Ask HN: Why do such strong opinions exist in our industry?

The trending frameworks might be because we usually get tired of working with the same technology over and over again. The moment of learning something new is always exciting and once it wears out I guess that starting with "the new thing" recreates that feeling momentarily so we start rationalizing about it. There is also in my opinion the big problem of being able to do the same thing with many different tools, it always generates this kind of conflicts of which one is better, but at the end of the day, any tool can be "the best one" depending on the team.

I think you have to follow the money. When people have invested a lot of money into something, and loan payments are due every month, people tend to "make decisions" like "we are going with xyz framework" or "we've always used java so why migrate".

Once this decision is made at the top, the middle managers and aspirational employees believe in the upper echelon's dogma because their jobs depend on it, and going against the organization might = getting fired.

Conversely, when you look at self organized teams that are just making a hobby project, they tend to go with "we will do whatever we want to do given the reality of the tech landscape".

Strong opinions exist in any industry. However, few industries have as many things going on (and actively vyying for your attention, whether for $$ or mindshare or otherwise) as software.

If you go deep in a niche you're going to learn things. Those things may be contrary to the standard thinking. But in turn if you're right they become the standard thinking.

I don't know if it has to do with the CS background of people, but i find it baffling too (physics background). A TON of decisions are given very little thought or justification even though it should not be hard to measure stuff. It borders or being similar to management.

Because there is no evidence either way.

There’s no single source of truth. Experiences are posted on social media and forgotten.

What evidence exists that definitively says JavaScript is better than perl? Or golang is definitively better than python.

There is no definitive evidence.

You can't even prove that rust is definitively better than assembly language.

Hence where no evidence exists, decisions without evidence are made constantly.

You also need to realize that this not baffling. Collecting hard scientific evidence for every freaking detail is incredibally hard and time consuming. It's part of why the medical industry moves so slowly to get medicine out.

In fact any problem with the word design means that you are working on a problem with no definitive evidence pointing to a definitive solution. You are literally designing a possible solution out of thin air. There's a whole segment of people who do this, they are called ux and ui designers and a large portion of the the content they spit out is not evidence based at all.

> Namely, why do opinions tend to completely trump evidence for technical decisions?

Because evidence that isolates the effect of the technical choices under discussion from all outside effects and is directly and unambiguously applicable to the context in which it needs to be applied basically doesn't exist for most technical decisions in our industry.

The evidence which is clear tends to be on questions which are distant from business value, the questions the get directly at business value tend to have unclear/ambiguous evidence.

Wisdom takes time. Effort and time. A certain striving few are willing to maintain.

Perhaps confidence, warranted or not, gets more attention (including upvotes and reshares) than being justifiably cautious about drawing a conclusion? In theory this will both reward confidence and make it appear like most people are doing it.

It seems tough to avoid. You need a system, and we don't really have one yet.

We had a term where I used to work - Promotion Oriented Architecture. In most places progress in the engineering org chart is heavily influenced by delivering software for complex usecases. Unfortunately IME most developers correlate complexity in the problem domain to complexity in the solution domain. This becomes a very strong driver for adopting novel solutions, even when a simpler one would have sufficed.

A second factor is the difficulty in measuring success. Once a solution is adopted it is very hard to demonstrate the opportunity cost incurred by not selecting a simpler solution. And the costs themselves may not be evident until long after delivery. And since measurement is hard, the lack of data also makes it difficult to articulate effective counter arguments to gratuitous complexity

There's too much going on in this industry. Nobody knows everything. When a question comes up (Angular vs React? Lisp vs Haskell? C++ vs Rust?), there are very few people who know both sides really thoroughly, even for a question that only has two sides. So we wind up with opinions, because nobody knows enough to do otherwise.

And if you don't know both sides, you're judging a side that you know less well against a side that you know better. That is, you're judging your experience of X against your opinion of Y. When you're talking to someone who knows Y better than X, you can get fireworks.

Because this isn't an engineering discipline, there's no science, it's all pop culture. Alan Kay:

Computing spread out much, much faster than educating unsophisticated people can happen. In the last 25 years or so, we actually got something like a pop culture, similar to what happened when television came on the scene and some of its inventors thought it would be a way of getting Shakespeare to the masses. But they forgot that you have to be more sophisticated and have more perspective to understand Shakespeare. What television was able to do was to capture people as they were. So I think the lack of a real computer science today, and the lack of real software engineering today, is partly due to this pop culture.

Basically everyone roots for their favorite team.

Software developement is an unconstrained field. Computers can do almost anything you dream up. You don't need expensive tools, space, raw materials, or lots of people to make it. Contrast with building a car, where strength of materials, wear and tear, safety requirements, etc. make the design space kind of small.

In other words, there are fewer decisions that can be resolved by "option A is the only way to meet constraints x, y, and z simultaneously."

Because it's art moreso than science or engineering, thus much less subject to constraints like physical limitations.

Compared to the constraints of the physical world, the universe of computability is essentially limitless.

On top of that, there are almost zero objective facts in the field. Faster and smaller are about the only factual statements that can be made about computability. Everything else is an opinion. With no physical truth to disprove wrong opinions, people get accustomed to putting forward passionate arguments without every worrying about being proven wrong.

Finally, it usually takes significant effort to even have a few sample cases to compare opinions when it comes to coding. And usually it's not under scientific controlled environment.

There are no two decades long arguments about "does a ball roll down a hill". Noone is taking a strong opinion against. Before it would get that far, someone would just put a ball on a hill and see the end result. You can't do that with software, the problem space is too large and there is no little bearing to an objective way to measure to results, or they often aren't generally applicable.

Programming is (in a lot of cases) a "personal endeavour", as generally 1 computer has 1 keyboard for 1 person to use. This typically means it's a personal experience to program software, which can have a stronger impact on you. Some experiences in programming are really good others are really bad...

"Strong opinions come from strong experiences."

Dunning-Kruger effect.

I think its also worth looking at how people usually learn computer science.

In my limited experience, most programmers learn real world software development using the internet.

Some might learn CS theory in formal education, but most real world tools and frameworks are just taught by random people on the internet, or worse, giant corporates that have business interest in promoting the use of X. (Flutter, React, etc.)

Internet cannot possibly provide the scientific temperament and rigour that formal education in say, Physics, is able to provide.

The general consensus among the public seems to be that internet has made CS education highly accessible to the masses. This is of course, very convenient for countries like India, where higher education is often unaffordable and unpractical.

What's not so clear however, is whether this is enough for a real engineering discipline to exist.

Network effects and platform thinking create a high sense of urgency: time to market is considered crucial.

This in combination with a high level of innovation means there is a big risk perceived in taking the time to truely evaluate and test new technologies.

And in the end, because most of this is venture capitalist driven, long term effects are not really important: a road to a rewarding exit (ipo or company purchase) is what really drives most investment decisions.

There are multiple reasons why strong opinions exist.

But at least one of them is quite valid. We have a lot of garbage in our industry. People perceive that correctly and call it out. Simple as that.