On Dec 30, 2005, at 8:03 AM, tin gherdanarra wrote: > I'm somewhat incontent with the hyperspec. It lacks examples > and the wording reminds me to that of CCCITT-lawyers. I guess > the traffic on comp.lang.lisp would drop by 50 % if there was > a better hyperspec. So here is an idea. > > I could robotically leech the hyperspec from lisp.org and > convert it to some sort of XML. Another volunteer sets up a > wikipedia-like wiki (using the wikipedia software, if that's > possible) and runs an XSLT over my XML if he does not > like it. As soon as that thing is online people can start annotating > the hyperspec, clarifying sentences, providing examples and > gotchas, crosslinking, etc. In other words, the current > hyperspec is just a starting point for a better, shinier > reference.
Do you have examples of the kinds of things that need clarification? The Hyperspec is hard to understand because it's a standard--it's *supposed* to be written in a fairly legalistic fashion. Just be glad they decided not to try and produce a "formal" specification in denotational semantics or something. But I--as someone who's spent a *lot* of time poring over the Hyperspec--am hard-pressed to think of ways to change or annotate it that would make it much better without radically changing what it is. > BUT: There are legal issues involved. I'm pretty sure lisp.org > does not like robots eating their bandwidth. What's more, > there are copyright issues involved because every mirror > of the hyperspec claims a different copyright. It's confusing; > the only thing that is obvious is that the hyperspec is not > in the public domain or open commons. > > Anybody in the know here? Trying to produce a "new" hyperspec is fraught with all kinds of trouble, legal and otherwise. Not least of which is, for better or worse, the text of the Hyperspec is the text of the ANSI standard. (Though the examples are non-normative.) Producing a document that contains largely the same text as the ANSI standard but with some changes just seems likely to muddy the waters. And if you're going to produce a lot of new text it's not clear you need to start from the Hyperspec; you could just write your own tutorials. At any rate, your problem with the Hyperspec seem to be one that many others have had in the past: namely, it's not a tutorial. Which is fine since it's not *supposed* to be a tutorial, just a more accessible version of the language standard. If the problem you're trying to solve is that Lisp newbies are directed to the Hyperspec simply because there aren't good alternatives, then it seems the only solution is to produce the alternatives. Unfortunately producing a comprehensive language reference or tutorial is a *lot* of work. More power to any who wants to try (and it *is* an excellent way to get a deep appreciation of the language) but it's not clear that it's the best gardeners-sized project. I'd much rather folks look for ways to improve existing resources such as the CL-Cookbook or maybe producing a stand alone ANSI standard errata. (Though I suspect the latter is not actually that interesting--there are a few known "bugs" in the standard such as (non-normative) examples that don't work right and a few examples of cut-n-paste editing gone awry (c.f. PROG2) but not that much stuff that's actually of interest to anyone but the most pedantic language lawyer types.) Anyway, I'd suggest that you think about what problem you're really trying to solve and create a project page of the ALU Wiki explaining what the problem is and how you think a few gardeners could help solve it and see if there's anyone who wants to help you out. (If you're real problem with the Hyperspec is a lack of examples, maybe you can produce a set of examples organized according to the structure of the Hyperspec. I'm not sure I buy that that'd be that useful but maybe I just lack imagination.) -Peter -- Peter Seibel * [EMAIL PROTECTED] Gigamonkeys Consulting * http://www.gigamonkeys.com/ Practical Common Lisp * http://www.gigamonkeys.com/book/ _______________________________________________ Gardeners mailing list [email protected] http://www.lispniks.com/mailman/listinfo/gardeners
