Am So., 24. Nov. 2024 um 19:55 Uhr schrieb Quinn <[email protected]>:
… I get why you chose xml for the claims (xslt and namespaces), but I think I would prefer to use json anyway 😅… On 2024-11-24 22:59, Markus Peröbner wrote:
…There are many reasons why I build everarch but my personal most important ones are: … 2) You probably know that 10 years later you might do things differently than today. The properties you write with perkeep into your JSON are pretty much fixed. Changing their meaning can hardly be done without rewriting attributes in probably lots of permanodes. I hope that with the everarch setup I will only need to introduce a new XML namespace in 10 years and then adjust my XSLT transformation to tell how to map the old data into my new way of thinking. We will see if this will work out :D
This comment by Markus reminds me of an interesting design issue for archives which are supposed to preserve data for decades and centuries into the future. You have to assume that the software which reads the archive's metadata a century in the future will not be the same software which wrote the metadata today. The original source code will probably have been lost, the programming language and libraries will have moved on. You have to assume that the future reader will need to reverse-engineer an implementation of your archive's format.
I like a comment I read once that JSON objects tend to represent a program's in-memory data structure. Future software which aspires to reverse-engineer JSON objects need to figure out the generating program's data structures. XML documents tend to be more self-contained. They tend to describe the semantics of the information being stored, rather than the data structures of the generating program. The XML language is designed with an assumption that the reading software is different than the writing software, and that the document may persist through multiple generations of reading software.
Yes, JSON is currently a popular format, and XML is out of fashion. However, I believe that archive metadata expressed as a document with XML markup will probably communicate the semantics of their content more effectively to future software writers than will a JSON object. Whatever the description language, I think the archive designer would do well to think of the needs of the reverse-engineer a century in the future.
Best regards, —Jim DeLaHunt -- . --Jim DeLaHunt,[email protected] http://blog.jdlh.com/ (http://jdlh.com/) multilingual websites consultant, Vancouver, B.C., Canada -- You received this message because you are subscribed to the Google Groups "Perkeep" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion visit https://groups.google.com/d/msgid/perkeep/2d9c8ecc-f278-46bc-a5cd-8d2ccedebb93%40jdlh.com.
