Re: G-Wrap 1.9.9 released
[email protected] (Ludovic Courtès)
| Newsgroups | gmane.comp.programming.g-wrap.devel,gmane.lisp.guile.gtk |
|---|---|
| Message-ID | <[email protected]> |
Hi, schemer <[email protected]> writes: > I just noticed that you have added the following code in file > compat.scm in version 1.9.9. > > ;; Hack to make WCTs nullable (null-ok) by default, as in G-Wrap 1.3.4 > (define-method (make-typespec (type <gw-wct>) (options <list>)) > (next-method type (if (memq 'non-null options) > (cons 'caller-owned (delq 'non-null options)) > (append '(null-ok caller-owned) options)))) > > It seems the module (g-wrap compat) gets loaded irrespective of > whether you want to have version 1.3.4 compatibility or not. As a > result it becomes impossinble to define non-nullable wct's. I just run "bzr pull" and noticed that Rotty has already fixed it. :-) The `NEWS' reads this: - Fix WCT nullable behaviour for 'new' (i.e. non-compat) wrapsets; there is now a SRFI-39 parameter 'gw-wcts-nullable?' controlling the nullable behaviour; this parameter is #f by default, and set to #t when generating compat wrapsets. However, wouldn't it make sense to remove the automatic dependency on the compatibility layer instead? "bzr blame" on `g-wrap.scm' seems to imply that the `module-use!' clause at the bottom has always been here, in which case removing it now would probably make people unhappy. Thanks, Ludovic.