Questions on the Future of Open Source
gist.github.com
gist.github.com
Large corporations tend to stay away from independent projects altogether. For example, my project has several thousands of stars on GitHub and gets tens of thousands of downloads. I've interacted with well known startups on many occasions but I've only once interacted with someone from a very large and well known corporation which was using my project and it was quite strange because when I asked if I could mention the company on my customers page, they told me that they have a policy against using their name as an endorsement... Yet when I look at other 'mainstream/corporate' OSS projects, their website is covered with big brand names.
I think that the problem is both that big corporations are less likely to use independent projects and also that if they do use them (out of desperate necessity), they don't want to admit it unless they know that other corporations are also using it.
Also, corporations are not homogenous entities. Perhaps other people in the company would have had the authority to say yes.
From what I've seen, there's very little incentive for any company to reveal the technologies -- FOSS or otherwise -- that they use, unless it's to attract job candidates or to position the company as a technological leader in preparation for a funding round. In either case, it's gonna be hype technologies that get on the website: Hadoop ten or so years ago, NoSQL five to seven years ago, some kind of distributed log three to five years ago, and so on.
Two possible reasons:
- Risk assessment: In 2 - 5 years, will there still be further development, bug fixes, especially of security bugs. Taking over responsibility for all those abandoned projects is not an option.
- Cultural differences: Large corporations understand requirements of other large corporations better. Like long-term maintenance commitments.
Legal departments in large consultancy companies have been working on "open source policies" internally (mostly on the "using open source software" angle, less on the "contributing" angle).
And those policies are specific/hard/complex enough to discourage public endorsement.
For instance, use of GPL-based software is strictly forbidden (for good reasons, when you take the perspective of the company); unless a specific review by legal allows it, in which case, you've already been through several layers of management with a convincing case for each of them.
The idea many people held (and still hold) is that if your project succeeds you'll somehow be inevitably rewarded. But in general this idealism is not really justified. The emergence of distributed computing 'exploiting' open source projects is just another example of reality along the same lines as the person who released an AI art generator and then was surprised when somebody used that generator to produce a piece of art that auctioned for $400k, yet chose to give exactly $0 back to the creator. Distributed computing working to sort of 'centralize' free software is mostly just another example along the same lines. Give something away for free and people will use it, monetize it, and you will generally see nothing particularly substantial in return.
It seems relevant to this thread.
"Corporate interests in open source and dev culture "with Zed Shaw" - http://changelog.com/podcast/300
A lot of the tools and libraries that I use every day are largely written and maintained by a single developer or a small group of developers. Some of these projects started before I was born and will keep on truckin' after I die. This sort of development has its ups and downs, but it's a good life in the long run.
are you very young and do you intend to die soon?
BSD derived works (including parts of MacOS, Windows and GNU/Linux, and numerous embedded devices like routers etc.) are still going strong today.
Most of the code will have been gradually replaced over such a long time, but there are libraries which can trace a lineage that far back. libc and libm come to mind.
Quite a few well known, non-trivial applications still used today were, perhaps surprisingly, first released in the 1970s. Examples: grep, diff, cron, awk, make.
Companies of all sizes can still benefit from releasing software as open source, so it's not going anywhere. But of course the way big companies behave is not getting unnoticed and is definitely going to reflect in future license choices away from permissive licenses for smaller companies and startups.