OT: Bug versus Feature

Jourin · #1 February 7th, 2002, 02:12 AM

Whether something is a bug or a feature is based off the rule of unintended consequences. You design something that results in an unintended consequence. If that consequence is positive, it is a feature. However, if that consequence is negative, it is a bug.

Using a specific example: Drones were designed with no "move to" capability resulting in the unintended consequence that drones can not do recon. Since I think that the majority of people believe this consequence is negative, this consequence is a bug.

Yes, drones were designed with no "move to" capability, but that does not mean the design didn't cause a bug.

Bugs can be caused by
Sloppy keystroke entry (syntax)
Poor programming design (logic)
Poor concept design (unintended consequences)

Resident Alien 2 · #2 February 7th, 2002, 04:38 AM

quote:
Originally posted by Jourin:
Whether something is a bug or a feature is based off the rule of unintended consequences. You design something that results in an unintended consequence. If that consequence is positive, it is a feature. However, if that consequence is negative, it is a bug.

Using a specific example: Drones were designed with no "move to" capability resulting in the unintended consequence that drones can not do recon. Since I think that the majority of people believe this consequence is negative, this consequence is a bug.

Yes, drones were designed with no "move to" capability, but that does not mean the design didn't cause a bug.

Bugs can be caused by
Sloppy keystroke entry (syntax)
Poor programming design (logic)
Poor concept design (unintended consequences)

I would have to disagree with calling current Drone functionality a bug. Disagreement about a design decision doesn't make it a bug.

The designers vision of Drones was attack only, he designed it that way, he coded it that way, it works that way.

Unless there's a user requirements specification we signed off on to say that drones should have "go to" capability then it's not a bug in my usage of the phrase "software bug".

Phoenix-D · #3 February 7th, 2002, 05:53 AM

"unintended consequence that drones can not do recon."

Who says it was unintended?

In any case, that will be my sole shot into this thread. I really doubt this will decide anything at all

If you really want to get something done, email MM and *politely* *ask* about adding the move-to function to drones.

Note the politely.

Phoenix-D

Mephisto · #4 February 7th, 2002, 10:56 AM

First: The definition of �bug� is not correct IMHO. It does not depend on the consequences if something is a bug or not. Else the graphics of SE4 are a bug. Or is there anyone out there that will really argue that the graphics of SE4 can compete with state-of-the-art graphic? I know, this was done deliberately but your definition only ask for the consequences and these are negative � it is not as good looking as it could.

Second: Drones were designed not to recon. Aaron KNOWS about it. It was done as it is nevertheless with a seeing eye.

Third: Ask MM to change it if you don�t like it.

dmm · #5 February 7th, 2002, 06:27 PM

Jourin,

I totally agree with you that one should at least be able to mod drones to have "move to" ability. And I think you are right to be disappointed.

However, I don't think "bug" is the proper term. I would call it a "design mistake" or "error in judgment" or "concept problem." I think MM made a poor decision, but I don't think they cheated anyone.

Your definition of "bug" is too broad. Someone might say that guns without integral trigger locks have a "bug" because a child might shoot it with bad consequences, contrary to the designer's intention. And if the designer fixed that, then someone else might cry "bug" because a child could still get hold of the key. And if the designer then made the gun so it only could be operated with adult-sized hands, then someone else might cry "bug" because now midgets couldn't use it but teenage boys still could. So then the designer fixes the "bug" by putting fingerprint ID into the gun. Finally someone sues the gun maker because a properly registered and IDed adult gun owner kills his neighbor's yip-yip dog with it at 2 am. Clearly the gun was defective since it produced consequences that someone didn't like.

Is McDonald's coffee "buggy" because it is served piping hot, and some idiot might try to drive with it between her legs?

You need to remember that Aaron is a normal human being with professional pride. He is not likely to respond well to accusations of cheating people, and may very well become obstinate about the issue. Hence, people are getting on your case, telling you to cool it, shut up, act nice, etc. I think that with time, if enough of us ask politely and often, that Aaron will give you what you want.

Gryphin · #6 February 7th, 2002, 07:21 PM

I have done 3 years of telephone support for Microsoft. I presently do desktop, (in person support) for computer Users.
The following definition of “Bug” is the one accepted by the vast majority of computer professionals and Users in general.

Bug: A feature within software, firmware, or hardware that does not function as the designer of that product intended.

I can state that it in no way has anything to do with weather or not the user of that product likes the feature.

I agree with: "Make a polite request to MM".

Jourin · #7 February 7th, 2002, 09:04 PM

I am not saying that MM cheated me, and I really do not want to debate whether drones have a bug or not. I was trying to define the difference between a bug and a feature. I was trying to educate people so they would not make false or inflammatory statements, nor assume that I was making false or inflammatory statements.

I am an application and processes designer. Although I started as a program writer and work closely with them, I no longer actually write the code. I am like an architect who designs the house that the builders build, or in this case an application or process that the coders code.

If I design something that is programmed and coded perfectly, but the result has a negative impact on the process, then I have designed a bug into the system. Just because the program was coded perfectly per the design doesn't mean the program doesn't have a bug.

If I design a house with a grand cathedral ceiling, that is built exactly to specifications but the result is undo stress upon a weight bearing wall that makes the ceiling unstable; is the house not defective? Does the house not have a "bug?"

The problems with most software applications and computer games are all design related. The application or the game has a bug because of poor design not because it was coded poorly.

The statement that a bug can not exist if the program was coded per design is false, invalid, and incorrect. A bug is a problem whether it is caused by poor coding or poor design. In fact, most bugs are caused by poor design which is why they are so hard to find.

If you don't accept this definition that is your choice, but then you are letting all the designers off the hook and do not complain when you receive poor quality product. I am in no way implying that SEIV is poor quality.

The only time I was not polite was in replying to what I perceived was a flame attack.

geoschmo · #8 February 7th, 2002, 10:21 PM

Jourin,

You can attempt to change the perception if you wish, but you did accuse MM of cheating you when you stated that you had paid for apple pie and received apples instead. How else could that possibly be taken? And most reasonable people would have taken that as an impolite remark. And it was the reason for your being warned, not because of your opinion about the facts of the discussion.

Your analogy of the house is flawed. SEIV gold is not unusable because drones can't "move to". You are simply wanting them to do something they were not intended to do. It really is as simple as that no matter how hard you try to change the terms of the debate.

This is not a result of poor coding. It was a concious and direct decision on the part of the programmer. Given enough input from the customers, MM may infact decide to modify drones to allow them to "move to". We will just have to wait and see.

Things such as the fighter stack bug, and the counter intel bug were bugs. They were a result of mistakes in coding, or the result of unintended consequences.

Drones only fit into that Category if one accepts your definition of the word "drone", which requires by nature that they be usable in those ways. That is circular logic. You cannot prove anything that way.

Geoschmo

[ 07 February 2002: Message edited by: geoschmo ]

PvK · #9 February 8th, 2002, 12:21 AM

Ya, I think Jourin means "design flaw", not what most would call a "bug." Trying to stretch the term "bug" to include flaws "by design" is likely to just cause confusion.

PvK

Mephisto · #10 February 8th, 2002, 12:33 AM

quote:
Originally posted by Jourin:
The only time I was not polite was in replying to what I perceived was a flame attack.

I don't know if you took my post as a flame attack. If so, I would like to apologize. I never intended to flame you. I'm sorry, if I offended you.

The whole "bug" things resolves around which persons view on the working mechanism counts. You say, it is yours/the customers, I would say in the first place the view of the programmer/designer. Of course it would be wise of the programmer to unify his view with the view of the customer...