Arc Forumnew | comments | leaders | submitlogin
3 points by akkartik 4672 days ago | link | parent

This was a lot of fun. Among many other things, I found out that Ron Garret (lisper) wants to get Joe Sixpack to host his own webservices -- his Google, his Facebook, etc. -- so that he controls his data and so on. It was mindblowing to realize someone else (and someone respected) had this crazy idea as well and wasn't rejecting it out of hand.

In response, waterhouse pointed me at http://unqualified-reservations.blogspot.com/2011/10/persona..., which is most entertaining.

---

Another highlight for me: idoh shared why he prefers arc to racket:

1. Less constrained defmacro macros that do what you mean. This is made possible by ignoring worries about hygiene.

2. A webserver that transparently reflects changes at the repl. This is made possible because mutable state isn't thread-private by default.

It was a welcome reminder of the fundamentals.

(Continuing my rant from http://arclanguage.org/item?id=15587)

All this effort goes into engineering constraints that at least a significant constituent of users (those new to programming, or those just looking to get a facebook app out asap) are actively hindered by. I think the pressure to over-engineer/industrial-strength-ify comes from two concerns:

a) those who want to make the code go faster, and

b) those who are afraid of what nameless other programmers might do and how to avoid stepping on each other's toes. (See, for example, http://arclanguage.org/item?id=8495)

I have a lot of sympathy for a) after my recent experiences, and I think we should try harder to work on performance without compromising Core Lisp Principles (tm).

To those concerned about b) I say, "go use Java" :p When I program in lisp I expect to have just a handful of collaborators, and I expect to know the antecedents of any libraries and modules I use. And to have lots of unit tests for maximum paranoia. Trust, but verify.