duration of the ports freeze

Ade Lovett ade at FreeBSD.org
Mon Dec 3 02:20:31 PST 2007


On Dec 03, 2007, at 01:18 , Aryeh M. Friedman wrote:

I'm probably going to regret joining this thread, but quite frankly,  
the amount of horse being thrown around has gotten way out of hand.

> I have practical knowledge here in working with different dependency
> management systems (which is essentially what the ports system is)...
> now that being said much of it is based on seeing how different people
> have solved the problems raised by the methods you site as being  
> "invalid"

Fine.  Prove your practical knowledge.  Take a small subset of the  
ports tree, x11/xorg springs to mind, bang around a few ideas, show  
why they're "better", and we'll take it from there.

Interminable threads about "why <X> sucks", "why <Y> is better", "why  
don't you do it with <Z>" gets nowhere.  Even a cursory glance at the  
archives of this mailing list would show that.

> That is the purpose of the survey I sent out to start to uncover what
> exactly the strengths and weaknesses are and decide if the weaknesses
> are sufficent to warrent any kind of re-engineering.

Unfortunately, as with most surveys, it suffers from a fundamental  
flaw in that it is self-selecting.  It has gone out to that tiny  
subset of folks that:

	(a) use FreeBSD
	(b) use FreeBSD ports/packages
	(c) subscribe to ports at FreeBSD.org
	(d) feel like filling out a survey

By my reckoning, just those 4 points have deselected at *least* 99% of  
folks that would potentially benefit from *any* kind of reworking.

Hopefully, y'all will take this as constructive criticism from someone  
that (a) actually really does give a damn about FreeBSD/ports and (b)  
spends a lot of time, in conjunction with others, doing heavy-lifting  
infrastructural changes that aren't eye-candy.

Around here, action, and not words, are taken much more seriously.   
We've heard the words before (albeit dressed and dolled up in a myriad  
of different way).

Y'all are *not* going to get a (potential) rewrite of the ports system  
right first time.  18k+ ports, 4 different OS versions, at least 5  
"useful" architectures.  You do the math.

Set up a wiki somewhere.  Announce it to the community at large (hint,  
that means more than sending mails to @FreeBSD.org mailing lists).   
Put up some proposals (at this point, it really doesn't matter how  
hair-brained they might be).  Let folks contribute in terms of  
editing.  See what comes out of it.

Highly restrictive mailing lists are *not* the right medium here.

-aDe



More information about the freebsd-ports mailing list