[rfk-dev] code status

Peter A. H. Peterson pedro at tastytronic.net
Fri Oct 14 10:35:08 CDT 2005


Quoting Leonard Richardson:
> The reason I wanted some NKI in the binary was so if you just found a
> 'robotfindskitten' binary lying around somewhere and ran it, *something*
> would happen. 

Yeah, agreed.

> If you actually passed in some NKI on standard input, or
> used the environment variable or /usr/share/robotfindskitten/ or
> ~/.robotfindskitten/ then those NKI would replace the ones in the binary,
> not add to them.

I see! Brilliant.

> I would be more or less satisfied if the hard-coded binary NKI were just
> descriptions of what the NKI file format was and all the places you could
> put it.
> 
> "Put text strings in a file, one per line."
> "Put the file in ~/.robotfindskitten/nki."
> "Install the file into "/usr/share/robotfindskitten."
> "Pass the file as standard input when invoking this program."
> etc.

Either way -- that could be funny and non-bloaty. But not as
"robotfindskitten Classic (original formula)"-ey. But as long as
existing NKI in files replace the hardcoded ones, I'm happy.

Peter

-- 
Peter A. H. Peterson, technician and musician.
 ---=[ http://tastytronic.net/~pedro/ ]=--- 


More information about the rfk-dev mailing list