Turning developers into production line workers is precisely what needs to happen. The way we produce code is so ruinously expensive that nearly everyone in this industry accepts the tradeoff between developer velocity and product quality. The only way out of this is to automate programming so thoroughly that only the creative necessities remain.
> The way we produce code is so ruinously expensive that nearly everyone in this industry accepts the tradeoff between developer velocity and product quality
I've seen too much over engineered crap to believe this. In many case the over engineering was an attempt at making us more like production line workers in the first place with the "architect" breaking things down into cookie cutter steps.
There is also the problem with user expectations, there are several "production line" like environments available but no one likes the trade off of inflexibility.
> Turning developers into production line workers is precisely what needs to happen
I actually agree with this sentiment, although many others to hate on you on this thread.
And it will happen, just like what happened to labor workers who used to make a good living working for Ford at production line. You can already see plenty of evidence when you see how all developers are forced to use package managers and follow "coding guidelines", and so forth. These efforts will all lead to a more and more structured way of building things, which then can be replaced by machines.
And these naive and passive programmers who thought they were set for life by taking a coding bootcamp and getting a programming job will suddenly find that it's not as lucrative anymore.
The point I was trying to make was: Don't be these people who live a passive life and do what others say, thinking just because programmers make a lot of money, all you gotta do is "become a good programmer" and you'll be set for life. You should have a meaning in life. Otherwise you'll find yourself in the similar situation as these Ford production line workers.
The only other science that's cheaper is math. Now I know plenty of human endeavours that bring absolutely nothing of value, and only serve to concentrate wealth, and cost everyone worldwide billions of dollars.
Are you aware of the military budget of the greatest country on Earth?
Ruinously expensive - what an apt description, just not for computer science.
In comparison to virtually any other design work. The most expensive commercial buildings ever constructed are in the neighborhood of 5-15 billion USD. Developing a Linux distribution from scratch would probably exceed $10 billion in developer costs [1]. Just think about that. Linux, not even the largest codebase known, compares with the most expensive civil engineering projects in history.
That sounds extraordinarily cheap, actually. The most visited mall in the world has something like 80M visitors/year. Linux has literally billions of daily users, and it powers everything from their phone to the public transport network they use.
Software development scales poorly. This was noted a long time ago, an I’ve not seen anything that fundamentally falsifies the Fred Brooks view. So to make the expected cost of software lower, shouldn’t we be looking to smaller-scale, more individualistic ways of working?
