Turns out feature new branches from Main have a lot of them. If I find 5 bugs in a month I'm doing pretty well. I just found 5 in 30 minutes and I only tested about 10% of the new features. But when your landing page is effed, your image and user confidence goes down the drain. Just saying. I love breaking things and sending 'em back to their Integrated Development Environments to rework.

Are you sure that they are not being sent to AI to get "fixed"?

1 hour ago, jdc2000 said:

Are you sure that they are not being sent to AI to get "fixed"?

No and for all I know AI wrote the heap in the first place.

Software designers are notoriously bad at handling "unplanned events". Error handling routines are an art form unto themselves. The concept of "what happens if the expected doesn't happen" isn't taught in schools.

1 hour ago, maddmaxx said:

Software designers are notoriously bad at handling "unplanned events". Error handling routines are an art form unto themselves. The concept of "what happens if the expected doesn't happen" isn't taught in schools.

That was definitely the most fun part of programming, trying to see what could go wrong. Turns oot, quite a bit! There was always the boundary conditions for a start.

11 minutes ago, EVOO said:

That was definitely the most fun part of programming, trying to see what could go wrong. Turns oot, quite a bit! There was always the boundary conditions for a start.

It requires a lot of thinking to imagine what can go wrong then you design something and know ahead of time what goes right.

In 1981, I spent a long time writing the programming for "Castles of Darkness" - one of the first animated adventure games. Find me on "the Giant List of Classic Game Programmers."

I had to think of all the stupid things those playing the game would try, which was hard considering I created and designed the game and knew all the right things to do.

I didn't have a big beta team to test it so I had to make sure the program wasn't going to end up in an endless loop, etc. because someone did something unexpected.

1 minute ago, MickinMD said:

It requires a lot of thinking to imagine what can go wrong then you design something and know ahead of time what goes right.

In 1981, I spent a long time writing the programming for "Castles of Darkness" - one of the first animated adventure games. Find me on "the Giant List of Classic Game Programmers."

I had to think of all the stupid things those playing the game would try, which was hard considering I created and designed the game and knew all the right things to do.

I didn't have a big beta team to test it so I had to make sure the program wasn't going to end up in an endless loop, etc. because someone did something unexpected.

I remember being proud when I managed to crash a corporate time sharing computer using a terminal with a mouse ears modem around 1981 or so just by entering some number it didn’t like.

9 hours ago, Dottleshead said:

No and for all I know AI wrote the heap in the first place.

Obviously AI will just handle it all very soon. You are wise to be retiring soon.

AI can only do what it's human programmers programmed for and what it had available to learn from. It's already clear that AI makes mistakes more often than was planned for.

In many places AI has already proven itself to be a bridge too far and has been halted.

Ironically here is the Google AI take on the stopping of AI projects. https://www.google.com/search?q=Where+has+AI+been+halted&rlz=1CAVARX_enUS1131&oq=Where+has+AI+been+halted&gs_lcrp=EgZjaHJvbWUyBggAEEUYOTIHCAEQIRigATIHCAIQIRigATIHCAMQIRigATIHCAQQIRigATIHCAUQIRigATIHCAYQIRiPAtIBCDc3ODdqMGo3qAIAsAIA&sourceid=chrome&ie=UTF-8

40 minutes ago, maddmaxx said:

AI can only do what it's human programmers programmed for and what it had available to learn from. It's already clear that AI makes mistakes more often than was planned for.

In many places AI has already proven itself to be a bridge too far and has been halted.

Ironically here is the Google AI take on the stopping of AI projects. https://www.google.com/search?q=Where+has+AI+been+halted&rlz=1CAVARX_enUS1131&oq=Where+has+AI+been+halted&gs_lcrp=EgZjaHJvbWUyBggAEEUYOTIHCAEQIRigATIHCAIQIRigATIHCAMQIRigATIHCAQQIRigATIHCAUQIRigATIHCAYQIRiPAtIBCDc3ODdqMGo3qAIAsAIA&sourceid=chrome&ie=UTF-8

You're sort of calling the game after the first pitch? We might not have even gotten all the players on the field yet with AI, and this is the pitcher warming up his arm and he batters swinging weighted bats still.

3 hours ago, Razors Edge said:

Obviously AI will just handle it all very soon. You are wise to be retiring soon.

2 hours ago, maddmaxx said:

AI can only do what it's human programmers programmed for and what it had available to learn from. It's already clear that AI makes mistakes more often than was planned for.

In many places AI has already proven itself to be a bridge too far and has been halted.

Ironically here is the Google AI take on the stopping of AI projects. https://www.google.com/search?q=Where+has+AI+been+halted&rlz=1CAVARX_enUS1131&oq=Where+has+AI+been+halted&gs_lcrp=EgZjaHJvbWUyBggAEEUYOTIHCAEQIRigATIHCAIQIRigATIHCAMQIRigATIHCAQQIRigATIHCAUQIRigATIHCAYQIRiPAtIBCDc3ODdqMGo3qAIAsAIA&sourceid=chrome&ie=UTF-8

1 hour ago, Razors Edge said:

You're sort of calling the game after the first pitch? We might not have even gotten all the players on the field yet with AI, and this is the pitcher warming up his arm and he batters swinging weighted bats still.

I agree with Max here. For now. Then I think I agree with RE. You will always need an overseer -- someone who understands things -- and I think for our product base we are a ways off from the RE outlook. Maybe decades. There is just way too much complexity and not all of it can be learned from input. Yet. Until then I might have luck for the next 3-4 years working where I am. But when they let me go, whether I am in the financial position I'm shooting for or not, I'm calling it a career. Nobody is interested in paying my old man salary and even if they agreed to, they're going to expect unreasonable things from me. And honestly I don't GAF. Let AI take over the jobs and then let AI buy their products while the rest of us prepare for the breakdown of society. Wait. That's happening now. Are you prepared?

2 hours ago, Razors Edge said:

You're sort of calling the game after the first pitch? We might not have even gotten all the players on the field yet with AI, and this is the pitcher warming up his arm and he batters swinging weighted bats still.

The game will be over before he's ready. The best part of AI will be in the guidance packages.

Back in 1999 I was the senior contractor tester for a major product for the US Post Office. I found bugs by the dozens. In April 1999 I was told not to submit any more bug findings as the product was going live in June 1999. Instead I was told to just write them down and they would be looked at after the release. It was for an international money transfer scheme and one of the bugs that I found was that a postal employee could steal money in such a way that it was untraceable.

So knowing when the excrement hit the rotating impeller a lot of it was going to hit me, I found another job. Fortunately the program never launched as it violated the charter of the USPS as other private companies, like Wells Fargo, were already doing that kind of money transfers.

The USPS was such a bureaucratic clusterfrack that it made the military look sensible.

5 hours ago, MickinMD said:

I had to think of all the stupid things those playing the game would try, which was hard considering I created and designed the game and knew all the right things to do.

Yeah... this.

We had a billing problem. A few thousand accounts had errors for meter rental charges, after a billing system update. Our dept got the opportunity to fix the problem.

I went to the training session and found out the process was 100% manually done. We were expected to search thru thousands of rows in an Excel file, find the individual customer data, copy that and paste it into another spreadsheet for some more calculations, etc... to correct the meter rental, figure out if the was over or under charged and provided the data our billing dept needed to fix the account.

More than a few people didn't have the Excel knowledge to do this. What a cluster fk. Soooo many ways this could go wrong

The training was on a Friday. I interrupted the training and told them don't do this. I'll write macro to make this much easier and more importantly have a process that will reduce the number of potential errors.

I wrote an Excel macro over the weekend to do nearly all of the tasks in Excel.

There was maybe 4 or 5 user input fields and a button to click that would run the macro.

OMG... having 60 different users, it took maybe 3 hours before I got the call, 'Your macro doesn't work.'

I had to call the user and ask what happened?

I ended using dropdown box for some inputs to eliminate the many creative ways someone can enter data.

Our 10 digit account number that can have leading zeros was an interesting input to validate.

If I recall I issued a few updates. Each version had more error trapping and better validation methods.

It's amazing how some people cound enter data in unique way that could blowup the macro.

The macro saved hundred of hours of manual work and eliminated many possible errors.

I opened 3 more today and I haven't even gotten to the good stuff yet. Some Dev tried to close one as not reprodicible so I sent him a video of the whole damn thing showing it was worse than the original bug filed. Let see if that doesn't shut him up or at least a call to arms. Don't be closing chit as "not reproducible" because you aren't trying to hard and I'm going to stick your nose in it. Anyway, it was good because I clearly exposed not only issues with the software but confusion w/ the process on how to file these sub-main branch bugs. Then we had a meeting about it and I advised our whole team on how to look for bugs in our graphical user interface as I had folks asking me how I was able to find them. Boss man was on the zoom video too and it was being recorded. It was great. For not getting much positive feedback, it sure was good. I asked a couple folks for feedback and they all said I did really really well. Those are my peers and honestly they're the only ones I really care about.

12 hours ago, MickinMD said:

It requires a lot of thinking to imagine what can go wrong then you design something and know ahead of time what goes right.

In 1981, I spent a long time writing the programming for "Castles of Darkness" - one of the first animated adventure games. Find me on "the Giant List of Classic Game Programmers."

I had to think of all the stupid things those playing the game would try, which was hard considering I created and designed the game and knew all the right things to do.

I didn't have a big beta team to test it so I had to make sure the program wasn't going to end up in an endless loop, etc. because someone did something unexpected.

This. When your the programmer you naturally have an inherent conflict of interest and your thinking about inside out. The external user is just the opposite and often the will try things to explore how the product works and can go down paths the programmer never dreamed of.

20 hours ago, Dottleshead said:

For now.

LITERALLY my POINT!!

We're in the infancy of consumer AI and we're calling the game as a loss for it??? WTF??

Use AI today vs AI six months or a year ago, and you see FAR greater improvement than you do in your or my skills in that period. There's a progression to something new like this stuff, and it is getting better - way better.

20 hours ago, Dottleshead said:

Then I think I agree with RE.

DARN TOOTIN'!

20 hours ago, Dottleshead said:

And honestly I don't GAF.

And THIS is the real nugget of truth here! We got ours. I'm ready to roll off into the sunset counting my blessings I escaped it all by a hair.

3 hours ago, Razors Edge said:

And THIS is the real nugget of truth here! We got ours. I'm ready to roll off into the sunset counting my blessings I escaped it all by a hair.

The timing should be really good. They can lay me off or they can fire me. I can reach an uncomfortable point and I can call it a career. Lot of options. But the one thing that is not an option is going back or seeking work in the same field after any of those outcomes. I'm not going back even if I don't have enough cashola. I'll change my lifestyle first. But for now I'm milking that tit.