|Home » Rants
To Bug or Not to Bug
Week of 11 March 2000
There has been a lot of talk in the media lately about the supposed 63.000+ bugs in Windows 2000. It's all based on yet another internal MS document that somehow managed to get itself 'liberated'. The document, an internal memo, was authored by Marc Lucovsky, one of the development leaders for Windows 2000.
The relevant passage:
Our customers do not want us to sell them products with over 63,000 potential known defects. They want these defects corrected. How many of you would spend $500 on a piece of software with over 63,000 potential known defects?
ZD had a field day with this one, as did the anti-MS forces. In MS' defense, on the other hand, we hear that 'these bugs are not that important' and that 'most users will never see these bugs anyway'. All in all, the arguments on both sides are rather weak.
What's important to remember (and evidently so easy to forget) is the internal memo that sparked off all this commotion in the first place. The language used in the memo is unequivocal; and so the only remaining question is whether the position taken was because of pure product fanaticism or because of a real and present danger.
For both alternatives there is more than ample evidence. Microsoft are truly 'betting the bloat' on this one (and about to lose most ungracefully according to radsoft.net) and the bug reports and bug fixes pouring in for Win2K, if not a tidal wave, do show signs of the product being a bit 'weak'. The first major bug fixes were announced and made available even before the product was officially released; further security alerts and bug fixes appear almost daily. For a product touted as the best ever from the Redmond company, this is surely a bad start.
Hardware compatibility and general performance seem to be the overriding issues, however - and it's difficult to find the objective truth with all the hype now flooding the market. Everywhere there are highway billboard ads, bus ads, TV and radio ads, all bespeaking the unprecedented stability of Windows 2000, all urging the general populace to toss out the old and run out and purchase the new. Yet it remains good advice to 'sit back and wait and see'.
- Most truly independent authorities still recommend waiting a while, until at least one or two official service packs are released.
- In either case, the cost of getting the right hardware to run Windows 2000 cannot be underestimated.
Windows 2000 is much slower than its predecessor, and there doesn't seem to be much reason for this at all. Amongst the fanatic claims being heard now, which everyone should watch out for, is 'Windows 2000 is just as fast as NT 4 - you just need more RAM and a faster CPU'. The portent of any grown human being making such a statement seriously must be meditated - and understanding the forces at work will give a good clue to just what the consumer is up against.
Windows 2000 is not going to necessarily work with the same hardware that ran NT 4, and it will not necessarily run the same software that ran without a hitch on NT 4 either - including MS Office modules and the MS development tools. The current list of incompatibilities seems endless. It cannot be stressed strongly enough that checking for hardware compatibility is of vital importance.
Don't upgrade to Windows 2000 just because you're curious - not unless you have a lot of money to burn and a throw away turbo machine to burn it on. Don't count on smooth operations on your 'test' machine for anything - don't use a machine that's simultaneously vital to your own business, hobbies, etc. You've too much to risk, and too much grief to go through when things go wrong (and odds are things will indeed go wrong). But if you have such magnificent hardware standing around collecting dust, and if you have a cool $500 to toss in the lake, then by all means - have a look.
Whether any of the 63,000+ bugs come out and bite you is hardly the issue.