Monday, September 23, 2002

This makes me nervous. Changing your data model because your application code changes seems like putting the cart before the horse. Your data model should be well thought out. Any changes to the data model should also be well thought out. This is especially true when multiple applications share access to a database. Changing the data model because of one application's changes is extremely risky. I've actually heard this time and time again, but I still don't buy it. This is our programmer laziness coming through, and it is one area where I believe we should not be lazy. [All Things Java] This is entirely dependent upon what you think the relationship between your code and your database is. When you say "changing your data model because your application code changes seems like putting the cart before the horse," you're implicitly taking the position of the primacy of the data model relative to the application, which in turn implies that the application and data model are distinct. Another point of view—and one that I prefer—is to treat the application object model as primary. The data model is subservient to the application, and in fact is ideally derived from the object model (hence "O/R mapping.") The database is just a means to persist objects. There shouldn't be multiple applications accessing the data, unless it's through the objects from which the data model was derived. The objects will change; the data model will change with them. The great pre-planned never-changing data model is a myth anyway. Now, of course, there will be exceptions to the general rule: a legacy database with legacy apps manipulating the data. So it's important to be able to go from schema to object model as well as vice-versa, but in new application development, this should be the exception rather than the rule. 10:09:48 PM    comment []   I've been hearing a lot about FOAF, which is an acronym for Friend Of A Friend. It's an RDF-based file format that lets you walk a network of people who are friends. It's a lot like a network of blogrolls. [Scripting News] Yep. Once again, the benefit here is really about consistency of representation and applicability of general-purpose tools (say, to do constraint satisfaction to support scheduling among the various F's in a FOAF network, each of whom might have an iCalendar-compatible calendar). Hmmm. Maybe the point here is that RDF is about self-describing data? Whatever the case, it's becoming clear that talking about "metadata," as if that had a context-free meaning, is a non-starter. 9:49:50 PM    comment []   Shelley Powers raises some interesting questions re whether RDF has a place in syndication. She says that RDF is trying to build a persistent database (aka the Semantic Web) and RSS is trying to flow news that has a short lifespan. I had not heard this point before. It's worth thinking about. [Scripting News] My immediate reaction to this is twofold: first, that I'm not sure that I agree that syndicated news content necessarily has a short lifespan, and second, that I'm not sure that I agree that RDF is necessarily trying to be persistent. Indeed, so far, most instances of RDF that I've actually consumed have been as ephemeral as a web page, since most RDF I've consumed has simply been displayed as part of a web page. To my mind, this is almost exactly the same durability as most RSS feeds that I read, which are also displayed in web pages. I'm inclined to think that the temporal issues described are orthogonal to the representation selected. RSS is about syndication; RDF is about machine-readable metadata. It's up to the individual deployer to determine whether and when the two meet. 9:41:03 PM    comment []   My windows belong to me, damnit.. Mike: Why don't people understand that opening windows for me is a bad idea? It's one of the original usability sins. Erik: I spawn new windows, and I like it. So stop your whining. Charles: I'm on Mike's side. [The Desktop Fishbowl] The solution to this is really quite simple: don't return to the offending site. 9:03:59 PM    comment []

© Copyright 2002 Paul Snively. Last update: 10/3/02; 8:09:08 AM. This theme is based on the SoundWaves (blue) Manila theme.