SQL can be difficult to factor out repetition of code/concepts. For example, you cannot reference repeated concepts by name, having to use nesting instead, which sometimes results in duplication when the references are not really tree-shaped. Time to dump SQL for better relational languages that don't try to mirror COBOL synax.

Further, most Xbase dialects lets one define a UDF while most RDBMS do not, at least not without special previlegdes.

I am not saying Xbase is perfect, but sometimes it gets around problems found in SQL and/or MS-Access for custom chomping beyond simple queries. The industry tends to force one to use Big-Iron SQL in combination with MS-Access, and those two are both half-ass.

My current gig requires a lot of custom chomping of data, so I need better tools. My partners do a lot of shit by hand that I would rather automate somehow. I cannot compete with their manual mass-copy-and-paste efforts because I am not used to so much manual stuff. I am not geared to be keyboard clerk. The problem is that it is tough to sell them on more automation and factoring because they are so fast and used to manual typing. They get edgy when I try to introduce scripts. I think they fear being automated out of work, but won't admit it. It's not like there is not plenty to do now, but maybe in slow times the scripts could result in staff reduction.