> We can write libraries that the next generation will reuse.
We could, but unfortunately we definitely do not do this as much as we could.
> hat's why software development, which "takes so long" according to the OP, still manages to go much faster (in terms of economic growth) than your beloved civil engineering :-)
If you took away the roads society would collapse, instantly. It is getting to the point where if you take away the internet (or even just a substantial fraction of it) for more than a few days that quite possibly the same will happen. That comes with a degree of responsibility that I do not see reflected in the quality of the product of our industry (including quite a few of those libraries).
There are exceptions to this but not many.
Economic growth is great, if we can keep that afloat in the long term. Given the stakes and the degree to which even the largest entities are failing to take responsibility for their products and services that growth might be the precursor to a very harsh lesson somewhere in our near future.
For every crash, data breach, security issue, hack you have to look back and wonder whether the speed of delivery turned out to be worth the risk (and then you have to discount that even further to take into account that what you read about is only a very small fraction of what actually goes wrong...).
> For every crash, data breach, security issue, hack you have to look back and wonder whether the speed of delivery turned out to be worth the risk (and then you have to discount that even further to take into account that what you read about is only a very small fraction of what actually goes wrong...).
Those risks are pure externalities, unfortunately. The answer to the question for the management of pretty much any company is a resounding "yes, it was worth the risk" - they earned truckloads of money over the years before the breach, and by the time shit hits the fan, the management team probably changed twice already.
Nobody cares, because nobody has to - business incentives are actually strongly aligned against caring.
It sounds like you want to hold software companies liable when their products get hacked. Is that really how other industries work? If a terrorist blows up a bridge, should the bridge builder be sued?
> It sounds like you want to hold software companies liable when their products get hacked.
If the hack is directly related to the quality of the product then yes, by all means. And they could (try to) take out some kind of professional insurance against this.
> If a terrorist blows up a bridge, should the bridge builder be sued?
No, but if a manufacturer of a vault leaves a backdoor in their vault then they should be liable (and they most likely are).
Now this is where the analogy breaks down somewhat, in the 'real' world the lock manufacturer is not liable for the damage due to a break-in, but the presence of a high quality lock will reduce your insurance premiums (at least, they do where I live).
So that's how shop owners deal with this, they take out insurance and the insurance company will try to make a stab at how big the risk is that you're going to be the subject of unwanted attention and price their premium accordingly. Any measures you take to make sure that you are not going to burglarized will be taken into account.
Contrast that with a terrorist blowing up your house, the chances are spectacularly small (vanishingly small even) so besides that not being an interesting risk to insure against you likely will not be worried about it either.
On the other hand if your bridge builder will attempt to sell you an explosion resistant bridge then it had better be that (in practice this will simply depend on the size of the explosion and no bridge manufacturer will say that that is acceptable use of their bridge and likely if you sued them the judge would side with them unless they made specific claims about such suitability).
We could, but unfortunately we definitely do not do this as much as we could.
> hat's why software development, which "takes so long" according to the OP, still manages to go much faster (in terms of economic growth) than your beloved civil engineering :-)
If you took away the roads society would collapse, instantly. It is getting to the point where if you take away the internet (or even just a substantial fraction of it) for more than a few days that quite possibly the same will happen. That comes with a degree of responsibility that I do not see reflected in the quality of the product of our industry (including quite a few of those libraries).
There are exceptions to this but not many.
Economic growth is great, if we can keep that afloat in the long term. Given the stakes and the degree to which even the largest entities are failing to take responsibility for their products and services that growth might be the precursor to a very harsh lesson somewhere in our near future.
For every crash, data breach, security issue, hack you have to look back and wonder whether the speed of delivery turned out to be worth the risk (and then you have to discount that even further to take into account that what you read about is only a very small fraction of what actually goes wrong...).