On 12/11/06, Ron <rjpeace@xxxxxxxxxxxxx> wrote:
Statements like these can not be reasonably interpreted in any manner _except_ that of presuming the results: "I expect that you'll discover, if you actually do these tests, that this belief (that using arch specific compiler options lead to better performing SW) is fairly much nonsense." "...IMO a waste of time..." etc The correct objective response to claims w/o evidence is to request evidence, and to do everything we can to support it being properly gathered. Not to try to discourage the claimant from even trying by ganging up on them with multiple instances of Argument From Authority or variations of Ad Hominem attacks. (The validity of the claim has nothing to do with the skills or experience of the claimant or anyone else in the discussion. Only on the evidence.)
/shrugs, this is not debate class, I just happened to have barked up this particular tree before, and decided to share my insights from it. A lot of the misunderstanding here stems from legacy perceptions about how cpus work, not to mention the entire architecture. If somebody produces hard facts to the contrary, great, and I encourage them to do so. also, some people posting here, not necessarily me, are authority figures. :-) merlin
