The state of software companies is pretty terrible. I have been on the acquisition side as well as the development / end-user side and it’s mind-boggling knowing the exorbitant costs with bare minimum value delivered, yet companies just keep paying whatever they’re told it costs, until it’s comically astronomical and the customers have to tell them to get bent. Yet still, software vendors keep changing their licensing structure until it meets that comically astronomical figure and pushing customers away.
Enterprise software licensing, support contracts, and technical account managers (TAMs) often run into hundreds of thousands or millions annually per organization. Yet, in practice, support tickets go unresolved or ignored, even for large clients.
The software quality of our most expensive products is extremely poor and unreliable, almost across the board. Many products suffer from bugs, outdated features, or incompatibility issues that disrupt operations. In development roles, this means wasted time on workarounds, custom patches, or integrations that shouldn't be necessary. For a non-small organization, this scales up to significant productivity losses and hidden costs in overhead.
These companies actively alienate us, the customer, through their business practices. Changes like aggressive licensing shifts (e.g., moving from per-core to per-employee models) force reevaluations and migrations and eroding trust (i.e. Oracle with Java, VMWare fiasco). This isn't isolated—it's a pattern where short-term revenue grabs risk long-term relationships, yet companies seem unfazed.
This jacks the entire ecosystem up. These practices stifle innovation by locking customers into suboptimal tools, increase overall IT spend industry-wide, and contribute to employee burnout in dev and ops teams.
It seems like it’s a race to the bottom. The strategy is to create an ecosystem with high switching costs and vendor lock-in. It just doesn’t seem sustainable, yet- it keeps truckin’ along.
Charlie Munger once said something to the effect of "show me the incentives, and I'll show you the outcomes".
There is very little incentive to produce high-quality software, to not alienate your customers, and to support the thing you already sold. Those things cost money. Money paid for those things is money not paid to shareholders, and that's the ultimate incentive in our system.
They've got you by the balls, and secretly, your CEO thinks their CEO is a genius for thinking up and implementing that business model. Pay up.
> There is very little incentive to produce high-quality software, to not alienate your customers, and to support the thing you already sold.
I think some of what I'm trying to portray is that this should be the incentive. Either do it or you don't have a customer. Yet, the customers don't hold this standard.
As an individual consumer, if I don't get what I pay for, I return it or can even submit a charge-back. Is it not irresponsible of business management to not do the same?
It seems like these problems are related to software and a certain size of corporation that is selling it. VMWare isn't hoping to make a few $million, they want/need to make a few $billion. And it is absolutely a race to the bottom when you can get Proxmox for free. Nevermind Harvester and some of the other projects out there are are doing kinds of similar things.
I was a beta tester for VMWare way back when. It was one of the first pieces of software I bought out of college. It was like manna from heaven, I could commit to Linux and have a backdoor for Windows and I needed it, and I did from time to time. I also did security testing over the years and once you've joined a machine to a Windows domain, it never can be made the same again. Vmware enabled that business without a spare laptop or spending tons of time rebuilding it. I've maintained the license since then, 25ish years. Bought it for the Mac too. I can't think of a worse transition than the one they're doing.
Software, done right, is both extremely hard and expensive. Hardware was cheapened by China/Asia but it is not happening for software (theirs generally sucks and they lack many fundamentals). Europe completely lost the race.
The current breed of managers in the US have decided to fire developers, abuse customers (you have nowhere to go) and burn all the money on AI (they believe it’ll solve all their problems).
Morale will remain low until an alternative spawns. Kinda with electric cars. Europeans, Japanese and Koreans are now forced to up their game and lower their prices.
You're just saying things you want to be true, "Asian software" doesn't suck and Europe didn't lose.
Just because "all" software companies have American entities doesn't mean you "won", that's just what happens when a jurisdiction let's companies do anything even if it's detrimental to society as a whole.
Most (all?) of the tech stack depends on American companies starting from Operating Systems, to Servers, to SaaS, to Cloud, to Software running on most businesses, etc. What you are saying is meaningless when they got you by the ba&&s.
Yes American companies have been good at capitalizing on IT and a lot of companies are "by the balls" of Microsoft, but much infra is opensource. Linux runs the world.
Certainly, the are a fair number of contributors in Europe and both SUSE and Canonical are European based. I still think it's hard not to think of Linux as fairly US-centric however if only because of how many large US companies use (and contribute to) Linux.
Valid, I've noticed that a lot of non-US contributors work for American corporations too. It's a bit ideological for me ("OSS is global") so I think I might gaslight myself into believing it's more equal than it is.
I'm happy wherever the contributions come from either way but I will never call Linux US-centric!
Lennart Poettering(German, works for American corps) comes to mind as an example, though not a kernel developer.
Most people do not interface with Linux on a bare metal, self-assembled server. They use AWS, GCP and Azure. There is Alibaba Cloud but it is so bad, I can't even properly signup/signin.
> Linux runs the world.
The world infra runs on top of Linux. Linux is open source. Most of this infra is American.
Another way to think of it: AWS Stockholm is a pipeline to navigate Sweden regulation/laws/banking to funnel money out of there. The core is still American/controlled in America. Otherwise, people would have just hosted in a Swedish hosting platform.
Swedish power, Swedish backbone, Taiwanese chips and boards, and an American control plane.
But yes you're right, it's an American service offering that I for cost reasons and I avoid big cloud like the plague. USA surely knows how to charge for their services and lock customers in
Agreed. My org can take some of the blame but with our big vendors:
10s of millions on Adobe - admin side craps out all the time, terrible outsourced support (easy to escalate past them but still annoying that we sometimes have to start there)
10s of millions on Microsoft - Laissez-faire attitude about when they will update stuff during renewals and true-ups. Not specific date. Little help.
Broadcom - PE play (as we know from Broadcom) to squeeze every penny out. Have been through several account managers now. They won't sell through VARs anymore, only direct, and to get anything beyond VCF you have to be on a very special list that supposedly the CEO personally approves (which is beyond mind-boggling).
Starting to see some post production software be licensed based on having limited amount of contiguous time zones which is also crazy for a global company
Note: Media and healthcare industries may be SOL because there are lots of content and cybersec 'requirements' for having private clouds. Curious what others in those industries are doing
A lot of it is the cost of doing business. We'll see how it goes when MS moves to the consumption model they're proposing.
My pet peeve is core-based licensing for products such as database engines. For that matter, any kind of capacity licensing tied to some variant of Moore’s law inevitably results in the vendor holding their product’s face under water as the tide rises around them.
As a random example, SQL Server Standard Edition is limited to a “very generous” maximum of four sockets, 24 cores, or 128 GB of memory.
That’s just slightly bigger then a laptop these days!
The problem here is that Microsoft kept their license limits as constants relative to a reality that moved exponentially. They would have to have applied “inflation”, but they just saw their sales figures go up and up… and nobody will rock that boat!
Inevitably they’ll keep choking their product until it turns purple and dies. It’s a force of nature, there is nothing anybody can do do counter this naked corporate greed that is enabled by accidental mis-pricing. This can never be corrected, except by letting products die and be replaced wholesale in the market.
This is another "beginning of the end." All eyes are on this situation and how much push back it gets. If there is little resistance, others will certainly follow suit.
People love to talk about management skills and how things should be, but in my experience the autonomy to have that freedom is greatly lost upon the manager. At every level of management you'll be barking down orders as a transitory and the main difference is how you do or do not get buy-in from your reports - and often it is next to impossible to gain that buy-in (more work! longer hours! emergency pragmatic fix for an administrative problem!), which is why you're the one presenting the task.
The thing that seemed to work best for me was reaching out to dev shops in my area and starting a conversation even if they didn't have a position posted. Go to meetups and talk to people, join their discord, etc... and job fairs, as lame as they seem, produced contacts for me that are still reaching out "just in case" even though I found a position eventually.
What's old is new again. Get those soft skills warmed up and go out and shake hands.
Job fairs are super underrated. When my spouse and I were a bit in flux years back for where we were going to live long-term, we ended up living in the city where we met which is very much not a tech hub - only a few big companies that hire engineers - despite spouse applying to various places in SF, Denver, and Seattle.
He landed the local job first by going to a career fair that I randomly discovered on Twitter. Affected the whole trajectory of our life.
I believe they're the current "cheat code" to the hiring process. If you get called up for an interview after talking to a company at a career fair, there's a good inclination that you have already passed or at least have a leg up on the "culture fit" part of the interview.
With hundreds to thousands of people spamming every job posting under the sun the second it is posted, just getting your credentials read by a human is a major barrier right now that career fairs help alleviate.
Belated reply for posterity: another cheat code for job fairs is pre-researching the companies there.
Typically, attending companies are listed (or you can ask an organizer for a list). Do some quick searching and type up a summary.
Company name. What they do. 1-2 questions. Emphasis on companies you've never heard of!
You'd be amazed how many doors open when you walk up to a recruiter for a small, specialized company and know exactly who they are and what market they're in. (Remember: that poor person has probably spent all day explaining that to everyone else)
Recently went on the market after my employer of 16 years went to shit. Got lucky with a hybrid gig that is really WFH 98% of the time after 120 applications and 26 interviews spread out over 15 companies. It was a roller coaster that I do not recommend.
