Did IBM make a $6.4B blunder by buying HashiCorp? In some ways, IBM paying a cool $6.4 billion for HashiCorp makes perfect sense. HashiCorp's infrastructure-as-code (IaC) tool Terraform is very popular and would work well with Red Hat Ansible. And, yes, I've heard the joke about how if you put them together, you'd get "Terrible." Seriously, though, Terraform and Ansible …
Why strategize and waste limited brain cells trying to do something modern, like actually use Open Source ?
When you're IBM, you do things the old way. You spend lavishly on companies because their product "fits" one of yours. Integrating an open source project doesn't show up in quarterly earnings, not soon enough to make a bonus's worth. Buying a company allows you to pump your balance sheet, gives the CEO something to put on his CV and lands a few cocktail parties and bonuses for the high-level peons involved.
Just integrating open-source code means that only the programmers get to share a beer in their success. Where's the fun in that ?
Do not assume that IBM's motives are always evil. They probably are highly dependent on Hashicorp which is #1 in the USA; It's a key component of the Linux open source and cloud community, and remember that IBM runs a cloud that is trying to catch up "To the big 3". And they see that HashiCorp is being managed just terribly, horribly, the company is floundering, the CEO is a lunk who is pissing off all the customers, and HashCorp's cloud offerings are being abandoned in droves, and this company that IBM depends upon is endangered. If the price goes low enough they might buy it just to fix the shitstorm that HashiCorp created, all by themselves. They might buy it just so nobody else buys it to shut it down (Broadcom) or twist it to fit their own needs (Broadcom).
To be fair to IBM, they were a very early corporate Linux supporter. I still remember when the iconized "IBM loves Linux" started to appear everywhere, and the $1b pledge sounds like peanuts today but was enormous back then.
I have no clue why they did what they did to RedHat, but let's hope they don't make the same mistake twice and revert the BSL. It'd be good for everybody.
System/360 operating system has been a gold mine for IBM since the mistake was released in the mid 1960s. They recognize that whomever controls the operating system that runs the world, controls the world, and can ask for whatever tax they want, up to a point. So of course IBM could see that the world was going towards Linux, so it would be important to purchase the #1 Linux commercial operating system provider.
"HashiCorp's programs are ideal fit for IBM/Red Hat's software lines, but why buy the company when the software's free and open?"
Presumably this is so that IBM can turn the open source software into closed source software to try to make $$$.
I guarantee that this will fail since a fork of the software will be made that will remain open source. Redis tried to pull that misguided stunt a few weeks ago and the only thing that happened was that the open source Valkey was created instead and will now by used by Redis' former customers.
"And on the same lines, with Oracle yet again, there's OpenOffice and LibreOffice."
That was a case of closed source being released into the open, not an open source product being acquired. Oracle inherited OpenOffice from Sun. Sun had acquired the German company Star Division —creator of StarOffice— in 1999, and released Star Office, together with its source, forming the original source code base of OpenOffice and LibreOffice.
I guarantee that this will fail since a fork of the software will be made that will remain open source. Redis tried to pull that misguided stunt a few weeks ago and the only thing that happened was that the open source Valkey was created instead and will now by used by Redis' former customers.
To be more precise: "... and will now be used by Redis' former non-paying users".
There's no particular reason for Redis' paying customers to move to the free fork. They could have used the free version of Redis previously, and chose to pay instead - for various reasons that made sense to them (e.g. support availability, supply chain or licensing policy reasons)
It remains to be seen in the long term, for products like Terraform, Redis and Elasticsearch, whether they remain viable without participation from the wider community. On the one hand, they will certainly lose contributed code, maybe some good testing and bug reporting, and exposure to potential future customers. But those were already companies that had full-time staff doing the majority of development, presumably concentrating on the needs of their paying customers more than the wider community. Especially in the case of Hashicorp under IBM, they now have hooks into all the largest corporations with the biggest pockets - the ones who are happy to pay for everything.
Not sure IBM cares too much about the command line terraform which was the only open source bit. It always struck me as being an 'open core' project of sorts. It's HCP Terraform or the enterprise stuff they're after, none of which was ever free (unless you've less than 5 users but IBM doesn't care if that's you). True, most terraform adopters only ever ran the terraform cli tool and built their own management around it with extra bits of scripts and CI/CD platforms but in my opinion the terraform tool itself is not of much use without a management process and supporting infrastructure around it except in the most simple of cases. So from IBM's perspective I reckon the merger of ansible and terraform into a unified platform with a web front end, lots of graphs, and reports for management types will be a winner for enterprises. For everyone else there's the forks.
The reason,of course, that everybody is rolling their own rather then pay the somewhat silly fees for TF Cloud is that a) the fee structure is silly, and b) TF Cloud's functionality isn't all that hard to emulate (I once did a passable imitation in three days). Note, too, that one of the reason that this open source stuff is so popular with the admins is that the senior ones have all been bitten hard by the commercial stuff at some point. Not worth $6b IMO.
Either that, or lump in a bunch of other failed business decisions together with this one during a re-org, then talk about how "changing market dynamics have forced us to make difficult decisions about where former synergies can no longer be leveraged across our solution portfolio" while they cut the team
how in the world terraform and ansible are good for each other? it's clear as day they are competitors, each having about a third of the market, IBM heavily invested in Ansible over the last few years after RH acquisition, and now having TF too, it's like having a split brain. They now have two tools that do exactly same thing. This is definitely a bad move for IBM and even worse for TF, since IBM will play with it and will throw it away like it does everything else.
They are good only in comparison. I have many complaints about Terraform, but I also have many complaints about Cloud Formation, and that one only runs on AWS anyway and sometimes I don't want to lock myself into Amazon. I started out with the servers as pets paradigm and I still have a lot of them managed that way, but there is a reason to use IaC tools. When doing so, support for a lot of components often gives one tool the advantage over another one with better syntax. Hence, Terraform and OpenTofu are the ones that tend to get used just because they are generally capable.
Your analogy is bad. Terraform, for all its faults, works. You can build lots of things successfully with it. Do I wish it was different, having had to use it? Yes, I do. It might be nice if I could define my own functions, because it's a functional language and trying to use one when you can't use the central part of the philosophy is a pain. Still, it is capable of doing the job that people want done.
Could it be built with better syntax and semantics? Again, yes. Probably many people have. Their versions, though, don't get adopted as much because they don't have the premade structures that Terraform and OpenTofu can rely on, so if you're using systems that the original author wasn't using, you have to do the foundational work. Meanwhile, Terraform has the network effects meaning that most companies will build the necessary components for their own stuff rather than expecting their users to do it. Terraform is often used by people who aren't programmers in their own right and would find it difficult to write that stuff themselves.
Can you fix this and produce a better tool that also has broad platform support? Brilliant. People will enjoy that. If you can't build that and all you can do is complain, you may not get anywhere. If you can't even understand the thing well enough to understand why your complaints aren't someone's main focus, you definitely won't get anywhere.
Sorry, they do different things... though the article doesn't really make this clear and I'm not buying the "they have 30% market share" each thing either as they aren't competitors.
Ansible - configuration management, configures your infrastructure (real competitor : something like Chef or Puppet).
Terraform - infrastructure as code, creates your infrastructure (real competitor : something like AWS Cloudformation).
There are points where there is crossover but not on the whole.
Create the infrastructure first with Terraform, then configure it, install software etc., with Ansible. You could create custom images first with software preinstalled and use them with Terraform to create your infrastructure, but in some situations that may not work as well.
So, yes, you can just use the recently-branched free and open source version of the various Hashicorp software; or I suppose branch it yourself off the last version provided under the open source license. But, Hashicorp has a paying customer base (that, if IBM's products indeed fit in so well with Hashicorps, they may be able to get some extra sales to, or with this wider portfoloio get new customers as well.) And they have developers that have an intimate knowledge of the code base (from what I saw in the past, Hashicorps software was largely developed by them, not like they are just coordinating the patches and almost all development is from outside contributions.)
With VMware on the ropes and people in the process of migrating away from their products in droves, this is a good time to acquire something just like this.
Did they overpay? That I don't know, that does sound like a lot of money. But this is also a lucrative market.
The only way they could absorb whatever new developments the open source community comes up with on the fork is if what they absorb it into is also under an open source license. So, first off, switch the order of your “absorb the fork, open source their version again”.
And then the next step, turning it closed-source again... That's only possible one of two three ways:
1) Rip out all those contributions. (Provided they were contributed under a proper Copyleft license. How's Tofu licensed?) Then they wouldn't gain anything from the whole exercise.
2) Acquire copyright assignments from all contributors whose code they want to include. (Again, same caveat as above.) You see that happening? I don't.
3) If OpenTofu is licensed under some idiotic ultra-“free” license. (How is Tofu licensed?!?) It should be obvious enough by now that this serves no purpose except enabling precisely the scenario you describe, so... Honestly, if the Tofu people have been so utterly moronic, I can't even find the energy to be sorry for them. Nobody but themselves to blame if it's so.
Whichever IBM business unit now owns this (let's say Software Group) will absorb its new revenues into its own and the figures will look good. The acquisition cost will be absorbed into IBM as a whole (and IBM does still have deep pockets). That business unit will look good, and shareholders will be appeased, and may be spun off at some point, with IBM divesting itself of staff, legacy products and stable low-spending customers in the process, IBM recoups its costs, gets rid of a bunch of overheads at little or no cost, and makes a bit on top, keeps looking viable as a business and the shareholders (which includes the senior execs are rewarded). Its a race to the bottom.
Uh, normally modus operandi is to make insanely complicated additions of dubious value. One of which might add value to customers, by accident. Then IBM throws every consultant they can muster selling licences and consulting services, as the only people that semi-understand the arcane conditions and uttered incantations as to which it will work .... sometimes.
HashiCorp somehow proved that you can run a huge Open-Source project while never listening to your community, never integrating any of their contributions, and doing an absolutely terrible job of running a cloud runner service (Terraform Cloud) based on your beautifully functional but also very-twisted programming language, Terraform. Hashicorp managed to piss off absolutely everyone. At least 4 companies were launched to create a decent CI/CD cloud runner service (env0, spacelift, pulumi(sorta) , and so forth), and the rest of the people who were angry that their open-source contributions were never accepted, gladly forked the software saying, "Fine, you don't play by Open Source rules, you lose all control of your software, and that begins NOW." Their CEO who is apparently just a terrible manager then in an act of desperation tried to shut the barn door once the horse got out, by changing to a BSL ...
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
