I'll bet there are some fascinating counter-consensus views at the edges on what the future holds in these domains.
Let's hear them!
What do you think is true in your domain that most of your peers don't?
I will say that HN is becoming more of an echo chamber than I experienced even a few years ago. I downvote and upvote but I'm starting to feel sick of myself every time I do because the observed effect is that it largely makes people feel bad or radicalizes/entrenches the opposition.
Perhaps my most contrarian view I feel comfortable sharing is that I think almost any form of cancelation is a net negative for society: it sends the canceled group or individual underground where any extremist views are amplified more than they would resident to whatever platform canceled them.
I think we as humans don't realize how badly we're doing in the hyperconnected world. Our brains can't cope. No one is forced to sit in the same room with someone different so tolerance (actual tolerance, interpret that how you like) is in the gutter.
I think the only way we can really do this thing, survive and prosper in kindness, is by retreating to dramatically smaller communities with local networks. That's my dream.
Cancellation or isolation does not help. I my opinion, education (or proper dialog) is the only option we have.
Independent of whether cancellation is a net negative or not, I don't think the data backs this up.
Maybe you see the point: Asking for data and citations is pretending to be rationale and civil, but usually it's just another form of sealioning. No, I don't have data to support that.
I observe people stating false facts on this site every day.
However, I'm pretty certain that when someone says, "do you have a reference for that" or similar, what they really saying is "I think you're a fucking liar". So really, they're being antagonistic and dishonest about it. It's passive aggressive.
In either case it tends to be easier to just search the information yourself and show what you found
Like as a whole, it’s quite bad. Best we got unfortunately, but making it a “golden standard” is just hilarious to me.
1. The best approximation for writing maintainable and fast code is to be a craftsman trying to write beautiful code. To find joy in the art of coding, not race towards making something that works.
2. Machine Learning and AI is significantly overhyped and overfunded, even more so than crypto.
3. The current iteration of the internet, speficially social services, was a great technological advancement for humanity, but it was a step backwards towards making our lives happy and fulfilling. We did not evolve to deal with this technology and we're becoming more and more miserable. This thing may be our downfall.
4. Related to 3. I think the internet and its power of communication had a negative effect on science. Science is now mostly social and status game. This has always been true to some extent, but the ability for instant global communication and comparison has made it worse. It's regression to the mean when everyone wants to work on the current global hot topic, which has only been possible in recent decades.
5. The whole gender equality movement and cancel culture from the US is a bunch of BS and just people trying to make careers out of it, or people frustrated with themselves and their own lives. And a great way to exact revenge for emotionally unstable female persons.
6. Cars suck and should be abandoned worldwide outside of car racing for fun.
7. In a few hundred years we will look at today's science the way we're looking at religions today
8. Crypto is good, even it's just a big global casino. I believe we need a big global casino and gambling should be legal and fun. But people should be educated about gambling and probability theory.
9. Probability Theory and Bayesian statistics should be the most important subject in middle/high school
10. Governments and countries make no sense. Nobody would come up with this stuff if we were to design to system from scratch today with the internet in mind.
I could go on, but let's stop here.
People like to say investing is just gambling but gambling in games of chance is not at all like putting money into a market full of asymmetric information, irrational players, and fraud.
Probability theory won’t help you parse the outcomes of crypto markets
But you are right that the casino analogy is bad since games inside a Casino are fairer on average and easier to model probabilistically. Still, that doesn't change my point about thinking of investments as gambles.
I'd like to see someone argue for something ridiculous, like "open source is a fad and more control of IP will be better for society" (I don't believe this) or some other truly contrarian and interesting position
This take I find particularly hilarious because if you look back at the science from a few hundred years ago.. it is pretty recognizable as science! Newtonian physics.. pretty good! A future where we think of today's science as akin to religion is a world where Newtonian physics is obviously seen as the same.. that seems wildly implausible!
Who's "we"? Just because the West is secular does not mean the entire world is. Ironically, scientism is the West's latest religion.
Spoken like a born and raised urbanite.
Waterfall is a perfectly good methodology in many cases.
China and Russia aren’t trying to take over the world.
We are under-treating some mental health issues as an over-correction to historical bad treatment practices.
Some important cultural glue has been thrown out with the bath water as traditional religion has declined.
Quality of life is generally great and getting better all the time.
And since my comment also included China, I see their assertiveness in their region (e.g. blue water navy and island building) not as a prelude to world domination, merely regional preeminence, driven by defensive security concerns.
I’m not condoning it, I’m just saying I think they are playing the same bully game that every large power has done historically.
So I don’t think it’s a straw man at all, but rather the entire basis of NATO.
The US is also militarily active far outside its territory but I also don’t believe this is a world domination plot either.
I don't know what you are basing this on. NATO concern is that multiple of its members are threatened by Russia directly. Now ever more so after Russia explicitly demanded to "roll-back" NATO to its 1997 state.
> And there may indeed have been some truth to this in the part of the 20th century when communism was a globalist totalitarian ideology.
"Some"?
> So I don’t think it’s a straw man at all, but rather the entire basis of NATO.
Entire basis of NATO is security of its members. Including eastern European ones. Care to provide reference to any NATO treaty/document that states that NATO is only concerned with threats of "global domination" (whatever you think that means)?
> The US is also militarily active far outside its territory but I also don’t believe this is a world domination plot either.
Of course it is? Or it was, since US successfully achieved it in early nineties and then found out that they don't really want it anyway? US, with some notable recent exceptions, has been targeting status of "world police" in some form since the end of WW2.
Experiencing christianity (EU) has led me to discover some of the most vile, exploitative predatory behavior towards the frail and weak I could imagine. I wish for the total disappearance of religion in every country, it's social cancer, not a glue. If it keeps people together, it's with fear.
* Enterprises don't like being locked in to single, large providers. (See examples of IBM, Oracle)
* New providers will attack the high margins of AWS and others, expanding the market. (See Cloudflare for such a provider)
* Building applications in the cloud by combining lots of SaaS services delivers faster time to market but results in complex integrations that are costly to manage and difficult to reduce costs. They also have serious drawbacks related to data privacy. (This looks like a new form of tech debt.)
* Kubernetes abstracts out platform differences and lower switching costs between cloud and self-managed models. As Kubernetes runs in more places it will open up more choices where to operate.
In summary the current cloud land-grab will change. I would bet on new ways to enable self-management and diverse economic models of software operation.
Things like Webpack and Babel make life harder, not easier, and so do most frameworks, because you end up spending half your time trying to get around the stuff that's made for large-scale - a lot of React is like trying to use nanotech to make a coffee table.
Almost all social networks are garbage and make the world awful and are very bad at their stated goal while being very good at their actual goal, which is creepy, invasive marketing.
We probably peaked as an industry when Google Reader came out. :-D
I don't think that's much of a contrarian view though.
Are you just hinting at the fact that these industries are slow to evolve, or is it something specific about the languages?
My team have in the last decade spearheaded initiatives like adopting Agile, C++ (subset) use, Git (as opposed to SVN/proprietary) and extensive use of modern code hosting and CIs and we've met with a lot of resistance in our company (and those we work with). I've also interviewed lots of experienced embedded engineers, e.g. automotive, who have never heard of Agile or used Git.
I think it's a fear of change (risk) and also that the embedded engineering domain is so closely tied to hardware development, which is even slower to change.
As someone not working in those industries, it would seem an obvious choice to move to a safer language, so there must be something I'm missing?
The second reason is lack of tooling - specifically Functional Safety toolchains that can be used for ISO26262 projects. There are plans by Ferrous to develop one, but it will take years to gain any adoptions [1].
tens of thousands of dollars and several months of QA per release means that thanks very much but we’re going to keep using what we have because yours looks nice i’m sure but we don’t care.
PS: I am saying this as a former head of product management...
You have to build the organization to be resilient to the types of incoming requests that it will receive. Some sales leads could -- intentionally or accidentally -- lead your organization and product down a path that limits future potential or causes cultural damage.
One of the ways to counter this is simply to spend more time collecting requirements and discussing the topic both internally and externally. Benevolent partners shouldn't want to rush to determine the shape of a feature.
Another slightly-less common counter would be to make the entire process more transparent. That could be difficult for some partners who might want to collaborate on the basis of creating a unique advantage that they'd like to achieve ahead of competitors. However, increasingly, I think that the cost to hide or obscure such plans is becoming greater than the benefits.
(and yep, none of this suggests anything about whether product manager roles are required or not: I think that's more to do with the size of the organization and whether specialization is required. my point is mostly that, regardless of scale, the organization needs to guard itself against misaligned incentives)
1. if these great UX and tech people do the product management themselves?
2. who does product and customer discovery?
3. Is it highly likely that the UX and tech people will know how to do product and customer discovery well - for example ask questions and carry out research the right way?
- they can be useful on their own without providing a path to something much more advanced
Do they claim to do this?
The UIs are terrible and limiting, the smarts become obsolete or will fail faster than the rest of the hardware, and will be costly if not impossible to repair/replace. And then there's the ads, surveillance, locked features, and so on.
Other applications are like trying to stick a round peg into a square hole.
-t.thoughtLeadersInTech
(it seems likely true, and with a success rate like that, improvements rather than migration of existing codebases would make sense. just keen to find some supporting success stories :) I found some cfengine ASAN examples but nothing hugely conclusive yet)
https://static.googleusercontent.com/media/research.google.c...
The recent trend towards decentralization of tech is great, but if data is not portable then its just a different centralization.
For ex. Matrix/Element is great as an example of decentralization, but since you cant do much with the data if you want to go to something else, its still lock in.
javascript is a bad front end scripting language for light use cases:
To expand on the above a little: it's not nostalgia, the Internet was better when most of the general populace didn't have access to it. Gatekeeping works. It was better when people had to dig a little and learn the mores of a community before joining it. Mobile phones and easy access to the Internet destroyed all this.
Specialized training, certifications and real world experience are a better use of time and money.
In any field, or just for people that want to develop software?
I’m only familiar with the tech industry (software, marketing, design, etc.).
I've known lots of people who called themselves full stack developers. I've only known one person who I would trust to do full stack development well. And he calls himself a mobile developer.
I should probably add that I have met a lot of those developers who wonder what they are doing, and cannot do a simple Google search about the topic they are supposed to learn. I'm scared of what will happen in the next few years for the companies who employ them. As an example, if you're learning Python, the first task would be to install Python and read the doc instead of being confused in various forums.
Is my first sentence a counter-consensus opinion?
- Building what your users say will make you create a product that nobody wants to use. People don't know what they want or need. If you want to create good products, listen to your users but don't built what they say
Sure, things can work okay server side if you're with a top 10 cloud provider, but support gets abysmal quick as you move further from the top.
[1] I'll define this as "the majority of home users perform the majority of their day to day web actively almost exclusively over IPv6"
Now I have better service (reliable 1gb fiber) but no IPv6. Once providers are better motivated, I predict a jump.
Agree with issue on server cloud provider support though, sadly.
- Managers must also be programmers
- Managers should set timescales for projects, not the people under them
Probably one of the biggest factors I’ve seen in predicting the success of a software team.
Well it certainly moves fast, compared to other industries that have been around for 100s or 1000s of years. But I feel technology is moving fast mostly on the surface, whereas what is beneath has changed much much slower if at all during my career since 2005. If you have solid fundamentals, you can just ignore and let the surface waves pass by until you need something and only then pick it up.
There is no need to constantly keep up with the framework/library du jour, better invest your precious time to deepen knowledge somewhere.
- Slack has an enormous cost on productivity by having everyone exposed to way more communication than is necessary across an organization. There has to be a significantly better way.
- Relatedly, tech companies in general should be doing significantly more to protect their engineers' time, and the ones that get good at this are going to have outsized impact, but it requires top down leadership from people whose job it is to be in meetings all day. The default at most companies seems to be to have days filled with meetings as a senior+ engineer, and it's swimming up river to fight back against this.
I have just watched history of The Secret of Monkey Island (https://www.youtube.com/watch?v=9F9ahZQ7oP0). What struck me was how relatively cheap it was (if I remember correctly it costed 130000$ and main team was only 3 people). I now that it was a different time, but even if You take the inflation (of money and requirements) into consideration its still a miracle.
And one of the reasons given for this project success was that Schafer and the rest was left alone to do their thing - I would not say they were protected, but at least nobody important enough paid attention to them to start messing with their creative process.
This is somehow consistent with my general observation - that in knowledge based work only small subset of people do the real work.
Apple's shift to Swift has caused a lot of software to be rewritten. It's not obvious that the reduction in software correctness bugs which Swift provides offsets the kind of bugs that are the result of rewriting software. Apple's software rewrites have introduced very little tangible benefit to the end customer while creating lots of software gremlins which are very annoying to them. Worse, focussing engineering resources on rewriting existing apps in Swift vs pushing the apps forward in terms of UX and features have allowed competition to catch up and surpass Apple in several areas that do matter to customers.
In hindsight, there will be a lingering question of if it would have been better for Apple to have continued to evolve Objective-C further in ways that could have provided similar technical benefits to Swift's introduction. At best the shift to Swift will have been seen as a necessary evil that exposed Apple's flanks. At worst, it will have been considered an unnecessary technical exercise that may have been started as a way to retain top technical talent, but even failed long-term in that respect.
Swift isn't alone in this respect. I think TypeScript is another example of a language which started out claiming to offer a better experience than Javascript only to become very complex. I'm sure that this added complexity helps some programmers some of the time. But just like CISC was a dead end and all computer chips are essentially RISC-based today, I suspect the pendulum may have swung too far in terms of adding features to some of these developer-friendly languages.
But, it's a contrarian view for a reason. I'm just somebody who fell in love with the simplicity of Objective-C almost 20 years ago, before the iPhone was a thing. So I may be looking back at things with rose tinted glasses.
However, it is far easier to read and learn than objective-c. All of our new software is written in Swift and we will never go back to objective-c.
2) We are drowning ourselves with the need to measure everything and killing ourselves by ignoring what we can't measure.
3) Efficiency is the bane of technological systems.
Some/Many employers are really attached to controlling their employees -> so they want to see them on time, at their job etc.
Nobody will rebel, when told to come back at work and stop remote working.
A lot of people are already so dependent on the system that they will fight to protect the idea: "Oh but it's important for social life to actually go to work". As if, it is literally impossible to meet people in other places than work.
Next recession you'll see even more companies go remote as the recession forces cost cutting and rent/commute and their related insurance costs are cost centers that can be removed.
Remote working is a cost cutting measure. Only those companies that can afford to return to the office are doing so.
I bet next recession companies extend their life by ditching the office and in extreme cases laying off staff and replacing with staff from lower cost of living areas who can work remote.
Bingo. Remote working is a cost cutting measure. Only those companies that can afford to return to the office are doing so.
Most of them refer to themselves as Advanced after taking their first course, even if they have no practical experience on anything other than the type of systems they have at home.
"I just treat it like a black box" really means "I'll just run Nessus and call it a day."
In the real world, in private. It was always techno-Utopianism to think that mass access to communication would correlate with propagation of wisdom.
GET requests should not modify server state outside of caching and should be idempotent!
Its like the difference between being a farmer with a plow and having an army of farming robots. The productive multiplier is astronomical.
To gain an appreciation for natural productive multipliers you could play games like Factorio.
Bless, or curse, capitalism.
After some 10 years as a developer I met the first person who actually knew how to practice TDD. It was eye opening how fast they could produce quality code.
I can't be sure, but what they did was reproducible (albeit slower) by all of us at the job, and I've been trying to learn it properly and get to the point where I can teach it ever since. Not there yet.
And then there's the flip point that unpopular opinions usually garner lots of popularity in the right groups and it isn't typically courageous to hold them.
One point is about correctness, the other point is about social popularity.
And to reinforce that, your assertion that reading the most downvoted comments are worth your time and implicitly where the correct views are is poor.
There is a big bubble in commercial launch.
"Urban Air Mobility" is another bubble.
The problems we have in developing and acquiring military technology have nothing to do with "innovation" or a lack thereof.
“In your domain” is the important part.
OP, thanks for including this so as not to start a flame war.
For those expressing contrarian views about social issues, foreign affairs, etc., OP isn’t looking for armchair opinions.
Avoid engineering practices from Facebook or Google. Both companies have monopolies, an army of developers and can afford to rewrite software on a whim.
We could probably have a fully autonomous auto fleet with current/available technologies, if we adapt the roads, rules and such.
Human-like autonomous driving might not happen but getting from a to b doesn't strictly require that.
You can do a lot with LIDAR, but you can't recognize e.g. traffic lights with absolute certainty.
"Loving one's neighbor as oneself" does not mean ceding money and power to distant, unaccountable bureaucrats who tend toward corruption over time.
Humanity is grappling with the benefits of piling up large "pattern of life" databases. Technical people love to optimize, but humanity is all about the entropy in the system.
The need to balance the conveniences to which we've grown accustomed without Big Brother (public or private sector) squeezing the humanity out of everything is perhaps THE issue of the 21st century.
The various other concerns, e.g. climate, are peripheral.
(a) Markets work.
(b) Markets fail.
Yes, it's possible (IMO rationally necessary) to believe both. Yes, both ideas will draw predictable attacks.
People probably aren't going to share the truly interesting insights that deletes a business competitive advantage they have or say something that risks getting them cancelled.
Does anyone here ever walk away from computers for a minute?
Here's a contrarian view that's sure to piss off almost everyone here:
Keeping and manufacturing animals for the purpose of human emotional support is a mental illness.
What if a solution to the homeless drug addiction problem is to give them animals to take care of and be responsible for?
(It's funny because I used to think the same when I was younger)
--------
Protestantism is dead, it just doesn't know it yet.
Protestants have a Bible that differs from Catholics/Orthodox - it excludes the Deuterocanonical books (like Maccabees). The central tenant of Protestantism is Sola Scripture - that the Bible is to be the determining factor on whether a doctrine is biblical or not.
John Calvin (one of the founders of Protestantism) says in his Antidote to Trent that Catholics added the Deuterocanonical books to the Bible as these books prove doctrines like Purgatory and exorcisms. Or to put in the inverse, that IF the Deuterocanonical books were in the Bible, doctrines like Purgatory and exorcisms would be biblical - but the Deuterocanonical books are NOT part of the Bible.
The problem Protestantism is facing is that their justification for their canon (excluding these Deuterocanonical books) has fallen apart with the findings published concerned the Dead Sea Scrolls in the last 20 or so years (especially by DSS researcher Emanuel Tov).
Protestantism's justification for excluding the Deuterocanonical books tends to boil down to one of two arguments - pointing back to Jerome or pointing back to a Jewish canon prior to the time of Jesus.
Jerome's arguments for excluding the books boiled down to two things, since disproven: 1) That Jesus and the Apostles quoted exclusively from the Proto-Masoretic text (which excluded the Deuterocanonical books) against the Septuagint (which included the Deuterocanonical books) whenever the two texts differed, and thus Jesus affirmed the Proto-Masoretic as original. However it's now universally accepted that Jerome was wrong and Jesus and the NT authors vastly preferred (and quoted from) the Septuagint [e.g. Matthew 21:16 and Psalm 8:2], with only a handful of the 300+ NT citations of the OT preferring the Masoretic. 2) When comparing the Hebrew text in Jerome's day (AD ~400) to the Septuagint text, Jerome noted many discrepancies. Noting that the original text of the OT was Hebrew, Jerome asserted that the discrepancies must have been from errors in translation of the Septuagint - that the Septuagint was a poor translation of the Hebrew text of his day. The Dead Sea Scrolls prove the existence of multiple ancient Hebrew text traditions, including a Proto-Septuagint Hebrew text, showing that the Septuagint was a good translation of this Proto-Septuagint Hebrew text, not a poor translation of the Proto-Masoretic text.
And the Jewish canon argument has no standing on anything. There is a lack of evidence of when the Jews finalized their canon - specifically their "writings" section. All existent evidence among rabbinic writings indicates it did not finalize until after the Christian era (when the Christians had already settled on the Septuagint's canon). Many Protestant scholars (see Sundberg) and Dead Sea Scroll researchers (see Tov) have written much on this, as the DSS provide further evidence of the lack of an established Hebrew canon at the time of Jesus.
Protestants are in for a reckoning. They insist on their scriptures having ultimate authority, but which scriptures? Their justifications for excluding the Deuterocanonical books have fallen apart, and there really aren't any substantial arguments remaining. Eventually they will have to deal with this - their doctrine of Sola Scriptura, and their innate push to seek for truth, demands it.
No, it's not. There were from very early on many different Protestant traditions, and sola scriptura isn't even part of all of them, much less the primary and universal foundation of Protestantism (Anglicans and Methodists, for instance, are Protestants, but have only ever held prima scriptura, not sola scriptura.)
> Protestants have a Bible that differs from Catholics/Orthodox - it excludes the Deuterocanonical books
Many mainline Protestants include the Deuterocanonical books in their Bibles, and in their lectionary, but give them lesser theological roles.
Anyhow, some Protestants hold to strict literalism as fundamental despite direct contradictions in the Bible. I think you misunderestimate the ability of people to maintain a religious tradition in the face of even flat logical contradictions in fundamental doctrine, much less a lack of clear justification for doctrine.
The values themselves don't matter much (whether it's about being carbon neutral, or getting vaccinated/wearing mask, or UBI): they like the power.
That's one of the problems with asking people for their "contrarian views".
but the 'climate change will kill us all in ~10 years' is absolutely idiotic.