BRFR Cake Stop 'breaking news' miscellany

[icowden]

Yup. Hands up anyone who has used a new IT system that worked first time. Hands up any one who has called IT support to be be told, Oh it should be working.

To be fair, when you test an IT system it all works perfectly. Then you let the users on and they do completely batshit stuff that no-one thought of testing. "Oh it should be working" is a euphemism for "oh god, what have they tried to do now?"

 

[C R]

To be fair, when you test an IT system it all works perfectly. Then you let the users on and they do completely batshit stuff that no-one thought of testing. "Oh it should be working" is a euphemism for "oh god, what have they tried to do now?"

That means that the test scenarios were not addressing actual user interaction with the system. That's a failure of test coverage.

 

[bobzmyunkle]

That means that the test scenarios were not addressing actual user interaction with the system. That's a failure of test coverage.

Mmm, somewhat idealist.

completely batshit stuff

In my experience the users will always outbatshit even the most comprehensive test plans.

 

[First Aspect]

Mmm, somewhat idealist.

In my experience the users will always outbatshit even the most comprehensive test plans.

There is, in my experience, a disconnect between the people using a system, and the IT people implementing it. I think our own IT dept could work here for a century and still have no real understanding of our workflows. As such, the cart leads the horse.

 

[C R]

There is, in my experience, a disconnect between the people using a system, and the IT people implementing it. I think our own IT dept could work here for a century and still have no real understanding of our workflows. As such, the cart leads the horse.

As developer and now product owner, one of my biggest frustrations is the inability and unwillingness of many developers of all levels of experience, to try and understand what the users want from the system. This also applies to hardware design.

 

[First Aspect]

As developer and now product owner, one of my biggest frustrations is the inability and unwillingness of many developers of all levels of experience, to try and understand what the users want from the system. This also applies to hardware design.

Do you develop products for managing the logistics of fast moving human mass transportation devices? The jeopardy is relevant in this particular example.

 

[First Aspect]

Well what do you know, it turns out that nuclear power is deemed ideal for generating baseline power for a low carbon energy grid.

 

[C R]

Do you develop products for managing the logistics of fast moving human mass transportation devices? The jeopardy is relevant in this particular example.

Without going into detail, I am in that neighbourhood.

 

[First Aspect]

Without going into detail, I am in that neighbourhood.

Trains?

 

[PurplePenguin]

Well what do you know, it turns out that nuclear power is deemed ideal for generating baseline power for a low carbon energy grid.

It turns out the government has come to that conclusion.

 

[Ian H]

Do you develop products for managing the logistics of fast moving human mass transportation devices? The jeopardy is relevant in this particular example.

As long as they issue flight controllers with binoculars it'll be fine.

 

[icowden]

There is, in my experience, a disconnect between the people using a system, and the IT people implementing it. I think our own IT dept could work here for a century and still have no real understanding of our workflows. As such, the cart leads the horse.

Usually you capture the workflows or "user journeys" so that you can build tests that mirror the reality of what people are doing. The problem is that even when you do that there will be variances in the "journey".

For example the user journey might be "Arrived at office, logs on to computer, makes coffee, returns to desk and starts work". Most users adhere, then you get the one that insists on buying coffee, arriving at the office then logging onto the computer, and the one that arrives at office, wanders about a bit, makes a coffee, remembers they are meant to be working and logs onto computer and the one that arrives at office, practices juggling, logs on, makes a coffee, starts work.

Woah! Juggling? No-one mentioned juggling? Why are you juggling? Are you meant to be juggling?

Can we address the juggling issue? Yes, the system supports juggling but we haven't tested that because no-one has mentioned the need to juggle?

 

[Pblakeney]

Usually you capture the workflows or "user journeys" so that you can build tests that mirror the reality of what people are doing. The problem is that even when you do that there will be variances in the "journey".

For example the user journey might be "Arrived at office, logs on to computer, makes coffee, returns to desk and starts work". Most users adhere, then you get the one that insists on buying coffee, arriving at the office then logging onto the computer, and the one that arrives at office, wanders about a bit, makes a coffee, remembers they are meant to be working and logs onto computer and the one that arrives at office, practices juggling, logs on, makes a coffee, starts work.

Woah! Juggling? No-one mentioned juggling? Why are you juggling? Are you meant to be juggling?

Can we address the juggling issue? Yes, the system supports juggling but we haven't tested that because no-one has mentioned the need to juggle?

Can’t that simply be covered by logs in with a possible pause before work?
The reason for the pause is irrelevant to the IT system.

 

[First Aspect]

It turns out the government has come to that conclusion.

Governments plural. And not, I wouldn't have imagined, over coffee and without speaking to a few experts.

The latest announcement is sensible, in that it mitigates some of the cost overruns of learnjng about the one they are still building. It also conspicuously doesn't include building any more large scale plants or postponing decommissioning of the existing ones. Instead they are pursuing the modular reactor concept.

You may have misgivings about cost, but I don't see a better alternative. And don't say gas plus CCS.

 

[briantrumpet]

If it ain't broke, don't fix it.
Oh....

That's certainly the case with COBOL... but not sure that floppy discs and W95 are in the same category.