I was not replying to the point about lowering the bar due to market demand. Though, now that I think about it, if you were a very bright engineer, would you go into a lackluster career in civil that would start out at $45k and only find small increases in salary now that there is little demand for housing, or would you go into software? So clearly there is going to be brain drain. But I was not referring to market economics. See my point above (http://news.ycombinator.com/item?id=4412278).
I will say that my point did not suggest "horror" - as in horrifying you. Leaking sensitive user data is not, in my opinion, on par with a bridge completely structurally failing while traffic is on it. Unless of course the web app is for a bank. Which is the sort of security work that likely does not have the coffee fetching intern designing its protocols. But even then, I think as a society we are clearly willing to have many more cases of ID theft than deaths from bridge failure.
The comparison should be equal: are there bridges that require maintenance more often than is reasonable? Is their lifespan long enough? Are their piers sunk deeply enough to accommodate the scour of the river (we have several bridges here locally that are under restoration efforts because the scour was not adequately considered)? Was the water surface elevation of the design storm of the river accurately predicted to which the answer is realistically always no; there is no such thing as accuracy in backwater analysis - for before one even begins to model the river, one has to assume how MUCH water it could have - and that can vary widely, being a combination of not-always-comprehensive statistical analysis of historical precipitation, stream gauge data, and semi-empirical algorithms - stress on the semi.
Now, see my above comment about how civil plans actually get out the door, and I would say this is not a metaphor. We know from experience that if a "hot new startup" or even a stodgy old site (linkedIn, eHarmony, Last.FM) loses user data, it is not the end of the world, let alone the end of the startup. If a bridge fails and falls, the PE loses his licence. If the bridge fails and requires extra maintenance, the county and the firm lawyer up and the JDs make a lot of money in court.
I will say that my point did not suggest "horror" - as in horrifying you. Leaking sensitive user data is not, in my opinion, on par with a bridge completely structurally failing while traffic is on it. Unless of course the web app is for a bank. Which is the sort of security work that likely does not have the coffee fetching intern designing its protocols. But even then, I think as a society we are clearly willing to have many more cases of ID theft than deaths from bridge failure.
The comparison should be equal: are there bridges that require maintenance more often than is reasonable? Is their lifespan long enough? Are their piers sunk deeply enough to accommodate the scour of the river (we have several bridges here locally that are under restoration efforts because the scour was not adequately considered)? Was the water surface elevation of the design storm of the river accurately predicted to which the answer is realistically always no; there is no such thing as accuracy in backwater analysis - for before one even begins to model the river, one has to assume how MUCH water it could have - and that can vary widely, being a combination of not-always-comprehensive statistical analysis of historical precipitation, stream gauge data, and semi-empirical algorithms - stress on the semi.
Now, see my above comment about how civil plans actually get out the door, and I would say this is not a metaphor. We know from experience that if a "hot new startup" or even a stodgy old site (linkedIn, eHarmony, Last.FM) loses user data, it is not the end of the world, let alone the end of the startup. If a bridge fails and falls, the PE loses his licence. If the bridge fails and requires extra maintenance, the county and the firm lawyer up and the JDs make a lot of money in court.