From: Marko Mäkelä (marko.makela_at_hut.fi)
Date: 2004-11-18 15:35:13
On Thu, Nov 18, 2004 at 03:05:04PM +0100, Spiro Trikaliotis wrote: > > Producing metadata is at least as important as testing and sorting the > > contents of the archive. Maybe it would be a good idea to define the > > format of the metadata first. > > What about some XML scheme? *justkidding* Well, I was thinking of something simple, such as key=value pairs separated by two empty newlines. XML is not very readable, and tree-based XML editing tools may alter whitespaces around markers, which makes it pretty much impossible to efficiently use any version control system. (I think that the metadata should be kept in CVS or similar, so that changes can be tracked easily.) The only advantage of XML I can think of is that it defines the character set to be Unicode, encoded in UTF-8 or presented with numeric entities. That'd mostly concern entries originating outside North America or Western Europe. UTF-8 can be used with any text file, and it's downward compatible with ASCII. > > All sorts of lists could be produced from such metadata. > > This has the advantage to not suffer from the "format of the day" > syndrome. :-) Yep, and the source format can be very compact, which simplifies editing and version control. > > I don't think that HTML is a suitable format for collecting the metadata, > > unless the HTML is kept very simple and edited only with text editors. > > I do not think HTML to be a good format for such a thing, either. Marko Message was sent through the cbm-hackers mailing list
Archive generated by hypermail pre-2.1.8.