[elinks-users] Re: Elinks <local-file>

Jonas Fonseca fonseca at diku.dk
Thu Oct 16 07:32:31 PDT 2003


Jean-Yves Levesque <jyves at nortelnetworks.com> wrote Thu, Oct 16, 2003:
> * Petr Baudis <pasky at ucw.cz>:
> > Dear diary, on Wed, Oct 15, 2003 at 09:48:36PM CEST, I got a letter,
> > where Jean-Yves Levesque <jyves at nortelnetworks.com> told me, that...
> 
> 	[stuff deleted]
> >
> > That's strange, we should never call init_mimetypes_map() before
> > init_mimetypes()... right? It looks like init_mimetypes() wasn't called
> > although it should have been.
> >
> > It should be called through init_mime_backends() which should be called
> > init_mime(). Could you please check (try inserting debug("x") calls
> > there) if these functions are called or not? And, in turn, if
> > init_mimetypes_map() is called then?
> >
> 
> DEBUG MESSAGE at mime.c:24: init_mime.
> DEBUG MESSAGE at common.c:34: init_mime_backends.
> DEBUG MESSAGE at mimetypes.c:234: get_content_type_mimetypes.
> DEBUG MESSAGE at mimetypes.c:152: init_mimetypes_map.
> DEBUG MESSAGE at mimetypes.c:221: init_mimetypes.
> 
> For un unknown reason, get_content_type_mimetypes is called
> before init_mimetypes.
> 
> init_mimetypes is called from my modified init_mimetypes_map function.

Maybe your fix/workaround is not so bad. Make the mimetypes init
function be called always from init_mimetypes_map() making the
initialization even more ``lazy'' and startup faster. :)

-- 
Jonas Fonseca



More information about the elinks-users mailing list