12.29.04
More gratuitous linking to stuff other people wrote
-
Chris Double has a post about
debugging Sisc Scheme web applications hosted in Jetty. I lust after
that interactivity, especially the part about not taking down the webapp for
every little change. -
Chris Double weighs in with the
size of serialized continuation state: 5KB to 100KB. In our system
(not continuation based) I’ve seen it run over a megabyte, but that tends to
only be when we are saving query results on the users session… something we
generally try to avoid. Normally we see it in the same range that he does. -
Bill Clementson has
apparently been writing about
Continuations for complex web applications (or maybe it is
Synchronous Web Programming) for a while. I’d seen one or two of
these, but it looks like I know have a lot of good reading. As complex web apps
are what I am mostly programming, how to make the environment more powerful and
easier to develop for is a big issue. -
Bill Clementson also wrote about Domain
Specific Languages and
Metaprogramming
recently. At work we have recently been talking about using more dynamic
languages, probably Python, to generate a lot of code for the static
languages we work with, hopefully we can just remove some of that staticness
altogether. -
It looks like they put the entire Lightweight
Languages 2004
conference online that should be some viewing goodness. The talk about the
love-child of Rest and Continuations looks especially interesting– I might be
paraphrasing a bit. -
Successful Lisp is
available in print, but more importantly it is available
online! Near the end of chapter 4
(the “Simple Macros” section) he gives an excellent example of how common lisp
macros are really nice. This is one of the better introductory examples I’ve
seen. It doesn’t take very long to understand, and it does offer positive
improvement over what can be done with functions alone. -
Elegant Lisp
Programs: From the abstract it looks like we’re in for a rollicking good
time:
Call a program “elegant” if no smaller program has the same
output. … For any computational task there is at least one elegant program,
perhaps more. Nevertheless, we present a Berry paradox proof that it is
impossible to prove that any particular large program is elegant. … This
establishes an extremely concrete and fundamental limitation on the power of
formal mathematical reasoning.It is always fun to see just how
broken everything really is.