When I started working as a developer - a long time ago, we were under constant supervision. For example, one had to ask the architects team for every new library. I remember one of the "official" library was iText, probably because reimplementing its features was deemed to expensive. Yet, though Struts was available, we had our own MVC framework. Just like any other company at this time.
Since that time, a lot has changed. The Apache Web Server has become ubiquitous, even though some companies still cling to IIS. But more than that, most libraries that we developers use are Open Source. Most software that we use is actually Open Source. Open Source has effectively won.
Even more, some companies deliver Open Source software. Of course, in the Java ecosystem, names such as Red Hat’s WildFly, Eclipse, JetBrains Kotlin, SonarSource’s SonarQube or Google’s Guava are familiar. What’s amazing is that some companies which core business is not software also provide Open Source.
Below is a list of Fortune 50 companies that have a Github presence, and some information about them. I also included banks at the end. As well as Microsoft for good measure.
|The limits of the correlation between Open Source and Github are very clear to me. However, that’s an easy common denominator. This is a blog post after all, not a PhD thesis.|
# of repos
# of people
Most recent update
Highest number of forks
Apache 2.0, MIT, EPL 2.0
Python, C, Java, C#, Limbo
Swift, Python, C, C++, HTML
Forks from third-party repos
Apache-2.0, BSD-2-Clause, MIT
Apache-2.0, MIT, BSD-3-Clause
C, C++, BitBake, Diff, Shell
Mainly forks from third-party repos
Forks from third-party repos
MIT, GPL-2.0, Apache-2.0, BSD-3-Clause
Apache-2.0, BSD-2-Clause, LGPL-3.0
Scala, Python, C++, Clojure, Go
C, Java, C++, Python, Perl
Many metrics can be use to evaluate a project’s traction:
In favor of simplicity, the above table displays the number of forks
There are some takeaways (and questions) that arise from this limited data sample:
- The case of China
Interestingly enough, no Chinese-owned Fortune 50 company has any Github presence. There are some possible reasons for that:
- State censorship in recent times
- Business unwillingness to host company data on a foreign website
- Cultural unwillingness to use foreign software
- Business domain
Of course, software-related companies have a lot of projects and contributions on Github. That’s expected. But companies from other business domains are also represented.
- Eclipse Collections
- Only third-party repos
On the other hand, some companies are present on Github but only (or mainly) have forks from third-party repos. Compared to those that create their own repos, their contribution require another kind of analysis.
- Companies with no people
Some companies list 0 members. I frankly don’t know how to interpret that.
- Language repartition
From the above sample, languages are pretty well represented. It’s interesting that some exotic languages are used nonetheless (e.g. Clojure, Scala, Perl).
- Repos with no license
A huge surprise is the lack of license type in most repos. Some possible reasons include:
- No Open Source strategy. Or even worse, a completely Maverick approach drive by some developers. "Open Source is cool".
- No collaboration with the legal department
- A focus on the technical side of things. "What’s the usage of a license anyway?"
- A combination of the above
- No real license strategy
On the same side, some companies have as much as 4 different license types. Is that a lack of understanding? Or a lack of central license management? Or perhaps there are actually good reasons for that…
In all cases, the fact that Microsoft, a.k.a. the legacy king of the closed source model, is actually one of the lead organization on Github is a definite proof that Open Source has won. We software developers should be happy about that.