test please delete

[zharrt]

Useless bloody client.

- Completed development and handed it over for UAT
- UAT carried out by a designated person, and testing passed
- Module released into production
- Users then do some testing in UAT and tell me it failed
- No test script, no instructions, no details

Apart from one "failure" which is impossible to do. Not that it shouldn't do it, but IT IS IMPOSSIBLE TO DO THAT!!!

Never underestimate the unique ability a user has to break something in a whole new way

[zharrt]

Pssst

[norrahe]

Should have gone Nexus 4

Nah, Mr N has the Nexus 7 as much as I like them you can't actually purchase them in cloggers yet, to get his N7, I had to get cojak to order it and as I was in blighty at the time I had to pick it up off her.

This phone will do me for the next 3 years.

[norrahe]

Afternoon all

[norrahe]

Useless bloody client.

- Completed development and handed it over for UAT
- UAT carried out by a designated person, and testing passed
- Module released into production
- Users then do some testing in UAT and tell me it failed
- No test script, no instructions, no details

Apart from one "failure" which is impossible to do. Not that it shouldn't do it, but IT IS IMPOSSIBLE TO DO THAT!!!

Users gotta love 'em - No test script, who was the muppet doing the UAT?

[norrahe]

My cookbooks have turned up

Am currently working my way through all the books I've bought but never got around to reading.

[norrahe]

Pssst

We know

[TheFaQQer]

Users gotta love 'em - No test script, who was the muppet doing the UAT?

The orignal UAT guy had a test script. I wrote it for him, to cover all scenarios.

The rest of it has been ad-hoc - they got a report from someone which told them what to expect, and then when it didn't do what they expected, said it failed. The problem seems to be that the report is looking at the wrong data.

So rather than coming from a situation where they know what should happen, and then basing the success or failure on that, they started somewhere completely different where they hope something should happen, and failed it.

Without recording what they actually did. Or what they were expecting to happen. Or what actually happened. Or switching debug on so that we know what they did and what happened.

[TheFaQQer]

Never underestimate the unique ability a user has to break something in a whole new way

Nope, it's impossible.

My code only does one thing. It is impossible for it to do something completely different.

I don't even know how to get it to do the thing they say happened, because it's complex and I haven't looked at the APIs for years. It's not even as if it can inadvertently happen.

It can't.

[norrahe]

The orignal UAT guy had a test script. I wrote it for him, to cover all scenarios.

The rest of it has been ad-hoc - they got a report from someone which told them what to expect, and then when it didn't do what they expected, said it failed. The problem seems to be that the report is looking at the wrong data.

So rather than coming from a situation where they know what should happen, and then basing the success or failure on that, they started somewhere completely different where they hope something should happen, and failed it.

Without recording what they actually did. Or what they were expecting to happen. Or what actually happened. Or switching debug on so that we know what they did and what happened.

indeed

I once took over a project and asked the ex- pm for his UAT scripts, he said "he didn't have any as it was data entry software, you couldn't possibly script that"