System Design

The Babase system is designed to facilitate the retrieval, storage, and maintenance of the Amboseli Baboon Project data. Data integrity is foremost. Analytical power, ease of use, and low cost are secondary goals. The system consists of tables to store and organize the data, software supporting data validation and derivative data generation, stand-alone programs used to facilitate the entry and maintenance of the data, a minimal tool set supporting the maintenance of the Babase system software itself, and documentation. data are retrieved from Babase using the SQL language, the standard[2] language used to query relational databases. SQL is declarational as opposed to procedural; from a single SQL query (a single statement) the database determines how to best retrieve the data requested, no matter the number of tables or criteria required. SQL provides a single, powerful, interface for ad-hoc data retrieval and manipulation. Generic software provides the bulk of the user interface[3], traditionally the most complex and costly software component.[4] Consequently there are few stand-alone programs written specifically for Babase. The overall philosophy of the systems implementation is to keep the software as easy to maintain as possible while assuring data integrity. To this end, the system is comprised of as many generic components as possible and the design requires custom programming for only the most crucial features.

Babase puts as much intelligence as possible into the database itself, including automatic data validation and complex automatic analysis and storage of the derived data.[5] Babase extends its sometimes complex and rather abstract database structures with alternative, more familiar and user-friendly, means of accessing the underlying data[6]. These constructs are, in so far as is possible, made indistinguishable from the underlying data when querying and updating the database. Babase often generates derivative data for more ready analysis. This is, for the most part, transparent to the user. The end-user is insulated from implementation details, the number of interfaces (primarily SQL) the user must learn is minimized, and the user is free to work with the data structures that embody the conceptual model best suited to the task at hand.[7]

Data input is an example of how Babase incorporates generic programs. The prototypical way to import data into Babase is in bulk, via a plain text file having columns delimited by the tab character. These are easily produced by almost any spreadsheet program; it is expected that most data imported into Babase will be typed into a spreadsheet and then exported to tab-delimited text for upload.[8] The use of generic interfaces reduces cost, and minimizing the number of novel interfaces frees the end-user to concentrate on the task at hand.

Babase is designed to be accessed over the Internet, primarily via the web. Although there are exceptions[9] the majority of Babase is accessed via a W3C compliant web browser. Individually assigned usernames and passwords are used, along with encryption, to secure the database content. The Babase Wiki provides content for an the structure of the project's web site. Another example of Babase leveraging a generic program, the wiki allows project members collaborate, share information, and build the project's web site without programmer intervention.

Babase is built upon standards[10] and popular, widely deployed, Open Source and Free Software. This means, among other things, that the tools used to build and run Babase are very likely available to anyone free of cost, and that the skill-sets required for the system maintenance of and, to some extent, use of Babase are readily learned[11] and unlikely to become obsolete[12].[13] The Babase source code itself is Free Software[14] and may be downloaded by the public.[15]

Note

The database design attempts 5th normal form, no redundant data, no empty data elements allowed, etc. What we've actually wound up with is about 3rd normal form.



[2] More or less. The last actual SQL standard was issued a very long time ago. None the less SQL is pervasive and, although specific SQL statements may not always be, the skill set involved in SQL use is quite portable.

[3] There are many PostgreSQL user interfaces available, although at the time of this writing only 2, phpPgAdmin and psql, are installed on the Babase database server. Many of these front-ends must be installed on the local workstation. These may require that the Babase VPN be running before initiating a connection to the database. Some of available front-ends may be found via the PostgreSQL FAQ question regarding graphical user interfaces for PostgreSQL.

[4] It's those pesky unpredictable users. Computer software would be a lot easier to write if it weren't for users always messing things up and then insisting on knowing what happened.

[5] A process which, admittedly, sometimes conflicts with the notion of easily maintaining the software. On the other hand when done right this approach does wonders for data integrity.

[7] These features also free the user from software interface lock-in. The database may be accessed and maintained with the software of choice. Data integrity, in both raw and derived data, is assured. Significantly, these features are those that allow Babase to leverage generic programs, using them for the bulk of its user interface as opposed to building a custom, Babase specific, interface.

[8] Of course, because Babase has no designated front-end and so much data validation takes place inside the database itself, any program able to talk with PostgreSQL, the database engine Babase uses, can be used to import data into the database. So there are no real limits on how data must be structured for import into Babase.

[9] There are 2 Unix shell programs that provide peripheral utility; both do tasks that can be done with other tools but are handy to have automated. The use of these programs are documented on the Babase Wiki. Comprehensive documentation of these programs should probably be added to this document.

The Unix Shell Programs

babase-copy-babase-schema

Copies the entire content of the babase schema from one database to another.

babase-user-add

Adds a postgresql user, granting the permission to use Babase

There is also the ranker program, which runs on the local workstation and uses the Internet to communicate with the database. Developed separately from the rest of Babase, neither the source code management of nor the documentation for the ranker program is particularly well integrated into Babase.

[10] Actual standards, not de facto ones.

[11] Because open standards and the documentation for Open Source and Free Software programs are available, without cost; and because the inherently transparent and public nature of open standards, Open Source and Free Software leads not only to a wealth of good instructional material freely available on the Internet but also rounds out the basic requirements of a complete learning environment by ensuring that the software itself is available to everyone.

[12] Because once software is released and distributed under a Free or Open Source license it cannot be locked away and made unavailable, and because open standards are rarely changed in a backwards-incompatible way.

[13] Consequently the skills are rather widely available. The difficult part, as always, is finding the all of the relevant skills at once. For more on this see The Babase Program Code section.

[14] Presently licensed under the GPL Version 3 or later.

[15] Babase database content is not available to the public.


Page generated: 2024-05-31T10:37:49-04:00.