Mark Evenson

Hmm. The ABCL representations of float NaNs are deliberately non-readable objects as they get processed by the condition handlers present in a running program when they are encountered. I note...

@brown if you are looking to represent such a value aeth (via #lisp) points to . Otherwise I am inclined at this point to patch the ABCL compiler to emit...

I'm glad you found a work-around that doesn't require a new ABCL for you. I'm still trying to figure out the best way to patch the implementation to bring it...

> @brown if you are looking to represent such a value aeth (via #lisp) points to https://github.com/Shinmera/float-features/blob/master/float-features.lisp. As corrected on #lisp, it was noted that FLOAT-FEATURES doesn't currently support cross-implementation...

There should be a mechanism, but I don't understand where at the moment. I will try to patch things and certainly document for .

Patch has been applied as https://github.com/armedbear/abcl/commit/73341974907ce3f8e4e9626af87f5ef5e470e255 @alanruttenberg Please test and close?

http://abcl.org/trac/ticket/433

c.f.

As far as I remember, the "intendedClass" mechanism memoizes some successful result of runtime "Do what I mean" (DWIM) algorithms. getCannonicalName() should return the current memoized value. Someone should introspect...

I believe you don't have all the dependencies correctly assembled in "shared-things-1.0.0-jar-with-dependencies.jar". When I build the jar from , and try to instantiate the ``` CL-USER> (java:add-to-classpath "/home/mevenson/work/hl7v2-fhir-converter/build/libs/hl7v2-fhir-converter-1.0.1-SNAPSHOT.jar T CL-USER>...