[RFC] The Future of the ALFS project

George Makrydakis gmakmail at gmail.com
Wed Mar 1 15:44:23 PST 2006

Jeremy Huntwork wrote:
> Hold up there, bud. :) If you start coding anything in C++, I would very 
> much like to be a part of it. If you want, you can even forward me 
> patches and I'll add them into the ALFS repository, at least until such 
> time as Thomas sees fit to give you access.
> I had already been thinking of trying to convert the C code I already 
> have into C++ because of its ease of manipulating strings. Can we do 
> that or would it be advisable to start from scratch? In any case, as I 
> hope is obvious from my work so far, I'm deeply interested in alfs and 
> I'd very much like to be a part of any coding done.
> Are you willing to work with me on this?
> -- 
> JH

Yes, time factor allowing. I would not like to move your attention from what you are doing right now though, for it would be extremely counter - productive. For 
a series of reasons I think that C++ is the right way to go (C tends to be an extremely "write - only" language), especially if you want a clean start (it is a 
*new* alfs after all, getting everything out of the book without human interaction). There is a series of design issues here, and most of all a fair amount of 
_mandatory_ patience to put all of this together public preview - wise, avoiding premature constructive criticism. The final goal is to have the end product 
fully self hosted. But the first thing to do is to think exactly what we need (imho): a small functional core for parsing the book, like your parser.c attempt, 
like my bash based work so far. I would like to experiment with the "parsing" methods that could be applied to the book(s), since this is the main issue, not 
just use a ready - made parser for this (various reasons for this last one to be a no - go). If possible, some live discussion about this would be appreciated. 
My own view of the alfs project is not that of a "tool", but rather a of a "building framework", where every "package" is considered an "object" of its own right...

Thanks again,

George Makrydakis


More information about the alfs-discuss mailing list