Re: DragonflyBSD kernel clock improvements

Dag-Erling Smørgrav wrote:
> Perhaps you should be more open to the idea that if
> you don't understand it, it's probably very hard to get right,

That can certainly be true. But perhaps we should also be more open to 
the possibility that some code will have accumulated cruft over the 
years and might be better off with a re-write (I am not saying this 
might be one of those areas - I haven't looked).

I would also say that there is no need for us to pick on someone who's 
willing to sit down and re-write something in a potentially better* way. 
We tend to shoot ideas down with "show us the patches"; I don't think we 
should also shoot them down with "your patches will suck even if you do 
write them".

Let's wait with the judgment until we see the results. If the 
reimplementation works as well or better than what we have, then it's 
something that we can gain from. If it doesn't, well, we haven't lost 
anything, and the reimplementer will be all the wiser for having made 
the attempt.

I'm sure most of us (who code) have reinvented a few wheels each in our 
days, even if some were square ones, and done only for the 
joy/amusement/curiosity of seeing if it could be done any better. 
Personally, I see that as a driving factor in writing good software. If 
we always just settled for something that just works on the surface and 
is "good enough" for the common case, we wouldn't have very reliable 
systems, imho.

I realize there's a history between Matt and FreeBSD, but let's not pick 
on each other unnecessarily. Let's just work towards making each of the 
BSDs the best it can be. :)

Cheers,
/Johny, acting mediator (or is that meddler?)

*) Better being such things as improved code clarity, efficiency, 
accuracy, etc.

-- 
Johny Mattsson - System Designer ,-.   ,-.   ,-.  There is no truth.
http://www.earthmagic.org     _.'  `-'   `-'  There is only perception.