"Stephen J. Turnbull" <turnbull(a)sk.tsukuba.ac.jp&gt; writes: jv> I think that change is correct. Well, be my guest. ;-) As far as I can see, in my environment the two functions should exhibit identical behavior, and they don't. So I expect binary file insertions will fail for me again in the future :-( jv> That call to insert-file-contents-internal really should also be jv> insert-file-contents-literally, BUT i-f-c-l is broken now (it also jv> disables efs which it shouldn't). Which one? The one in `package-get-maybe-save-index'? It isn't; it's simply `insert-file-contents' (at least in 21.1.6). The reason is apparently that i-f-c-i needs the file-handler-alist temporarily bound to nil, which is what disables efs as I understand it. -- University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN Institute of Policy and Planning Sciences Tel/fax: +81 (298) 53-5091 __________________________________________________________________________ __________________________________________________________________________ What are those two straight lines for? "Free software rules."