- From: Keld Jørn Simonsen <keld@dkuug.dk>
- Date: Tue, 04 Apr 2000 13:28:45 +0200
- To: Tex Texin <texin@progress.com>
- Cc: Erik van der Poel <erik@netscape.com>, Misha Wolf <misha.wolf@reuters.com>, ietf-charsets@innosoft.com
On Mon, Apr 03, 2000 at 09:14:22PM -0400, Tex Texin wrote: > Keld, > I don't see any reason to make the problem worse, just because > past practices were bad. It is not just a problem that it increases > the size of tables, it is a problem for people diagnosing problems > to run into an infrequently used alias to have to go off and figure out > what this unusual code page represents. > > What is the advantage of having more than one name for it? It is not a problem, it is a feature:-) The case is that mime names are used many places also beyound the Internet realm. The alisas I have listed is to specify that the names used there are realy the same as this IANA registration, so that if X is talking about ISO_8859-14 then it is the same as iso-8859-14 and also what UNIX people use for a charmap under the name of iso8859-14. The ISO character set prople know it by its registration number in the 2375 registry so that is the iso-ir name. The idea with the aliasing was to record all the information on the charset in one place, including all the names it is customary known as. This facilitatets interoperation between different applications on a system, including internet aware applications. That was the original plan when I wrote rfc 1345 and I think in the greater perspective this is still valid. I don't think this complicates life, I just have my computer organize this for me, and whether there are 100 charset names or 500 does not matter. And it helps users avoid interoperability problems. Keld
Received on Tuesday, 4 April 2000 10:02:15 UTC