microdesign Consulting Inc.

Net Neutrality.

While I tend to agree with you about less government regulation when it comes to ISP there are some big problems.

1. They have recived billions of dollars in tax money to provide universal broadband coverage but have failed what they promised.

2. There is a real lack of competition in the ISP market. Cable companies are granted "franchises" and are given access to right of way for their "pipes" as to the phone companies. We are a captive audience.

The other thing we REALLY have to do is get rid of this "the conditions of this contract may be changed without notice" clause that is in every contract ISP. Yes when I got my High Speed internet I was told it was "unlimited" but now they have changed the ads and changed the contracts. And yes the line about Google getting a free ride line has been said time and time again by the ISPs. It is good that the courts have struck it down for now but will that last? I pay for internet access. I pay for x bandwidth and at no time was I told that I got X amount of data per month for that price. If the ISPs are going to market it like cable TV then they should be forced to live up to that marketing.

If I sit at home and watch Cable TV 16 hours a day I pay no more than if I watch it four hours a week.

The net neutrality issues was created not by the customer but by the ISPs make public plans that where clearly counter to what they advertised and what the consumers thought they where paying for. I would love to see this handled without regulation but the ISPs don't seem willing to back down.

And here are two stories about the Government making me crazy.

1. How I came to understand the $4000 hammer.

I work for a software company. Our product is used by court houses around the world so we thought we where used to dealing with the Government. One day we get a call from the USMC and they want to buy our software. Great we gave them a demo and they loved it. It was the best program on the market and they wanted it but it had to be put up for a bid. Two weeks later the bid came. It was a 50 pound box of paper work! After we filled out the bid we found that we lost. A competitor claimed the matched the spec but didn't and beat our price by $20 per copy. The kicker was the support was $200 per copy more a year! We challenged the bid and lost. Tow years latter the USMC came back and said. That other program didn't meet our needs so we want to buy your software! We got that bid but how much time, effort, and money was wasted?

And just to show that it isn't just the US.

The Canadian government also bought our software. One day I get a call because a set of updates that we sent didn't get delivered. They were held up at the border for taxes! The Canadian department of taxation wanted to know the worth of the package to charge the Canadian government an import tax! I explained that it was a free update and that the floppy was worth about $1.00 but that wasn't good enough. I finally lost it and told them that they where trying to figure out how much money to tax so the governess could pay the tax to the government! I told them that they had just invented Taxterbation and they hung up on me.

That day we set up a BBS so the IT people could download the updates. In the end I am not sure which is worse.

Government regulations or unregulated mega corps? Probably like most things in life it is a case of all things in moderation.


We do not need a new regulatory agency in order to enforce truth in advertising. For that matter, if you buy snake oil you may have some obligations; perhaps one should cancel the subscription when you find that the oil comes from lizards, not snakes. Or perhaps not. The real question is do we need a federal agency to remedy the problem?

Technology advances rapidly. Bureaucracies adapt slowly. The history of the railroads and the now abolished Interstate Commerce Commission provides many useful lessons.

Alas, Pournelle's Iron Law of Bureaucracy is stronger than Moore's Law.

And Brian Bilbrey, Chaos Manor Reviews Managing Editor, says

A note about network neutrality from my parochial perspective: I only expect my carrier to move packets. I expect those packets to receive equal treatment by the carrier. If I want more packets, I expect to pay for those. I pay my carrier for my allotment of packets, Google pays for its allotment of packets, Hulu pays for its connection, etc.

How the packets move between carriers is subject to the peering agreements between them. Everyone has already paid for their connection to the network (and voice traffic just doesn't involve that much data). Hulu, YouTube, and other video providers: that's a lot of packets ... but again, they've paid, as have we. If my carrier wants to charge less than for services similar to Hulu or Skype or Vonage, that's fine. But to slow down the Vonage or Hulu packets is wrong, since I'm paying for the packets. It'd be fine if there was really choice for many folks, but connection to the Internet is, in most communities, a duopoly: the CLEC and the CableCo.

If Verizon (to pick one carrier among many) really wants to sell 20 mbit/second service, in the expectation that most users are going to be doing traffic in the kilobits most of the time, they've made a good decision for 10 years ago. That was then. Now this usage model is outdated, but they continue to sell the plans, then penalize/throttle the users that actually USE the bandwidth that was advertised and "sold" to them. I think that's heinous.

Efficient use of networks involves distributed copies of data. I think the carriers are in a bind, because peer-to-peer networking has been effectively painted as the bad guy in the RIAA/MPAA war on their customers. If you can make a request of Hulu, and get some of those packets from Hulu, and others from other Hulu subscribers on your own ISP's network, then the ISP is saving money on the inter-ISP peering side of the equation. But P2P is now evil. They're stuck.

Technology and the market will sort all this out; but it took a hundred years to get rid of the ICC.

: Microsoft Antivirus products-


Actually Microsoft did have an anti-virus product back in 1993(?) with Dos 6. They'd licensed the Central Point anti-virus program and re-labeled it as Microsoft AntiVirus for Dos & Windows (respectively). I suspect Symantec's purchase of Central Point Software in 1994 had something to do with Microsoft's decision to get out of the anti-virus business by the time it released Windows 95 which did not include an anti-virus solution. Also, Microsoft may have gotten too cocky. This link points to a Microsoft Knowledge Base article where they claim that Windows 95 included some protective features from virus infections (link).

My church has an old Gateway 486DX2-66 computer w/ Win 3.11 that I setup back in '94 for them still running and I happened to notice it had Microsoft AntiVirus on it a couple of weeks ago. Your article brought it to mind.


Nathan Stiltner

I had forgotten that one. Thanks. I can report that I continue to find Microsoft Security Essentials to be working properly and well, and almost invisible, and I continue to be satisfied with it.

I got several messages to this effect:

Windows Security Essentials Update Now

Dear Dr. Pournelle,

There is a very simple way to update WSE. Open WSE and select the Update Tab. There is a large button that say's "Update", and clicking on that updates your computer immediately.


Bill Grigg

The problem is that when you get the "Out of Date" message, and click on that, you get a different screen when you try to open the program; and that warning screen does not have the "update" button. On the other hand, if you wait a few minutes with the system connected to the Internet it updates itself anyway. Thanks to all who responded.

An alternative to transferring Outlook pst files

Hi Jerry:

You mention the inconvenience of having to transfer a large pst file if you want to use Outlook on more than one PC. I am using Sync2PST as an alternative to transferring a pst file between my home desktop and a laptop that I can connect to my home network. Sync2PST creates a shared PST file on one of the PCs, and can log any new items or changes on either copy of Outlook in that file. Either copy of Outlook can then be updated to match the other. Sync2PST isn't cheap (and you have to buy a license for each machine), but it seems to work very well. You can choose which folders to sync, and you can run it, and sync, when Outlook is running. There may be other programs that do the same thing (and perhaps do it better and cheaper), but I have found Sync2PST makes it much more practical to use Outlook on two machines.

Richard Spencer

Thanks. I continue to do the xcopy routine, which works, and it's not that tedious...

Subj: Gigabit Ethernet: why bother getting a new router?

Towards the end of the second item in the Chaos Manor Reviews Mailbag for 28 Oct 2009 -- the report from Ed on his first homebuilt system -- Ed says, "I'm upgrading my home network from 100M to 1G Ethernet. Found a switch, waiting for the new router to arrive."

If Ed has enough ports on his switch for everything, does he really need a gig-capable router?

Even the fastest Verizon FIOS feed is only 50Mb/s -- well within the capacity of 100M Ethernet, no?

Connecting everything through the switch would let all in-house devices communicate with each other at gig speed, and the switch would handle the speed translation between everything else and the slower link to the router.

I guess, if within-router processing became the limiting constraint on overall throughput between the house network and the Internet, it might be advantageous to get a router with a faster processor, but would it not make sense to try the old router first?

If nothing else, it would be interesting to see how performance of the new router differed from that of the old. My theory says there'd be no measurable difference. Perhaps Ed would be so kind as to do a Little Experiment, and report?

Rod Montgomery==monty@starfief.com

We can ask him. Routers don't wear out...

I am about to replace a pre-N router with a new one simply because I can. We'll see if it gives any performance improvement.

Subject: Windows 7 64 bit and Wireless Network Adapters


Just wanted to fill you in on my struggles over the past week with hardware and Windows 7 Ultimate 64-bit. You may recall, I did the installation, and everything went pretty well. I started to experience a bit of problem when doing data transfers and accessing the Internet...and worst of all...playing World of Warcraft! The machine that I put Windows 7 64 bit on is in a remote corner of my house, so I connect wirelessly. It worked fine under Vista 64 bit with a Cisco (not Linksys..the REAL Cisco) Aironet A/B/G card, I never had a problem. When I ran the upgrade advisor, it refused to acknowledge the card, the Cisco website did not provide drivers, so I ran down and got a new Linksys card. It worked fine...I thought...but as time went on, there would be a lot of lag or pauses when moving a lot of data across the card, and sometimes the computer would lock up, requiring a hard reset. I found that Linksys did not support the card on Windows 64-bit. This started me on a weeklong odyssey of trying to find a wireless card supported by Windows 7 64-bit. The only two cards I found to actually come out and say they were supported (Intel and an off brand), and have users confirm they worked were mail order, and somewhat difficult to get. I tried several cards and their drivers...many supported Win7 32 bit, but was unsuccessful.

The final and very effective solution was a wireless bridge (Linksys WET610N). Since this unit sits outside the computer and uses the built Ethernet port, it works perfectly. Ever since installing it yesterday, everything has worked like it used to, and I'm pretty happy.

I thought I would share this in case anyone else ran into these problems...I'm sure in time hardware vendors will produce valid drivers for their wireless cards to run on Windows 7 64 bit, but right now the choices are very limited.

Tracy Walters, CISSP
Information Technology Consultant

Thanks. A wireless bridge is often the solution... I have one Windows 7 64 bit system, but my others are still 32-bit. I don't really notice a performance difference, but over time there will be more and more programs taking advantage of 64-bit capabilities. World of Warcraft works just fine on both, of course.

Good experience with LG monitor W2353

Hi Jerry,

My teenage son has a three year old MacBook with a 13" screen. Needing a birthday gift, I looked at monitors that would give him a larger viewing area and better upright posture -- he was constantly curled up over his MacBook like the hunchback of Notre Dame!

I was leaning towards an Acer based on your recent recommendations, but the store I visited had Acers side by side with LGs. The 23" LG model W2353 had more vivid colors and was brighter, offered three ports (D-SUB, DVI-D, and HDMI for future HDTV or Blu-Ray viewing) and a 2 ms response time (vs. 5 ms for most other monitors). It also had a matte screen -- many other monitors were glossy (I think the Acer was also matte). It was $239, which I thought quite reasonable.

He now has a display nearly 2000 pixels wide... and I am slightly envious!

The biggest headache was finding the correct mini-DVI to DVI-D adapter as Apple's mini-DVI ports have varied slightly over the years.

The monitor comes with an driver install disk, but the MacBook didn't need it. The monitor just worked. If you wish to use only the external monitor and not the MacBook's screen (so that it stays dark), but still use the MacBook's keyboard, waking a closed and sleeping Mac by unplugging then replugging in an external hard drive (or other USB device) seems to work -- a rather curious and obscure process for a Mac.


-John G. Hackett

Thanks. Monitor prices continue to plunge. I'm now considering a much larger one, perhaps 30", for my main machine; but I'm in no hurry.