[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Petname System

This page is part of the web mail archives of SRFI 84 from before July 7th, 2015. The new archives for SRFI 84 contain all messages, not just those from before July 7th, 2015.



Andrew Wilcox wrote:

My thanks to Tom Lord for his insightful comments and to Tony
Garnock-Jones for posting the references to the Petname System and
Zooko's Triangle.

 http://www.skyhunter.com/marcs/petnames/IntroPetNames.html
 http://www.zooko.com/distnames.html

It took me a while to read and absorb the material.  Now that I have,
I find the arguments of Petname system quite compelling and a great
improvement over my current proposal.

Interesting ideas. Definitely worth exploring.

However, the problem with scheme modules is that a standard one does not exist at all.

Can't we just get some sort of simple working module system in place and leave the notion of secure bindings to version 2?

Look at Java. It has a simple library system that doesn't obsess with library authentication and people can use it to distribute code.

Why can't scheme have something simple that just works?

Just get it out the door. Complicating things unnecessarily will prevent something usable (again).

Regarding petnames in particular, if "raw" (and unreadable) keys have to be in source code then things won't work, unless that becomes simply the distribution form only. So one can imagine that conversion tools can be made to convert to the more readable pet names.

But if that is just name conversion, then why can't the sources be distributed with the author's exact version (much more robust!), and an extra library "digest" be added to provide the binding from "author" names to library keys to the local user's petname? That way, source code is always directly readable.

Cheers,
Ray Blaak