emacs-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: display word wrapping


From: Jason Rumney
Subject: Re: display word wrapping
Date: Fri, 28 May 2004 08:57:43 +0100
User-agent: Gnus/5.110002 (No Gnus v0.2) Emacs/21.3.50 (windows-nt)

Juanma Barranquero <address@hidden> writes:

> On 27 May 2004 12:41:49 +0200
> address@hidden (Kim F. Storm) wrote:
>
>> It certainly sounds messy, and quite tricky to do right for
>> making a binary distribution for W32!
>
> Perhaps I've made it sound worse that it is.  I think Jason had it
> working fine on his setup (with gcc, IIRC) last time I exchanged e-mail
> with him about the issue (though that was more than a year ago).

Image support has always worked for me with both gcc and msvc
(non-optimising version though). I recall we made some changes to
detect broken versions of some libraries and refuse to use them, but
was not aware (had forgotten?) that the problems were still there.

I think the problems with MSVC's optimisations is likely due to
alignment differences in structs we are passing into the library. If
the problem is just with MSVC, we can just add a note to PROBLEMS,
since the official binaries will be built with gcc anyway.

> That's why I would advocate suppling our own (not necessarily
> compiled by us), either on the binary tarball, or as a .zip pointed
> from somewhere in http://www.gnu.org/software/emacs/windows/ntemacs.html.

Remember anything we distribute in binary form, we also need to
distribute source for. Rather than clutter the gnu.org servers with
non-GNU image library code just because some Windows users might want
to recompile their Emacs with a proprietary compiler, I think we can
just say that such people should recompile the image libraries
themselves too. That is the easiest way to ensure that the image
libraries will work.


There may also be problems with specific versions of image libraries
that affect a gcc build of Emacs as well. These cases we should try
to find out why and test for it when loading the libraries.





reply via email to

[Prev in Thread] Current Thread [Next in Thread]