3 ms·
I have read "Lisp 1.5 Programmers Manual”, “The Mythical Man-Month” and “The Meta-Object Protocol” by Kiczales. These all are definitely timeless classics. Th
by nabla9 7y ago
I have read "Lisp 1.5 Programmers Manual”, “The Mythical Man-Month” and “The Meta-Object Protocol” by Kiczales. These all are definitely timeless classics.
These quotes about Gregor Kiczales and AspectJ https://en.wikipedia.org/wiki/AspectJ https://en.wikipedia.org/wiki/AspectJ are good intro to MOP:
"In Lisp, if you want to do aspect-oriented programming, you just do a
bunch of macros and you're there. In Java, you have to get Gregor
Kiczales to go out and start a new company, taking months and years
and try to get that to work. Lisp still has the advantage there, it's
just a question of people wanting that." -- Peter Norvig
"I am reminded of Gregor Kiczales at ILC 2003 displaying some AspectJ to a silent crowd, pausing, then plaintively adding, "When I show that to Java programmers they stand up and cheer." -- Kenny Tilton
- jimbokun 7y agoAgreed, and I also think "annotation oriented programming" has become the bane of Java programmers everywhere. Seems simple when you first slap an annotation onto a class or method. But God help you when it stops working and you actually have to debug what it's doing. Figuring out what code is actually executing and what it's doing seems nigh impossible. (Which is all to say, Lisp is still far superior, because there is a straightforward process for figuring out what kind of code a macro will generate, or to run a macro and look at its output. Macros can become complex and convoluted, but that's still nothing to the mess that annotations in modern Java frameworks create.)
- therealdrag0 7y agoYou can debug an aspect too eh? Throw a breakpoint before a method call, step into it, "oh I'm in a proxy now" step step "oh I'm in this forgotten point cut I made 6 months ago, okay now I know." I have no exp with Lisp so not arguing just observing the one side of exp I have.
- jimbokun 7y agoBut you inadvertently reinforce the point I am trying to make. There is no way to reason about the code without actually running it. You can not read the code and understand what it's going to do.