Re: Summary and few tips

On Sun, 23 Jan 2005 09:11:39 +0000, Casey Allen Shobe <[email protected]> wrote:
> On Friday 21. January 2005 07:10, Fred P. wrote:
> [a lot about not coding properly, advocating not keeping one's long
> term goals in mind and writing a big dirty hack that "can be fixed
> later"]
> 
> Well, then we'd have another XFree86, and I don't think anybody wants
> that.
> 

So is the linux kernel. XFree86 was mismanaged. Linux kernel is not.
Did Linus ever think about the grand design of the unified driver
model when he started out? Linus did not even design Linux to become a
general purpose kernel. He started with a terminal emulator.

Are you familiar with the HURD? I can see the parallels here.

I once tried starting a project that did not go anywhere because I
wanted to get the initial design and implemenation to be as bug free
as as possible. I ended up building smaller projects so I can learn
enough to build the grand design. Now 5 years passed and all I have
are the small bits of my grand design. I was even scared to refactor
the pieces into the initial implementation because I can not verify
the correctness of the code.

Its better to start with the most simple and obvious design. Thats the
baseline. Then you test and refactor and so on. If we all wanted to
have enough experience coding so we get the main effort right the
first time, its going to take decades before we see any progress.

How many experienced coders are needed to have regular releases of Y?
Right now there seems to be 4 part time coders. Do we want a 36 month
cycle? What chance does this project have if modular xorg becomes
stable? I think we would all be so busy playing with our window
managers so that Y 1.0 will be pushed back another 10 years. Or never
released at all.

-- 
Blood is thicker than water... and much tastier
                                           John Davidorff Pell