Why critical thinking is more imortant than ever

I'm an optimistic sort of fellow and I really thought that promotions like this:


... were fewer and further between in our industry.

Now I know it's foolish to think that technical people (with experience and relevant knowledge) will ever write marketing blurb (so it is credible), but it stil pains me to read this sort of thing.

"Top 10 Reasons to Implement Automated Application Quality Management"

Let's ignore the 'Top 10 Reasons...' stuff and pick on some of the language used in the post.

Can we agree what quality means? The article implies that quality=performance (well, good, high, acceptable, reliable performance is a jolly good thing). Says not a thing about functionality, security, usability etc. etc.

I'm no fan of standards, but for heavens sake go look up what software quality is in a book or standard - say ISO 9126? Here you are: http://en.wikipedia.org/wiki/ISO_9126

And now, dear readers, read for yourselves the 'ten reasons...'. Assimilate and learn - how NOT to make your case.

One: "Your mainframe isn't going anywhere". Why would I "throw money at an ASQ tool?". Surely all I need to do is to observe my current system load (of 300+ applications) to identify the bottlenecks so I could fix/upgrade them. Ehe-huh. 0/10.

Two: "Processing Power Doesn't Come Cheap" - of course not. see 1 above why I should observe my current environment first.

Three: "Schedule Upgrades Rather than Scrambling for Them" - see 1 and 2 above. Why would you buy a perfomance tool when you ould use basic system utilities to capture past perfoamcne stats and predict future challenges?

Four: "Tuning Applications is simpler than Finding IT People" - the two aspects are not related. Tuning is a specialist task that needs good monitoring information. see 1, 2, 3 above. Tuning is a highly technical task - if you have the people, that's their job. If you lack people with those skills - hire them. A tool cannot ever help with that.

Five: "Improved Visibility" - er yes. Exactly WHAT does your tool do differently from basic system monitoring (as per 1,2,3,4)?

Six: "Fire Prevention, not fireFighting" - sensible - but what has that to do with your offering?

Seven: "Build Better Mousetraps" - the implication is that the offered tool can do better monitoring than the developer is currently using. In fact most developers do little (no monitoring or adjustment) about performance while developing. But having said that, their development tools probably provide better monitoring stats than any non-developer tool.

Eight: "Productive users=productive companies": Better quality will improve the lives and productivity of end users. But this phenomena is independent of HOW you achieve it. Irrelevant.

Nine: Better Response Times = Bettter Results". Excuse me? The response times of people are more likely to be measured in minutes or hours. I don't think a 0.1-0.2 second saving will really mean that much to most employees. Get some real data.

Ten: "Automated mainframe AQB Can Run in a Mainframe Environment". Maybe. But the spiel gives no indication of that the CA solution can actually do.

Our suspicion must be that it can do nothing for average end users.

Computer Associates. Shame on you.