Sorry, for some reasons I did not receive messages besides Wed, 29 Jan
2025 18:15:55 +0100, so I was unaware of comments.
On Sun, 26 Jan 2025 10:40:04 +0900 Charles Plessy wrote:
I was not aware that some tools use the order of extensions listed in
/etc/mime.types as an indication of priority.
It is unavoidable. Applications have to use some tie-break rule to
resolve ambiguity. Actually in python built-in mapping (having higher
priority) has been updated and my example reports proper media type
without modification of /etc/mime.types.
I am not sure that this
meaning is defined or conveyed in the media type declarations published
by the IANA…
And mime.types format is even more limited. It does not allow to convey
that both text/javascript and text/ecmascript both may have .mjs suffix,
but the former with "Intended usage: COMMON" while the latter "Intended
usage: OBSOLETE". Treating of entries order is left up to particular
application and almost certainly a part would pick first one and others
last entry.
I think that
text/javascript js mjs es
is OK. A separate text/ecmascript entry might be even better, but I am
not sure.
--- mime.types.orig 2023-02-11 13:45:22.000000000 +0000
+++ mime.types 2025-02-01 11:44:56.420858567 +0000
@@ -2072,6 +2072,7 @@
text/csv csv
text/csv-schema csvs
text/dns soa zone
+text/ecmascript es
text/encaprtp
text/enriched
text/example
@@ -2082,7 +2083,7 @@
text/grammar-ref-list
text/hl7v2
text/html html htm shtml
-text/javascript es js mjs
+text/javascript js mjs
text/jcr-cnd cnd
text/markdown md markdown
text/mizar miz
