Scripting Languages - purpose, roles, and metrics?

I've been gathering some source material on the "philosophical"
foundations of scripting languages - primarily the real and acquired
roles they have and where they seem to be going, and how well they
seem to work for certain purposes - and I am hoping someone here can
point me to good resources for this. I am not after specific details
on language X, just material that tries to analyze the role of
scripting languages in computing.


The reason I am asking here is that the vast majority of theoretical
scripting material I find tends to be shallow discussions, often
biased (even unconsciously) by advocates of a particular language or
paradigm. In the past, it looks like comp.compilers posters seem to
be people who both have solid understandings about the foundations of
everything in computing, and understand the reality of different
languages and coding environments having different uses.


To give you a feel for the kinds of things I am after, here are some
representative examples of key material I have found:


(1) Richard Stallman's "Why You Should Not Use Tcl" post (23 Sep 94,
visible root of the [in]famous "First Tcl War"]. Various sources,
including http://www.vanderburg.org/Tcl/war/0000.html.


Setting aside details specific to the immediate language and situation
surrounding this post, Stallman raised several important points about
what scripting languages mean in a larger context. What I find most
intriguing is that in the scripting languages which I use, every point
he raised has been an important problem at some point.


(2) Ousterhout's "Scripting: Higher Level Programming for the 21st
Century" (IEEE Computer, March 1998)
http://home.pacbell.net/ouster/scripting.html


Views on the future direction of scripting languages.


(3) Kernighan and Van Wyk's paper, "Timing Trials, or, the Trials of Timing:
Experiments with Scripting and User-Interface Languages".
http://cm.bell-labs.com/cm/cs/who/bwk/interps/pap.html


This dealt with one particular aspect of scripting languages (and not
always an important one) - the issues of measuring performance. Had
some valuable points about the difficulty of getting comparative data,
as well as some interesting generic algorithms to use for abstract
performance measurements.
--


[It's an interesting question. There's a bunch of internal tensions
in any scripting language. On the one hand, you want a really simple
syntax and minimal punctuation and no declarations so that tiny one
line scripts are in fact one line scripts. On the other hand, in any
successful scripting language, people will build programs large enough
that you want scoping, declarations, and the other baggage that comes
with a real programming language. I've never seen a really clear
discussion of the tradeoffs, much less a satisfactory answer. -John]

Return to the comp.compilers page.
Search the comp.compilers archives again.