Arbitrary Identifiers

[BetterRed]

I've noticed that these are written to the content OPF file thus:

...
<dc:identifier opf:scheme="LOWYPUB">islamic-state-propaganda-and-mainstream-media</dc:identifier>
...

Wondering whether they should be, given the only place LOWYPUB etc are defined is on this computer.

They haven't caused me any problems, but I only read EPUBs on Windows with calibre and sigil, and on Android with Moon+ (via ES File Explorer). I don't make any use of the metadata outside of calibre. FWIW: I strip the non-DC elements, and covers, out of the EPUBs with Modify.

BR

[eschwartz]

Why would they cause problems anymore than Amazon/GP/Kobo/B&N/Apple identifiers?

Which are just as meaningless depending on the reader application, yet certainly have use and aren't likely to get cut.

[BetterRed]

Originally Posted by eschwartz

Why would they cause problems anymore than Amazon/GP/Kobo/B&N/Apple identifiers?

Dunno - was just wondering and applying the precautionary principle.

Assuming the AMAZON/KOBO... APPLE schemes are 'well known' then maybe some crappy reader software gives a clickable link for them, but barfs on unknown schemes; Arbitrary Identifiers are by definition an unknown scheme.

It's in my nature to 'abhor' meaningless data. It's why I use Modify; and sometimes the Sigil MD editor to remove dc:description, so I'll use it to remove my Arbitrary dc:identifiers. Then I at least, can stop wondering whether there might not be some gremlin waiting in the wings. . .

BR