Hi Agustin, Agustin Martin <agmar...@debian.org> writes:
> On Mon, Mar 24, 2025 at 02:56:21PM -0700, Xiyue Deng wrote: >> Agustin Martin <agmar...@debian.org> writes: > ... >> > As long as no lexical binding is used, code should work the same in Emacs >> > and XEmacs. However, if lexical binding is ever used in this file, it will >> > not work as expected in XEmacs. I think setting it to nil should force the >> > same behavior in both Emacs and XEmacs. >> > >> >> Just trying to understand this better: does using "lexical-binding: t" >> already causing issue with the current XEmacs, or this is theoretical? >> If it's the former, how does it break XEmacs? I cannot seem to find any >> reference that XEmacs reacts to lexical-binding directives through >> searching. > > Hi, > > AFAIK current code does not use lexical binding, so currently there is no > difference in practice between setting lexical-binding to `t' or to `nil', > neither for Emacs nor for XEmacs. > > My concern is that if lexical binding is ever used in this file, behavior > might be different between Emacs and XEmacs (yes, this is hypothetical). > > Regarding current status of XEmacs, seems that there is some (very) slow > developent at > > https://foss.heptapod.net/xemacs/ > > but the real thing is that manpower behind Emacs and XEmacs are not > comparable. > > I read somewhere that people behind XEmacs planned some sort of lexical > binding implementation, but they did not like the Emacs way of declaring > it. So, the best way for current compatibility may be not to use lexical > binding when possible. > Thanks for providing the background. It's interesting to see that XEmacs development is ongoing, and it's sad to see that currently XEmacs in Debian is not in a good shape - it looks like it won't be part of Trixie. Hope someone will continue the work. OTOH, it's still unclear how things will become. As Emacs is moving towards using lexical binding by default, it would make more sense to go in this direction than to worry about a hypothetical incompatibility. Now that more and more Emacs Lisp source code is marking lexical binding as t, I guess XEmacs would not be going in the route that reject the directive. At best it will be a no-op, and hopefully it will provide some compatibility. Just my 2 cents. > Regards, > > -- > Agustin -- Regards, Xiyue Deng
signature.asc
Description: PGP signature
