home *** CD-ROM | disk | FTP | other *** search
- x-gateway: rodan.UU.NET from help-lucid-emacs to alt.lucid-emacs.help; Fri, 28 Aug 1992 09:38:58 EDT
- Date: Thu, 27 Aug 1992 13:00:44 PDT
- Message-ID: <kyle-9208272000.AA06497@thalidomide.lucid>
- X-Windows: It could be worse, but it'll take time.
- From: jwz@lucid.com (Jamie Zawinski)
- Sender: jwz%thalidomide@lucid.com
- Subject: Re: retained state from byte compile
- References: <LYSTAD.92Aug27120852@m2.dseg.ti.com>
- Newsgroups: alt.lucid-emacs.help
- Path: sparky!uunet!wendy-fate.uu.net!help-lucid-emacs
- Lines: 15
-
- In message <LYSTAD.92Aug27120852@m2.dseg.ti.com> Garr Lystad wrote:
- >
- > I wrote a little advise macro when I first started using Epoch.
- > One of my files advises an lemacs function (call it x) and the advice
- > calls a function I define in the same file (call this function of mine
- > y). If I do not load or evaluate any part of the file, but I do byte
- > compile it, the advice on x remains and I get an error because y is
- > undefined. That shouldn't really be, should it?
-
- It sounds to me like your adviser is a macro, and it is doing its work
- in the macroexpander function rather than in the emitted code. This is
- a no-no. It worked with the old byte compiler because it simply didn't
- compile top-level calls to macros.
-
- -- Jamie
-