[mcclim-devel] Scigraph's DWIM layer

[Timothy Moore]

Clemens Fruhwirth wrote:
> I'm interested in using Scigraph for generating histograms, but atm
> scigraph isn't in a good shape. The compatiblity layer for DWIM
> generates about 5-8 compile warnings (depends whether one is compiling
> to a fresh image or an image already containing dwim).
> 
> I would like to ask what's the future of the compatiblity layer. We have
> a lot read-time conditionals for genera or clim 0.9/1 in the dwim
> files. I'm pondering to remove these conditionals, as mcclim isn't
> clim-0.9 nor clim-1. Also there are some hacks for non-ansi CL
> environments. I'm temped to remove them too as I would have to wrap
> those things in #-ansi-cl. But that's not really useful anyway, as
> mcclim is ansi-cl.
What, and get rid of #FEATURE-CASE? :)
> 
> So what about that: DWIM as contained in the mcclim repository is a
> compatiblity layer for running dynamic window code on mcclim (and only
> on mcclim) in an ANSI CL environment. 
> 
> With a definition as above I could kick out ~25% of the DWIM
> code. Opinions? 
Although I found the Dynamic Windows code in DWIM interesting from a 
historical perspective, I think it's fine to strip it down to code that 
only supports McCLIM. I believe there are a couple of functions that are 
disabled in McCLIM; it would be worthwhile to investigate those and fix 
the problems in McCLIM.

Tim