1. 09 Nov, 2015 1 commit
    • mevenson@1c010e3e-69d0-11dd-93a8-456734b0d56f's avatar
      {decode,encode}-universal-time "time of the time" timezone semantics · d270f095
      Scott L. Burson
      
      The implementation of EXT:GET-TIME-ZONE provides the underlying
      mechanism.
      
      Both GET-TIME-ZONE and DEFAULT-TIME-ZONE have been elevated to
      supported status by virtue of promotion of their respective symbols to
      the EXTENSIONS and SYSTEM packages, respectively.
      
      ME: A mechanism for adding symbols to both SYS and EXT packages with
      DEFINE-SYMBOL-MACRO didn't work, but left as comment at the end of
      'time.lisp' for further work.
      d270f095
  2. 21 Oct, 2015 2 commits
    • mevenson@1c010e3e-69d0-11dd-93a8-456734b0d56f's avatar
      jss: ensure strings are used for comparison in class hashtable · e6137401
      Fixes a problem encountered in lookup for
      'abcl-asdf:AETHER.GRAPH.DEPENDENCY' which for some reason resulting in
      the class name asdf:AETHER.GRAPH.DEPENDENCY. However it was supposed
      to be just a class suffix.
      
      Thanks to Alan Ruttenberg.
      e6137401
    • mevenson@1c010e3e-69d0-11dd-93a8-456734b0d56f's avatar
      Upgrade to asdf-3.1.6 · 5ebfec09
      N.b. one will have to do a full clean of existing fasls for this
      change to compile.  Otherwise one gets errors about "wrong number of
      arguments for DEFSETF".
      
      N.b. the configuration directory under Windows has changed in
      incompatible ways.  We intend to "fix" this in a backwards compatible
      manner before releasing a version of ABCL based on this change but
      wish to get asdf-3.1.6 out into circulation as a first step.
      5ebfec09
  3. 18 Sep, 2015 4 commits
  4. 15 Sep, 2015 1 commit
  5. 13 Sep, 2015 2 commits
  6. 10 Sep, 2015 1 commit
  7. 09 Sep, 2015 2 commits
  8. 31 Aug, 2015 2 commits
  9. 01 Jul, 2015 1 commit
  10. 30 Jun, 2015 3 commits
  11. 12 Jun, 2015 1 commit
  12. 19 Apr, 2015 3 commits
  13. 14 Apr, 2015 7 commits
  14. 11 Apr, 2015 2 commits
  15. 09 Apr, 2015 1 commit
  16. 05 Apr, 2015 2 commits
  17. 01 Apr, 2015 1 commit
  18. 31 Mar, 2015 1 commit
  19. 12 Mar, 2015 1 commit
  20. 04 Mar, 2015 1 commit
  21. 26 Jan, 2015 1 commit
    • mevenson@1c010e3e-69d0-11dd-93a8-456734b0d56f's avatar
      Numeric tower repairs on promoting floats across representation boundaries. · 56638363
      Thanks to Massimiliano Ghilardi.
      
      If a bignum exceeds the largest representable single-float,
      EQUALP will raise a type-error when comparing it to a single-float.
      The same problem happens with double-floats.
      
      I am reporting this because my understanding of CLHS
      http://www.lispworks.com/documentation/lw51/CLHS/Body/f_equalp.htm#equalp
      tells that EQUALP is not supposed to raise errors at all:
      "Exceptional Situations: None."
      
      How to reproduce:
      
      ;; =========== double-floats ===========
      (= (ash 1 1024) most-positive-double-float)
      NIL
      
      (equalp (ash 1 1024) most-positive-double-float)
      ; Evaluation aborted on #<TYPE-ERROR {215F90FE}>.
      
      
      ;; =========== single-floats ===========
      (= 340282360000000000000000000000000000000 0.0f0)
      NIL
      
      (equalp 340282360000000000000000000000000000000 0.0f0)
      ; Evaluation aborted on #<TYPE-ERROR {1CFFC364}>.
      56638363