4 ms·

I think you've cleverly bundled together a couple of meaningless correlations. Yes, there's a grab bag (but, in my opinion, not much more than that) of importan

I think you've cleverly bundled together a couple of meaningless correlations. Yes, there's a grab bag (but, in my opinion, not much more than that) of important technological innovations which have come from large companies, but so what? You still have corporations with large amounts of disposable cash for R&D in a world without monopolies and intellectual property. The handful of innovations you can mention don't really have anything to do with monopoly or intellectual property, they just happen to have come from big companies - and there's really no evidence that those companies would simply collapse by being slightly more cooperative. The majority of important software innovations have come from the open bazaar, not the cathedral. The Internet; The Web; Firefox, WebKit; GNU/Linux, Unix; HTTP; SSH, SSL, AES; Python, Ruby, C, C++. Thousands of humble open source projects which have been distributed to and changed by millions of users. And even when large corporations do spawn projects, those that succeed are those that encouraged at least some degree of openness. Android from Google, Java from Sun. Open standards for computer hardware architectures and software have lasted and stood the test of time. Open code, projects, and standards have more than stood the test of time. We are actually living in a time of unprecedented open source success. The majority of computing systems run on significantly open source code. Millions of Android phones running Linux. Every serious server on the planet running on Apache or NGINX on Linux or BSD. A landslide percentage of users accessing them using Webkit, Blink, and Firefox. That's just the tip of the iceberg! Millions of lines of code are being updated publicly, in real time, by thousands of people across the world in collaboration. More so than ever before! Sure, maybe some people are a bunch of VC-backed business-savy Steve Jobs-worshipping iPhone mobile app developers now, but it's still sad to see someone say that tech innovation "happens in periods of monopoly" on HackerNews.

parasubvert 12y ago
"majority of important software innovations have come from the open bazaar, not the cathedral. The Internet; The Web; Firefox, WebKit; GNU/Linux, Unix; HTTP; SSH, SSL, AES; Python, Ruby, C, C++." The vast majority of those were not developed by an open bazaar, but rather by focused cathedral organizations: the early Internet, UNIX, C, C++, SSH, Firefox, SSL. The bazaar maintains them or cloned the success of others (SSH). I also think you're vastly underestimating the influence and power over millions that continues to exist from closed commercial software: Apple iOS, IBM zOS, pOS, and DB2, SAP ERP, Oracle database, middleware, and various HR/SCM/CRM/ERP suites, Microsoft Office and Windows / IIS. The big difference is that there is an order of magnitude more revenue and more people being paid to work with commercial software than with open source. That you claim all serious servers run Apache or Ngnix on *nix is profoundly ignorant of the millions of Windows servers and IIS servers out there, some of which power very large websites. I think Ngnix is a great server, but most of the large enterprises (banks, Telcos, etc) I've visted in my consulting run WebSphere or IIS and have never heard of HAproxy or Ngnix. This is not to say this is a static state of affairs. Certainly in many cases the open source alternative is a better technology or better project, sometimes even the better product. Just there's a big difference between an innovative PRODUCT and an innovative PROJECT. There are lots of the latter on Github, but other than saving people some money, they're not having anywhere the economic impact a PRODUCT has. There is scant evidence that open source leads to more innovative products. There's a lot of evidence it helps to lower the acquisition costs of supporting software for what usually wind up as partly proprietary products. All of the latest software kings: Facebook, Twitter, Amazon, Apple, Salesforce, Google - their primary money making innovations are closed source built on (mostly non GPL) open source . Their contributions of open componentry back to the community is welcome but limited. The only large and "purely" open company is Red Hat, and even they go through major hoops to make it inconvenient for others (eg. oracle) to use their open source code for their own purposes (ie. Offer a RHEL clone and loss lead on support). They remain innovative. But so is SAP. So is Oracle.
- AlisdairO 12y ago
  > but most of the large enterprises (banks, Telcos, etc) I've visted in my consulting run WebSphere or IIS and have never heard of HAproxy or Ngnix. WebSphere is an app server rather than a straight http server. The equivalent to Apache/Nginx from IBM would be IBM HTTP Server, which is an extension of Apache.
  - parasubvert 12y ago
    I didn't want to really get into the distinction (Apache is an app server too if you consider mod_*). Yes, IHS is common and they will often use that. As part of my broader point, IHS is a product, one of many Apache-based products. They're not using the open source project code directly. Even if IBM's "innovations" are modest, they're using it because IHS and the WAS plugin work best with WebSphere. I've managed to convince some to avoid this combo in favour of a simple HAproxy, but it's hard.
rayiner 12y ago
> The Internet; The Web; Firefox, WebKit; GNU/Linux, Unix; HTTP; SSH, SSL, AES; Python, Ruby, C, C++. The internet came from the ultimate monopoly: the government. It was a government project developed by government contractors. Unix and C were invented at Bell Labs when AT&T was a government sanctioned monopoly. AES was developed while the authors were at an university research lab. Stroustrup started C++ in his PhD work, and did much of its early development at Bell Labs pre-breakup. HTTP and WWW were developed at CERN, a government research lab, as was the first webserver. Its odd to use Python and Ruby as examples of the bazaar. They were developed by Matz and GVR, tackling the fundamental design decisions, before they were released to the community. Development doesn't mean innovation, though its important too. There is innovation in open source, sure, but making another OO scripting language isn't as innovative as inventing Smalltalk in the first place (which was done at PARC when Xerox had 100% of the US copier market thanks to its patents).
- grey-area 12y ago
  So far you've included university research (Google), large public companies (Xerox), state-sanctioned monopolies (Bell), companies taking advantage of network effects (Facebook), startups (Google's big innovation was in search before the company was born), and anything funded by governments in your definition of a monopoly. If you define monopoly in such a broad way it's really not surprising that most innovation happens there, because your definition encompasses most of human endeavour.
  - rayiner 12y ago
    You're misstating the categories. I'm making the distinction between organizations that are insulated from copying and competition and those that aren't. Xerox at the time of PARC wasn't just a large public company. It had 100% of the copier market and was insulated from competition by its patents. It was a large public company monopoly. Google and Facebook, circa 2014, also aren't your typical public companies. Google is protected from copying because they don't ship a product. Their algorithms are hidden behind a web-API, and they are kept essentially as trade secrets. Facebook enjoys network effects, which provide protection from competition and in antitrust economics are considered a factor that gives rise to natural monopolies. E.g. the Microsoft and Offices monopolies were built on network effects. It's obvious why government-sanctioned monopolies, and government and university research labs are insulated from competition. It doesn't matter if people copy their work, because they don't need to make profits in the market in order to secure their funding. I'm purposefully excluding start-ups and public and private companies that don't enjoy substantial protection from copying and competition. These make up the bulk of economic actors.

parasubvert 12y ago

"majority of important software innovations have come from the open bazaar, not the cathedral. The Internet; The Web; Firefox, WebKit; GNU/Linux, Unix; HTTP; SSH, SSL, AES; Python, Ruby, C, C++." The vast majority of those were not developed by an open bazaar, but rather by focused cathedral organizations: the early Internet, UNIX, C, C++, SSH, Firefox, SSL. The bazaar maintains them or cloned the success of others (SSH). I also think you're vastly underestimating the influence and power over millions that continues to exist from closed commercial software: Apple iOS, IBM zOS, pOS, and DB2, SAP ERP, Oracle database, middleware, and various HR/SCM/CRM/ERP suites, Microsoft Office and Windows / IIS. The big difference is that there is an order of magnitude more revenue and more people being paid to work with commercial software than with open source. That you claim all serious servers run Apache or Ngnix on *nix is profoundly ignorant of the millions of Windows servers and IIS servers out there, some of which power very large websites. I think Ngnix is a great server, but most of the large enterprises (banks, Telcos, etc) I've visted in my consulting run WebSphere or IIS and have never heard of HAproxy or Ngnix. This is not to say this is a static state of affairs. Certainly in many cases the open source alternative is a better technology or better project, sometimes even the better product. Just there's a big difference between an innovative PRODUCT and an innovative PROJECT. There are lots of the latter on Github, but other than saving people some money, they're not having anywhere the economic impact a PRODUCT has. There is scant evidence that open source leads to more innovative products. There's a lot of evidence it helps to lower the acquisition costs of supporting software for what usually wind up as partly proprietary products. All of the latest software kings: Facebook, Twitter, Amazon, Apple, Salesforce, Google - their primary money making innovations are closed source built on (mostly non GPL) open source . Their contributions of open componentry back to the community is welcome but limited. The only large and "purely" open company is Red Hat, and even they go through major hoops to make it inconvenient for others (eg. oracle) to use their open source code for their own purposes (ie. Offer a RHEL clone and loss lead on support). They remain innovative. But so is SAP. So is Oracle.

AlisdairO 12y ago
> but most of the large enterprises (banks, Telcos, etc) I've visted in my consulting run WebSphere or IIS and have never heard of HAproxy or Ngnix. WebSphere is an app server rather than a straight http server. The equivalent to Apache/Nginx from IBM would be IBM HTTP Server, which is an extension of Apache.
- parasubvert 12y ago
  I didn't want to really get into the distinction (Apache is an app server too if you consider mod_*). Yes, IHS is common and they will often use that. As part of my broader point, IHS is a product, one of many Apache-based products. They're not using the open source project code directly. Even if IBM's "innovations" are modest, they're using it because IHS and the WAS plugin work best with WebSphere. I've managed to convince some to avoid this combo in favour of a simple HAproxy, but it's hard.

rayiner 12y ago

> The Internet; The Web; Firefox, WebKit; GNU/Linux, Unix; HTTP; SSH, SSL, AES; Python, Ruby, C, C++. The internet came from the ultimate monopoly: the government. It was a government project developed by government contractors. Unix and C were invented at Bell Labs when AT&T was a government sanctioned monopoly. AES was developed while the authors were at an university research lab. Stroustrup started C++ in his PhD work, and did much of its early development at Bell Labs pre-breakup. HTTP and WWW were developed at CERN, a government research lab, as was the first webserver. Its odd to use Python and Ruby as examples of the bazaar. They were developed by Matz and GVR, tackling the fundamental design decisions, before they were released to the community. Development doesn't mean innovation, though its important too. There is innovation in open source, sure, but making another OO scripting language isn't as innovative as inventing Smalltalk in the first place (which was done at PARC when Xerox had 100% of the US copier market thanks to its patents).

grey-area 12y ago
So far you've included university research (Google), large public companies (Xerox), state-sanctioned monopolies (Bell), companies taking advantage of network effects (Facebook), startups (Google's big innovation was in search before the company was born), and anything funded by governments in your definition of a monopoly. If you define monopoly in such a broad way it's really not surprising that most innovation happens there, because your definition encompasses most of human endeavour.
- rayiner 12y ago
  You're misstating the categories. I'm making the distinction between organizations that are insulated from copying and competition and those that aren't. Xerox at the time of PARC wasn't just a large public company. It had 100% of the copier market and was insulated from competition by its patents. It was a large public company monopoly. Google and Facebook, circa 2014, also aren't your typical public companies. Google is protected from copying because they don't ship a product. Their algorithms are hidden behind a web-API, and they are kept essentially as trade secrets. Facebook enjoys network effects, which provide protection from competition and in antitrust economics are considered a factor that gives rise to natural monopolies. E.g. the Microsoft and Offices monopolies were built on network effects. It's obvious why government-sanctioned monopolies, and government and university research labs are insulated from competition. It doesn't matter if people copy their work, because they don't need to make profits in the market in order to secure their funding. I'm purposefully excluding start-ups and public and private companies that don't enjoy substantial protection from copying and competition. These make up the bulk of economic actors.