Thumbnail review not working in BETA-3 [fixed in Beta 4]

[Oliver]

This is the last I will write about this:

This is my point:

Must the user be informed that an IO Exception occurred? Yes!

We have NEVER, EVER had an IO Exception, this is the very first time. And we never thought we would, so obviously we wouldn't have dreamed of telling the user about this! And again: instead of implementing the methods to tell the user about this, I would have rather implemented handling the IO exception gracefully, without the user ever having to know about it.

We do have LOTS of information for the user when things aren't right: we show errors and warnings and LED sequences about all sorts of things that we expect to "go wrong" and inform the user of what's wrong. Those things can be compared to the warning lights in a car!

Staying in your analogy: You're expecting the car manufacturer to build in mechanisms to inform the driver that the noise he is hearing when going on the Autobahn is coming from a slightly out of balance wheel, or a never seen before material weakness where the Alloy of the 2nd cogwheel of the 5th gear was cooled slightly too fast. Or you know - an unknown life-form that has built a civilisation in your rear fender that gets scared over 100km/h and starts banging on your car trying to tell you to slow down.
You're not seriously expecting the car manufacturer to have sensors in place with a user interface that lets the driver diagnose the issue as precisely as possible to be able to tell the car manufacturer how he should change the design of the wheels they are less prone to require balancing, or to start changing the mixture of the Alloy, or how to build more comfortable fenders where those lifeforms could remain calm up to 143km/h (which is your preferred maximum speed). But that's the analogy of what you're doing here.

So no. Not going to happen.

[Unfoolishly]

This is the last I will write about this:

This is my point:

Must the user be informed that an IO Exception occurred? Yes!

We have NEVER, EVER had an IO Exception, this is the very first time. And we never thought we would...

I am sorry to ruin the party, but that is assumption based programming. And that is not good :(
It's a sign of immaturity in the development process.

Because you never thought you would ever get ill, you never take out a health insurance.
Because you never thought you would ever get a flat tire, you never place a spare wheel in your trunk of your car.
Because you never thought you would ever have a fire in your house, you won't install fire detectors in your house.
Because you never thought you would ever get a bugler into your house, you won't install a burglar alarm.
Because you never thought you would ever fail an exam, you just won't study for it hard enough.

You are either really smart, or really leaning too much on assumptions that do not life up to reality. And I don't mean that to offend you in any way!

Putting in my own life's perspective:
I never thought I would be involved in a car accident, but I did. Luckily I survived!
I never thought I would be scammed by an insurance company, but I did. Luckily I got out off that mess in time!
I never thought I would buy an Unleashed that after 4 years of development still isn't working out of the box, but luckily I still support you! And don't become bitter about it!
I never thought I would have to write this to you, but I just did ;-)

On a more serious note: that you THINK that something will never occur, doesn't mean it is good programming to just forget about it in code!
I have seen so much assumptions of really smart people, that actually did work out worst than they could have imagined themselves!

Remember the Pentium Floating Point Unit bug, back in 1994? See here.
That could not have happened in the first place. Intel CPUs are always perfect. Yeah, right! See here.

Don't get annoyed that your own logfiles show you that it DOES happen. I did not alter these logs, they came fresh from the Unleashed App, which your team programmed.
Excuse me for reporting a bug or exception to the case of which you and your team would never had expected in the first place to being EVER possible. But, reality does acknowledge that it is true!

Relax a bit more! Nobody is perfect, and nobody ever will. But you can fix it, and also prevent it.... mainly by NOT assuming too much and just program defensively.
Remember the first law of Murphy: "Anything that can go wrong will go wrong". See here.

Saying: THAT will NEVER happen, means you put yourself in a vulnerable position, because you now must admit it actually did happen! That means one of you is not telling the truth: you vs reality. Sorry, but reality strikes again!

Never say never. A valuable lesson learned today, kids!

And NO, I don't write this to be cocky or to be the "always right"-guy. It just seems to be very realistic not to think of the fact that something NEVER will happen because you don't WANT it to happen, or that you THINK it NEVER can happen! I NEVER wanted to be part of that car crash either, but it did happen!

Greetings,
Unfoolishly

[Unfoolishly]

This is the last I will write about this:

This is my point:

Must the user be informed that an IO Exception occurred? Yes!

We do have LOTS of information for the user when things aren't right: we show errors and warnings and LED sequences about all sorts of things that we expect to "go wrong" and inform the user of what's wrong. Those things can be compared to the warning lights in a car!

You are absolutely right on this.

But, even better would be some logging telling what went right, and what not.
Then the user can look back into some historic events that just happened and determine that it ain't the BT connection, and ain't the sending of the Unleashed thumbnail failing, but the internals of the thumbnail viewer due to IO Exception. That gives the user some rest that it is not due to some action he/she did or did not do. I was actually figuring out what I was doing wrong for 30 minutes, until I just gave up and thought: it ain't me, it's the App itself!

Please, DON'T shoot the messenger!

Greetings,
Unfoolishly

[Oliver]

We completely refactored this screen, and addressed several issues with the thumbnail view, so this should be fixed in Beta 4.