Opinion The true opponent of digital sovereignty is “enterprise IT” advertising and marketing, in line with one Crimson Hat engineer who ranted entertainingly in regards to the repeated waves of bullshit the trade hype cycle emits.
Throughout a espresso break at this yr’s CentOS Join convention, The Reg FOSS desk paused for a chat with a developer who was shocked however blissful to seek out us there. We can’t title them – we’re certain that they’d desire to maintain their job moderately than take pleasure in a second of fame – however we a lot loved their pithy abstract of how IT has confronted repeated waves of company bullshit for at the very least 15 years now, and the way they keenly and enthusiastically anticipate a large-scale monetary collapse bursting the AI bubble.
This vulture has been working within the tech discipline for some 38 years now, and the Linux developer we spoke with has been within the enterprise almost as lengthy. We each agreed that the late twentieth century – broadly, the interval from the early Nineties onward for a decade or so – had principally been one in every of pretty regular enchancment. Then, they steered, roughly following the 2008 credit score crunch, we have had some 15 years of bullshit in tech.
They referred to as out about half a dozen explicit cases of what they thought of to be bullshit know-how. We have been too busy laughing sympathetically to whip out a laptop computer to make notes, however as greatest as we are able to recall the sequence, they have been:
- Containers
- Kubernetes
- The “Cloud”
- Something in any respect “as a Service”
- The Blockchain – something, all the pieces, based mostly on it
- And now, arguably the largest and worst of all, “generative AI”
Including again a number of the moderately invective-laden commentary…
Containers: Certain, sure, they work, they’re helpful for testing. However they are not a deployment technique. You should not want them. Something you could run in a container, you possibly can simply run on the naked metallic, and if you happen to’re not competent sufficient to get – and preserve – that working, then you definitely in all probability aren’t competent sufficient to deploy a container both.
Kubernetes: If you happen to do not want containers, then you do not want one other vastly extra difficult device to deploy these containers. The probabilities are, you aren’t a huge multinational that should be capable to face up to ten million potential new prospects visiting your web site abruptly. It will not occur, so you will not lose any of that imaginary enterprise.
(That is generally referred to as the Use One Massive Server method, and in our humble opinion, it has nice advantage.)
The cloud: Nebulous by title and by nature. Who thought it was good to take all of your firm’s vital information and hand it to some web rando – in all probability the bottom bidder – trusting them to retailer the crown jewels, preserve them secure, and by no means ever peek at them. If that sounds cheap to you, perhaps it’s best to attempt promoting homeopathy.
(This may be summarized as There Is No Cloud – There’s Simply Someone Else’s Pc, and was being spelled out in clear language in 2015.)
Something “as a service” – it would not matter what: Infrastructure as a service – if you happen to want servers, purchase servers, or hire your personal non-public servers. No person else will ever care as a lot about your servers as you’ll. Platform as a Service – now you do not even get servers, simply OS cases. That is even worse. Software program as a Service? Now you do not even know what the server is, or the place it’s, or what it is working; you do not get software program, and you do not even know what information you might have or the way it’s saved – you are paying for entry to your personal stuff.
(The issues with this complete idea arguably return to Peter Deutsch’s Fallacies of Community Computing some 30 years in the past.)
The blockchain, and something constructed on the blockchain: the world’s slowest and most-distributed database. Cryptocurrencies? Hashcash on the blockchain. NFTs – URL shorteners on the blockchain, solely they’re longer moderately than shorter. Nugatory. Web3? Get ripped off, as a service.
Which brings us spherical to “generative AI” or, as we desire to time period them, giant language fashions, powered by the transformer algorithm. If The Monetary Occasions can clarify the way it works to a banker in a few thousand phrases and some minutes, it could possibly’t be that difficult or onerous to know, and it is not. It is predictive textual content turned as much as 11. It will probably’t even depend. As Daniel Stenberg, writer of curl, caustically noticed:
(This vulture laid out a few of his case towards it when Gentoo and NetBSD banned LLM bot slop in 2024.)
Truthfully, we will not fault any of this reasoning. We have appeared into the chronological sequence of the waves of promoting drivel, and it isn’t fairly how we anticipated. Though the earliest point out of Salesforce.com we are able to discover on The Reg is from 2002, once we referred to as it “comparatively new,” it was based in 1999. Maybe the primary mass SaaS providing to most of the people was Google’s Gmail in 2004.
Cloud computing within the sense of automated creation and deployment of VMs arguably dates to Amazon taking Amazon Net Providers dwell in 2002.
Whoever “Satoshi Nakamoto” is or was, their paper [PDF] introducing Bitcoin was revealed in 2008, though it did not come to The Reg’s consideration till 2011.
2008 was additionally the yr that the primary model of LXC (you possibly can nonetheless discover model 0.1.0 on the downloads web page) was launched. Docker debuted in 2013 however your humble correspondent had predicted that Linux containers can be the Subsequent Massive Factor a few years earlier, again in 2011. We reckon we referred to as it.
Kubernetes first appeared in 2014, though Google had been working “Borg” internally since round 2008. We nonetheless harbor a cynical suspicion that Mountain View threw it over the wall for no different cause than to distract the Penguinisti and preserve them busy.
So in chronological order, these are:
-
1999: SaaS
-
2002: Cloud computing goes mainstream with AWS
-
2004: SaaS reaches most of the people
-
2008: LXC frees software containers from FreeBSD’s Jail and brings them to Linux… and Bitcoin delivers full industrial-scale mass-production of the traditional rip-off of the pyramid scheme
-
2014: Kubernetes is loosed upon an all-too-willing tech world
-
2022: ChatGPT opens to the general public, or as The Reg calls it, one other AI to fill the world with kinda-true stuff
All proper, sure, moderately greater than 15 years. “A century of tech BS” appears a bit excessive when it is solely 2026, but it surely actually feels that lengthy.
Clearly there are various extra potential candidates, however we thought this was a superb high six. A few of the different contenders are extra area of interest, from the eternally terrible Jira to the challenge managers’ faith of Agile. Which of your bêtes noires did we miss? ®
