This page is part of the web mail archives of SRFI 83 from before July 7th, 2015. The new archives for SRFI 83 contain all messages, not just those from before July 7th, 2015.
On 1/16/06, Jim Blandy <jimb@xxxxxxxxxxxx> wrote: > > In other words, it leaves it entirely implementation-defined how the > header names are interpreted, and where the referenced files are > stored. I think actually the concern is not so much the filesystem names, as the module names themselves. C at least includes standard libraries such as "stdlib.h" leading naturally to the practice of <short-unique-lib-name.h> The current draft gives us only one example, scheme://R6RS, and it's not immediately obvious how that's supposed to be extended. Do we use scheme://mylib or scheme://myhost/mylib or scheme://myhost/myuser/mycategory/mylib ? Do we include the extension in the path for non-default libraries? Should we use http://... for extensions available on the web? I think perhaps Java is a better and more modern example to take from with an explicit naming hierarchy. It says nothing about how the package names translate to file names, and nothing needs to be said, because once you establish the hierarchy semantics, people on OSes with filesystem hierarchies will just translate that in the natural way. -- Alex