I believe the main thing that changed is that 10 years ago, "building software" meant you would work with (presumably reasonable) engineers. Nowadays, it means you do web busywork for some newly rich kid chasing a startup lifestyle, who may or may not have any idea about what's easy/difficult possible/impossible.
Also, the whole industry has been strongly commercialized. Before, people would share source code on the internet just for the fun of it. Nowadays, that is a surefire way for someone else to take your source code and sell it as theirs. I mean that's basically what Cloud providers do. They rent out access to open source software.
The insane pressure of money also makes sure that most software nowadays is not build to any reasonable engineering standard, even when it really should be (like Boeing MCAS). Instead, every piece of software nowadays is optimized for the sweet spot between sucking so bad that nobody buys it and not spending a penny more than what is necessary.
The goal of software development has changed, and I think $1 phone apps nicely illustrate the new commercial landscape. It's the bare minimum quality at the bare minimum price.
BTW, on a related note, most bicycles made in 2000 still had a much more stable frame than the 2020 models. The marked switched from costly steel to cheaper aluminum so that you can make $300 supermarket bikes. And obviously, the quality has suffered.
It appears that in general, there's always a race to the lowest possible quality in the hopes of reducing costs, thereby increasing profits. Does anyone have any suggestion how we could reverse this trend in general?
It was always like this except in the hobbyist world, and I think maybe you and some developers are confused because the corporate takeover of the web is relatively recent. When I think of software development in the past I think of Office Space (the movie), not some idyllic times.
Thee $300 supermarket bike (I have one and put over 40 miles on it weekly and it's help up fine for four years) provides access to folks who otherwise can't afford the fancy $1,500 aluminium/carbon machines. Fancy bikes are still getting made and even getting better.
Newly rich kid or VC funded startups distributes money from wealthy people to an every growing software industry. It allows people like me to work in software; something that was previously only available to elite educated people who happened to live in the correct zip code.
tl;dr: The pie has increased in size and become much more inclusive. Yes, there is a lot of low-quality pie out there. But there is also plenty of high-quality pie for those who can afford it. This is good for everyone except the very elite entrenched class.
I'm not sure I deserve the "elitist" label, but yes, I find this development very negative.
While I agree with your point that hobby startups distribute money and enlarge the market, I was trying to point out that people working under management with no experience will probably not learn their craft well. In my opinion, the old apprentice system used for jobs like becoming a carpenter would be quite appropriate for software engineers. But it only works if there are enough master level programmers around to teach everyone else.
As for the pie analogy, I don't share your opinion. When the market moves towards lower quality, that usually makes high quality more expensive or even outright impossible to acquire. Case in point, I'm not aware of any bicycle that is rated for 200kg+ for driver and luggage. Not only are there no cheap bikes at that stability level, but also nothing in the $2000+ premium range. So something that used to be easy to buy for everyday folks 20 years ago is now too expensive even for rich people. And all that only to drive the price of the cheapest bikes down from $500 to $300, which I presume won't make much difference to anyone because a good bike will last you 10 years, so it's <$1 monthly in either case.
Not only will it easily take that weight, it's designed to comfortably carry a passenger and cargo. Or the Yuba Mundo, there are a few other brands. And you can get them in electric, if your thighs aren't made of steel cable.
Yuba Mundo is $1800, which sounds like a lot, but with inflation that's well under a grand at the time when more bikes were heavy and over-engineered. With the nice side effect that they could carry a lot of weight....
Sorry about that, I didn't mean it to come across as a personal attack. I mostly agree with your paragraph about hobby startups. And I secretly lust after a nicer bike and might go for one soon :)
That's fair, but before I spent a lot of time in industry, I thought most tech entrepreneurs were genius engineer/business types.
I was shocked that a LOT of startup founders are well-connected rich kids who have no skills or ability to lead a tech company. I learned that, while saying that you have a huge trust fund isn't very cool, saying that you're a tech entrepreneur CEO is perceived as cool. So the rich family uses their connections to raise a few million (often indirectly) and suddenly the kid is an "entrepreneur", despite lacking any relevant skill sets. It's not like there's any expectation for these companies to make a profit anyway.
Yes, you can avoid these companies, but it's worth warning younger engineers that this is maybe 50% of tech startups. It's more of a problem with NYC based companies than SF based companies (as NYC has more old money and less emphasis on tech companies actually building real tech) but I've seen it in both cities.
This is kind of true. Young tech entrepreneurs are mostly either harvard/stanford/other top college grads, or theyre from wealthy families. But entrepreneurship has always been this way, its not a middle class career path, and arguably, tech has increased the number of middle class/raw talent trying to start their own companies.
If you were to look back at the early 1900s, or even 1980s, its much better today than back then
Ignore the downvotes. Initial votes are misleading, and if the comment is substantial, they'll stabilize to a reasonable value over the following couple hours.
> Cheaper does not eliminate the quality. They can both co-exist and indeed do.
It sometimes does, if it eliminates quality out of the market. Say a quality frame costed $500 before; then you have a new, cheaper technology that can achieve similar performance by cutting out more extreme loads / road conditions; as manufacturers jump at it to save money, suddenly the whole $100-$800 range uses just that, and the quality frame rises to $1500 due to collapsing demand. Can't say whether it's good or bad on the whole, but it annoys the hell out of me in cases where I could afford the quality product but I can't, because nobody is making it anymore.
Open source as an art is inclusive. Or at least as a social model for developing software, has greater potential than startups, which are exclusive to those with financially viable skills
A working culture tuned to movement of a minority of capital “haves” isn’t much different mathematically than a monarchy and his lords and barons, etc.
There’s a gentler temperament, but “work the jobs we will pay for, while deflating your buying power to maintain historical human narratives” isn’t exactly fostering free flow of capital, labor, and ideas.
I think you might be experiencing what it's like to have, er, experience.
Like, I could make the exact same complaint except I'd situate it about 10 years earlier, around the first dot-com explosion.
And somewhat earlier than that, I remember grizzled programmers from the 1980s who hand-tuned their C and assembly and who thought we were flagrantly wasting computer resources on garbage platforms like the web.
(However, bringing it back to the 2010s and beyond, you're right that open source development has been completely co-opted.)
Also, the whole industry has been strongly commercialized. Before, people would share source code on the internet just for the fun of it. Nowadays, that is a surefire way for someone else to take your source code and sell it as theirs. I mean that's basically what Cloud providers do. They rent out access to open source software.
The insane pressure of money also makes sure that most software nowadays is not build to any reasonable engineering standard, even when it really should be (like Boeing MCAS). Instead, every piece of software nowadays is optimized for the sweet spot between sucking so bad that nobody buys it and not spending a penny more than what is necessary.
The goal of software development has changed, and I think $1 phone apps nicely illustrate the new commercial landscape. It's the bare minimum quality at the bare minimum price.
BTW, on a related note, most bicycles made in 2000 still had a much more stable frame than the 2020 models. The marked switched from costly steel to cheaper aluminum so that you can make $300 supermarket bikes. And obviously, the quality has suffered.
It appears that in general, there's always a race to the lowest possible quality in the hopes of reducing costs, thereby increasing profits. Does anyone have any suggestion how we could reverse this trend in general?