What Happend to the Dream?

[wrmiller]

I feel your pain. For many years now, there has been an entire industry built by so called "experts" that have tried to come up with the grail to software development, putting out methodologies, processes, evaluation of the art of coding to make it more like a predictable activity, like a production line for some commodity. I've seen methodologies after methodologies coming and then disappearing. Use cases, Booch, now Agile. I'm not against some kind of planning but I'm just amazed by all the people who were claiming having discovered the holy grail of software development, charging huge amount of money for big books and conferences. Pfffft!!

Many years ago I designed a uber precise 6-DOF control system for DARPA using laser interferometers for positional information/accuracy in the millionths of an inch. Used Yourdon & Constantine Structured Design methodologies which allowed us to model real-time events as they happened. The project took over two years, with a team of 8 very senior engineers and myself acting as the lead design engineer. We spent almost 10 months doing the design from the original context diagram, down to each individual leaf. Then we wrote over 5 million lines of C code. We met delivery schedule, and spent the next four months chasing a few bugs. Then we froze the main trunk/code as we simply could not find any more bugs.

At my last job, I had software engineers that couldn't take a mathematical algorithm I would give them and convert that to a few hundred to a few thousand lines of C++ code without spending months and months debugging the code to some basic functionality, all the while espousing the 'greatness' of the current methodologies.

Fortunately, I am now retired and playing games, building custom competition guns, and generally having a good time. But occasionally I get frustrated trying to play a game where the bugs are SO obvious that a junior test engineer working for me would have likely found them prior to general release. But hey, what do I know?

 

[wrmiller]

Wait, you mean sprints aren't going to help us find the grail after all? I hear yeah about the methodologies and so much focus today is on the process rather than actually getting the right product. Company I'm at is also trying to apply Agile to hardware development - let's just say that it's turning out interesting.

Back OT: the dream is likely only still alive in the originator's mind. It certainly hasn't made it to the real world.

Boeing, the aircraft company, tried Agile years ago. To the tune of millions wasted. They eventually threw it, the baby, AND the bathwater out the window.

IF the dream is still alive in the originator's mind, that doesn't mean there is a clear-cut path to a functional software. Years ago, we had the Design (based on the requirements), which was the WHAT the design had to do. The Implementation Model was the HOW. There were many companies that floundered while trying to get from the Design to the Implementation. And in a rush (usually caused by Marketing...) the models were thrown away, and the engineers were just told to Go Do It.

 

[Ralph Vargr]

Boeing, the aircraft company, tried Agile years ago. To the tune of millions wasted. They eventually threw it, the baby, AND the bathwater out the window.

IF the dream is still alive in the originator's mind, that doesn't mean there is a clear-cut path to a functional software. Years ago, we had the Design (based on the requirements), which was the WHAT the design had to do. The Implementation Model was the HOW. There were many companies that floundered while trying to get from the Design to the Implementation. And in a rush (usually caused by Marketing...) the models were thrown away, and the engineers were just told to Go Do It.

I have friends there. The painful memories remain.

No discipline can overrule the 2nd Law of Thermodynamics, or Murphy's Law. The Truth Is Not There. There Can Be Only None. There Is No Spork.

We're hitting hardware limits. They have to go through the 5 Stages, before they admit that they have hit the human talent limits as well.

You Can Have It On Time.

You Can Be Under Budget.

You Can Get It Right.

With software, pick any none. *It* will pick *you*.

 

[sollisb]

Many years ago I designed a uber precise 6-DOF control system for DARPA using laser interferometers for positional information/accuracy in the millionths of an inch. Used Yourdon & Constantine Structured Design methodologies which allowed us to model real-time events as they happened. The project took over two years, with a team of 8 very senior engineers and myself acting as the lead design engineer. We spent almost 10 months doing the design from the original context diagram, down to each individual leaf. Then we wrote over 5 million lines of C code. We met delivery schedule, and spent the next four months chasing a few bugs. Then we froze the main trunk/code as we simply could not find any more bugs.

At my last job, I had software engineers that couldn't take a mathematical algorithm I would give them and convert that to a few hundred to a few thousand lines of C++ code without spending months and months debugging the code to some basic functionality, all the while espousing the 'greatness' of the current methodologies.

Fortunately, I am now retired and playing games, building custom competition guns, and generally having a good time. But occasionally I get frustrated trying to play a game where the bugs are SO obvious that a junior test engineer working for me would have likely found them prior to general release. But hey, what do I know?

This is what irks me so much too.. The level of dismissal and incompetence in release code these days astounds me. And even more worrying is the acceptance at root level by people paying hard cash.

 

[tarragon]

Many years ago I designed a uber precise 6-DOF control system for DARPA using laser interferometers for positional information/accuracy in the millionths of an inch. Used Yourdon & Constantine Structured Design methodologies which allowed us to model real-time events as they happened. The project took over two years, with a team of 8 very senior engineers and myself acting as the lead design engineer. We spent almost 10 months doing the design from the original context diagram, down to each individual leaf. Then we wrote over 5 million lines of C code. We met delivery schedule, and spent the next four months chasing a few bugs. Then we froze the main trunk/code as we simply could not find any more bugs.

At my last job, I had software engineers that couldn't take a mathematical algorithm I would give them and convert that to a few hundred to a few thousand lines of C++ code without spending months and months debugging the code to some basic functionality, all the while espousing the 'greatness' of the current methodologies.

Fortunately, I am now retired and playing games, building custom competition guns, and generally having a good time. But occasionally I get frustrated trying to play a game where the bugs are SO obvious that a junior test engineer working for me would have likely found them prior to general release. But hey, what do I know?

Give broken english turnip collectors turned FDEV interns a break. They are surely doing the best they can, for CVs sake and they are surely passionate. Normal sw/game design engineers went to work on new games or left for companies that care about their released products beyond marketing it.

 

[Goose4291]

https://media.tenor.com/images/0489fb2f025d80cb993ac1e2712682fa/tenor.gif

The dude has made one reasonably good game, botched another and made a whole bunch of broken promises about his latest one.

You obviously have not seen the Wing Commander movie or the giant mess that was made of Freelancer.

To be honest I'd say the entire wing commander game franchise (with the exception of Armada) were fantastic games.

Freelancer is solid for its time and oddly still going due to its modability as an online game.

Im not surprised to see you posting negative stuff at anything thats not the one true space game, seen as youre comments are the first on any star citizen/no mans sky articles on facebook and almost exclusively never positive.

 

[Goose4291]

Do you have a reference for that ?

I dont know about the eight year shareholder thing. But I do know the ten year plan thing is based on a misconstruded Braben quote from an early lave radio episode where they asked him about elite 5 and he said (something along the lines of) that he didnt see an idea of having an elite 5 but instead he hoped development would be continuing on ED 10 years from now.

 

[Nervous Energy]

Same same old

Pvp ruins games unless the game has be designed for it. ED was not... DEAL WITH IT.

FDEV have 10 year plan we are at the end of year 2 and it is still a mess. With zero direction apart from some deluded 10 year plan.

Would I recommend this game to anyone yes only on two point.

If you want to show off and enjoy VR

If you want to play a SPACE game in approx 4-5 years time.

 

[Harry Hates Golf]

Jeepers....All I know is that I like Elite: Dangerous.

 

[KING5TON]

FDEV have 10 year plan we are at the end of year 2 and it is still a mess. With zero direction apart from some deluded 10 year plan.

Not that there really is a 10 year plan but FD consider it to be year 5 not year 2. The 10 year plan that doesn't exist started in 2012 in Kickstarter, not release (and we're nearly 3 years in from release anyway).