Hacker News Reader

I would add one word to this:

> There was nothing anymore preventing web development from becoming so complicated

The early web was much simpler because of the resource constraints of the 90s and early 2000s. Pages only started serving 10 MB of JS when browsers got to the point where they could chew through 10 MB of JS at a barely acceptable pace.

Also:

> Managers don't really feel like pointing out "the project I'm managing doesn't really deserve a team of developers, you should take them away from me."

As a techie watching corporate politics from the sideline, this was a particularly depressing insight. I used to wonder why big companies require 50 employees to do the work that small companies need 10 employees for. I do not wonder anymore.

Resources are allocated to their constraints. You spend as much money as you have, as much time as you have, you hire as many people as you can afford, and so on.

It may be because people don't think on an absolute time scale but rather a relative one; if you make 1 million in revenue and you only have 3 employees, you might feel like you should hire more. What happens if you start making one billion in revenue?

Read Bullshit Jobs by David Graeber if you haven't already.

Also companies aim to maximize the margins.

If profit is currently $800,000/year and adding 1 software developer at $120,000/year salary (+benefits, hidden costs...) can lead to $800,001 profit, the company would do it.

And that's actually rational behavior. It's also a very simple and very materialistic explanation of why humankind evolves, bit by bit. It works the same way even for discovering new medicine or any kind of technology.

The new thing must be at least a bit better than the old thing. Over time those improvements compound and you end up with this: https://www.youtube.com/watch?v=fPF4fBGNK0U

Also known as "Parkinson's law".

Also the name of the little funny book where you will find the origin of "bikeshedding".

More pleasant and timelesd a book than "bullshit job", IMHO.

What's the full title of the book? I searched on Amazon and there are a few titles containing "Parkinson's Law". Thanks.

https://www.amazon.co.uk/Parkinsons-Law-Pursuit-Progress-Bus...

Yes that one indeed.

It’s because you can do a lot of work manually if you have 40 extra drones. If there’s only 10 of you you need to think about things and automate.

So , is it a result of the VC funded bubble? If that's the case it will not outlast the next bubble burst maybe.

While I don't think the phenomenon would completely vanish without the VC funded bubble, I have to think that it has exacerbated it significantly.

But, you know, I guess we'll find out sometime in the next couple years.

In the early 2000s I worked for a company whose IT infrastructure was primarily Microsoft based. Maybe about 10% of the company had Macs, and we typically had to fight to get them. The Windows boxes required a huge amount of IT support per capita, while the Macs pretty much just worked and the Mac users didn't need much help. Not to mention that the Windows boxes were notoriously malware-prone and needed constant patches and the Macs didn't. I wondered for a long time why the company didn't just switch entirely to Macs. After all, most of the corporate apps were web-based and they worked fine on either platform. It took me a while to figure out exactly what you wrote above: The CIO didn't want a small, lean organization. He wanted a huge organization with a huge budget, because that's what looks good to C-level executives. Hence, Windows was the officially-sanctioned desktop solution.

I really did enjoy my brief times working on embedded devices. Design conversations are quite a bit easier when physics is on your side.

I wish the IoT space weren't such a dumpster fire. I should go back to my plan to run a website on a cluster of nanoPi's.

Good point that lower bandwidth was once a restraint on this.

I love your comment, but I think it does miss out on some important concepts.

I used to think when I first started my company that I could just do every part of it far better than anyone else. I would push on my employees pretty hard and often denigrate their efforts. As the company is continued to grow, I now have 180 full-time employees. The reason I have all these people is that they can do enormously more work than I can and most of them are now better at doing their individual job than I am.

I think your comment visits are tendency as tech people to view ourselves in the best possible light, and to view other people as not as good. We especially tend to view projects we do as big and very complex and project done by others as not impressive.

The question is, are you auditing the final work being done by your "teams" and asking whether it could have been done more efficiently / simply?

> view ourselves in the best possible light, and to view other people as not as good

Remembered me of this cognitive bias: "Overestimating Our Own Importance" [1]

[1] "Psychology of Intelligence Analysis", Chapter 11. Available at:

- https://www.cia.gov/library/center-for-the-study-of-intellig...

The other thing is, programmers love re-inventing things. Its the standard programmer mindset of wanting to work on problems that are one or two levels of difficulty over their currently capability. And so, they fuck shit up, because only someone whose been in the weeds, knows the ins, outs and the corner cases, is capable actually doing a good job. (IMHO, $0.02, YMMV etc)

I'm not sure if I fully agree with your point about developers or managers not wanting to suggest they aren't needed. You're argument relies on the idea that most business needs could be sufficiently met by static html doesn't really fit how businesses works. If a business in an industry has additional functionality on their page that gives them an edge beyond just text on a page, then it is needed. Additionally, when extra functionality is needed the whole industry is going to compete on that. Most businesses don't have enough extra funds to support unnecessary or non-competitive functionality development, so while there might be something better they could be doing with their money, it's almost never the case that what's being developed is so over complicated or over architected that it would be a better financial move to build a static page or site than what they're doing because losing that competitive edge often results in the company, division or product line failing to get enough sales to continue.

You do make something of a good point about nothing preventing web development becoming over complicated, except you seem to have forgotten how much that costs. Unless there is a perceived marketable difference, people will rarely push for it to be developed at all, and given the size of the average development team's backlog that pressure to develop also has to outweigh everything in the backlog as well.

The honest truth is that web development has become so complicated because that's the equilibrium point of the supply of labor to the demand for advanced features. Developing the same features from scratch using static web development methods takes exponentially longer as a function of feature complexity, and the ability to use frameworks and other more complicated tools reduces that down to being merely linear.

Web development isn't complicated because nothing is stopping it from being complicated, it's complicated because consumers want features that they assume should be simple, but are in fact very complicated.

While this is occasionally true, I believe it is most of the time not true. Many, many features (e.g. endless scrolling) get pushed out to consumers who don't want it. By and large, consumers of websites want text and images. The push to put embedded video, lots of ajaxy updating of content, etc. has happened on countless newspaper, magazine, and other websites where all the consumer wants is the words and the images. Many features, like allowing comments, are actively disliked by the vast majority of users. There are, certainly, a few corners of the web that need more than static html, but >90% of the websites in existence that are _not_ static html, should be, and the consumers would like them better that way.

I don't believe consumers secretly yearn for largely text based pages, and that's part of the reason why web development has become so complicated. The fact is, 'you' may believe that a simple HTML page will suit > '90%' of consumers, but until you can prove that objectively, then you'll find that there's a whole industry dedicated to determining what really does suit consumers better.

And it's a multi billion dollar market now.

This is merely my opinion, so take it for what it's worth, but I think it's a reasonable answer to the question in the thread title. Also, why is there so much confusion over tools? Probably because there is so much confusion about what people do and do not like in a webpage / web application.

There's more art to software development than many practitioners would like to admit.

"I don't believe consumers secretly yearn for largely text based pages"

I sure as hell do.

Joe Sixpack doesn't give a fuck either way.

I'm sure there is some research that validates that endless scrolling increases user engagement. High user engagement leads to high user exposition, something that marketers are attracted to and spend lots of money to put their ads, and there you go, more profit due to this feature

This seems demonstrably untrue. Basic html development is vastly cheaper than “modern JavaScript” which would give any company w an HTML-only strategy a huge advantage. Why aren’t HTML-only companies dominating in every market?

You mean like amazon or ebay or paypal or yahoo? Or like how facebook was so html that they wrote their own php interpreter?

you mean the same facebook which created react.js? If you think any of those companies is not betting heavily on JS, you're out of touch :)

maybe, but i dont think they grew because of JS. they grew despite it maybe ;)

maybe, but it's impossible without it

None of those sites are static html. Not even close.

html != static

Static literally == not using JavaScript to manipulate the DOM. Once you're stuck using JS for core functionality, you really want a decent programming language.

I think people are complaining about two separate things - actually-static-ish websites like newspapers, shipping 10MB, and web apps like Facebook shipping 10MB. The latter is an unavoidable outcome of shipping Serious Software; the former is mostly driven by ad networks, not development ease.

I think static could have multiple meaning, and the one the original comment meant is ambiguous. There are "static site generators" which technically generate static sites, but could still use JavaScript.

Yeah we re talking about html, not static. Plus, iirc, it is actually most SPAs that are ‘static’, i.e served as plain index.html containing all the Javascript.

I love this comment. So fucking dead-on in my experience.

Not to mention, saying any of those things would make you a "bad culture fit" and potentially break the "no assholes rule" if you argued the point.