The arguments made in this article are misleading in my opinion. The article claims that the Facebook patent grant is no big deal because: (1) Yes, it has d

The arguments made in this article are misleading in my opinion. The article claims that the Facebook patent grant is no big deal because: (1) Yes, it has defensive termination, but defensive termination is standard. (2) Even Apache's own Apache 2.0 License has defensive termination. (3) And while Facebook's defensive termination language is on the broad side, it is similar to IBM's CPL 1.0 (an OSI approved license). The first two points are completely true, but the third is misleading at best. First, IBM's CPL patent grant triggers termination if you sue a project contributor for software patent infringement. Facebook's patent grant terminates if you sue ANYONE if ANY PART of your patent claim relates to ANY product or service of Facebook or ANY of it's corporate affiliates. That's about as broad as you could possibly get. Does that mean that suing a Facebook shareholder like Microsoft for patent infringement totally unrelated to Facebook could invalidate your React patent grant? Who knows? I bet your lawyer hates you for asking. Second, the article claims that the "CPL is already approved by the Open Source Initiative as compliant." Using the CPL 1.0 as a point of comparison of what is a normal OSI-approved license is completely silly. If you check the OSI site, the CPL is technically classified a "Superseded license" which means it's not intended for further use. It was formally replaced by the Eclipse Public License (EPL) about a decade ago. Most notably, the newer Eclipse license has a much less strict patent termination clause. So why did Eclipse go through all the work to re-write their license and replace CPL? This is important - I'll cut and paste directly from the Eclipse Public License FAQ [1]: "Why was the EPL written? [..snip..] it addresses concerns some Eclipse Foundation members had with how the CPL deals with possible patent litigation." In other words, the CPL 1.0 had the exact same patent grant criticism lodged against it - so Eclipse ditched it and re-wrote the license and deprecated the earlier license. So saying that "Facebook's license is normal" because it's a more broad version of a license that OSI suggests you no longer use is pretty misleading in my opinion. [1]: https://eclipse.org/legal/eplfaq.php#WHYEPL https://eclipse.org/legal/eplfaq.php#WHYEPL