hello Ross and many thanks for replying, On Tue, Sep 24, 2013 at 08:47:19AM -0400, Ross Singer wrote: > In this array of arrays approach, where the tag name is just another value > (as opposed to a key), you cannot take advantage of JsonPath, thereby > eliminating almost any possible way of querying this data in those > databases. I have to admit MARC::MIR is so easy and fast to use from indexed flat files so i never felt the need of Json KV storage for the moment. Also: when i need to expose data for end-users, i use MARC::MIR::Template so a MARC::MIR like this: - [001, PPNxxxx ] - [200, [ [a, Doe], [b, john], [b, elias], [b, frederik] ]] - [200, [ [a, Doe], [b, jane] ]] - [300, [ [a, "i can haz title"], [b, "also subs"] ]] with a template like this: 001: id 200: [ authors, { a: name, b: firstname } ] 300: { a: title, b: subtitle } 700: [ auth_author, { a: name, b: firstname } ] 701: [ auth_author, { a: name, b: firstname } ] will give me a structure like this (which i think is much simpler to query because end users aren't used to MARC grids): authors: - { name: Doe, firstname: [john, elias, frederik] } - { name: Doe, firstname: jane } title: "i can haz title" subtitle: "also subs" id: PPNxxxx But yet, i fully understand the leading ideas behind marc-mij. thanks again marc -- Marc Chantreux Université de Strasbourg, Direction Informatique 14 Rue René Descartes, 67084 STRASBOURG CEDEX ☎: 03.68.85.57.40 http://unistra.fr "Don't believe everything you read on the Internet" -- Abraham Lincoln