Bob, your post sounds backasswards to me. "I'd argue that a flaw in one Unix implies a flaw in another Unix, just as much as a flaw in IIS3.0 implies one in IIS4.0 implies one in IIS5.0. If it is incorrect to say to lump all Unixes together, then all Windowses shouldn't be lumped together either. Frankly these statistics don't tell us very much at all." Who's made these implications? A flaw in Solaris can, may be, and very frequently is different from what may affect BSD which can, may be, and very frequently is different from what may affect other Unices. The reason these two are broken out is because they are much more distinctly their own system than other Unices like HP/UX, etc... I've never heard your bizarro IIS implication, but it's certainly different--if IIS 3 has a flaw by no means does it mean 4 and 5 do, but since they are direct descendants, generally speaking just adding new features, then that makes sense. But Solaris and BSD have substantially forked away from the original ATT Unix for sometime--development happens rather independently--what is synchronized are individual tools like gcc or apache. As for Windows, knock yourself out, break it out version wise even though in reality there are 2, but go ahead do it--it's appropriate. More people should now that Windows 95% is actually more like 5%, 5%, 15%, 40%, 30%.
And besides, doesn't the data lump Unices, Solaris, and BSD TOGETHER as "other?" A pretty standard practice that I've seen a million times before.
This post was edited by sodajerk on Friday, August 16, 2002 at 21:16.
|