[racket-dev] new package system collections and conflicts

From: Neil Van Dyke (neil at neilvandyke.org)
Date: Wed Dec 3 11:32:30 EST 2014

I don't think I need charity.

I thought the vision for the new package system had already been 
explained adequately.  I would be very interested to learn how the model 
is well-suited to third-party developers like me.

But -- I mean this constructively -- I'd be happy if someone simply came 
out and said "this model is great for core developers, we still have to 
figure out everyone else, and maybe the model isn't great for everyone 
else".  The reason is that I've looked at the new package system 
seriously 5-6 times since it was announced, and I keep being told that 
the model is intended for non-core people like me, and that someone else 
knows my needs better than me.  Open source reuse was an especial area 
of interest to me, the package system is very important, and I've given 
the benefit of the doubt 5-6 times now.  (This has actually stalled most 
of my public Racket work, one way or another, for about 2 years.)

I'm not harshing on Racket; just on how the new package system was 
sprung on non-core people, and the narrative.  It doesn't look typical 
of Racket.  Racket is usually in the position that it could say "we have 
a better idea" (on, e.g., module system sophistication, various syntax 
extension mechanisms and mixed languages support, various aspects of 
DrRacket, the related pedagogic projects, etc.), and usually that 
doesn't have to be said, because the superiority of Racket is 
immediately apparent.  That's why I've been a Racketeer for 13 years and 
counting.

Neil V.


Posted on the dev mailing list.