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 2008 Jerry E. Pournelle, Ph.D.

February 25, 2008

Continuing the discussion of the New Mac:

Experience With AppleCare


I've been reading of your migration to the Mac platform with interest. I thought you'd be interested in my experience with AppleCare.

Since I returned to the Mac platform back at the end of 2004*. I have not had the need to call Apple for tech support. However, I noticed recently that the battery life on my 2.16 GHz Macbook Pro, purchased in January 2007, was deteriorating drastically. I could pretty much watch the battery life percentage indicator in the menu bar drop second-by-second. I was lucky to get an hour and a quarter out of it, and that's with the Bluetooth module and AirPort turned off, and screen dimmed.

Because my MBP is my primary machine, back in December I sprung for the AppleCare extended warranty. While I mostly used the machine while plugged in, I do sometimes need to run it off battery power. So, I called AppleCare last Thursday afternoon (2/15/08). It turned out to be one of the most pleasant experiences I've had with tech support from any vendor.

First, I was greeted by an automated system which asked me a few questions which required a "Yes" or "No" answer. After about a minute on hold I was then forwarded to Colin. It's a sad sign of the times that I was happy to get a native English speaker rather than someone claiming to be "Steve" or "Bob" but who's real name is probably something like "Mujibar."

Anyway, after getting my name and MBP's serial number, Colin walked me through a few troubleshooting steps and decided that I should get a replacement battery. He needed to get a supervisor's override to put in the order, which required me waiting on hold for about 5 minutes. But at the end of the call he confirmed my shipping address and put in the order. I was told that it was a bit late in the day to get the replacement battery shipped, so I should expect to see it Monday 2/18/08 along with a return shipping label for the old battery.

To my surprise, the battery arrived Friday the 16th, one day after my call to Apple.

Note that under the terms of AppleCare, laptop batteries are considered consumables so unless they are defective, Apple isn't obligated to replace them. But based on the reported behavior and information in System Profiler, it wasn't difficult to convince them to replace mine.

It was refreshing to get a competent tech support representative who spoke English well, and not have to throw a fit to get good service. I'm impressed.

* My first Mac was a Mac Plus bought in January 1997, but I primarily used PCs running Windows, Linux, or FreeBSD from 1992 through 2004.

-- Dave Markowitz

Thanks. I will probably get AppleCare when I bring in a MacBook Pro to accompany Imogene, but we'll see.

Subject: Mac Sculpted keyboards

Jerry, Most Apple stores and most other Apple authorized distributors have discontinued the sculpted keyboard, but I visited the University of Denver book store recently and found they had a nice stock. Denver is probably too far to travel for just a keyboard, but you might check local college bookstores for stock, or better yet perhaps a reader knows of a cache somewhere.

Dan Brantley

Thanks. I should look for one, but at the moment I am making do with the new flat keyboard, and next I will install the Microsoft Comfort Curve.

Mac OS X Time Machine and SuperDuper!


I use a partitioned Firewire drive for my MacBook backup. One partition is for Time Machine. The other holds a bootable clone of the internal drive. I use SuperDuper! to make the clone because its Smart Update feature saves time, but you can use the Mac's Disk Utility to Restore the internal drive to the external partition.

In Time Machine, I exclude the entire System folder, the Downloads folder, and any Parallels virtual machines, which are already backed up on the bootable partition. This frees up a lot of space on the Time Machine partition. The Windows XP virtual machine alone takes up 10 to 15 GB, and it changes every time Windows is used, so a few changed copies would quickly fill the Time Machine partition.

Running SuperDuper! once a day is enough to keep the folders excluded from Time Machine adequately backed up for my purposes. Actually, once a week would be plenty.

Should the internal drive fail, it would be easy to boot from the external drive and continue working, and then to restore everything to a replacement internal drive.


Thanks. I haven't decided on my final setup, but I am very fond of Time Machine – at least of the concept. I haven't really had to rely on it yet.

Apple Network Shares

Dr. Pournelle:

You can have network shares automatically reconnect when you restart your machine by creating an alias. You do this by establishing the connection to the network share and provide credentials. You then create an Alias. That is located under the file (I think) menu item. Highlight the network share on your desktop, then create the alias. Leave both items on the desktop. The next time the Mac starts it should reconnect to the network share.

At least it works in my environment. I am mostly Windows trained and only deal with the Macs when there are problems and am by no means proficient.

Ray Thompson

Thanks. The downside is that it takes a long time and the computer is not available while the connections are established.

Kaypro Z-80 CP/M


I too had a Kaypro Z-80 CP/M machine. I wrote my master thesis on it as well as a lot of papers in graduate school. I still have it in a closet. Every couple of years I pull it out and turn it on. It still has a great keyboard and a great little text-only display. If I only had a way to move text files from it to my iMac....any ideas?

-- Dwayne Phillips

I put the question to the advisors:

Alex Pournelle says:

A quick check of the Google Machine shows multiple companies who still offer disk format conversion services. I typed "CP/M disk conversion" but your mileage may vary.

The first (because they pay Google for placement) hit was Computer Convert, link here, who seems (from 10 seconds' browsing) to be fluent with stuff us old guys remember fondly.


Peter Glaskowsky adds:

Well, let's see. Mounting a network drive over WiFi would probably be easiest.

Wait, no...

I second Alex's nomination of a commercial data-conversion company. They aren't cheap but if Mr. Phillips can afford it, that's the easiest way. It's usually a good idea to give them a non-critical disk to try first before mailing off your collection of irreplaceables.

This company is recommended on the FAQ for comp.os.cpm:


They provide the service, and created a program that can read some old formats on IBM floppy drives.


If the service route is too expensive and you can't copy by floppy, there's always RS-232. I don't find Hyperterminal on my Vista tablet, but there are probably many terminal programs out there that'll still work on the Windows side. If Mr. Phillips doesn't have a terminal program on his Kaypro, there's always PIP...

That FAQ is here, btw; see question 12.


. png

I have a bunch of old 8" floppies, but I suspect the 1's and 0's have become ½'s and unreadable. I suspect that may have happened with the 5 ¼" as well, but a commercial service does seem like the best idea.

Another tale of Vista:

"Copying a file seems like a relatively straightforward operation"

Dear Dr. Pournelle:

Have you read the February installment Mark Russinovich's blog on Microsoft TechNet (link) ?

This seems to confirm that, while severely unit tested, VISTA was never fully exposed to the real world conditions of a working network before release. How else can you explain why the copying or moving of large numbers, or large-sized, files across a network (or even on a single machine) can cause so much disruption to VISTA, leading to so many confusing results, and ambiguous error messages? As the blogger writes: "Copying a file seems like a relatively straightforward operation"! Yet these aren't rare failures in unusual situations. The blog acknowledges it to be an error in design, reproducible in practice, but never exposed in testing,

More worrying, while Russinovich claims that VISTA file management has been fixed in SP1 to be as good as, or maybe better than, Windows XP, some of the follow-up questions and comments suggest that the problem may have only been half-fixed.

I dearly look forward to seeing VISTA SP1 in March, but I do not expect it to clear up all of my complaints about VISTA! "Reboot daily", "Save regularly and often", and "Read after write, to confirm copying" were all good axioms in the early days of desktop computing and, to the VISTA user, they are just as important today.

Alun Whittaker

Peter Glaskowsky notes:

I don't agree with Mr. Whittaker that Russinovich's piece "seems to confirm that, while severely unit tested, VISTA was never fully exposed to the real world conditions of a working network before release." To the contrary, it surely was, with widespread testing within Microsoft and by many beta testers.

Some people don't like the results, but I'm reasonably happy with Vista and have now installed it on all of my Windows-compatible systems.

. png

I have no real data but I would be astonished if Microsoft had not done extensive field tests. Nevertheless, I continue to recommend windows XP for all those not of an adventurous caste of mind. There is nothing in Vista that any of us really needs. And under no circumstances let them 'upgrade' your XP machine to Vista.

We'll see. Eventually they will get it right, and Vista can be fun, but I still have networking problems and mysterious glitches that fix themselves. I'll leave Vista on Roxanne, but I won't put it on any of my other machines. (And of course most of the work that used to be on Roxanne is now done on Imogene the iMac.)

A comment on Kindle:

Hello Jerry,

I got my wife a Kindle for Christmas. For several years she has been asking why she can't buy a portable computer to read books on, and when I saw the ad about Thanksgiving I decided to surprise her. My 15 year old son told me I shouldn't buy version 1 because by next year version 2 will have color, etc. Ah, but next year I wouldn't manage the surprise.

Diana has read several books on it and I have worked through one. It is clunky. The buttons are way too big. There is no way to pick it up without bumping one. I also don't understand why the on-off switch and network switch are on the back. Every time I flip it over to get to them I bump one of the page buttons. The book I read was non-fiction and some of the tables and outlines were distorted. Diana says it works better for fiction. It is a new born, but it does work.

Ad Astra,


My own experience has been far better. I soon got used to picking up the Kindle without hitting buttons, I find it readable, and when I am in a waiting room I do find it a good way to have lots of reading matter. It is a bit over priced, but it's on the right track, and I am glad I have it.


Your comment [in View from Chaos Manor] about being awakened by a wrong number reminded me of a thought that I had a number of years ago.

At some point in time it will be desirable/necessary to revamp the 3 digit area code plus seven digit phone number. When and if this happens an extra check digit should be added. Then, using a modulo 7 or 11 scheme similar to that used with credit card numbers fumble fingered wrong numbers would be detected at the originating exchange and a message provided that the number requested was invalid.

Bob Holmes

Excellent suggestion. I wonder if The Phone Company has thought of it?

And finally a discussion by the Advisors:

Bob Thompson said:

Patentability Of Business Model And Software Patents Comes Under Court Scrutiny

[Techdirt article link]

This one could be big. If things work out as I hope they do, stupid stuff like the Amazon 1-Click patent will go away and software patents will be, if not entirely eliminated, at least significantly reduced in scope, and a lot of patent trolls will find themselves looking for new jobs. Given that big software patent holders like Microsoft have arguably suffered more than they've benefited from software patents, I suspect that many of them will look at this case with mixed emotions, but all of them must at least see the benefit of getting the software patent monkey off their backs.

Robert Bruce Thompson

Peter Glaskowsky comments:

I think it's highly unlikely this case will resolve the open questions regarding business-method patents. The most likely result here is that at least some claims of this particular patent will be ruled excessively vague because the specific methods for making the decisions required in the claims are not disclosed in the claims.

For example, claim 1 requires selecting a "fixed rate corresponding to a risk position of said consumer" but doesn't say how to do it. Selecting the best rate appears to be the essence of the patent, but it isn't part of this claim. This is analogous to claiming an "alloy suitable for the construction of an airplane" then saying you have a patent on airplane construction.

Although I haven't read the patent in this case, it's likely it has dependent claims that lay out methods for selecting these rates, and those may well be valid. (Similarly, there were a great many valid patents that claimed the use of specific alloys and tempers of aluminum based on specific requirements for aircraft construction such as resistance to stress corrosion cracking, etc.)

And I'm certain this case will not provide a basis for re-examining software patents, where the rules have already been debated at great length and have settled out fairly sensibly. Software patents already have to claim a specific connection between abstract algorithms and concrete results, and that isn't likely to change.

. png

I fear I must agree with Peter, but I do wish something could be done about patent trolls.

And that should do for this week....