Re: [Sbcl-help] Logical pathnames as source locations for, user defined functions

[Robert Goldman]

On 27 Jun 2023, at 8:19, sbcl-hel...@lists.sourceforge.net wrote:

But honestly, even though I do occasionally use logical pathnames

myself, I think of that usage as a personal preference that I really

can't recommend. Like a strong-smelling cheese or something; nobody

strictly needs the stuff, some people like it, some people hate it, and

it may not be good for you to use every day. :-)

ASDF has put me firmly in the some people hate it camp!

But this is because LPNs are not sufficiently well standardized to give reliable results across implementations. If you want your code to be portable, I recommend staying away. If you are just working on SBCL, then this isn't such a problem.

Another thing that I dislike is the rule against underscores and mixed-case. One day I had to put together a system that contained files named by someone who was primarily a C (or Java?) programmer and the results were not pretty.

If you can get away with using asdf:system-relative-pathname that might be the easiest way to go. I'm not sure what happens with your build-and-relocate case, though.

Now that most of the non-POSIX filesystems are dead, and ASDF figured out how to exploit *load-truename* to find source files without (much) explicit configuration, LPNs have a lot less reason to be than they did when the ANSI spec was written.