What comes after open source? (2019) | Hacker News Reader

What comes after open source? (2019)

denisnazarov.com

55 pointssimonpure34 comments

I've always found the, uh, "IP-flexible" environment of Shenzhen to be a fascinating insight into what a fully open source world would look like. From what I've seen (please correct me if I'm wrong!) no energy is wasted on licensing or who you should ask for permission, because what are they gonna do to you? And little effort is wasted on locking things down, because they'll just crack it anyway. If you can make something you just do it, and then you ship it and hopefully make some money before someone copies it. And the rate of progress is incredible! Mostly because of the huge amount of hard work done by the people working in in the industry here, but also because there's less time wasted on these regulatory questions.

Yeah, I think IP creates a local-maxima where it seems essential to have IP to have innovation. But, it’s not at all clear to me that it doesn’t simultaneously restrict innovation: all the locked down walled gardens, content silos and such make it significantly harder for people to recombine inventions in unforeseen ways.

Patents certainly cause unforeseen damage all the time, sometimes with damages with far too many zeros to be considered at all reasonable (like what happened to Blackberry a few years back). My personal belief regarding software is that if a skilled practitioner of the art can come up with a particular solution when given similar design constraints for solving a problem, the solution should not be considered novel and worthy of patent protection. Sadly this is not anywhere close to how patents actually work in the real world.

I'd agree with your definition of novel except for the addition "in a minimal timeframe or effort" because else of course you can replicate the same outcomes with about the same effort.

The rate of progress is fast, but how are the wages and opportunities? Does this progress translate into progress for the people doing it or is it a rat race to nowhere? There are many ways to goose growth and progress that are bad for people.

Services are bad and libraries are good. (Extra proof: the latter can be turned into the former if you really want, but not vice versa).

Cryptocurrency doesn't factor into any of this.

As a programmer I agree, but for a business a service might have a nice mote and offer the right incentives to develop and sell a service.

A crypto service is the worst of both worlds, in this case the income just goes to the miners.

And the enegy companies that sell the power for mining, and the chip makers that sell chips for mining.

But chips and power-plants are harder to create then just copying software.

From my understanding, mining is generally pretty difficult to do profitably.

And the African children mining the metals!

Following it that way is like following the glass maker in the standard broken window fallacy.

And in the parable of the broken window fallacy, the glassmaker does benefit from the broken window, but the broken window is bad for the economy overall. It's a similar situation here, at least in the short term GPU makers are profitting from increased demand, as well as scalpers and people selling secondhand GPUs, but for the economy (and in this case the environment) as a whole it is probably unhealthy.

So, the broken window fallacy is definitely relevant but doesn't really refute my claim that someone is benefitting from it

True. I had interpreted your comment as implying that the existence of beneficiaries beyond the transaction validators themselves makes the system as a whole be beneficial.

We have open source. Crypto allows for open state. Blockchains are databases where everyone is a root user.

Nobody is a root user. Try rm -rf / on a distributed blockchain.

Fair. I'm trying to think if "Blockchains are 'no-delete' databases where everyone is root" makes sense?

You need to account for consensus among many

Well, I think "crypto" here means "universally viewable and add-only modifiable" database which feels like there might be a use for?

ed: "consensus-driven publicly-auditable append-only database" is actually a pretty decent generalization of "crypto", thanks.

Is "consensus-driven" that necessary? I'm thinking how one might argue that it's fundamentally an implementation detail of "append-only," meaning that you don't really have to say it out loud, like "electricity-driven" or "cryptography-powered."

A centralized database operating off a write-once-read-many drive would be append-only and not consensus-driven, and also not really what anyone means by crypto.

A whole lot more efficient though.

Basically Git(hub)

Other people have to agree on what you put in there. There are rules. It's not an append-only buffer of arbitrary bytes that everyone is forced to replicate for no reason.

Not to be confused with the identically-named article by Steve Klabnik from around the same time: https://steveklabnik.com/writing/what-comes-after-open-sourc...

Wow... this article (the not-the-OP one you linked to here, from the guy I mostly know from his missionary work for the Rust ecosystem) represents a complete misunderstanding of what the Free Software movement is prioritizing... it was painfully disappointing (but I guess not surprising) to read this :(.

The core disagreement between the philosophy of Free Software advocates and the world of modern Open Source developers is that the former is fighting for the rights of end users to analyze and even specifics modify the software they use and to achieve this they are willing to severely limit developer freedoms. This is a move that many developers don't like, but this article is somehow missing that disagreement entirely and us claiming that Free Software is somehow missing the point because developers actually want "[a list of things involving the production of software]" that the Free Software people are going to point out does not in any way help achieve freedom for users.

I fully understand that the free software movement believes that they are doing things for the rights of end users.

I believe that too, and believe that companies (handwavy) are looking forward to permissive open source licenses becoming de rigeur so that they can build and bundle projects cheaply without contributing back (proprietary-style hacks behind closed doors will suffice to ship; production costs reduce, and, strategically, air -- developer time -- to the FOSS ecosystem becomes gradually easier to stifle).

Easy, return of shareware and public domain.

That is what dual licensing is, only with rejuvenated names for a newer generation born after FOSS became mainstream and found out money needs to be earned somehow.

Sadly, I think this will be gone forever soon! If you are on Linux, everything might continue for a while, but for Mac and Windows, the writing is on the wall ... there is a near future where nothing will be installable on your computer unless it is "authorized" by Microsoft or Apple.

Marketplaces may be examples of open services. Such as the smart contract-based Uber or Amazon.

I'm not sure if I agree that crypto is the future, but I agree with most of the rest.

Is disk space cheap enough now to have a “GitHub for state”?

Telekommunisten.

Reply on news.ycombinator.com