Comment by notauser | Hacker News Reader

Comment by notauser | Hacker News Reader

Parent Full thread

It's not quite that straightforward.

Think about it from the other side of the fence. My experience was with weapon systems but to a certain extent it does apply to roads.

Because of THE RULES only fixed priced bids are acceptable.

As a contractor you are being asked to quote on the complete price for a new product build, even though you haven't designed, tested or built anything like it before. The price you quote needs to include the cost of upgrading a bunch of existing equipment to support your new product. You have no idea what the current state of the existing equipment is.

You have three choices:

- Quote a huge price that fairly covers every possible risk, and get told it's too expensive.

- Quote a small price and caveat the hell out of everything and get blamed when the initial conditions or complexity turns out to have been misunderstood or misrepresented.

- Go bankrupt.

You actually even see the same decision being made by normal employees at work. "Sure boss, I can do it in three weeks, as long as Dell delivers on time" would be an example of option (2) in action.

So, when forced to choose between three honest options with (potentially!) unfavorable outcomes and a dishonest one it's ok to be dishonest?

The contractor has an incentive to identify caveats, the purchaser has an incentive to pin down those caveats before agreeing to the purchase. Where is the dishonesty in that? It is the purpose of the contract.

The dishonesty is in undercutting another party that is willing to work without trickery by using a bunch of paper manipulation to effectively obtain a much higher price than the original bid.

That both parties collude in this makes it no less wrong.

By "a bunch of paper manipulation" you must mean a legally binding contract?

If something is specified in the contract, I find it difficult to see how this can be called "trickery", it is a magician telling you in advance that he may hide the coin up his sleeve.

By 'a bunch of paper manipulation' he probably means some of the practices mentioned elsewhere in this discussion - such as hiring someone whose full-time job is to find loopholes and contradictions within complex contracts, so they can charge more money; or grossly misrepresenting the likely cost of a project, with full intention of raising the price regardless of ultimate conditions. The misrepresentation in the latter case is making it out that a clause applies to exceptional circumstances, when its actual purpose is to pad the contract under any circumstances.

You seem to think that each party should simply expect the other party to try to get as much from them as possible, through whatever misrepresentations and legal strategems they can get away with. Maybe this is normal in the areas where you work - I guess Jacques and I both come from backgrounds where this would be abnormal, and in fact disgusting behavior.

The government also has people who audit contracts looking for places where the contractor has committed a technical breach.

In my experience breaches on either side tend to get netted off fairly amicably. I might swap you a change request that you want for signing off a performance deviation I can't avoid, that kind of thing.

However the contract still serves a purpose - it provides the ground rules for that kind of trade. It does sometimes happen in bad faith (usually when the contract has a long and sticky history already) but most of the time it's just a more formalized version of goodwill. And that formality is necessary when you are talking about millions of dollars per change.

IANAL but "misrepresentating" has a specific meaning in law so that to do so would render a contract void.

I do not "expect the other party to try to get as much from them as possible", I just expect that should one party fulfill the obligations of the contract, that the other party fulfills theirs.

The point of having a contract is that I can do business without invoking any kind of ethics or morality based on a background I presume to share with another party, or my own personal definition of "misrepresentation".

Or at least, I can advise the other party of my definitions and expectations in advance of them entering into the contract of their own free will.

See: http://news.ycombinator.com/item?id=1379829

The sad consequence of people choosing option 2 is that legal contracts are hard to understand. For purpose.

Whole mafia of lawyers earns big bucks thanks to making these contracts even more convulted.

Outsmarting somebody by clever legalese isn't moral in my book.

The truly fundamental flaw is deeper than that. The fundamental flaw is that the bid process is intrinsically built on one party making a future commitment of results. It is impossible to actually commit to future results; there's too many confounding factors, which I really shouldn't have to list to anyone here. This impossibility is the foundation of the system. When you put something impossible at the foundation of the system, the entire rest of the system chokes and bends around it in a most distressing way.

As notauser correctly observed before I could beat him/her to it, this is virtually identical to the way most of our jobs work. Management asks for commitments of dates and times, and the mere fact that this is literally impossible does not stop them. I am not kidding or being metaphorical about that "impossible", I mean it literally. And they do have their reasons, but the mere fact they have reasons does not prevent the entire system from getting horrifically distorted as it builds itself around the impossible primitives, producing the nonsensical and nonfunctional results we are all familiar with.

The really, really, really key insight of Agile is to reduce the amount of impossibility built into the base of the system, and the smart Agile-using engineering team lead will take every advantage of that they can. Unfortunately, Agile really only works when all sides are capable of cooperating; it can survive internal-type rivalry, but where everyone is still in the same company. I can't imagine how to make Agile work in the face of hostile agents, which is what this contract situation is.

If you can give a probability distribution instead of a fixed time, you may be able to make the commitment less pathological. Of course, as time goes on, the standard deviation of your distribution should go down, since more and more is known.

This is a game in the game theory sense. All the other strategies are dominated, except option 2. If all else is equal (and it isn't always), only companies that choose option 2 can survive by winning bids and keep playing.

Reply on news.ycombinator.com