By: Ireland (boh.delete@this.outlook.ie), May 2, 2017 11:52 am
Room: Moderated Discussions
Linus B Torvalds (torvalds.delete@this.linux-foundation.org) on May 2, 2017 11:53 am wrote:
>
> Does that mean that the 1% is unimportant? Hell no. The 1% is what makes or breaks the whole thing.
> It's the 99% that is largely unimportant and people just want it to work. How you do it is almost
> immaterial. Maybe you do it with good voltage scaling. Maybe you do it with really aggressive clock
> gating. Maybe you do it with turning off high-performance parts and turning on low-performance ones
> (at various granularities - it doesn't have to be about a "core", it could be at any level).
>
Most engineering systems are architect-ed, in some version of the above. It doesn't matter whether you're looking at networking, whether you're looking at operating systems, or whether you're looking at CPU's. It always boils down to something like this. The problem is, is that the vast, vast, vast majority do find the 99%, to be the most intriguing. It's probably because if one wants to 'appear' to be doing engineering, as opposed to stopping, and figuring out actually how to go about it, . . . it's a lot easier to wade into the 99% 'end of the pool', get all wet and drenched in all of that, . . . and it appears like you are ploughing through the workload and being exceptionally productive.
I agree though, it's the one percent stuff (the stuff which human intelligence can actually be applied to, and that which human intelligence was designed to deal with), that is the stuff that is demanding of most attention, most of the time. There's a problem though, with that. Many of us in our daily lives, working underneath management who are ultimately out of their depth. The result of that, is that the guy who might target the one percent, that's actually the one percent that needs to get sorted first - is also the guy, whom your typical incompetent manager - is going to single out, make an exhibition of. Because the same typical, incompetent manager also has a habit of 'showing off', and putting on a sort of circus for their mates. They point to the guys, who wade into the ninety-percent, and points to 'charts' and rubbish that demonstrate the size of their workloads completed, . . . . and it makes any guy who tries to separate out the one percent, and giv it it's due attention, look really bad.
The purpose of engineering really, at it's purist, was to turn the table around, so that at some stage sanity could prevail. Because really, it's only in focusing on the one percent, and one can obtain the correct signals that all other pieces (in the ninety nine percent need), in order to fall into line, fall into place. It doesn't matter here, whether we're talking about operating systems, networking systems, central processing units. It's universal. Engineering as a discipline had to be invented, because something had to prevail that would 'correct' the natural bias of the human being, to simply show off to other human beings, rather than trying to identify what really is the problem.
So like, this isn't really anything earth-shattering.
It's not that people are being purposefully dumb, or are being purposefully silly, or any of those things. It's that the human brain operates in that way, most of the time, for most people. Engineering as a whole discipline isn't something that was designed in sync with how the human brain operates. It was designed to recognize that the brain, often does the wrong thing, and we need to identify that, when it happens. And brought down, to it's most basic nuts and bolts - that is what it amounts to. The really clever brains out there, know that one percent of the problem, is by far and away the most important part of the problem to solve (and also, well within the reach of human being's capabilities). But the really clever brains out there, also know that a lot of other human beings are really dumb - and often, for the sake of an easy and prosperous life - the easiest thing to do, is lean with the dumb majority people, rather than against them.
The most valuable contribution of engineering to the human race - was that it didn't do the clever thing. It actually tries to do the dumb 'smart' thing. And generally speaking, good engineers end up paying the ultimate price for that, and just get their asses fired. Unfortunately, it is the human condition, and I'd say that it is as old as caves and cave men/women.
>
> Does that mean that the 1% is unimportant? Hell no. The 1% is what makes or breaks the whole thing.
> It's the 99% that is largely unimportant and people just want it to work. How you do it is almost
> immaterial. Maybe you do it with good voltage scaling. Maybe you do it with really aggressive clock
> gating. Maybe you do it with turning off high-performance parts and turning on low-performance ones
> (at various granularities - it doesn't have to be about a "core", it could be at any level).
>
Most engineering systems are architect-ed, in some version of the above. It doesn't matter whether you're looking at networking, whether you're looking at operating systems, or whether you're looking at CPU's. It always boils down to something like this. The problem is, is that the vast, vast, vast majority do find the 99%, to be the most intriguing. It's probably because if one wants to 'appear' to be doing engineering, as opposed to stopping, and figuring out actually how to go about it, . . . it's a lot easier to wade into the 99% 'end of the pool', get all wet and drenched in all of that, . . . and it appears like you are ploughing through the workload and being exceptionally productive.
I agree though, it's the one percent stuff (the stuff which human intelligence can actually be applied to, and that which human intelligence was designed to deal with), that is the stuff that is demanding of most attention, most of the time. There's a problem though, with that. Many of us in our daily lives, working underneath management who are ultimately out of their depth. The result of that, is that the guy who might target the one percent, that's actually the one percent that needs to get sorted first - is also the guy, whom your typical incompetent manager - is going to single out, make an exhibition of. Because the same typical, incompetent manager also has a habit of 'showing off', and putting on a sort of circus for their mates. They point to the guys, who wade into the ninety-percent, and points to 'charts' and rubbish that demonstrate the size of their workloads completed, . . . . and it makes any guy who tries to separate out the one percent, and giv it it's due attention, look really bad.
The purpose of engineering really, at it's purist, was to turn the table around, so that at some stage sanity could prevail. Because really, it's only in focusing on the one percent, and one can obtain the correct signals that all other pieces (in the ninety nine percent need), in order to fall into line, fall into place. It doesn't matter here, whether we're talking about operating systems, networking systems, central processing units. It's universal. Engineering as a discipline had to be invented, because something had to prevail that would 'correct' the natural bias of the human being, to simply show off to other human beings, rather than trying to identify what really is the problem.
So like, this isn't really anything earth-shattering.
It's not that people are being purposefully dumb, or are being purposefully silly, or any of those things. It's that the human brain operates in that way, most of the time, for most people. Engineering as a whole discipline isn't something that was designed in sync with how the human brain operates. It was designed to recognize that the brain, often does the wrong thing, and we need to identify that, when it happens. And brought down, to it's most basic nuts and bolts - that is what it amounts to. The really clever brains out there, know that one percent of the problem, is by far and away the most important part of the problem to solve (and also, well within the reach of human being's capabilities). But the really clever brains out there, also know that a lot of other human beings are really dumb - and often, for the sake of an easy and prosperous life - the easiest thing to do, is lean with the dumb majority people, rather than against them.
The most valuable contribution of engineering to the human race - was that it didn't do the clever thing. It actually tries to do the dumb 'smart' thing. And generally speaking, good engineers end up paying the ultimate price for that, and just get their asses fired. Unfortunately, it is the human condition, and I'd say that it is as old as caves and cave men/women.
