Software code is like an ever expanding gas, with script code tending to expand faster than most other kinds. Part of this expansion is due to ordinary maintenance, but it has costs. Fortunately, these can be identified and reduced, often with some serendipitous results.

Web scripts, particularly html, grow over time for a number of reasons, mainly having to do with maintenance as changes get piled on top of changes. Hopefully, these changes are highlighted with comments, ideally explanatory, for those who follow. The worst are those that remove a section or block from interpretation, but remain in the script for various reasons, none reasonable when version control is in effect.

All this extra stuff increases the KLOCs (thousand lines of code), much beloved by managers without technical expertise. This crude measure of output encourages measurement by weight rather than quality or density. But the last two are hard to measure, while the first one lends itself to qualitative activities, often a natural output of the version control systems themselves.

It’s always better to have comments, at least up to a point where contradictory information starts getting in the way of clarity. Not so the extra white space, with multiple tabs and such that programmers like to use in profuse quantities.

The costs of all this lint shows up in two major areas. The first is in the maintenance activities themselves, as the thicket of lint and KLOCs start to obscure the quick and the working. Another cost component is the increased time it takes to transmit a scripted page, and then to interpret it, either by a web browser or a search engine indexer.

Transmission times tend not to concern designers and developers too much, if at all. On the surface this indifference can make sense in this age of increasing availability of high speed pipes, and the tacit assumption that more pipes are lighting up every day and that they will only get bigger. But the lint component of delivered content can easy tip 20%, and often exceed 30%. This is increased transmission time, and everyone knows that time is money, and maybe having to get another server ready.

Reducing transit time by 20% means moving that portion of line costs straight to the bottom line, resulting in a return on investment that can be quite impressive. Search engines don’t have to do as much work either, which can result in deeper indexing.

All it takes is a vanishingly small one time investment in a program that removes the lint from the script code before it goes out, either to a web site server repository, or inline with the output handler. This works not only with straight html, but cascading style sheets and javascript sections as well. And, doing this has the side benefit of protecting these exposed intellectual assets by making reverse engineering very difficult because these long strings are really hard for a human to read.

© Copyright 2009 Chuck Brooks for FutureWare SCG

A Word From Our Sponser

Excel Excel When Importing  Spreadsheets Into Indesign, Pagemaker Or Quark!

Tags: html lint, kloc, klocs, protecting intellectual property, software complexity, software maintenance, software roi, web design

This entry was posted on Tuesday, September 29th, 2009 at 0:50 and is filed under Software That Works. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.