By far the best way to teach programming is the first programming language taught to me when I was 4 in 1983 -- LOGO. Extend it with the features that you want your salepersons to know -- function calls like drawCircle() would be easy. It's very straight-forward, and the vast majority of concepts can be boiled down into something suitable for the turtle-friend.
I do bug repair for a year, minor tweaks and cosmetic changes for a year, reasonable functionality adjustments for six months, and non-major aesthetic changes for six months. But it's different in my world because I don't go through a spec first at all. I basically do that effort after the build instead of before.
In your case, I'd say that two months isn't too short a period for real world testing, but it definitely requires the client to actually do their real-world testing directly. If you don't intend to pressure them into doing it immediately, but to allow it to come up naturally instead (which I'm not sure is any better, and I can see how it's worse) I'd say that up to four months is perfectly reasonable for a client paying your full price.
Also as someone who sits in a chair programming all day, I too have wondered. But my wonder was quite brief. I discovered the other side of things.
I can easily, right now in five seconds, convert my desk into a standing desk. My monitor is a large 30" that telescopes, so it can orient appropriate for standing. Raising the keyboard and mouse, and grabbing a touch screen or air mouse is just as easy. But standing desks are great for 5 minutes of work. They royally suck for 5 hours of programming. After about 30 minutes, my balance becomes my only focus. Turns out that keeping my fingers locked onto a keyboard position is the opposite of balancing. So it just doesn't work for very long.
But about ten years ago, I solved the problem the other way. I don't sit either. I bought a real "chair". It's a Global Concorde Executive 24HR chair. Technically, it's designed for a security guard to be 350 lbs and to sit in it 24/7/52. But in practice, it's a fantastic programming chair when you orient it the way I have.
I don't sit on my buttocks. I sit on my back. The chair is tilted all the way back, and an ottoman holds up my legs. The result is that I'm reclining in a V-like shape, so most of my weight is actually on my back, and the front of my neck holding my head perfectly vertical. I'm 5'8" and 195 lbs, by the way. Parts of me are well toned, my limbs especially. Others not so much.
So in the end, as long as my wrists, back, shoulders, and neck feel perfect, that's the most I can hope for ergonomically from my workstation. Also the TrulyErgonomic keyboard, blank, in dvorak mode is awesome.
Here's hoping sporty driving is exercise enough for the rest of me. Yeah, I know, more sitting. What can I say? I've got a great set of cheeks under me.
...how it's easier to hold up my phone and press a few buttons while someone else holds up their phone and presses a few buttons than it is to hand cash to someone. Tell me again, and tell me really really loudly.
this whole cyber-bullying is nonsense. It needn't ever have been anything more than standard slander and libel laws which already existed.
I certainly can't say that I made it through my childhood without being bullied. But minus the actual bruises, I'd never suggest that I'd be better off without the bullies.
Quite frankly, the amount of adult insulting I've received from family, friends, and clients for having spent a real amount of money on trees is for more offensive than anything from my school days. Probably because I've long since learned to ignore other people's opinions.
It would seem that argument is long dead. Looks like they were right to fear VCRs. It may have taken quite a while, but a this point, the VCR certainly did lead to DVR, PVR, and ad skipping in general.
nah, those roads aren't main roads. they are west virginia beautiful mountain roads, long and winding. they are a tourist attraction for car clubs, certainly. but that's about it.
don't get me wrong, ask me what I want, and I want a road. but it's certainly not worth the expense. You've got some roads that go six thousand feet up a mountain, to reveal nothing more than a look-out point and an out-house (seriously). Must have cost a nice ten million dollars to build the road. I'm just saying that money could have been spent elsewhere.
Don't get me wrong, I absolutely love driving your mountain roads that go from nowhere to nowhere and have absolutely zero traffic for hundreds of kilometres. Certainly a beautiful way to waste money. This seems like the very same thing.
Mobile devices such as these have been around for about 5 years. That makes them new technology, especially in government circles. What's going to happen after the 90 days? Will the next order be to improve the sites to support the next big mobile browser? Oh wait, that's what this is -- wasn't it just two years ago that he ordered everybody to make their services available online in the first place?
Government's always been required to make things available to the widest audience. If everyone could access the government services from a desktop, that'd suffice. It needn't be better than functional. You don't need to pay your taxes from your shitty smart phone -- especially because 10 years from now your smart phone won't be so shitty.
accidents would go down, intentionals would go up. It's very easy to make a driverless car screw up, remotely.
you still wouldn't get work done during the commune for all of the reasons that passengers don't today. that's just stupid. if you can do your work from a safe car seat, then your work isn't very important or doesn't require much focus in the first place.
the big change would be something way different -- parking lots. why park nearby when I can send my empty car back home for the day. similarly, why have a car of my own when the car isn't forced to stay with me. car "pooling" can actually be exactly what it says, on a city-wide level. pay a subscription, have access to a car any time. pay more, get a better car. the taxi industry in new york city would change drastically.
but most importantly, you'd lose out on all the fun of driving, if, like me, you enjoy driving.
but laws that do get created, tend to last way longer than they ever should. and laws that should have been created take way longer to create than they ever should.
so the pace at which government gets things done is slow. making 150 years a drop in the bucket.
whereas science is now growing crazy fast, and in 150 years, the number of new materials, techniques, processes, careers, and opportunities will change at least a dozen times, making 150 years a very long time in terms of change.
Oh hell yeah! I meant it as an example of one of the ways that we could go about proactively thinking about managing these scary things, as opposed to concealing the research.
Each of those four items are the potential subject of nightmares and downfalls, but each and every one of them is a guarantee -- all eight.
Imagine the year 2150. Distant to any human life, not at all distant to government, mediocre to construction (some city construction projects take 70 years), and eons to technology.
In your 2150, can we spot genetic defects before birth? Of course. Can we select babies for the life that we want? As in can I choose the embryo with athletic skills over the embryo with mathematic skils? I'd sure hope so. It sounds dangerous today, but it's only dangerous in advance, like everything. By the time it's ubiquitous, it's just another form of choosing your child's academic goals. It just starts even earlier.
Same goes for the other six in your 2150. I'd sure as hell hope that we can read minds to some extent by then. But just like the polygraph didn't destroy interrogations, and the mouse didn't destroy the keyboard, and television didn't kill radio, and the plane didn't kill the car, it won't be the only form of communication.
As for police states reading minds, that's the ethical equivalent of humane execution. It's already a police state, it's already killing people, I'm not worried about the mind reading.
Geoengineering is absolutely required in order to live anywhere but terrestrial land. Period. So it's guaranteed to happen. And it'll happen quite suddenly the day before it's required. And by the time it can be used to "undermind the political will to fight it" it'll be so easy to do that it'll be a part of normal construction.
Nuclear weapons don't kill people. People's mistakes kill people. But people don't kill asteroids. Nuclear weapons kill asteroids. That's another period, by the way.
I like how bird flu wasn't one of the top four, having inspired the thing in the first place. But that's the same concept. Of course we're going to have a major outbreak of something. We've had it before. Everyone's so worried that this time, with common means of global transportation, it'll be much worse. I think that they forget one thing. In probably under an hour, every airport and every border can instantly have screeners for whatever the current outbreak is. We have TSA and border and customs security everywhere nowadays. It'd be easy to suddenly, and globally, halt anyone displaying symptoms, or quickly test everyone as a part of transportation procedures.
My point is that, as a civilization, we can't not have those things. Being scared of the research in advance is stupid. Focus on being scared of the initially flawed execution of that research. Work on that while the research is underway. We have M.A.D. for nuclear weapons. That's already worked a few times. It's dumb, but it worked. I'm stunned, but it worked. That's the sort of thing that we need for the rest of them. A Nash equillibrium for each one.
welcome to inspiration, the gloriful wealth of optimism. that would be the first third of the scientific method from elementary school. Have a question, think about it, take a guess, and start exploring to gather evidence. From there your form a hypothesis and test it. We're currently stuck on that whole evidence thing. And it'll take a while.
So why is this a surprise to anyone? That we haven't found evidence of earth-like life yet? They do realize that things speed up quite quickly if we ever do.
I'm sure they just don't want to risk the enterprise market on day one. See how the device works in the consumer world first. the service pack will add more features if the general product works out, maybe that'll include enterprise features.
and, by the way, arm sucks. that's the definition of arm: for lower requirements. get the 86 tablet, and you can have whatever you want right away.
With absolutely zero social networking outside of these posts on slashdot, I'll love that feature if I can control it. But I've long been requesting the ability to filter search results based on the reputation of the source. Sometimes I'm searching the equivalent of "what would the general population think is the...", and wikipedia results are great. Other times, I need a real medical journal or newspaper result. And other times still I'm searching on behalf of a seven year old, and want the lowest level of result imaginable. Give me a slider, and I'll be happy.
the only thing that was wrong here is that the passengers weren't wearing a seatbelt when they were told to be wearing one.
there is absolutely nothing wrong with an experienced pilot choosing to take a sudden action to a perceived immediate threat. the entire point is to not think it through, because there may not be enough time.
to suddenly drop the aircraft by 400 feet, with thousands of feet to go isn't dangerous (functioning aircraft presumed). It's even within normal altitudes bands.
I don't think anyone would have preferred no action taken to a perceived threat.
so, no seatbelt, no safety. big surprise. I hope those were mostly u.s. and zurich passengers, I don't want to believe that my fellow canadians would avoid wearing seatbelts just for fun.
Right, so you can't unit test for correct but undesired results. Because the only thing that makes them undesired is the scenario at large, which has nothing to do with the input parameters to that function, nor anywhere near it.
So what unit test are you going to write? There's nothing to test. Yes when the input is six, the object will turn left. Yes, that's correct. But the input shouldn't have been six. Yet the previous function output six was also correct. It saw a six, and output six.
But it turns out that your entire design, while 100% correct, bet on seeing a six resulting in turning left. And that wasn't correct. It wasn't correct because when that number's a six, a different number needs to be checked, and if that number is an eight, the object should turn right. It's not an exception, it's not a boundary case, it's knowledge of the world to make a real-world decision about human beings. And it's totally something that you didn't miss, it's just something that wasn't originally known at all.
But the result is that your software makes the wrong decision, the wrong choice. and no unit testing can ever find that. because it was a choice. and you can never validate a choice, you can only co-roberate it. That requires making it again differently. And since you didn't know the scenario before, you won't know it again.
As I've said, you've never written functions that depend on the insides of other functions.
I don't at all mean your string library. I mean one business-logic function affecting another business-logic function. For example, write a neural network algorithm, and process different node types using different functions. Notice really quickly that the internal actions of one function doesn't break the other function but drastically changes what it does. It's a correct result, but an undesired result.
That's what makes the unit tests useless. When the result is totally correct, but completely undesired.
See, that's how clients react. And since you didn't get the money ahead of time -- they wouldn't give it to you -- you just wasted all of your development time, and your sales effort time, and you got nothing. That's a hell of an opportunity cost.
Oh, by the way, congrats, you still need to pay your employees.
Oh, and you also won't get that client again in the future.
I think you're missing the idea that the project and it's requirements and the ways in which features work drastically changes from one day to the next. Think about it as an entirely different project, that you won't be paid again for. Trying to fabricate a reduced example is rough here, but imagine buliding a product catalogue web-site for a grocery store, with foods and a set of tax laws and organizational structures and delivery.
Six months later it changes to a furniture store, with photography, different tax laws, distributors, fabric options, colour options, show rooms and delivery.
It's the same client and the same project because the client isn't actually either of those. The client manages delivery for all sorts of businesses. Of course, the client didn't plan to have this web-site do both in the beginning. And not only were you not told that this client does more than just food delivery, but you'd have lost the sale had you even suggested it -- because that wasn't what the client was looking for originally, and so you were competing with much lower offers and ideas.
Oh, and now the site actually needs to do both, because the client does both. And the food site was so good that they changed their entire business around to flex that advantage, and now they want the same for their furniture delivery.
So I hope you didn't base your design on the tax laws, or how many pictures per product, or how many types of picture per product, or whether or not the product has colour options, or fabric options, or size options, or how they get grouped and categorized, or how they're stored in the database, or administered by the client's staff, or anything else, because now it's not only all changed, but it's going to change again.
Of course I based my design on most of those things -- especially the whole options and grouping things. The difference is that I can literally shift around code and change it all very quickly, with the client on the phone. I can then charge them what they expect -- because they don't see the difference in the two businesses, at their level, it's the same business. So whereas someone else would quote them a number and a timeframe that would result in the client eventually not going ahead with it, I get the go-ahead over night. Even if the dollar amounts are the same, I can get the go-ahead without a quote, without a time-frame, and without the client shopping it around. I can also show them what it might look like on-the-fly.
That's the business that I'm in. It's for the "delivery service" company, that treats food and furniture the same way, even though I need to treat them differently.
I don't write doom engines. you're trying to attribute my words to every single project in the world? You think that's sensible? My words and my opinions and my experience are valid for my world, the projects that I do -- oh and since I get to hand pick the projects that I do, they all fit, every time.
You might want to try taking other people's advice for what it is -- their experience. So sorry that you'll need to adapt it to your environment. If you want someone else's experience to come with an API, you'll need to pay them for the consulting report first.
Oh, and it wouldn't be difficult to conceive of a toolset allowing you to walk through a doom level, and when staring at the wall, drag a texture to it. Or when picking up a gun, changing which gun it is. Or when an enemy walks left, adjusting their AI to walk right instead. Rewind ten seconds and try again.
Maybe you can't think of ways to do anything. Others can think of ways to do some stuff. In my world, for my projects, I've designed a language that lets me live test and live develop everything. I had to invent a lot up front.
Slashdot Mirror
By far the best way to teach programming is the first programming language taught to me when I was 4 in 1983 -- LOGO. Extend it with the features that you want your salepersons to know -- function calls like drawCircle() would be easy. It's very straight-forward, and the vast majority of concepts can be boiled down into something suitable for the turtle-friend.
I do bug repair for a year, minor tweaks and cosmetic changes for a year, reasonable functionality adjustments for six months, and non-major aesthetic changes for six months. But it's different in my world because I don't go through a spec first at all. I basically do that effort after the build instead of before.
In your case, I'd say that two months isn't too short a period for real world testing, but it definitely requires the client to actually do their real-world testing directly. If you don't intend to pressure them into doing it immediately, but to allow it to come up naturally instead (which I'm not sure is any better, and I can see how it's worse) I'd say that up to four months is perfectly reasonable for a client paying your full price.
Also as someone who sits in a chair programming all day, I too have wondered. But my wonder was quite brief. I discovered the other side of things.
I can easily, right now in five seconds, convert my desk into a standing desk. My monitor is a large 30" that telescopes, so it can orient appropriate for standing. Raising the keyboard and mouse, and grabbing a touch screen or air mouse is just as easy. But standing desks are great for 5 minutes of work. They royally suck for 5 hours of programming. After about 30 minutes, my balance becomes my only focus. Turns out that keeping my fingers locked onto a keyboard position is the opposite of balancing. So it just doesn't work for very long.
But about ten years ago, I solved the problem the other way. I don't sit either. I bought a real "chair". It's a Global Concorde Executive 24HR chair. Technically, it's designed for a security guard to be 350 lbs and to sit in it 24/7/52. But in practice, it's a fantastic programming chair when you orient it the way I have.
I don't sit on my buttocks. I sit on my back. The chair is tilted all the way back, and an ottoman holds up my legs. The result is that I'm reclining in a V-like shape, so most of my weight is actually on my back, and the front of my neck holding my head perfectly vertical. I'm 5'8" and 195 lbs, by the way. Parts of me are well toned, my limbs especially. Others not so much.
So in the end, as long as my wrists, back, shoulders, and neck feel perfect, that's the most I can hope for ergonomically from my workstation. Also the TrulyErgonomic keyboard, blank, in dvorak mode is awesome.
Here's hoping sporty driving is exercise enough for the rest of me. Yeah, I know, more sitting. What can I say? I've got a great set of cheeks under me.
...how it's easier to hold up my phone and press a few buttons while someone else holds up their phone and presses a few buttons than it is to hand cash to someone. Tell me again, and tell me really really loudly.
this whole cyber-bullying is nonsense. It needn't ever have been anything more than standard slander and libel laws which already existed.
I certainly can't say that I made it through my childhood without being bullied. But minus the actual bruises, I'd never suggest that I'd be better off without the bullies.
Quite frankly, the amount of adult insulting I've received from family, friends, and clients for having spent a real amount of money on trees is for more offensive than anything from my school days. Probably because I've long since learned to ignore other people's opinions.
Guess where I learned that lesson. Guess when.
It would seem that argument is long dead. Looks like they were right to fear VCRs. It may have taken quite a while, but a this point, the VCR certainly did lead to DVR, PVR, and ad skipping in general.
nah, those roads aren't main roads. they are west virginia beautiful mountain roads, long and winding. they are a tourist attraction for car clubs, certainly. but that's about it.
don't get me wrong, ask me what I want, and I want a road. but it's certainly not worth the expense. You've got some roads that go six thousand feet up a mountain, to reveal nothing more than a look-out point and an out-house (seriously). Must have cost a nice ten million dollars to build the road. I'm just saying that money could have been spent elsewhere.
Don't get me wrong, I absolutely love driving your mountain roads that go from nowhere to nowhere and have absolutely zero traffic for hundreds of kilometres. Certainly a beautiful way to waste money. This seems like the very same thing.
Mobile devices such as these have been around for about 5 years. That makes them new technology, especially in government circles. What's going to happen after the 90 days? Will the next order be to improve the sites to support the next big mobile browser? Oh wait, that's what this is -- wasn't it just two years ago that he ordered everybody to make their services available online in the first place?
Government's always been required to make things available to the widest audience. If everyone could access the government services from a desktop, that'd suffice. It needn't be better than functional. You don't need to pay your taxes from your shitty smart phone -- especially because 10 years from now your smart phone won't be so shitty.
or you could wake up on the way to work. or you could try desperately not to vomit -- being nautious as a passenger.
accidents would go down, intentionals would go up. It's very easy to make a driverless car screw up, remotely.
you still wouldn't get work done during the commune for all of the reasons that passengers don't today. that's just stupid. if you can do your work from a safe car seat, then your work isn't very important or doesn't require much focus in the first place.
the big change would be something way different -- parking lots. why park nearby when I can send my empty car back home for the day. similarly, why have a car of my own when the car isn't forced to stay with me. car "pooling" can actually be exactly what it says, on a city-wide level. pay a subscription, have access to a car any time. pay more, get a better car. the taxi industry in new york city would change drastically.
but most importantly, you'd lose out on all the fun of driving, if, like me, you enjoy driving.
It was someone's job to validate the resume. That person didn't do the job properly. So that person loses their job. That makes perfect sense.
As for the person having been hired with a false resume, that person may still be a good candidate. Could be the right person for the job.
The first person, having done the hiring, didn't make the mistake of hiring the wrong person. Made the mistake of not knowing what was being hired.
Why do I sound as though english isn't my first language? It's been a stressful week.
hence government taking a long time to do things. hence 150 years is short for them. and really, this isn't the point.
but laws that do get created, tend to last way longer than they ever should. and laws that should have been created take way longer to create than they ever should.
so the pace at which government gets things done is slow. making 150 years a drop in the bucket.
whereas science is now growing crazy fast, and in 150 years, the number of new materials, techniques, processes, careers, and opportunities will change at least a dozen times, making 150 years a very long time in terms of change.
Oh hell yeah! I meant it as an example of one of the ways that we could go about proactively thinking about managing these scary things, as opposed to concealing the research.
Each of those four items are the potential subject of nightmares and downfalls, but each and every one of them is a guarantee -- all eight.
Imagine the year 2150. Distant to any human life, not at all distant to government, mediocre to construction (some city construction projects take 70 years), and eons to technology.
In your 2150, can we spot genetic defects before birth? Of course. Can we select babies for the life that we want? As in can I choose the embryo with athletic skills over the embryo with mathematic skils? I'd sure hope so. It sounds dangerous today, but it's only dangerous in advance, like everything. By the time it's ubiquitous, it's just another form of choosing your child's academic goals. It just starts even earlier.
Same goes for the other six in your 2150. I'd sure as hell hope that we can read minds to some extent by then. But just like the polygraph didn't destroy interrogations, and the mouse didn't destroy the keyboard, and television didn't kill radio, and the plane didn't kill the car, it won't be the only form of communication.
As for police states reading minds, that's the ethical equivalent of humane execution. It's already a police state, it's already killing people, I'm not worried about the mind reading.
Geoengineering is absolutely required in order to live anywhere but terrestrial land. Period. So it's guaranteed to happen. And it'll happen quite suddenly the day before it's required. And by the time it can be used to "undermind the political will to fight it" it'll be so easy to do that it'll be a part of normal construction.
Nuclear weapons don't kill people. People's mistakes kill people. But people don't kill asteroids. Nuclear weapons kill asteroids. That's another period, by the way.
I like how bird flu wasn't one of the top four, having inspired the thing in the first place. But that's the same concept. Of course we're going to have a major outbreak of something. We've had it before. Everyone's so worried that this time, with common means of global transportation, it'll be much worse. I think that they forget one thing. In probably under an hour, every airport and every border can instantly have screeners for whatever the current outbreak is. We have TSA and border and customs security everywhere nowadays. It'd be easy to suddenly, and globally, halt anyone displaying symptoms, or quickly test everyone as a part of transportation procedures.
My point is that, as a civilization, we can't not have those things. Being scared of the research in advance is stupid. Focus on being scared of the initially flawed execution of that research. Work on that while the research is underway. We have M.A.D. for nuclear weapons. That's already worked a few times. It's dumb, but it worked. I'm stunned, but it worked. That's the sort of thing that we need for the rest of them. A Nash equillibrium for each one.
welcome to inspiration, the gloriful wealth of optimism. that would be the first third of the scientific method from elementary school. Have a question, think about it, take a guess, and start exploring to gather evidence. From there your form a hypothesis and test it. We're currently stuck on that whole evidence thing. And it'll take a while.
So why is this a surprise to anyone? That we haven't found evidence of earth-like life yet? They do realize that things speed up quite quickly if we ever do.
oh look, a reason to avoid it. I was looking for one.
I'm sure they just don't want to risk the enterprise market on day one. See how the device works in the consumer world first. the service pack will add more features if the general product works out, maybe that'll include enterprise features.
and, by the way, arm sucks. that's the definition of arm: for lower requirements. get the 86 tablet, and you can have whatever you want right away.
With absolutely zero social networking outside of these posts on slashdot, I'll love that feature if I can control it. But I've long been requesting the ability to filter search results based on the reputation of the source. Sometimes I'm searching the equivalent of "what would the general population think is the...", and wikipedia results are great. Other times, I need a real medical journal or newspaper result. And other times still I'm searching on behalf of a seven year old, and want the lowest level of result imaginable. Give me a slider, and I'll be happy.
the only thing that was wrong here is that the passengers weren't wearing a seatbelt when they were told to be wearing one.
there is absolutely nothing wrong with an experienced pilot choosing to take a sudden action to a perceived immediate threat. the entire point is to not think it through, because there may not be enough time.
to suddenly drop the aircraft by 400 feet, with thousands of feet to go isn't dangerous (functioning aircraft presumed). It's even within normal altitudes bands.
I don't think anyone would have preferred no action taken to a perceived threat.
so, no seatbelt, no safety. big surprise. I hope those were mostly u.s. and zurich passengers, I don't want to believe that my fellow canadians would avoid wearing seatbelts just for fun.
Right, so you can't unit test for correct but undesired results. Because the only thing that makes them undesired is the scenario at large, which has nothing to do with the input parameters to that function, nor anywhere near it.
So what unit test are you going to write? There's nothing to test. Yes when the input is six, the object will turn left. Yes, that's correct. But the input shouldn't have been six. Yet the previous function output six was also correct. It saw a six, and output six.
But it turns out that your entire design, while 100% correct, bet on seeing a six resulting in turning left. And that wasn't correct. It wasn't correct because when that number's a six, a different number needs to be checked, and if that number is an eight, the object should turn right. It's not an exception, it's not a boundary case, it's knowledge of the world to make a real-world decision about human beings. And it's totally something that you didn't miss, it's just something that wasn't originally known at all.
But the result is that your software makes the wrong decision, the wrong choice. and no unit testing can ever find that. because it was a choice. and you can never validate a choice, you can only co-roberate it. That requires making it again differently. And since you didn't know the scenario before, you won't know it again.
As I've said, you've never written functions that depend on the insides of other functions.
I don't at all mean your string library. I mean one business-logic function affecting another business-logic function. For example, write a neural network algorithm, and process different node types using different functions. Notice really quickly that the internal actions of one function doesn't break the other function but drastically changes what it does. It's a correct result, but an undesired result.
That's what makes the unit tests useless. When the result is totally correct, but completely undesired.
See, that's how clients react. And since you didn't get the money ahead of time -- they wouldn't give it to you -- you just wasted all of your development time, and your sales effort time, and you got nothing. That's a hell of an opportunity cost.
Oh, by the way, congrats, you still need to pay your employees.
Oh, and you also won't get that client again in the future.
I think you're missing the idea that the project and it's requirements and the ways in which features work drastically changes from one day to the next. Think about it as an entirely different project, that you won't be paid again for. Trying to fabricate a reduced example is rough here, but imagine buliding a product catalogue web-site for a grocery store, with foods and a set of tax laws and organizational structures and delivery.
Six months later it changes to a furniture store, with photography, different tax laws, distributors, fabric options, colour options, show rooms and delivery.
It's the same client and the same project because the client isn't actually either of those. The client manages delivery for all sorts of businesses. Of course, the client didn't plan to have this web-site do both in the beginning. And not only were you not told that this client does more than just food delivery, but you'd have lost the sale had you even suggested it -- because that wasn't what the client was looking for originally, and so you were competing with much lower offers and ideas.
Oh, and now the site actually needs to do both, because the client does both. And the food site was so good that they changed their entire business around to flex that advantage, and now they want the same for their furniture delivery.
So I hope you didn't base your design on the tax laws, or how many pictures per product, or how many types of picture per product, or whether or not the product has colour options, or fabric options, or size options, or how they get grouped and categorized, or how they're stored in the database, or administered by the client's staff, or anything else, because now it's not only all changed, but it's going to change again.
Of course I based my design on most of those things -- especially the whole options and grouping things. The difference is that I can literally shift around code and change it all very quickly, with the client on the phone. I can then charge them what they expect -- because they don't see the difference in the two businesses, at their level, it's the same business. So whereas someone else would quote them a number and a timeframe that would result in the client eventually not going ahead with it, I get the go-ahead over night. Even if the dollar amounts are the same, I can get the go-ahead without a quote, without a time-frame, and without the client shopping it around. I can also show them what it might look like on-the-fly.
That's the business that I'm in. It's for the "delivery service" company, that treats food and furniture the same way, even though I need to treat them differently.
I don't write doom engines. you're trying to attribute my words to every single project in the world? You think that's sensible? My words and my opinions and my experience are valid for my world, the projects that I do -- oh and since I get to hand pick the projects that I do, they all fit, every time.
You might want to try taking other people's advice for what it is -- their experience. So sorry that you'll need to adapt it to your environment. If you want someone else's experience to come with an API, you'll need to pay them for the consulting report first.
Oh, and it wouldn't be difficult to conceive of a toolset allowing you to walk through a doom level, and when staring at the wall, drag a texture to it. Or when picking up a gun, changing which gun it is. Or when an enemy walks left, adjusting their AI to walk right instead. Rewind ten seconds and try again.
Maybe you can't think of ways to do anything. Others can think of ways to do some stuff. In my world, for my projects, I've designed a language that lets me live test and live develop everything. I had to invent a lot up front.