I guess that in some way a lot of companies are taking from open source without giving enough back. Open source is much cheaper than proprietary software to produce, a team of a few guys do the work of what would be a much bigger team in a proprietary software, and there is no need to have a profit, paying the developers is enough, so... if companies using open source would give back even a small amount of money...
But maybe some global "fund" should be used or something like that, I don't think that giving to the project you use is going to scale (we don't want a few guys rich, and nothing for all the rest, otherwise it's a company). But then bureaucracy starts to be an issue, who selects the right projects? And so forth. Not trivial.
Or maybe the solution is that if you are an enough large company, hire N open source developers. Like charity there should be a non written rule that most tech corps should follow. Like the 1% rule: hire 1% of your work force to write open source (and probably this would end being a largere return about branding than many ADs large companies run for more money than that).
I don't think resources are really the problem. There's a ton of money out there that pays for full-time open source developers. Basically all the core contributors of all the "serious" projects (i.e. things shipped in high-revenue-producing contexts) are already on staff somewhere.
It's true that there's a much larger population of people (like me) who want to work full time on open source and would take a job if it was offered. But that just gets to what the linked article is talking about: companies don't generally "staff" free software projects with new hires. They hire the projects that already exist.
So basically the solution, if you want one of these jobs, is to get off your butt and make something people want. Then give it away. :)
> There's a ton of money out there that pays for full-time open source developers.
I don't know... I'm not sure it's really a ton. And a great deal of it comes from companies that derive their revenue from something else - in many cases, proprietary software. Ubuntu, for instance, is all about Mark's millions made from the sale of Thawte.
> So basically the solution, if you want one of these jobs, is to get off your butt and make something people want. Then give it away. :)
And then hope it's useful to someone who makes a bunch of money via proprietary software or something else that's scarce, and that they hire you. Why not just make something proprietary of your own to be more in control?
For some definition of "ton" anyway. Developers are cheap, comparatively. But the point remains: all the major kernel developers are paid for. All the big X.org people too; the Gnome people (though I'm told that investment is lacking on KDE); virtualization folks behind Xen and kvm; Android obviously comes right out of Google's payroll, likewise Chrome and Mozilla products.
Really, if you were handed a big check and told to build a team of open source people, you'd have a hard time doing it without poaching from Red Hat, Google, Novell or Intel.
And the final point was exactly what I was getting at. Getting yourself a free software gig requires pretty much exactly the same things as getting yourself a funded startup. In some sense it's a subset. Whether you want to make proprietary things or not is, I think, a personal decision.
My last big open source effort, Hecl, weighs in at something like 100K lines of code. LiberWriter, my first real "pay for it if you want to use it" project, doesn't quite pay me enough to live on, yet, but earns decent money at around 10% of the lines of code.
I did make some money consulting on Hecl, so it's not a total loss, and I had fun with it, and learned a lot. But in terms of the money, there's simply no competition: LiberWriter wins by a huge margin, even though it's simpler code that doesn't do as much.
The big difference is with LW, people pay me directly. With Hecl, it was a matter of hoping someone would come along and pay to have some work done on it. It's just so much easier to deal with the former in terms of cash coming in.
Isn't that just saying that LiberWriter is a more successful product than Hecl? I mean, if Google or whoever wanted to base a service on Hecl, they would have hired you, right? My analogy was to startup funding, not sales, but I think it still holds.
Certainly if you make something people want, people will pay you to develop it. I don't see a real problem with the model, though for obvious reasons the incentives are going to drive more people to startups than to new open source projects.
Several companies used Hecl code, including one that was later sold to BlackBerry, so there was some demand for it commercially, just that apparently the code was good enough or clear enough that they didn't need any further involvement from me.
My point is that for people to start hiring you at a high hourly rate, there's got to be a burning demand, whereas with a product or app, the costs are much lower, so it's easy to spread them around.
> Certainly if you make something people want, people will pay you to develop it
Only after a certain critical mass, that is a lot higher than what it is for a proprietary project, where all sales feed back into the project. I mean, I certainly don't pay people to develop Linux, Postgres, Ruby, Rails, gcc or any of the other projects I use, although I do try and contribute back where possible.
BTW, strange but true story: I probably made the most money off of Hecl doing a BlackBerry port for... of all people, PASOK, the Greek center-left political party. This was all before the country ran out of money, and I did get paid.
That's the problem IMHO. Company get sold for X, and did substantial business with Hecl? Get x/200 or alike and give it to the Hecl developer as a donation.
It's not too hard... Open source coders are likely to write more open source software, so it's a good investment for the collectivity.
My company, probably like many, uses a majority of open source software but rarely contributes back. As an aspiring open source developer (several projects in the works, not yet released), this has always felt wrong. Honestly, I haven't really raised it much with my company, though. Guilty as charged, although we've made some movement to release a few bits of code under an open license.
Interesting thing is that we currently give 1% of our profits to an environmental program through Patagonia called 1% for the Planet. Maybe we need a 1% for Open Source where companies can be encouraged to contribute developer or $$ to open source projects. Anybody want to work on that?
"Interesting thing is that we currently give 1% of our profits to an environmental program through Patagonia called 1% for the Planet."
I think you are already doing it right, honestly. If you devolve 1% to whatever non-profit / good cause IMHO it's already enough. The real problem is that most companies will not do anything like that (1% is a lot), otherwise a few 1% to environment, a few 1% to open source, the world wold be a different place, entirely.
IF open source is The Way, it wouldn't need subsidizing. It sounds like all utopian arguments: If Only everybody came around to the right way of thinking, we'd all be better off.
yep, but it's strange because it's hard to think at the whole IT economy of today without open source. The idea of startups itself was completely different. What Unix? buy a license, with proprietary hardware maybe, and then purchase the web server license, and the compiler...
So soon or later it's gonna happen maybe as everything is posing on this base.
I wrote this after the 'PG has lost the plot' thread. I didn't like the title, but the content was fairly good. People really seem to have trouble visualizing "that which is not seen":-)
Bravo and bingo. These are the exactly the frustrations I tend to feel as an unpaid, open-source coder.
And the next level of the phenomenon is that the entire software industry has shifted its business model to avoid having to compete with piracy or open-source. There are now, to my knowledge, two major for-profit companies building their own PC operating systems: Microsoft and Apple. There are four or five building mobile operating systems: Apple, Microsoft, Blackberry, Google, (maybe) Symbian. The number of for-profit IDEs and programming platforms is still mostly healthy, but much, much smaller than it once was.
"That which is not seen"? That is, the effect of this privative of profits on the real world? Almost whole fields have shifted over into the domain of unpaid, spare-time open source work. Thousands, millions of man-hours are expended for no salary and no equity.
Think what would happen if all the open-source programmers (paid and unpaid) went on strike. Now you're realizing the economic value of open-source code that goes completely uncaptured by its producers.
Yes, the result is that less gets done, because everyone has to find day-jobs, and that the software industry has changed almost entirely from product business models to service business models.
Now, another scary thought: what if someone stole the code and databases to your web-service? You say that your contacts and relationships are what really makes your web-app worth something, but frankly, that's a load of crap. I'm sure there are some starving Chinese coders who would love to steal your code and your data, bring up their own web-service identical to yours (but in Chinese), get a few million users, and then translate theirs back into English for North American and European usage.
The "relationships and business expertise" excuse sounds much like what everyone said when America sent its manufacturing overseas: "We'll do all the high-value services and design and development work here." Well, no, the manufacturing was itself high-value and the other high-value points in the supply chain eventually followed it, and we're China's bitch now.
There is an imperative to enable value-creators to capture some portion of the value they create as cash. Otherwise the entire foundation of not only capitalism but all possible market economies breaks down into mush.
I think it might be instructive to think of what the world was like before the idea of letting people who create intellectual products maintain some ownership of (and therefore be more able to make a living from) their work:
- Serfdom was still legal in most of Western Europe.
- Professional artists occupied the same social rung as prostitutes.
- An intellectual professional's primary option for making a living off his (that pronoun being sufficient for the era) work was patronage.
- The vast bulk of intellectual output served the primary purpose of glorifying the political and religious Powers that Be.
Of course that doesn't imply that the current copyright and patent law isn't enormously out of hand, or that it doesn't largely function to the detriment of creative professionals thanks to the grotesque ways in which it has been amended over the past three or so centuries. But the basic idea is admirable, and it can and should be salvaged.
I can play this game too: you know what the software industry was like before Microsoft and Apple? EVERYTHING was open source. When you bought a program (if they didn't throw it in for free with the bloody expensive hardware) you ALWAYS got source code. People still sell open source software; I've never understood why people think that commercial software is incompatible with open source. It isn't.
Everything was open source, but "everything" wasn't much. There was no VisiCalc, no WordStar, no Sierra adventures. Bill Gates's packaged software business model allowed a lot of new software to be created.
your own statement explains why it was like that: because software was just a small commodity passed along the real stuff (the hardware). Not much of an industry there...
??? What utopian world did you live in? Before Microsoft and Apple was before the Internet. Open Source was a dream, everything was proprietary, software engineers were just Coders.
What does it even mean to say "When you bought a program..." in the context of Open Source? Of course if you paid for it, it isn't open source.
The very GNU GPL itself explicitly allows code licensed under it to be sold for money. What makes you think that code transferred for money cannot possibly be Free?
Everyone always says this, but the majority of the community is very much against you making money on it and will destroy any chances if you making any sort of profit by releasing the source for free (which is allowed under the license).
What times you are talking about? Professional creators - like composers, poets, etc. - existed long before Berne Convention. Of course, then conditions were different than they are now - J.S. Bach had to write a new cantata every week as the condition of his employment, and also teach Latin and singing to the young. Of course, what was good enough for J.S. Bach is definitely can not be good enough for a person named something like "Jay-Z" or "50 Cent", as the latters' art is much more valuable for the society than Bach's.
As for serving Powers That Be - most of the creative people do it right now, only Powers changed. Some time ago, most disposable income was in the hands of aristocracy and later rich merchants. So they were served. Now due to huge technological advances pretty much everybody has disposable income - so artists cater to everybody. Some of them still cater only to the richest, of course - everybody has his niche. But except for the distribution of the disposable income, not much changed.
Now, another scary thought: what if someone stole the code and databases to your web-service?
This exactly.
It's easy for a community oriented around mostly around things like webapps, SaaS, enterprisey stuff to hop on the anti-IP bandwagon. It just so happens that there's no convenient ways of pirating their software.
It doesn't "just so happen". SaaS and web-apps arose deliberately on the marketplaces as a way of producing software that could not be pirated without actually getting past firewalls and cracking private servers.
Nothing would happen if somebody stole their software. Exactly nothing. You think if you somehow copied Zuckerberg's hard disk in 2004, you'd have the billion-dollar business now? No, you'd have a pile of 8-year-old useless code. The code doesn't matter, the people building the company and writing that code to accomodate specific (and rapidly changing) needs matter.
As to 2, If you code, it's a very hot job market out there right now.
You make a lot of what seem like handwavey notions that OSS coders should get paid or go on strike or something like that without really convincing why.
Most professional coders know what they use daily has some strong OSS roots (then there are other branches which have little involvement there). A subset of them do work on them.
The fact people make stuff in the evening that happens to be OSS is often a sign of they like to program not some economic crime.
OS kernels. Linus used to work for Transmeta (sorry if I misspelled the name) as a chips-and-assembly guy before they got enough charity funding to put him on full-time Linux kernel coding, remember? He started the kernel when he as a Master's student in Finland, too. It took until the late 2000s (after having started in 1992 or so) for the creator of Linux to be put on full-time payroll working on Linux.
>As to 2, If you code, it's a very hot job market out there right now.
It's a very hot job market but very, very specialized. I cannot tell you how many things I've applied to and never heard back, gotten interviews with and been turned down for having the wrong skills (ie: C, C++, Java, Scala instead of Ruby on Rails, HTML+CSS, iPhone, Android), been turned down from for no given reason whatsoever, been turned down from because I had admitted to applying to PhD programs (and that was simply the coolest company I ever loved interviewing with that turned me down like this, and a HN-featured start-up too!), and just generally not had offers forthcoming.
Honest to God, if you'd like to see my resume so you can tell me what the fuck's wrong with me that's creating this pattern, I'll send it to you, have at it.
Oh, and then there's the companies with broken jobs websites. There's nothing I hate more than trying to apply for a job I'd love and finding myself facing a page that says, "Oops, an unexpected error has occurred!"
Just a data point for you to consider - US and China have roughly equal manufacturing output now, despite US being in recession last years and China having 4 times more people. That is if we compare whole population, if we compare manufacturing sector it is 10 times difference in employees. So US workers right now are on average about 10 times more productive than Chinese workers (note I do not mean to imply anything bad about China or Chinese people - it's just statistical facts that do not reflect on anybody personally :). I would not describe this situation as "we're China's bitch" just yet.
Prooflink: http://www.manufacturingdigital.com/news_archive/tags/us/chi...
Of course, China would improve. I hope so would US.
Then imagine the Indians, or the Germans, or (to a much lesser extent) the Israelis. The point is, what if someone says, "My comparative advantage will be that I steal your code and localize it for people who don't speak your language"? Baidu is Chinese Google, Wallah is Israeli Yahoo. Once you've got web designs out in public, or important algorithms like PageRank in published research papers, these supposed "trade secrets" to running a web-service are actually dead easy to copy/steal.
The only reason Google doesn't feel the pain from Baidu is that Google never depended on China as their primary market. While we're talking about the Chinese versus other nationalities, by the way, the Chinese currently share the attribute of many rising superpowers (including, in its day, the USA): complete disrespect for "intellectual property" as subordinate to actual production.
Baidu and Wallah did not "steal code" from Google and Yahoo. You can argue they "stole" ideas of search and portal, but that would imply there's ownership of these concepts and it belongs to Google and Yahoo - which is wrong on both counts, both ideas existed long before their implementations by current kings of the hill, and nobody owns them.
If you think copying Google is "dead easy", you're well on the way to become a multi-millionaire. I've tried recently a number of emerging search engines, and none of them is as good as Google, despite the fact that copying PageRank is supposed to be dead easy. On the other hand, I'm sure Google's understanding of Chineese or Hebrew or Russian is not perfect and specialized engines can do much better (I know about Yandex on Russian market, but I'm sure there are more).
There's a place on these markets for much more than one player, so treating everybody being in the same search engine market as Google as "stealing intellectual property" makes zero sense. It's like treating everyone that makes smartphones as "stealing Apple's intellectual property". Maybe some hardcore Apple fanboys think that way, but this is not useful for anybody but them and Apple, certainly not for the consumers or for the public.
As for "respect for intellectual property", I don't see why Chinese should necessarily respect artificial mental constructs that were erected by Western politicians in order to produce some utilitarian value. These things are defined as utilitarian, the only reason that is being argued of why somebody can use the idea and somebody else can not is: "To promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries". If Chinese do not agree that this would promote Progress and useful Arts for them, why they should feel obligation to respect it?
>As for "respect for intellectual property", I don't see why Chinese should necessarily respect artificial mental constructs that were erected by Western politicians in order to produce some utilitarian value.
I don't either. But you've missed my fundamental claim: lacking any kind of "intellectual property" enforcement, pretty much nothing stops people from treating the code for web-services the exact same way they treat the code for video games or office suites: avast ye maties!
I didn't miss it - I actually addressed it in another comment. While stealing your website code would be unpleasant, it doesn't really matter that much. First, backend website code is not protected by copyright because it's not published - only the design is. Second, do you really think Facebook became Facebook because they found right magic color and right column widths to be successful, and if only you could use the same design you would be billionaire? This is obviously nonsense - Facebook success is composed of many factors, design being very very minor part.
As for DDG, I used it for a month to test, and unfortunately had to go back to Google - for my loads, its quality proved to be inadequate, Google found what I needed in more cases than they did. It may be because usually I search for very easy and obvious stuff. I would like to say my coworkers "why, you still use Google? Man, it's so 2000s... It's 2012 now, use something that's not ancient!" Unfortunately, it's not the time yet that I could do that.
But wouldn't you need to distribute for copyright protection to kick in? Copyright is exclusive right to distribute. If you just stole the code and keep it, you can be liable to various crimes, but I don't see how copyright infringement can be one of them.
As someone who has been paid to work on OSS, let me say that it is absolutely not about skill. Most notably, Google and Mozilla pay plenty people to work on Firefox and Chromium. Even Apple has some OSS projects (Webkit and Darwin come to mind).
Maybe its better to say that "there is no need to be a superstar programmer" to work on OSS. The companies you list aren't known for hiring unskilled workers :-)
Indeed, but they're by far not the only ones working on OSS. They're simply the most recognizable. I've worked on OSS for companies with far less stellar track records when it comes to hiring.
It was a bit tongue in cheek - I have actually had paying gigs working on open source in the past.
But the point about where the money comes from stands - Google's money comes from their proprietary stuff.
I actually think something like Google is a pretty good way to produce open source software, by the way, but it's made possible by the money made elsewhere.
I'm not sure it was necessary to refer to the "I'm not good enough" thing. It really has little to nothing to do with your point. Because you've just undermined your own example (Guido Van Rossum / Python / Google) above, it weakens your position.
If I get it correctly, your point is that one requires copyright for open source software to be viable, by opening up a revenue stream to support the (free) open source part. This is true.
You can make money form open source on either:
1) Selling alternate, source closed licenses. (MySQL for example)
2) Selling support, consulting. (Red Hat, many others)
(1) requires copyright. (2) maybe less so, but without the open source licenses, which require copyright law to work, the software wouldn't exist to begin with either. Red Hat uses trademarks (which really are similar to copyright) to reinforce their position. So does Mozilla.
I know you can make money from open source. It's difficult though.
The title was something of an attempt at linkbait (albeit not a very good one), but also a way to point out that effectively, it is harder to get paid to work on open source software.
How are they similar? Trademarks just prevent people from using the branding, and just represents a minor hurdle (see CENTOS), while copyright (with a restrictive license) would prevent people using almost anything.
If trademarks are the main protection, the main barrier to entry is branding, not technology.
Interesting perspective but I have to say there are quite a few companies working the whole "open source the code, charge for consulting, etc." thing. Just in data alone, Cloudera, Data Stax, Basho, 10gen and Hortonworks come to mind.
My point is that thousands of people get paid to work on open source projects. The project might not be their original creation but they still get to give back by producing something that everyone can use. Isn't that really the point anyway?
FIWI, I have personally not heard of a single one of those companies.
To continue the OP's concept, these are likely companies with employee counts in the single or double digits, whereas closed source companies have employees in the 5-6 digit ranges, reinforcing the "I'm not smart enough to get paid writing open source" concept the OP presented.
The problem with consulting is that you first have to have a product. That can take a lot of time and money to produce. Then, if you start consulting, it's all about billable hours. Hours spent fooling around making the product better may not translate, at all, into more money for the company.
This business model is fairly popular in education. We have tons of open source projects where we COULD run it for free, but sometime we pay for hosting, or consulting, or customization. It seems to work well too.
For those interested in sustaining themselves by making open / libre software, I plan on dedicating my next summer to starting a cooperative for hackers. The basic idea is that if we can create a decentralized ecosystem for exchanging currency, ideas, and goods in an open and cooperative manner, we can solve our needs as hackers without resorting to non-free alternatives. You can find my early planning here: https://github.com/SirDinosaur/hackercoop, and I'd greatly appreciate any questions, comments, critiques on how to make my / our dream possible.
In the UK: Taxes are complicated but mostly solved. Either your employers deduct taxes before you get your money or you are self-employed and pay an accountant to sort it out; and people you buy things from add taxes before you buy stuff. If you are rich you get an accountant and practice various forms of avoidance (legal, but those loopholes may close soon, and of dubious ethicality) to evasion (not legal).
Missing from the above description are people involved in barter economies. There are several barter economy systems in the UK. The UK tax collectors (HM Revenue and Customs) are clear: barter trades can be taxable.
Suggestion: Find an accountant that would like to work as part of a co-operative. The collective either has one big joint account (like a business) or lots of individual accounts - either way good accounting will legally minimise tax burden and help with other stuff.
Awesome, thanks for the suggestion. Do you suppose the accounting could also be automated, with help from an accountant, as part of the system that facilitates exchanges? My current thinking is to create a system that can be used by anyone, not just those within a single collective.
The article is confusing. He talks about intellectual property, copyright, and open source. He seems to confuse intellectual property and copyright (or at least, use them interchangeably), and seems think that open source and copyright are at odds. That open source licenses rely on copyright indicate this isn't the case.
Maybe I just misread the entire thing, but I honestly don't see the message he's trying to get across. It feels like their is a massive amount of context that is missing from this.
I'm paid to work on Open Source full time and have been for the better part of a decade.
I don't really agree with anything the article says. Open Source development is a segment of the industry. Getting a job in Open Source is really no different than getting a job in Game Development, Robotics, or any other niche market.
You generally have to (1) be willing to accept less pay (2) be more flexible about relocating (3) have specialized skills and experience or just get really lucky.
An interesting perspective. I too would like to spend more time writing open source software. The main reason I do not, and something which I think holds true for a lot of people, is that even if our company wanted to open source some of our software it's fairly difficult to do so and of questionable value.
Many of the components and practices in our software are quite specific to our business and environment. They'd just be useless or, worse, an impediment to other people wanting to reuse the software in a different context.
Even if we were to refactor parts of our code base to produce some "clean" components that didn't rely on the specifics of our business or other processes, there are many cases where that would just add to the overhead of maintaining the software for us.
Of course, another factor is how ending copyright might redirect a lot more resources into open source. Without being able to benefit from rents that come from keeping proprietary code, how many companies would switch to using and improving open source alternatives?
It's very easy and simple to pay $100 to use some bit of software. It's way harder to actually contribute to it meaningfully.
Software markets solve a coordination problem. Not perfectly, but reasonably well. Same thing goes for many other kinds of information goods. No way would I pay an author to sit around writing a book, but I'm happy to give them $5 for a copy.
Oddly, in the age of the internet, it is easier for someone to do open source at least part of the time for their employer. With the internet, a lot of the money for web based companies is not from the software itself, but the services the software provides. This allows for the company to open source a large chunk of the code base and still make money to pay the developers. (see things like 37signals/RoR)
This throws the absurdity of capitalism in our faces: Guido, a guy who developed software that benefits hundreds of thousands even by conservative reckoning, can't work on it full time.
But maybe some global "fund" should be used or something like that, I don't think that giving to the project you use is going to scale (we don't want a few guys rich, and nothing for all the rest, otherwise it's a company). But then bureaucracy starts to be an issue, who selects the right projects? And so forth. Not trivial.
Or maybe the solution is that if you are an enough large company, hire N open source developers. Like charity there should be a non written rule that most tech corps should follow. Like the 1% rule: hire 1% of your work force to write open source (and probably this would end being a largere return about branding than many ADs large companies run for more money than that).