I have been thinking about how to improve the testing process on the new project and it’s not gonna be easy.

They talk about wanting to get Test Director, or Mercury Quality Center as it is now called but I think it’s just talk. They have sold this software to a big company and come next spring the software will be running live on three times more computers. That should be enough to make them take this stuff very seriously, which they do, but in a bit of a hokey way.

They are a small company but about to get a bit bigger but they still think small. The reasons for not getting Test Director apparently was that Mercury don’t support small companies very well. Well they might be right about that as I have only evere used TD at big compainies. But how much support do they need from Mercury? Once they have it setup I’m not sure they’ll need any support. I have never setup TD from ground up but surely we are just talking about the underlying database which can be anything from Access to Oracle to SQL Server, so what’s the big deal. Get that lot in place and I can teach them how to use it in 3 10 minute lessons.

I know that sounds a bit short but I have done it before. If you leave people alone with TD then they will muscle there way around it but very inefficiently. First week give them a 10 minue lesson on the Test Requirements tab of TD. Don’t let the new users touch anything else, just teach them how to write test conditions as test requirements with references back to source documents and leave them doing that and onlt that for a week.

Week 2 give them a 10 minute lesson on test scripts and how they are written to prove the previously input test requirements. Leave them for a week writing test scripts to prove all of the test requirements and doing nothing else. On the third week give them a last 10 minute demo on test sets and how to create them and also the reports that can be run and the horrendous printed reports, which hopefully may have been made better in the last 18 months since I last used TD.

I think another excuse that the company have is that they don’t have time to do all the TD stuff, but with their current work practices they never will. They don’t seem worried about taking on additional staff so what they should consider doing is getting two more staff with TD experience and unleash them in the direction of TD for a few months to get it setup and generate the TD documentation for all of the additional functionality that is being written right now for the new client. It is obviously not feasible for them to write test scripts etc for everything that is already live, howveer buggy it is. Once the test scripts have been written and used and they see how comprehensive it is then they could maybe lean back a bit a write test requirements and scripts for other new functionality that has been released so far this year.

I think another problem is that one of the top testers is actually a user rather than an IT person and all of her testing is user acceptance type of testing. Trying to sell her the idea of something as almost downright tedious as TD is ain’t going to be easy. She prefers the play with it till it falls apart type of testing. Hmmm…that makes it exploratory testing, which is the new kid on the block isn’t it? Yes it is but so what, what we need here is structure not playing around. Got a feeling we’ll just muddle through this one. Oh well I can muddle through as good as the next person.

Just checked my Sitemeter stats for today (13:00 GMT) and their count of visits to this site has missed at least 80% of the hits. Worse still, I have just had a look at the Sitemeter who’s on page, who is currently viewing the blog, and it says there have been no visits for the last 20 minutes. I also checked this page 20 minutes ago with the same result. If I look at my blog log file, which registers all visits, including spiders and feeds as well as normal visits I can see a lot of activity. Am I paying per month for this the more professional version of Sitemeter. Why do Ibeat myself up with this garbage. The staff at Sitemeter can justify these lousy stats, but I betcha they don’t use their own product to accumulate visits to the Sitemeter site.

Test Director questions

I look in the SQAForums every now and again. They have numerous threads on different subject of testing. When you look at many of the questions posed it is pretty obvious that a lot of the posters have absolutely no software testing experience at all and ask questions so wide that they need a complete brain dump on the subject to answer the question. There was one question either in the interview questions thread or the Test Director thread where somebody wanted to know answers to a TD interview. Sounded like he was going to an interview at a company which used TD and he wanted answers to any possible questions. Didn’t he have any knowledge of the product? Has he never used TD? If not then why were this company interested in interviewing him? Maybe they had just bought the TD product and wanted to hire people with TD experience. Sounds more like the blind leading the blind to me. Shudder.

Oh to be a fly on the wall at that interview.