On Mon, Feb 23, 2009 at 4:23 PM, Brian Harvey <[email protected]> wrote: >> Now, how much would it bother you if you had to ask for case- >> insensitivity in your programs? > > Not at all, provided I can ask for it in a REPL startup file and have > it apply to the REPL itself and to any file I (load ...) from the REPL > (I guess not including libraries that I load implicitly). > > We /are/ going to be able to load startup files, right? That's what's > always kept me from using DrScheme. (And yes I know about teachpacks.)
During the comment period, I suggested that there be a standard way to switch case-fold on and off, and that the default be implementation dependent. This still makes sense to me as users of a particular implementation will want that implementation to retain its default behavior, and writers of portable programs will just add the boilerplate stating their preference. But my suggestion was soundly ignored (whatever that means). GJS subsequently suggested to me that case-folding should be just one of the standard attributes appearing at the beginning of a file, character encoding being another. (Does R6RS provide a means for specifying encoding? I haven't checked.) I tend to agree. It would be nice to have a generic mechanism for declaring file-specific attributes that are necessary for interpreting the contents of the file; that would allow extensions such as mime type. I _don't_ particularly like the use of #!case-fold for this reason, as it's too dynamic. I don't see any reason why it's a good idea to change this in the middle of a file. I much prefer to the model of "attributes" for the entire file. _______________________________________________ r6rs-discuss mailing list [email protected] http://lists.r6rs.org/cgi-bin/mailman/listinfo/r6rs-discuss
