When I was looking for how to say relational database in Esperanto - rilata datumbazo I came across the alternative datenbazo. What's the difference and what speaks for the use of datumbazo in favor of datenbazo?
According to PIV, datumo (the basis of datumbazo) and dateno (the basis of datenbazo) are synonyms, so I would assume datumbazo and datenbazo are also synonymous. However, PIV does not contain datumbazo, and says that datenbanko is preferable to datenbazo.
However, I personally rarely hear datenbazo, and komputeko.net, which is generally a good source for computer terminology does not contain it. Also, searching datumbazo in Google yeilds 5,920,000, while searching datenbazo only gives 228,000 (datenbanko only gives 6,500), which tells me it is more commonly used.
You could, then, argue for whichever you like the most - there's nothing about relational databases in the Fundamento. I assume that PIV prefers dateno/datenbazo/datenbanko is because it's easy to assume that datumo = dat (date) + um + o, which doesn't really make logical sense.
Personally I have never heard the word datenbazo, to my recollection. Datumbazo strikes me as being very common and is the word that I always use for this.
I had never known the word dateno, only datumo, so to my eyes "datenbazo" looked almost like the word "dato" receiving the accusative of direction (-en) and joining with "bazo", which created (to my mind) kind of a strange nonsense word. I wouldn't say this is a strong argument against the word dateno, its a legitimate synonym after all, but if you're going for easy comprehension I though I'd just mention that for me (and so I'm guessing for the average Esperanto speaker, though of course I could be wrong), datumo is a more quickly recognized root.
The origin of the word "data" is in the plural form of the Latin word "datum" (a given thing → many given things). From that it was taken to denote "identifier given to a day", a datum, many data.
When computer programming was invented, one needed a word for input that was given to the programs. The chosen word was the Latin-based "data", the given things.
However this created a couple of issues in some, especially Indoeuropean, languages.
- What is the singular form of the programming related word "data"?
- Does the word "data" relate to programming or dates or is just a Latin word?
There is no excuse to repeat this mix-up in Esperanto. Therefore the root dat/ (and thereby its derivatives datumi and datumo) should be restricted to day/date (according to that interpretation datumbazo is a collection of date stamps ,say, in a logging system). For data in the general programming sense one should use a completely other root, like daten/ (so datenbazo is a database capable of containing different kinds of data).
When it comes to whether use bazo or banko, many people (me included) within the computer science make a separation between these. A databank is a general term for storing data, while it can be and usually is somehow ordered for the data to be programmatically retrievable, it's not as ordered a database, which has a separate system (a database management system) for defining, manipulating and retrieving data.