Windows 7 on its way :|

[Davros]

Reguarding small hardly noticable bugs that dont impact on the running of the program, I can see your point here and maybe they shouldnt qualify for a refund if the program does what its supposed to do. and it should be classed as working thats not to say its ok to ship with them you should try to get rid of them
but thats not what is happening a hell of a lot of prgrams do ship with huge bugs that definately do impact on the working/enjoyment of the program and that is plain wrong...

 

[Albuquerque]

"what if you didnt know"

Unlike other software companies I stand by my product and i guarantee it to do whatever its supposed to do
If you pay me to write you a program to do A, B + C then my program will do A, B + C guaranteed or your money back
"I didnt know" is not a defence and if it means i take a financial hit then i have no choice but to take it the only other option would be to rip off my customers are you saying thats preferable ?

You're assuming too much...

What if your program doesn't work because another piece of software (that you have nothing to do with and isn't yours) is installed on the machine? How could you have known to test for that?

What if your program doesn't work because another piece of software (that you have nothing to do with and isn't yours) replaced a DLL that your application depends on? How could you have known to test for that?

What if your program doesn't work because a hardware conflict exists even though your application doesn't even interact directly it? How could you have known to test for that? Or better yet, what if a driver update suddenly exposes something that you fat-fingered eons ago, but that worked in 100% of your testing -- but was still wrong?

Here's where it gets good, and you can't just scape-goat out of these:

What if the customer PURPOSELY enters hugely-errant data in hopes of exposing a bug? How could you know the thousands of ways a person could hypothetically do such a thing, and then how would you test against them?

What if someone who doesn't even OWN your software decides it's popular enough to target for remote attacks? What are the millions of ways that someone may try to attack your software, and then how would you test against them?

I want you to answer every single one of these from the viewpoint of your wonder-world where you're going to reimburse the customer 100% of their purchsae price in case of a bug, and we'll see if you're getting it or not.

 

[BRiT]

What if Davros got a clue as to how the real world of software product development worked?

We wouldn't still be having this conversation.

 

[AlexV]

What if Davros got a clue as to how the real world of software product development worked?

We wouldn't still be having this conversation.

Going by his line of reasoning, what are the chances of that?

Davros-products with no faults my arse...there is a cutoff point for bug squashing/fault eliminating in every darned area, not only in software development. There are admissable errors/variations from base value/whatever, because after a point, nailing down every little delta that appears simply is too expensive/not worth it(knowing about diminishing returns values nothing if you choose to discard it so easily). So goodluck with your plan of turning the world into a place where nothing ever comes out from an assembly line, and if something ever comes out, about 55 years after being designed, it's so expensive due to irrelevant bug/fault tracing, and so outdated, that it brings about 0 money for the one making it. Come on man, you must be able to see that you're being dense just for the sake of arguing...can you not?

 

[Davros]

"Come on man, you must be able to see that you're being dense just for the sake of arguing...can you not?"

If you buy something it should work properly - is that so unreasonable ????
you want me to make some consessions ? ok if the bug/fault ect is small/unnoticable doesnt impact on the function or user enjoyment of the product then maybe thats acceptable

but thats not exactly whats is happening is it, programs (escpecially games) are shipping with bugs so big they render the game unplayable (there is no way in hell you'll ever get me to accept that is ok)

my favourite example is lomac it didnt even run at all under 9x not even a menu screen (and before you say it lomac was not a xp only title win 9x was fully supported)
funny thing is one of the guys associated wiith lomac used to post in a forum i visited and a while later he was interviewed and he was anoyed that people complained about it yet they didnt understand what was involved in making a flight sim

 

[Frank]

In Davros' defence: most programs have a very limited scope. For example, I've written a program that is used world-wide, it uses an English user interface regardless, and only works on Windows. And it does work. It meets all the specs and has never crashed so far that I know of.

Is it a bug that the UI uses English in China? Nope. Because the specs state that it does so. Is it a bug that it will skip actions when needed resources are unavailable? No, again, for the same reason.

Requests for change come along regulary. Is it a bug that some nice to have additional features aren't implemented, or that people complain it doesn't do things as they want it to, merely how the manual states they are done? Of course not.

So, yes, programs can certainly be bug-free. Even if they don't meet any random expectation. Because no program does that.

 

[Davros]

thank you frank (or is it digiguru?)