Background: I'm looking at parsing certain X12N messages. There's a Perl program on CPAN to do X12 parsing ; the program itself isn't useful, but the X12 messages are defined via INI files. There's also a Python X12 parser with the X12 messages defined via XML files. I don't really like …

I hate to complain, but there's a Python syntax rule that has slightly cramped my style. The problem stems from a need to describe a really complex and extensible structure.

We need to describe X12N messages, which have a rather complex structure, and the structure can be customized. While the …

See Sean McGrath's Spot the warning signs in configuration file design for thoughts on the ever-gnarly problem of how to provide a workable set of configuration parameters to a program.

A programming language has all the advantages as a language for encoding parameters. I see why it's rare -- too many …

See Spolsky's Talks at Yale, Part 1 of 3 for a fairly typical summary of formal methods in software development. Here's the part RL liked: "Now, if the spec does define everything about how the program is going to behave, then, lo and behold, it contains all the information necessary …

In Lenses That Distort Our Software , I noted that the relational model of data can distort our view of the underlying reality that software models. In that posting I was whining about how meaningful structure is sacrificed to force fit information into the rows and columns of the relational world …

A taxonomy of pipeline alternatives is a big, hairy deal, involving lots of alternatives. While your basic API methods are relatively few in number, each language and platform introduces new options. Without some focus, it's challenging to produce a big, fluffy description of all potential alternatives.

However, the interesting thing …

First, the confusion. A couple of comments asked for code, offering to "fix" the problems. Specifically, offering to optimize or improve the PL/SQL. The code was already posted, but I failed to make that clear. Here's a recap of the various blog entries on this subject.

1. PL/SQL vs …

There appear to be four responses. Two were numbered, the third is a rambling thought on scalability, and the fourth is a thought on management's discomfort with the truth.

Physical Tuning.

"...databases can be tuned by creating appropriate indexes, removing others, modifying the physical database schema to suit the most common types of queries, etc. So it isn't clear whether the PL/SQL figure could be better if the database was optimised by a good DBA."

While a good …

While the question of speed came up recently, it ties in to a long-standing position of mine. I ran some realistic benchmarks, and I'm much happier with my architecture now.

Background.

Years ago, at a Siebel conference, I heard a heretical comment. Specifically, they use the RDBMS as simple, flat …