On 2020/02/04 21:15:44, dak wrote: > NLGTM > > This patch just punts on using an internal function, and that could be equally > well be done without involving syntax-case. > > An alternative proposal that just relies on a single external symbol in order to > achieve the original design is given as > > Tracker issue: 5735 (https://sourceforge.net/p/testlilyissues/issues/5735/) > Rietveld issue: 549510043 (https://codereview.appspot.com/549510043) > Issue description: > Rewrite define-session and define-session-public macros The byte > compiler of Guile-2.x is not able to compile anonymous > functions/closures into constants. Rewriting the macros not to rely > on such constants by moving the local functions into their own > function definitions is the easiest expedient. > > https://codereview.appspot.com/553480044/diff/557280043/scm/lily.scm > File scm/lily.scm (right): > > https://codereview.appspot.com/553480044/diff/557280043/scm/lily.scm#newcode135 > scm/lily.scm:135: (acons (quote name) (make-session-variable (quote name) value) > lilypond-exports)) > This is actually a purely textual replacement rather than anything more complex > or hygienic. As such it does nothing that cannot equally well be achieved using > defmacro: the problematic issue with the byte compiler was the internal function > that was used for _avoiding_ a wholesale textual replacement. > > So drawing in a module that is known to be buggy and unmaintained in Guile-1.8 > does not even serve a purpose here.
closing in favor of https://codereview.appspot.com/549510043/ - you have my permission to skip countdown :-) https://codereview.appspot.com/553480044/
