[not_a_shill]

You are cheap. You're still basing your decision on the cost of using the product instead of the value it provides you.

This isn't targeting you specifically though, most comments in this thread are stereotypical developers too.

[jchw]

No, I am not basing my decision on the cost of using the product. I'm basing my decision on factors that have nothing to do with money.

I think a perfect example of this for me is IDA Pro, for which I own a license. When Hex-Rays switched over to a subscription model, I did not renew. I still use the old version today, and it works perfectly fine for my needs.

The difference between buying it on subscription and not is whether or not I own it. I really own this copy; it's mine whether or not I'm connected to the Internet, and it will work in perpetuity until the sun burns out. That's how exchanging money for software is supposed to work.

[erickhill]

That's how it used to work, back when you could go to MicroCenter and pull a box off a shelf then take it home and hope to hell you had the right drivers installed.

But it's also why when I run across a company that shockingly lets you buy their product outright I fully support their efforts and invest in a purchase. It really is cool to sometimes be off WiFi somewhere and use software to your heart's content.

[not_a_shill]

>you might come to realize it's not really worth the amount of money you could've made by selling it for $5/mo,

I wouldn't have brought this up if you weren't trying to give the impression that money is a factor.

I interpreted it as "I still wouldn't pay $120 because even though it returns more value than that, it isn't worth that."

Edit: (that being said, the one time purchase cost of this for a "lifetime" license is $900, which seems more like a way to push subscriptions than a serious attempt at selling a one time purchase)

[jchw]

Yes, but the qualifier is important.

>just one reason why developers don't like subscriptions.

The truth is that I don't even consider paying for new subscriptions anymore unless there's a good reason for it to be a subscription. I don't want to have to "trust" someone. Will they exist in a year? Will the software get worse and leave me with no choice but to use the "new" version? The test here is that I genuinely don't think I would use software that asked for a $0.05/month subscription either. That's just not how I want to use software.

If we want to talk about money, though, then there are still good reasons to not buy a $5/mo subscription. It sounds like it's a good enough deal, and in fact, I might even agree with you. But there is absolutely nothing that guarantees that this will stay the price forever in perpetuity. In fact, even a literal guarantee that enforces this would be worth toilet paper because if it ever came to a head, if the company went bankrupt, it would take down the activation servers and potentially leave you without your software.

So playing it from the top, let's say you get a new-fangled productivity software. Everything's great and you're only paying $5/mo.

- Then they release a "new" version. The new version breaks all of your workflows and you have to relearn everything. It's now awkward for the way you were using it.

- Then they add new features that are only available on a new $20/mo subscription. Turns out $5/mo wasn't really sustainable after all, so they need to adjust the price. Slowly over time, pressure mounts on the grandfathered old accounts to do something that will push them up to a higher tier.

- Then it all goes badly: Some venture capital and/or acquisitions later, and the software goes broke. Now what? Well, if you own it, nothing. If you don't, tough luck: All of your workflows are broken again, and you have to go back to doing things the old way.

For tools that are core to the stuff I work on, I do not play games like this. I do not care how sure you are that this won't be you. I don't want re-assurances, I want control over it.

[LapsangGuzzler]

Every developer understands how subscriptions work. A product may provide value today, but there’s no guarantee that the payoff will be worth the cost in the long run as the company increases its prices.

With a subscription, you’re agreeing to let a company tax you to do your job in the hope that it pays off instead paying upfront one time for a product.

[not_a_shill]

>but there’s no guarantee that the payoff will be worth the cost in the long run as the company increases its prices.

You can cancel a subscription

[jchw]

It's a problem, though, to become reliant on something and then experience a rugpull. A lot of people seem to think that you pay for a product and then in exchange you get increased productivity, and the only "cost" is, well, the cost. But it's not. Once you start becoming reliant on something to do your job, you orient your workflow around it. There may be alternatives, but it costs your time and effort to switch. It differs per tool: imagine switching text editors though. That's a very non-trivial endeavor for many software developers, even people who use fairly stock editors will have become pretty integrated into how their setup works, with hundreds of hours of muscle memory and re-orienting yourself to the "mental model" it has.

That's exactly why people selling products love subscriptions so much. It's not JUST reliable income. It's sticky. If I own something for real, as in it works offline with no activation garbage needed, I can just not pay for the new version if I don't like it. Maybe the old one becomes obsolete eventually, so it's not risk-free, but nonetheless. But, if a subscription fee increases, my options are typically 1. Go to hell.

Am I exaggerating? Well, yes and no. For one thing, I have really truly experienced the rugpull. It doesn't even have to be malicious, it can be as simple as "company disappears". The more established and trustworthy a company has proven to be, the more willing I am to put up with some loss of control. But monthly subscriptions to random products I see on HN? No thanks.

I'm not even asking for a perpetual lifetime license, which I already find sketchy from a sustainability standpoint. I am fine with the old-school "upgrade license" concept employed by say, VMWare and friends, and the less DRM-heavy the solution is, the better.

[bombcar]

I’ve not seen the research but is suspect the average cancelled subscription goes far beyond “when it was no longer worthwhile” before it is cancelled.

Which reminds me I have a few airline card to cancel.

[LapsangGuzzler]

Right, but what if your subscription tool that you want to cancel is pipelined into your workflow in such a way that you’re paying a technical or operational cost to cancel beyond just severing the account itself, i.e. vendor lock-in?

This is why it’s often better as a solo dev to not rely on these tools in case you have to jettison them and break your workflow.

[warent]

Another way of wording what you just said: "What if cancelling the subscription is more costly than keeping it?"

Uh... then keep it, because that means it's providing value to you.

[Dylan16807]

If the insulation in my building was asbestos, it could be more costly to remove it than to leave it there and pay for regular testing and extra maintenance. But that doesn't mean it would be providing value.

Switching software has a cost. Avoiding that cost is not value.

[not_a_shill]

Then factor in switching costs or find a way to reduce them when initially making a purchase.

Y'all are twisting yourselves into pretzels trying to justify why you can't easily cancel a subscription.

[bgirard]

You have to recognize the incentive of developers working for a large company:

I can either 1) pay for it out of my own pocket, 2) expense it monthly which after wasting time in expense reports basically goes back to 1 where I pay for it myself, or 3) go through the software procurement process. I've done all 3, and none of these really yield a personal benefit. 1+2 are annoying, and 3 takes time away from your main job and it's hard to get the licensing right. And at scale I think game theory does well at modeling behavior for why developers wouldn't want to go through any of these options.

For a freelance developer I'll speculate that for a dev that bills hourly the benefit aren't great, but it's still value passed on to the customer so it might be worthwhile.

If I ran my own freelance business and charged a flat fee I would 100% buy this tool if I did a lot of CSS work.