The end of MS-DOS: When Window Manager began, MS-DOS was Microsoft’s most important product, and Windows 3.x ran on top of it. On May 3, 1993, InfoWorld’s front page reported that the new DOS 6.0 mysteriously corrupted hard disk files on a number of PCs that the InfoWorld Test Center had tested.
After Microsoft technicians flew down to critique the tests, InfoWorld was forced to report the following week that one-third of the corruption was not the fault of DOS. No one could explain, technically, why the other disk files were scrambled. Competing publications clucked that nothing was wrong with Microsoft’s cash cow, DOS 6.
I was the first journalist to explain, in my column of June 7, 1993, exactly how a new disk-compression utility in DOS 6 corrupted files. And on August 23, I disclosed a free utility that showed a similar flaw in DOS 6’s new disk cache.
The outcome: Microsoft was forced to recall all copies of DOS 6 from store shelves, at enormous expense. The problems were corrected in a new release called DOS 6.2 — the last stand-alone version of DOS that Microsoft would ever ship.
The end of DR DOS: DOS 6 had been created to stop a competing product, Digital Research’s DR DOS, that offered disk compression (later the operating system was sold to Novell and then to Caldera). My November 15, 1993, column reported that Microsoft tweaked the beta version of Windows 3.1 to display an error message when it was installed on anything other than MS-DOS, in a successful effort to keep PC makers from shipping DR DOS.
The outcome: Microsoft paid Caldera a settlement last year, reported to be at least $150 million.
Undocumented features: I revealed in my November 14, 1994, column that Windows 95 included several secret functions used by Microsoft applications. InfoWorld ran a front-page story the following week quoting such then-major companies as Lotus and WordPerfect saying they would’ve used these functions had they been disclosed.
The outcome: Microsoft released a white paper explaining that its applications used the hidden functions only in unimportant ways. This would later become more grist in Microsoft’s antitrust case.
Removal of Internet Explorer: At the height of that antitrust trial, my March 8, 1999, column announced a free utility that removed Internet Explorer from Windows 98. Microsoft had spent millions arguing in court that IE was inseparable from Windows, but the utility was so simple that even a federal judge could use it.
The outcome: The utility’s author launched a successful online business, embeddingwindows.com. The company has stripped Windows down so it runs in 32MB of flash memory and supports streaming video, among other things.
We’ll conclude this series next week.
Making an impression
I’ve always wondered about the effectiveness of “affiliate” programs offered by e-commerce sites. A research firm, Affiliate Metrix, has released a reasonably scientific study that provides detailed answers.
The study, which surveyed more than 100 online merchants with affiliate programs, updates the old 80/20 rule, estimating that 95% of sales that were acquired through affiliates came from only 5% of those affiliates. Other specific findings from the study include the following:
- The majority of merchants reported that 80% of their affiliates had not yet produced a single click-through to the merchant’s site.
- At the same time, a significant minority of merchants reported that more than 20% of all sales come through affiliates. This bump in sales could mean the difference between profit and loss for some e-commerce sites, and it’s an inexpensive way of acquiring new customers.
- 38% of the merchants reported that their affiliates generate more than 500,000 impressions per month. The other merchants reported fewer page views than this, but even the lower counts can still be significant for many merchants.
- Affiliate programs take time. Within the first 12 months of operation, 21% of the merchants had signed up as many as 2500 affiliate sites. After a year or two years of operation, however, 93% of the merchants had more than 2500 affiliates.
The service costs $US79 for an unlimited period, but it requires as many as 24 hours for human approval of applications. There is a no-questions-asked, 100-percent-money-back guarantee, however. See the site.
Livingston’s latest book is Windows Me Secrets (IDG Books). Send tips to Brian Livingston.