[developers] Bug fix not merged with LKB code

Stephan Oepen oe at csli.Stanford.EDU
Fri Mar 24 14:33:21 CET 2006

hi bernd,

> i just re-identified a bug in LKB that i already fixed two years ago.

i'm sorry to hear your earlier patch got lost!  i must admit i have no
memory of that earlier message.

> Maybe somebody can verify that this change is OK in the LKB environment
> too and can propagate it into the CVS. Thanks to anyone who'll do it.

your proposed fix would actually break in the LKB context, because the
value of *mrs-package* _is_ :lkb there.  it is only when using the MRS
code that *mrs-package* is :mrs.

fortunately, the MRS code already provides:

  (defun vsym (str) 
    (intern (string-upcase str) *mrs-package*))

i.e. the intention is that all MRS-specific symbols be created through
vsym().  thus, the general fix for the package problem should be using
vsym() instead of intern(), i believe.

ann, can you (a) confirm the above is correct and (b) apply the fix to
make-mrs-feature()?  i see there are a few more usages of intern() in
the MRS code.  it would probably be good to look at those too then.

                                                      all best  -  oe

+++ Universitetet i Oslo (IFI); Boks 1080 Blindern; 0316 Oslo; (+47) 2285 7989
+++     CSLI Stanford; Ventura Hall; Stanford, CA 94305; (+1 650) 723 0515
+++       --- oe at csli.stanford.edu; oe at ifi.uio.no; stephan at oepen.net ---

More information about the developers mailing list