A large portion of the software and IT industries have refocused in recent years on “cloud computing”, a term that refers to running things over the internet instead of on premises. I believe this buzzword originated from the practice of representing the internet as a cloud on old network diagrams.
I’m not a huge fan. I tend to be something of a Luddite anyway, but I have inherent skepticisms about new technology. Running your infrastructure over the internet makes a lot of sense if your business is based in Florida and one bad hurricane could drive you into bankruptcy, and it also makes sense if you are a small business and don’t have $50,000 for equipment, but beyond this, it’s all window-dressing. A lot of people are switching to the “cloud” because everyone else is, and this creates the inevitable domino effect of companies trying to claim they have the same services as everyone else, or risk being shunned by consumers [we are the architects of our own idiocy….].
However, one claim that repeatedly surfaces is this idea that hosting in the cloud can save you money because you can scale down when business is slow and scale up when business is fast [okay, technically it’s often called scaling “in” and scaling “out”. Whatever]. And there’s definitely some value to that. I once interviewed at a company that used this model quite effectively.
But there’s still a heavy cost: once plugged into one vendor’s system, you are basically trapped forever. If you think that Amazon and Microsoft want to make it easy for you to switch over if you aren’t satisfied, think again.
This is just like with databases. Every major database vendor wants to you use their database. But I’ve never known a company that was paying for Oracle and was happy about it.
And to be fair, every company wants to offer something unique, so treating databases or cloud providers as a “detail” the way Robert Martin might advocate simply isn’t possible. But you’d better hope you pick the right vendor, because once your code base is hooked into one, it isn’t going to change without a massive effort.
And that’s what gets me. Handing over your freedom for a slice of savings pie just doesn’t make much sense to me unless you really need the features. It’s nice that cloud services are constantly being updated, whether that involves hard drives, ethernet connections, security updates, etc., but that still ignores the fact that some on-prem servers have been running for decades and performing perfectly fine. If you’re trying to run a cloud-streaming service, yes, new technology makes a big difference, but if you’re just serving up text or basic services, I dunno, man, I like the idea of owning my hardware and not sending a check to Micro-sucks every month.
Which brings me to the whole monthly-payment business model:
It’s usually not a good deal.
Companies know this. I repeat this story a lot, but when Adobe had pretty much added every feature the average user needed to the Photoshop/Illustrator/etc. suite of products, they knew that fewer people would have any need to pay for upgrades, so they switched to a subscription model to force people into monthly payments. Microsoft knew this, too, and has slowly been phasing out lifetime Office software purchases in favor of their Office 365 service, which charges a monthly or yearly fee.
And the sad part is, after 2-3 years you could often have paid off the software when it was a one-time fee. But you’ll probably keep paying anyway, because that’s how it’s structured. We can talk all day about how subscription models are cheaper for startups and beginners, but it doesn’t change the fact that they want to be cheaper so you are stuck paying dues your whole life. “They throw us pennies, hoping we give them dollars”.
This is what I’m talking about when I say corporations provide great value, but then they go way beyond and find every means available to screw you. I’m kind of with the liberals on this one: corporations are not to be trusted. Making profit the single most important goal of your business causes so, so many problems.
There’s a certain TV show my parents like to watch. The first four or five seasons were produced on DVD, but now those have been phased out and the almost only way to buy these seasons is to either have a monthly subscription to certain streaming services, or to buy the electronic version, which is way more expensive than the DVDs ever were and require you to lock-in to either Amazon or the channel’s service, whose future is quite tenuous.
And yeah, the world economic forum made a big stink by saying, “You will own nothing and be happy”, which of course pissed a lot of people off, and rightfully so. I half suspect this was just a marketing stunt to draw people’s attention, but there’s no doubt that, in general, businesses want you to be stuck paying monthly because it makes their cycles much easier to predict and generally earns them a lot more money. The people at the top don’t have to be lizards from another planet: Bob and Sue down the street like the idea, too. “Passive income, baby!”
And none of this means that paying monthly for something is the devil. A lot of the expenses that go into running these services are assholes like me – software developers, system administrators, network administrators, and the rest of our ilk, who want a slice of that green pie. I don’t write SQL for free [I really don’t. C# maybe, but not SQL, goodness no]. But it’s important to remember that a lot of this stuff would pay for itself in short order.
One exception might be gym equipment. It takes a lot of $30 monthly payments to pay off a $3,000 treadmill, just sayin’.
Which brings me to my final point: just because I don’t like the powers-that-be doesn’t mean monthly payments are necessarily bad. You really have to question if you are getting screwed or not, and you aren’t always getting screwed.
$10/month, $15/month for music streaming, Netflix, or YouTube? Not a bad deal if you are actually getting good use out of it. $50/month for software you rarely use that used to be a one-time $600 fee and has free and open source alternatives? Not so much. And as far as office software goes, I use Libre Office instead of Microsoft Office. Ain’t nobody got time for that crap.
So it really depends.
I hate paying $15/month for my GPS service, but then I also find myself in some pretty wild corners of valleys. But buying a stand-alone emergency beacon is about $300. This of course means that a few years of paying $15/month would actually pay the device off, but I like having the GPS and emergency beacon features together. I’m torn on this one. I really should buy the standalone, though. The greater your monthly expenses, the generally greater your financial fragility.
Of course this does bring up rent, though. Some people online, bitter about the high price of houses, cynically quote the world economic forum on this one, too, but it’s important to realize that even industry experts say you should keep a house at least 5 years to pay off the expense of the realtor (lol). Then, it would seem, the best gains only come 30 years later when the house is actually paid off. If it were just a matter between owning a house outright or paying rent, owning the house would almost always be better, but when people talk about owning a house, they actually mean “making a down payment then taking out a mortgage”, forgetting that the down payment might have a pretty steep opportunity cost. So it’s a little more involved then just treating rent payments as a monthly expense. You could even think of it as cheap insurance against needing to pay $10,000 out of pocket for a major repair, which is more a question of when and not if. If it takes 30 years for something to truly pay off…I dunno, man, that’s fundamentally different than paying monthly in perpetuity when something would otherwise pay itself off in two years, but as always, this topic is huge and I’ve only written one paragraph on it.
Of course, with so much of the tech industry switching to “cloud computing”, I have to wonder if I will get swept up in this, too. Will all the big money be in cloud technology? Will washed-out old C# developers like me find ourselves stuck in musty backrooms, tapping out code to load on our ancient 2020s server racks for a dwindling client base that only pays us because their managers are equally old and crusty?
Haha. I don’t know. Azure DevOps is actually pretty nice, I just loath the day I have to write some sort of container access code and deal with that garbage. Of course, I probably won’t be in software long enough to really care. Cloud will likely be something I learn to security test, but that’s different than fundamentally seeing it as a good thing. All I know is that I’ve bought a lot of networking equipment as of late and it’s all really cool. I suspect a lot of other people feel this way, too. I’ve really enjoyed learning this stuff. Cloud doesn’t offer great benefits for every business, so I suspect while it will continue growing, it will also plateau at a certain point. The cloud makes it easier for end users to know nothing, but I think there will always be good jobs for those who are interested in the guts of computers. Just because administration might be done from some remote console that hides all the details, doesn’t mean the right attack can’t take down the whole cluster behind the scenes, so there will always be a need for people to know how to prevent that from happening.
And personally, I’d rather own my hardware.