The Assassination of CentOS Linux by the Coward IBM RedHat

3 min read

Let us get a few things out of the way first: I am (and have been) a Red Hat Certified Engineer for the better part of a decade. I had been a Red Hat advocate for longer than that - encouraging both employers and clients to pay for licenses, entitlements, and then subscriptions as the model changed over time. I have also been, for smaller employers and clients, a CentOS advocate - knowing that if I can get the organization to grow in their technology use and their trust in both me and linux as a stable, enterprise platform, especially during the Microsoft FUD years, converting them from CentOS to RHEL would be an easy sell.

And it was.

And now I am ashamed. I am ashamed of Red Hat and I am ashamed of the CentOS team and its board acquiescing to the idea that rescinding a community commitment is acceptable. Arbitrarily slashing a commitment to stability and even more importantly security is a horrendous betrayal.

Today The Register published excerpts of an interview with CentOS board member and liaison for Red Hat, Brian Excelbierd, in which he states that CentOS discontinued long term support and abandoned its community commitment for CentOS 8 because Red Hat pulled the funding rug out from under it.

Well that is just moronic. And to steal from Aaron Sorkin, “Post hoc ergo propter hoc.”

For those of us who do not speak latin it is a logical fallacy that translates as “After this therefore because of this” or if you want to get really deep, affirming a conjunct.

If CentOS had gone to Amazon’s AWS and Microsoft’s Azure, both organizations which have a vested interest and have money to invest, and explained to them that Red Hat was pulling the funding rug out from under the project, my perception is (and full disclosure, I am an employee at an AWS consulting partner) that AWS and Microsoft’s Azure business unit would have worked to help with the shortfall.

That apparently did not happen, and that could be that “IBM Cloud” is a competitor, but it could be also that the decision was rendered very quickly and nobody asked because while, according to Excelbierd, Red Hat appreciates open source independence, in actuality CentOS has effectively been a wholly owned project since 2014 when the intellectual property was sold to Red Hat.

I am not about to say that Red Hat can not do business in a way that is productive financially for Red Hat, but every time someone at Red Hat says “this was not IBM’s decision” the phrase “methinks thou dost protest too loudly” runs through my head, and the massive public relations push that has been made saying “this is great for the community! It is basically the same” by Red Hat employees or associates reinforces that.

Let us not pull punches: I get that CentOS is a money suck. It reuses something that is built for profit and gives it to the community for no monetary transaction, or “free as in beer.” That is bad for Red Hat which, as an organization, wants to profit off of Open Source. The margin, I think, is now the question.

Technical analysts will tell you that it is really hard to make increased margins for the failing parent (and let us not believe that IBM is anything but an ever so slowly dying behemoth) if 90+% of your potential customers are using a “free” version of your software.

But this is not the era of business machines in the 60s, 70s, and early 80s, nor is it the era of paid-for language compilers to build software of the late 80s and most of the 90s. The thing that people pay for now is trust and security. It is not trust OR security, it is both.

The assassination of CentOS was performed by Red Hat - whether by direction and/or influence of IBM or not - by the cowardly undermining of trust that the community had placed in it.

So shame on you, CentOS, shame on you, Red Hat, and shame on me for believing you would keep your word.

Image of Stephen Sadowski

Stephen Sadowski

Leader focusing on quality, delivery, technical debt management, and leadership education about DevOps and SRE practices