Blog

go back to the blog

What is the point of a Gate if there is a chuffing great hole in the fence?

  • 12/08/2011
  • 6908 Views
  • no comments
  • Posted by EuroSTAR
-->

Last year, as I drove through the beautiful Alentejo on holiday, the current Mrs Simms pointed out a rather ornate wrought iron gate built on the side of the road by a dirt track. The gate was closed, and the track split into three, one (if the gate was open) would take you through the gate up to what we supposed was a ‘Monte’ somewhere above. The other two  split left and right around the gate and then joint the first in its ascent. There were no walls or fences, just the gate! Now a gate that serves no purpose might sound strange to some people, but not to testers. Oh no, we are only too familiar with such events.

We take time to review the requirements and give each an acceptance criteria  or test result that it has to achieve before it can be marked as a ‘pass’. We create a test strategy that states that for this or that phase of testing there will be a set of entrance criteria that needs to be met before something can be considered ready for test and a set of exit criteria that that have to be met before we can consider the phase completed. We may even hold readiness review and quality gate meetings.  But time and time again we see the criteria failing and compromises or exceptions being made. What’s the point of a gate, if the blighters can simply navigate around it with ease? In fact I have recently had cause to ponder this very situation as we near a quality gate on my current project with no hope of achieving its criteria and the fence hurdlers, tunnel diggers, hedge cutters and gate crashers start to rehearse their arguments and voice their demands to be let through.

First of all, just like the gates on my dirt track, even if they are ineffective, the very fact that they are there presents a message, By stating in advance what the acceptance criteria is, it gives a clear goal to aim for. Knowing what is being aimed for is the first step towards achieving. So in my case,  I have set some fairly stringent entry criteria for a round of testing that is coming up. The supplier knew what the criteria were and, although they won’t be able to meet them, at least they knew what was expected and what to work towards. Without the gate, we would simply have been delivered a set of code that might or might not have had a suitable level of maturity.

The fact that we had specified the quality gate criteria in advance, also means that everyone is clear that they have failed. The supplier can’t say that they have delivered what was expected, and the customer cant suddenly deicide that they want a lot more than has been delivered. There was a measure, the gate has shown that the measure has not been met and although we will navigate around the gate it is clear on what basis we proceed.

Also I believe that having gates at regular intervals along the path does give you the opportunity to track progress. I have written elsewhere about the tendency for suppliers to go ‘Red October’, so that you really don’t know where the software is at until it suddenly pops up. The gate means that the supplier has to stop and present the current state of their software at predefined intervals. Involving the right people in the gate review means that you can get a quick decision on any issues and decide if you will or will not allow the software to pass through to the next stage, or whether the gate will bar progress until the criteria has been met.

The gates we have set are governed by ‘Readiness Reviews’, so along with the question around what criteria has been met and what has failed, we also look at all the elements we have to have in place before we can proceed with testing.

Typically the gate will look to review the following elements:

• Risk & Issues
• Release Note
• Entry criteria
• Environment build and availability
• Data
• Resource

So, even though my fence may be full of holes, and the gate can be pushed open fairly easily, the very fact that I have a gate at all adds some value to the process. The stronger the fences and the more secure the gate, the more value it adds.

Tony Simms is the principal consultant at Roque Consulting (www.roque.co.uk).

You can contact him via email at; [email protected]

Blog post by

go back to the blog

eurostar

Leave your blog link in the comments below.

EuroSTAR In Pictures

View image gallery