You've made the mistake of expressing an unpopular opinion, so you're getting down voted. I strongly dislike proprietary software, but you raise valid concerns that should be considered instead of being downvoted into silence. (edit: at the time I wrote this comment, the parent appeared to be headed for oblivion, happily that appears to have improved)

I don't know that open sourcing Java killed Sun. More likely it was Intel coming close enough to SPARC performance at a lower price, and Linux coming close enough to Enterprise reliability also at a lower price. But it's certainly true that open sourcing Java and Solaris didn't save Sun.

Likewise, open sourcing Mozilla didn't save Netscape. In fact, I would bet the vast majority of companies that start out proprietary before deciding to open source ultimately die because they're already unhealthy and doing it as an act of desperation.

If a company isn't designed from the start to support an open source model, it's very hard to make the transition and survive. And we have no right to demand they try. Just like everyone should respect the license F/OSS authors choose, everyone should respect Wolfram's right to stay proprietary.

Of course it's also important to remember that proprietary software vendors are more likely to yank the carpet from under their users without warning. Instead of demanding the vendor open source, users should just find and support open alternatives.

Providing Java for free (no matter whether open or closed-source) was a big strategic blunder for Sun: it commoditized CPU architectures by providing an abstraction layer over them. Write once, run anywhere! The problem was, Intel chips were cheaper. So people just ran their Java workloads on their existing x86 PCs instead of buying expensive Sun hardware. Sun should have never written or released Java: it consumed resources, and hastened their demise.

> it commoditized CPU architectures by providing an abstraction layer over them

I think that was part of the point - let people run their existing applications on Sun hardware. Yes it was more expensive but it was also more capable. Emphasising running on lots of smaller 'commodity' instances wasn't yet the done thing at the time.

You are right that it was the point. It was simply a bad goal. Sun hoped it would allow people to migrate to their hardware, when instead it enabled them to migrate away, or to never have to choose.

I think it had little to do with that. Every year Sun had bigger and bigger computers like we needed bigger monolithic servers. Run an ISP and need 100,000 user telnet accounts? Bigger computer! Then, seemingly overnight, the problem is solved by buying a hundred blades instead, and paying nothing for except hardware.

Sun was riding high on the dotcom era where every new company ran out and bought an E10000, and maybe one with less specs for a development server. I've heard stories of these being shipped before payment just so they could declare the sales early -- once the bloodbath started, it Sun was left holding the bag.

BSD and Linux on commodity hardware killed everything at Sun and Java really had nothing to do with it.

I think Java had quite a lot to do with making those BSD and Linux boxes run the kind of software people were using for ‘e-commerce’ back then.

> You've made the mistake of expressing an unpopular opinion, so you're getting down voted.

It's never worth saying this on HN. "downvoted" has temporal dependence, and so right now, for example, the comment you're replying to is not at net-negative. Just comment on what they said, not on how other HN readers are voting what they said.

What's wrong with that comment is that people think it is wrong for some fairly obvious reasons. for example: there is no causal link (which is also the point of the comment originating this thread). Examples of software needing to be open source to succeed, like other languages, the Android OS, etc. are easy to find.

Even your more qualified point has loads of counter-examples, that companies that did not start with their software as open source find it hard to succeed at open source: Google's search engine and much other software is not open source. Many companies mix their approach to open source with proprietary systems and products. As did Sun.

That Wolfram's justifications are weak is not necessarily a knock on their decision. But the specific point that "High-level languages need more design than low-level languages" seems particularly odd. There are high level languages where all the widely used implementations are open source and few cases of them becoming chaotic. I suppose there is survivorship bias in that observation but there are also obvious reasons why no language user community would have a lot of patience for a proliferation of variants.

Netscape exited for ten billion dollars. It didn’t need saving because it was by any reasonable measure a grand slam home run. The exit was so lucrative that Andreesen is the first name in Andreesen Horowitz.

This doesn’t make a good argument that Microsoft is evil, so Silicon Valley created a mythology in which Netscape failed.

Netscape did not bet on open source. It didn’t fail. And the browser has a de facto monopoly anyway in the form of Chrome.

A successful public fundraising round is an indicia of success for prior investors. It is not an indicia of long-term success for a company as a going concern in terms of profits.

I don’t disagree.

AOL bought Netscape (with stock) and the founders, Clark and Andreesen liquidated their Netscape stock. So did the investors. So did vested employees such as Zawinski.

Or to put it another way, if Netscape failed, then so did Viaweb. I don’t think the YC founders would describe Viaweb that way.

I guess it depends on what the subject is, and how one defines success. As a product, Viaweb was a failure: it gained no market traction and Yahoo! didn't even use it for very long after buying it. PG and company might not consider the business a failure since it made them rich, but that's like saying cryptocurrency is a success because it made early Bitcoin investors rich. The jury's still very much out as to whether crypto will be a success among the other well-recognized successes in the business world such as telecom, oil, steel, and automobiles.

My understanding is Yahoo Stores was the Viaweb code base for a fair number of years. [1]

As a product, Firefox is still around. And Netscape disappeared in large part as an unsuccessful ground up rewrite as the brand continued following the company sale.

Bitcoin is not a company. Participants had different objectives. Those whose objective was to hold had success. Those whose objectives were political perhaps less so.

Entrepreneurs are in it to make money. Exits are how startups do it.

Telecom, oil, steel, and automobiles are state industries. Even if political correctness often demands pretending they are not…it wasn’t called the arsenal of democracy for nothing.

[1]: https://news.ycombinator.com/item?id=5739392

Yes agreed, as I mentioned in this other comment I think of it as Google and Microsoft successfully "commoditizing their complements", while Sun and Netscape failed to:

https://news.ycombinator.com/item?id=29605638

And to be clear I would like an open source Wolfram -- in fact it's almost a requirement before I spend the effort to learn it! (I have never learned a proprietary language outside of work or school)

I just think it's obvious why it's not open source -- because they have a working and sustainable business model! And on the other side, I think it's great to actually pay for what you use, rather than having something free that's filled with dark patterns or irrelevant ads, etc.

It's a hard problem, and it's obvious that we still don't know the answers :)

Open source projects seem happy to yank the carpet from under their users, depending on what you mean. In the Python world, I'd put Python 3 and pip 21 in that category.

> The argument is that open source licenses are incompatible with the kind paid license they're offering their customers.

I don’t see them making that argument in all these points.

For example they say

> Crowd-sourced decisions can be bad for you

Well ok that’s a valid point so… don’t do that. Crowd sourcing design decisions or code is not a fundamental part of open source. You can have open source but do all the development internally if you want to.

I think they’re confusing source licensing with a development methodology.

> I think they’re confusing source licensing with a development methodology.

Which is why as nitpicky as it sounds, we should really be making an effort not to use the term "open source" to refer to both.

There are dozens of edge cases that you can argue; is redistribution allowed, is attribution required, etc. - open source is a blanket statement and is perfectly suitable as a term IMO.

Open source is a well defined term that means a very specific thing. The definition is exactly 1 sentence long. It's not a blanket term at all.

Something is Open Sources if the code is published under a license that grants the user the right to study, change and redistribute to anyone and for any purpose.

There is other similar things, like source available, or stricter subsets like Free Software. And there is many things associated with open source, like collaborative development in public, community building and so on.

> redistribute to anyone and for any purpose.

This is certainly not a universally accepted definition. ‘Open source’ has long been a generic term for any situation in which the source is available. People may want it to be a term of art with a technical definition, but that isn’t how it is used in general.

> Something is Open Sources if the code is published under a license that grants the user the right to study, change and redistribute to anyone and for any purpose.

That's how some people define open source. Other people use it to mean being able to see the source.

They are wrong. The term for that is source available.

There is absolutely no reason to have any doubt about the true meaning of the term when this definition is on every search result on Googles first page, in Wikipedia, when it's that name of a non-profit that defines it like that and when the page that literally is called opensource.org defines it like that.

Having a domain name doesn’t mean you own a term.

Find me a reputable source on the same level as Wikipedia, the open source foundation or red hat that defines open source in a different way.

This is not about owning a term, it's simply about the fact that term had a definition that people agree on. The fact that some people don't know about that definition or that some people don't agree with it doesn't change it.

A term with no agreed meaning has limited utility. We have an official definition. I don't understand how the community benefits from making the definition more ambiguous and murky. Is there a group that's heavily invested conflating open source with source-available? I just don't understand who wins in that scenario.

> We have an official definition.

Lol how is it 'official'? Who officiated it?

You can read the history of the term Open Source here: https://opensource.org/history

The folks who coined the term were among those that led the OSI.

> The “open source” label was created at a strategy session held on February 3rd, 1998 in Palo Alto, California, shortly after the announcement of the release of the Netscape source code. The strategy session grew from a realization that the attention around the Netscape announcement had created an opportunity to educate and advocate for the superiority of an open development process.

> Two of those present at the Palo Alto meeting (Eric Raymond and Michael Tiemann) would later serve as presidents of OSI, and other attendees (including Todd Andersen, Jon “maddog” Hall, Larry Augustin, and Sam Ockman) became key early supporters of the organization.

These folks went on to create the Open Source Definition: https://opensource.org/docs/osd

The folks who created the term started an organization named after the term, created an official definition, and registered the domain. If that's not enough for us to have an official definition, what does it take? And what do we gain by arguing about this? Reading your posts in this thread (and elsewhere), my impression is that you're invested in insisting that "open source" doesn't have that definition, but I can't figure out what your motivation is.

FWIW, my motivation is to create an environment where terms have precise meaning so we can discuss complex issues (like copyright, licensing, development models, etc.) effectively. If we're wrangling over terms and language all the time, communication becomes much more difficult.

> The “open source” label was created at a strategy session held on February 3rd, 1998

You can see the term gaining popularity prior to 3 Feb 1998, and there are references to for example BSD being 'open source' as far back as 1990.

https://books.google.com/ngrams/graph?content=open+source%2C...

Via https://www.arp242.net/open-source.html.

BSD fits the modern open source definition.

Martin's article pleads with readers not to be pedantic when using the term, but offers no alternative definition, beyond 'let people use it when they can read source code'. Perhaps that's the proposed usage you and Martin prefer?

If I'm understanding your position correctly, perhaps you would agree that Ocarina of Time has now been released as open source?

https://arstechnica.com/gaming/2021/11/reverse-engineering-t...

If so, the term is uselessly vague, as it says nothing about the rights of the licensee. In this sense, 'Free Software' has much more meaning...perhaps RMS was onto something!

> offers no alternative definition

If the OSI picked a term that wasn't simply descriptive they could trademark it, then they could enforce their definition and there would be no ambiguity. That would be better for everyone.

Without delving into how distinctive a mark has to be, that ship has sailed: I see no path to changing the term now.

> You can see the term gaining popularity prior to 3 Feb 1998

“Open source” has older non-software uses, most notably in the intelligence context, where it picks up sharply in the 1980s:

https://books.google.com/ngrams/graph?content=open+source+in...

Did you look at all the references in that article? They aren’t about the intelligence term.

Lua is an example of open source licensing with closed development process. SQLite is similar (though public domain instead of OSS code). It's a pretty poor justification to say OSS itself is bad for Wolfram based on that.

> I think they’re confusing source licensing with a development methodology.

I'm a suspicious, cynical person, so I think that's deliberate.

Conflation is a very useful rhetorical trick.

> The company that created Java went out of business and was sold to Oracle for a fraction of its original worth. So Java was a business failure -- or at least I'd be interested to see someone defend Java as a business success.

The name Sun Microsystems is one that I, and most people, associated with microsystems - with workstations (and servers). Sun was founded in 1982 with Bill Joy as a founder, who was already a renowned programmer from CSRG. Andy Bechtolsheim had an exciting hardware design. Vinod Khosla and Scott McNealy risked starting a new company, and had enough chops to manage it into a success. Sun IPOd in 1986. Sun almost bought Apple right before Java was released.

Java was released in 1995. Sun Microsystems already had over a billion and a half dollars in revenue a quarter, which was pretty good at the time.

One reason for Java was protective - Sun wanted a language people could run on workstations and servers that it had some say over. A free, open source language was part of that, it helped in its adoption. Java was more of a defensive move, that Sun wouldn't be locked out of some kind of language ecosystem controlled by Microsoft or Netscape or other companies.

Sun wasn't a software company, they were a hardware company that released a product to defend their hardware from being locked out of competitor's ecosystems. They also weren't in a position to put out a computer language and force programmers onto it, being free and open source, and the ability to run on multiple platforms helped in its acceptance. Being able to run on multiple platforms was the whole point, they didn't want people to get locked into the Microsoft ecosystem with the release of NT 3.51 etc.

This is a bit meta, but why does everyone refer to a company's revenue instead of profit or margins? Ignoring operating costs, if you manufacture 1M units for 20M dollars and sell them for 19M, you have a revenue of 19M despite losing money, so revenue says almost nothing about a company by itself

Probably because it’s the biggest number. If someone wants to make an argument that a company can afford to do thing X, it’s common to just point to the revenue and say, “X has a revenue of $200M a year! Why couldn’t they do it?”

> The company that created Java went out of business and was sold to Oracle for a fraction of its original worth. So Java was a business failure -- or at least I'd be interested to see someone defend Java as a business success.

Sun and its peers declined because PCs running Linux became good enough; selling commodity hardware is a different business model (https://yarchive.net/comp/sgi_market.html) and it's incredibly difficult to move down market.

Oracle acquired Sun because they didn't want Java be owned by a competitor (https://www.reuters.com/article/us-oracle-alphabet-trial/ora...). Java was also why IBM offered to buy Sun; it seems hard to consider Java a business failure when it's most of the remaining value of the business.

If their argument is that they can't make money with open source, then they should say that, instead of this list of non sequitors.

Sun went bankrupt for more than just Java, it was a complete mismanagement of everything, and they only opened up Java on the last days, until Java 6 it was pretty much only a free beer thing, no source to play around.

Additionally Java has been a major success to those that played into the ecosystem, people keep forgetting there were several (some are still around) commercial implementations of Java with features not available on the free beer JDK like AOT compilation.

The way Sun managed Solaris and SPARC hardware had more impact on their wallet than Java.

Java wasn't a business failure. SPARC was a business failure, and Sun had an incompetent CEO.

Java ultimately kept their declining hardware and operating systems afloat for longer than they would have been otherwise.

How about just sticking a proprietary, "source available" license in the source and including it on the product download? Developement stays private, control stays in their hands, and it's still proprietary.

>The company that created Java went out of business and was sold to Oracle for a fraction of its original worth. So Java was a business failure -- or at least I'd be interested to see someone defend Java as a business success.

Just as Java is an afterthought to Oracle's database sales today, Java originally was an afterthought to Sun's workstation sales. Java was never intended to make money directly but rather as a conduit into the money-making side of both owners.

Sun was already on life support when they tried the open source strategy. Their failure isn't related to open source.

ahem it obviously was a business service model success as Oracle bought ti for exactly that reason!

Counter: Android is open source.

because it's a vehicle to deliver other, profit-making products. I.e. Just come up with a different business that makes hundreds of millions of dollars and use that to pay for Wolfram !

Covered by point 10 (quid pro quo) - Android is both a Commercial Trap (effectively requires Google Play services) and User Exploitation.

Java was killed by Microsoft using their famous Embrace, Extend And Extinguish method. It's on the wikipedia page as an example of the EEE method.

The fact that Java remains at all today is a testament to the power of open sourcing something.

And ironically Google gets a free pass from SV cheering crowd for trying to do the same.

It's harder to root for Oracle than it was Sun.

No one else bothered to buy Sun, so.