I’d imagine payment software … historically … not have an unnecessary middle man.

I used to work in the payments industry, and it’s middlemen all the way down. Visa/MC/etc. don’t deal with small fry. There is an enormous amount of regulatory overhead (which is not a bad thing here, fraud is rampant and it’s a cat and mouse game). Unless you’re buying from a heavy hitter like Walmart etc., that business is going through at least one layer of transaction processing before it gets to the issuer. The smaller the business/processing traffic, the more likely it is that there’s several hops in the chain.

Example: Processor A has direct connections to V/MC (which involves hosting V/MC’s hardware in a secured datacenter with multiple redundant network connections, and paying for the privilege). Unless a client does $X in volume, the overhead outweighs the revenue from that client. Clients that process less than $X still need servicing, so Processor B contracts with Processor A (hosting Processor A’s hardware, paying for dedicated data lines, etc.), dealing with smaller clients that add up to at least $X so Processor A is happy. Processor B may have a volume floor as well, in which case the chain continues. Each hop takes a percentage, and generally the clients with the lowest volume pay higher fees per transaction. Now add in other card types like Amex and Discover, debit transactions (which require dedicated hardware to decrypt/encrypt the PIN to verify before passing the transaction), EBT, etc. Only the largest processors are going to have direct capability to handle all of that. More typically, mid-sized processors contract with multiple other processors to cover the spectrum… or just not worry about Discover or something.

tl;dr there’s a whole lot more to it than just rolling your own point of sale software.

I daily drive linux, but I don’t work in a field that involves computers much if at all. I’ve always tinkered with whatever OS I had installed. It was OS 9 then OSX and macOS as a kid. Then windows once I had my own computers, and now Linux. I jumped the windows “relatively” early. As in a good bit before copilot and such, still definitely “recent” on a broader scale. I’ve been on Linux for over a year full time.

I’m now looking into helping other people around me adopt Linux and just FOSS in general. A friend and I have talked about opening a tech repair shop that also offers custom system/home/network builds. Would love to see more Linux being used in local businesses.

I find Linux reacts better to my tinkering. Or at the very least gives me an actual error message to work with when I tinker to close to the sun. I dove right in with the Arch minimal installer, and built my system from the command line. Inevitably my first install had some jank, so I’m trying to make a more refined version in NixOS to see if I like the paradigm shift.

I’ll also toss in that it was actually Syncthing that got me into Linux and also inspired trying NixOS. I got very fed up with the clunky options for running Syncthing on windows. That among other reasons sent me to Linux, and once I started learning more the idea of using Syncthing to manage NixOS configs across all my machines started to bounce around my mind. Syncthing already kind of gives this feeling that all my devices are just one big distributed file system. Carrying that over to the actual OS completes the process of making it a completely distributed single system, simply with different interfaces for accessing it.

I love how this very specific topic is posted in an Autism community and not any of the tech communities and yet it’s getting potentially better engagement than if it was posted to a Linux/IT community

But onto the subject at hand, I work at an MSP at the moment (and probably not for long due to management issues) and the one thing I’ve noticed is that the MSP industry is all about risk, and deploying something they’ve not deployed to customers before is a huge risk. They’d much rather work with “the devil you know” than take the risk on something they haven’t worked with yet. Commercial vendors also have the benefit of being able to hawk stuff onto their support to free up your own techs to take more tickets (and therefore make more money) plus of course the extra cost to customers is either a non-issue or a plus due to more margin (because everything an MSP sells you is sold with a margin, usually ~20%)

I will say, this experience has further solidified my belief that paying for outsourced services will cost more than doing it in-house for most businesses. About the only way outsourcing makes sense is if you literally don’t have enough work to hire one dedicated employee to do the thing.

I switched to Linux 15 years ago and I work in the legal industry as a computer guy, so I’ve been trying. The main obstacle in my experience has been liability insurance; if an entity like a small law firm uses Linux instead of Windows, their liability insurance rates increase; with Linux and other FOSS software, there is no deep-pocketed corporation to sue if a technical failure costs you money. Recent changes to Swiss law might help convince insurance companies that Linux is genuinely better enough that this will change, but insurance companies are the heart of all that is corporate and conservative, and famously hard to sway without mountains of evidence gathered on their terms by their people. Microsoft products are defective by design because institutions that only care about the bottom line do not want their users empowered.

work for a state wide university managing endpoints. if we exclude all the servers and iot stuff and just look at end user laptops in an effort to expand the use of linuxs, it comes down to support cost. we primarily offer win or apple as a choice with about a 60/40 split. theres a small specilized linux service for researchers outside the central service thats does most things as manual set ups by use case. I’ve pitched offering managed linux as a service a few times. i’ll keep doing it, but. it requires staff with knowledge in remote management of dynamically active device. who also understand the tools for supporting linux endpoints under any compliance req the spaces has. all of those tools are far more manual than the win/apple stuff. its hard enough to hire engineers for the apple side. you’ll have to grow your own, it’s just too specialized. all your documentation and training then need to be updated for a 3rd unique platform. its probably a 3yr undertaking at a few hundred grand per year in staffing to stand up. absolutely doable but it needs high level support in an org to do.

My work issues Windows or Apple laptops, your choice. But we’re not an IT/CS company really, at least not my division. 99% of the time we are just accessing web-based apps for paperwork/inventory management.

I use Linux at home but I don’t really care too much about using Windows for work. Not (truly) my computer, not my problem to manage.

I think there are perceptions of a lack of security/reliability with Linux desktop and the programs associated. Enterprise companies want to use enterprise products. Raytheon wants to use Solidworks, not some FOSS project maintained by 7 people in Romania. Whether or not this perception is well-founded, I don’t know. Again, not my problem.

Not sure what year you’re living in, but Linux is absolutely everywhere.

I guess what I’m saying is, if you want it to happen begin making it happen! Get a job in a position that gives you some creative freedom and begin embedding it everywhere you can.

We’ve come full circle (Linux in the Autism community)

To answer your question, it depends. Are you talking about it as a desktop OS?

I work on some routers and other equipment that utilizes Linux. But this is not uncommon in my industry.