Overview

The backbone of the UDF analysis environment is the UDF Distributed Archive. The existance of archives of various content allow sites using UDF data to seemless update their software, add content, and obtain data.

On the surface a UDF archive is an organized collection of files in association with a Server. A system may host multiple archives of various content. No matter the number of archives or their content, all of the archives on a single system are handled via a single Server application. The Server manages the archives, adding and removing content from it as well as takes requests from Clients for archived files and passing those files back to the Client. There are no restrictions on what an archive may hold, practically however, archives are limited in scope to conatin items which the Server understands and Clients know how to request and how to use once obtained.

At the present there are two broad types of supported archives; data archives and software archives. The first contains UDF data sets which can be promoted and installed by Clients as they need them. The second contains primarily software packages which can be promoted and used to update an installed system or add capability to it.

Using the word software as a description of one of the two archive classes is perhaps a bit narrow. The original intent was that it be a repository of software packages but in essence it can house any set of files which only need to be promoted to a Client system and their installed or simply unpacked. As such you will find documentation packages archived withing the software archive as well as other packages containing diverse information.

This document is divided into two sections. The first part covers aTool, a utility for the creation, management, and configuration of UDF archives. The second part is a conceptual overview of the UDF archive, going into the directory structure, the various sub-classes of archives, and the specific formats required for the items they contain.