🚀 go-pugleaf

In article <4d632h$qjr@daily-planet.nodak.edu>  writes:
> Ian S. Nelson (nelsoni@rintintin.Colorado.EDU) wrote:
> 
> : Eratostophenes?  I'll give up a few KB/sec on IDEA speed if it means that  
my
> : compilers and editors will run 10% faster.  
> 
>    Speaking of 10%, has everyone heard the news on Intel's Pentium 
> benchmarks?  They recently admitted that the advertized speed ratings 
> where erroniously high, by "an estimated 10%."  Intel of course insists 
> that it was all an inocent mistake, but I have to wonder:  EVERYBODY and 
> his cat knew that the Intel Specs were, not to put too fine a point on 
> it, bullshit.  Could Intel really be so stupid as to think their specs 
> were a reflection of reality in light of all the contradictory real-world 
> speed tests?  "Never assume maliciousness when something is more easily 
> explained by gross incompetence" is generally a good maxim, but I have 
> little faith in the honesty of Intel (Can we say "FDIV" folks?)

Agreed, and who is checking the last bits on the supposed FDIV fix and why is  
everyone so quiet on the P6 bug that causes IO errors. These are just the bugs  
that are so outstanding that they beat people in the face. Now that Intel  
publishes its "errata" the legal situation is that you bought the defective  
thing with the knowledge that it was defective. So don't expect a replacement  
this time. What about the errata/bugs/major-malfunctions that just quietly  
trash your data and files but only often enough to make you think that it was  
"user error" and you end up thinking, "gee, maybe I should take the Win95  
course for only $100 and the problem would go away". Of course, as Intel fans  
are so fond of pointing out these days, these chips are awfully complicated and  
maybe they all have bugs. While this may or may not be true, it is certainly no  
argument to use a chip by a manufucturer who couldn't even make something  
simple like a 80286 without a major malfunction and who tried to use  
statistical arguments to ignore the fact that its double precision arithmetic  
simply did not work. Don't get me wrong. I think that Intel based systems are  
great for home use and I own one myself. I would guess that Intel really  
doesn't need the professional and commercial sales to make a living. I think  
that its strictly a computer-penis-ego thing to push the idea that these  
processors are reliable commercial grade products. My concern is that  
professional and commercial users aren't making the distinction between  
commercial can consumer grade products and that all of the good stuff will  
perish in this blizzard of ignorance and horseshit. 


richard