Check this out for a new model of software development. Why hire full-time developers when your business is not software development? Why rent contract labor? Why not just buy the custom software?

Note that your own internal QA processes have to be top-shelf to do this. Internal staff can screw …

Software, and a software architecture, is a solution to a problem. Without a specific problem, software is a pointless exercise: statements in a formal language with no thesis or purpose. An abstract piece of software akin to the statement: "Time flies like an arrow." It parses (two different ways!) but …

All software problems are large or complex as a direct result of the tremendous reuse potential of software. Consider two alternatives:

• The problem is relatively well understood. Solution: locate an open source or commercial solution and use or adapt that solution.
• The problem is not well understood. Solution: study enough …

The topsoil is in the front yard, the garden is in the back yard.

The users form up a bucket brigade to use several pails and buckets to move the soil. It's backbreaking and slow, so they go to IT for a solution.

"We've measured the work activity very carefully …

Years and years ago I started using Delphi internet services. They had a complicated dialup arrangement through Sprintnet and Tymenet to a farm of VAX servers in the Boston area, part of Northeast Education and Research (NEAR) net. This was Delphi internet, character-mode, 9600 baud. I used Mac classic and …

The most common argument for introducing YAPL goes like this:

"There are way too many programming languages out there. And they are each flawed in some way (small and odd, or large and incomprehensible). So I, in my wisdom and will create YAPL, which will displace all these other, inferior …

A Question:

I would like to build a web browser interface that has capabilities like a spreadsheet.

At the same time, would like to use a tool like OpenROAD from Ingres/CA to do this.

It seems that XUL is the way to go but the entire Mozilla thing is …

The idea of "universal" processing via XSLT is moderately appealing. There's a big issue with poorly defined semantics, which always breaks a meta-data driven approache. The holy grail is perfect metadata from which all application software can be derived in a trivial way. I'm not sure this is ever achievable …

From an email I got recently, lightly edited. Responses to follow

Problem: The data layer of application architecture must be flexible enough to adapt to virtually any new data-driven functional requirement demanded by today's client..right...well to support such adaptability in client deployments, the data model must allow almost …

Mac #1 was a 1986 128K Mac, later upgraded to 1M Mac Plus, with external everything. Eventually a 70Mb SCSI hard drive. Bundled up with another just like it, it sold on eBay for shipping + a few bucks. I kept it running until 2001, when I finally shut it off …