On Tue, 14 Jun 2016 22:21:00 +0200, Edward Bartolo wrote:
> Irrwahn wrote:
> << apparently simply content with
> writing bad code >>
> 
> Bad code is not good for anyone let alone as an init. What I am doing
> it experimenting so that I further understand which lines of the code
> play the central roles and which lines play auxiliary roles. 

Felker's minimal init lives up to its name. It is as simple 
as possible without actually being broken. There are no lines 
playing "auxiliary roles", each and every line is as important 
as any other. In fact, if one were anal about matters, it could 
even be argued it is already a little bit too simple, due to 
the omission of one header include directive.

"Everything should be made as simple as possible, but not simpler." 

To learn something from making "experiments" like omitting code 
from a previously working program, one'd first have to know what 
tasks the program is actually supposed to perform, a.k.a. the 
specification. Otherwise you are in high danger of tricking 
yourself into false beliefs à la: "Hm, the system just booted 
fine, therefore it must be okay to leave that part out. Good job, 
I've just made the already simple thing even simpler!" 

That's not called learning, that's called self-deception.

Regards
Urban

_______________________________________________
Dng mailing list
[email protected]
https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng

Reply via email to