Nothing signals to me that the market for high quality software is utterly and completely broken than this piece of news. Sad to hear this - if the Omni folks are having difficulty what hope do other small shops have. I am always amazed at the creativity SW folks show when coming up with reasons/excuses for not paying for tools - the culture around this is very saddening. On the other hand my friends who are into electrical, woodworking or automotive take a deep pride in the tools they own. Many of these fields pay a fraction of what SW dev does.
I never fully understood why we do not have the "a good tool saves me x hours, and makes me happy while doing my work so I should pay for it" culture. SW engineers are pretty well paid and companies themselves have decent margins so this penny pinching tendency is upsetting.
> On the other hand my friends who are into electrical, woodworking or automotive take a deep pride in the tools they own.
I don't want to justify the "cheapness" of the software users, which is certainly excessive, but the feeling of ownership of a physical object is still very different from software.
For example, you can easily lend your toolbox to a friend, your hammer won't ever stop functioning through no fault of your own, you can re-sell or gift a chainsaw if you want to get a new one.
Since software is immaterial, we've ended up in a situation where these basic things don't work, people do not feel like they actually _own_ software, and thus feel less motivated to buy it.
I think the not owning the software subjective sentiment is made worse by the subscription model. Personally paying for Photoshop for example was more acceptable when it was "physically" an App in my Mac. I still have very old versions I can launch on vintage Macs or Emulators. As a kid I definitely remember owning and organizing my apps with their nice icons (the spatial file manager helped), thinking it was "real" things.
I think we're a middle generation that grew up before the internet that this makes sense for. The CDs are probably going at the next move, but I expect to keep my MP3 collection forever.
Maybe it was a mistake for the software industry to give up on the nice big colorful cardboard box with software inside it. Bought at a physical retail store, with CD-ROMs in it an a thick printed manual. Certainly gave more or a feeling of actually having bought something, different from just clicking a "buy" button in an appstore.
And I don't know about you but the first thing I do when I get a paper manual is go online to try to find a PDF version. Seems more like a downside than a value-add?
I don't think that professionals are likely to loan their tools to friends nor do I think they consider that option when they decide if they will buy a tool. It's tools that make their jobs easier or possible.
This reminds me a lot of presentation slides that I read a few nights ago. I was reading presentation slides from Carnegie Mellon professor Heather Miller about the funding challenges that open source software projects have (https://2019.programming-conference.org/details/programming-...). So many people and companies rely on open source infrastructure and productivity tools, and while there are plenty of people contributing to open source software in the terms of donations and development, there are key open source projects that aren't getting the support they need despite the wide usage and importance of those projects.
I think one of the key challenges of this decade in the software industry is figuring out how the development and maintenance of high-quality software can be funded in sustainable ways. There is clearly a need for high-quality software that is not ad-supported, is not subscription-based, and does not monetize user data. However, the days of shrink-wrapped software and the shareware business model are over. If we want high-quality software, then we will need to come up with an answer to this problem.
Vendors get recurring money and aren't incentivized to make sweeping changes to have major releases to get more money for the same product; consumers can manage cash flow by only paying while they need it, there is a lower barrier to purchase, and vendors have more predictable income.
As a consumer of software there are three MAJOR problems with the subscription model:
1) If it's for a piece of software I use on an ongoing basis the cost often rapidly eclipses what I'd have paid for standalone software, and it never stops rising.
2) There's usually some awkwardly contrived cloud tie-in, purely to enforce subscriptions, which breaks the software if your internet connection or the supplier's servers are down. This adds a completely unnecessary external dependency.
3) You're entirely at the vendor's mercy if they decide to change the software. At the very least, you're subjected to constant small breakages of your workflow, and at worst they change some major feature that you depend on and put you out of business for days or weeks until you find a workaround.
Edit: Also 4) The aforementioned awkward cloud tie-in often brings with it some gnarly privacy issues, with your sensitive files and private data being leeched off to the vendor's servers FoR YoUR ConVeNIeNcE.
I've found JetBrains and .. tele.. ? (screenflow) to both be decent compromises. Well... I guess screenflow isn't a rental - I upgraded through version 6 - I may upgrade to 9 now, as there's enough 'new' stuff (and I have a newer laptop) to make it worthwhile. I appreciated it not being a subscription, precisely because nothing changed for me - I kept using the same software for a few years, knowing I was missing out on some bug fixes (but also knowing I could upgrade to get them reasonably cheap - it was a choice on my part).
The JetBrains model is good because I'm able to keep using a version without paying for upgrades (but I have kept up anyway as the annual fee is so cheap for what I get).
The points you've made are why I don't typically get software that is truly 'rental only' or hits your other points. That said, I've not run in to too many in the past few years that are affected by those issues any more. Are there some in particular you're thinking of that have those problems?
Do JetBrains products still work if the company would go out of business or decided to undo the perpetual license? (I haven't upgraded my license since they switched to the subscription model.) [1]
[1] It would probably be difficult contractually, but such things happen. See e.g. TextDrive hosting lifetime accounts.
Your perpetual license only works on a historical version. It doesn't work with any updates. The incentive is for them to continue building products that make you want to upgrade. The perpetual license key works in a disconnected network so it doesn't matter if JetBrains went out of business. You just need to make sure you have a local copy backed up.
The "JetBrains model" is how all software used to work. You... buy the software. And then maybe you get free updates until the next major release. If you buy the next major release then that's another "12 month subscription".
All great reasons. I would like to add, also, that I wonder how many subs developers think people in flyover country (not the 500K/year SV country) can actually afford each and every month?
80/20 rule, probably. Most of their revenue is coming from a handful of places, and and increasing the money coming out of other parts of the market might not be worth it.
Easy to support a couple hundred whoppers who pay big money on the coasts -- and who also have the time and money to really learn the tool. Expand that out to 100,000 customers with lower costs and greater support needs and it doesn't make sense.
When you're talking mobile apps and microtransactions and we call these people "whales" then this is a bad thing. In a professional setting is it suddenly OK to treat your customers like this?
> If it's for a piece of software I use on an ongoing basis the cost often rapidly eclipses what I'd have paid for standalone software, and it never stops rising
High quality software has always required a recurring payment model. It's not viable to sell standalone versions to a new set of customers each year. In the past this was informal with upgrade versions every year. You could theoretically choose not to upgrade but software developers would have all kinds of tricks to "encourage you" such as not providing forwards compatible file formats.
What the subscription model does is align customers with the realities of software development. Yes the cost of software never stops rising but neither do the maintenance costs.
No it hasn't. The only thing that makes you think so is the unrelenting churn of modern web development. Back in the day if you bought a particular toolchain you could keep developing with it for as long as you wanted and the only thing that could force you to upgrade is the need for more features. In the past you'd get nice-to-haves at every upgrade but if you didn't opt for it you would be perfectly able to continue developing your current condebase using your current, paid-for tools.
macOS indeed requires additional effort. But regardless of that customers tend to get annoyed and stop giving you money if you don't ever fix their bug reports.
You fix the bugs in paid updates, and users decide whether the bug fixes are worth the cost.
It's a good balance—if you spend extra time on worthwhile changes, the users need to pay for that time. The users can also decide to stick with what they have if they don't think your changes were worthwhile, which may be a perfectly reasonable path.
I think you should look into the subscription model that Jetbrains has for their IDEs. It has none of the issues you described other than, inevitably, the recurring cost.
JetBrains products might be unusual in that the IDE is all about integrating lots of constantly changing external tools and therefore needs constant updates, so it's well-suited for a subscription model.
And for many people an IDE is something they can spend 5-10 hours a day in.
I own a fair chunk of the Omni apps, but even doing the diagrams for my thesis I didn’t spent anywhere near that much time in Omnigraffle. The incremental cost of all of the subscriptions every month, in perpetuity makes the marginal choice to get another one (eg 1Password) tricky.
The mental overhead of renting every piece of software I own would be exhausting, not to mention consistently expensive. Subscriptions come out of a different part of the my budget than purchases do, and it is an ongoing expense I have to factor is versus coming out of my funds once every few years, if I can afford it at that time. There are several pieces of software I would have gladly bought at 2-3 times their annual subscription rate, but instead simply moved on from. I get it. Software developers need to be paid, but I’m not renting every tool I want to use. I also understand I’m in the minority because I understand why there is a new version every few years and would gladly purchase those. Simply put, the overhead and uncertainty of a subscription model devalues the vast majority of software for me to make me find alternatives.
Same for me. I describe it in terms of relationship. Each subscription is another relationship with a third party that I otherwise don't care about, that I don't want to be in any ongoing relationship with. I just want to a) pay for the software, and b) use it to do the things I need done. Just like if I were buying a hammer.
A power drill is my favorite analogy. If I drilled holes once every 18 months, I’d gladly rent from Home Depot. If I found myself doing a weekend project every month, I want to own a drill. That drill comes with the expectation that it has an 18-24 month warranty (support cycle) and if my power outlet (OS) doesn’t change after that, I expect the current drill (current version) to keep working, essentially, until the hardware dies. However, I do not expect Milwaukee to add a light or whatever new feature to the drill, or change the plug if my outlet changes. After that 18-24 month period, I expect that new features, breakage, etc will mean I buy a new drill.
I know it is an imperfect metaphor, but it helps get my point across that valuing ownership doesn’t make someone cheap. It makes them unwilling to lease software.
I think subscriptions will be terrible for vendors in the end as well.
First, because they have to compete with the likes of Office 365, which offers an Office Suite, 1TB cloud storage, and 60 Skype minutes (at least the used to) for little over 5 Euro per month. Subscriptions for more basic programs will look overpriced in comparison.
Second, as subscriptions become the new normal, we will see services like Apple Arcade. I think Apple Arcade will eventually eat up a lot of revenue as casual gamers will move from purchasing games to the 'all you can eat' Arcade. If Apple Arcade turns out to be successful, what holds them from launching 'App Arcade' [1]? Vendors that don't get into the App Arcade will lose a lot of revenue, vendors that do get into App Arcade will have a steady source of income, but will also be completely at the mercy of Apple (or Microsoft).
[1] I know there is SetApp, but anything that Apple launches has a lot more visibility. They also have more clout to get high profile applications in such a service.
There are definitely some benefits to subscriptions, but returning to the GP, it’s sort of like renting your nail gun. And your compressor. And your tool belt. And your screw driver. At some point it makes sense to own it and stop renting. Renting is very expensive.
In the enterprise space, subscription is just higher cost maintenance with the additional risk of losing the functionality to some high stakes bullshit at renewal time or spending lots of time on some rat race to upgrade.
Enterprises love subscriptions for the same reasons why they prefer leases over capital expenditure. They look much better on the balance sheet and allow to spread the cost over time.
Oh sure, enterprise CFOs love them. They love money too, and software companies who aren't Microsoft, Oracle, etc should be wary. Your entire company could be rendered irrelevant by a minor modification to a subscription offering that is critical. Why should I absorb a $7 user/month subscription for 100,000 people ($8.4M) forever when I can spend $500k upfront and 20% of that ongoing to hire someone to build a functional solution using O365, GSuite or Oracle whatever?
Look at Microsoft's behavior with O365. Many companies will be fully locked into Microsoft across voice, security and identity verticals based on tweaks to the O365/M365 licensing models. You need Azure AD to meet your email compliance requirements, but that's bundled with Intune, which in turn has a co-bundle with Defender ATP, etc.
It's time for us to give up on the dream of having good software for free. That's what this whole thread is about. It's a wake-up call for us all. We have to pay for what we value.
There isn't enough wealth to go around for every adult
to have a $50+K/yr income. The developed world economy only works if most people are paid very little but woke anyway.
I was a delighted paying customer of OmniGraffle for a decade, and looked forward to each update.
I also payed for multiple version on iPad.
However at about the point that they released their own sync service, they just let it stagnate.
There are so many ways the iOS and MacVersions could be more modern and more useful on a day to day level, none of which require a massive re-engineering effort.
My conclusion is that they have just fallen behind the times.
> I never fully understood why we do not have the "a good tool saves me x hours, and makes me happy while doing my work so I should pay for it" culture. SW engineers are pretty well paid and companies themselves have decent margins so this penny pinching tendency is upsetting.
Don't we spend on software, though? Software spend is surely in the billions, if not tens or hundreds of billions a year. As consummate consumers of software, I'd think software engineers are more likely to spend money on software than the general public (and more likely to use open-source software as well).
I work in the embedded space and so I have used a few proprietary debuggers/IDEs/build systems. They've still got a lot of market share.
The huge problem with them is that they're proprietary. They don't have the wide usage of the open source equivalents, so debugging problems is hard. If you want to develop at home, at work and on a PC in a lab, you have to navigate licensing issues. If you want to do a build and test as part of your CI process, you have to navigate licensing issues. Then you move jobs and they've bought into a different, incompatible proprietary system and you have to learn it all over again.
The open source options are actually good enough, and the advantage of being able to have the entire build system on a new computer after running a few apt-gets and downloading some free software is huge. The skills persist across jobs, I can use them at home for my own projects, I never have to worry about licenses expiring.
I have a company in this space and our software components and tools were more a marketing tool for our services than a product based business. Most often a customer buy a software component and require to build a product around it. Doing this we got customers like Trend Micro, Symantec, and Dell.
Most of these started off with proprietary options that people left. These tools often need to solve lots of little problems for lots of different people, and it's difficult for companies to keep up.
But, these are all pretty easy to start for a single hacker who can recruit help to expand an open source version.
Took a meeting with a potential customer years ago. He was high on our technology and seemed certain to purchase. At our last lunch he told us their software company's CEO had a policy, they only used FOSS software. The irony of it hurt. Still does.
At my company I take a very different view depending on where in the stack the tech is used:
Developer tooling (Developer machines, IDE's, source control, CI, issue tracking, etc): Pay for the best.
The product: Only use FOSS. Give back by contributing and by making some of our work available as FOSS.
For the product, especially for the client-side, I just see it as too risky to include proprietary software. Even if the pricing is fine for right now, it can easily change later (got burned by both Google Maps and Mapbox years ago), and in some cases even prevent you from using different pricing models because of base costs it may add per user.
On the infrastructure side I'm more flexible, but even there:
1. Stick to commoditized services as far as possible, where the pricing is predictable. I'd include most cloud services in this, and most "enterprise" offerings not.
2. Avoid hard vendor lock-in if feasible, but only up to a point (don't build something from scratch just to avoid the lock-in).
This is what everyone’s approach should be. The product itself should be standing on the shoulders of FOSS giants. Everything else is just quality-of-life improvements and tools which should be paid for if they speed up your productivity or decrease your mental load.
Proprietary software lock-in in your product can generally be very damaging when things go wrong. I know of one company which was so throughly fucked over that they had to halt all feature development for 6 months while they were busy building a replacement. Cost them enough time and customers to a competitor. For non-product stuff, replacing proprietary software is a bitch but far less disrupting.
I have seen this myself - every software startup feels that everything they build upon must be a commodity and the only value add should happen at their layer.
I agree with your general point. We don't value time savings enough. However, I'd offer a few observations.
A lathe does what a lathe does every time. Even if the company goes out of business, there will be parts available for it for a while. No one at the lathe company can decide that it will stop performing feature X because it isn't popular or make enough money.
This leads into the second point. Companies (and individuals) have become fearful of external dependencies. So they push for FOSS so that they can avoid lock-in. Or they adopt an "invent here" mentality. Look at how many technology projects Google has cancelled despite the fact that they have infinite money.
I have used it for years and years, for normal stuff like software diagrams and business process and shit like that, but also for stuff like floor-planning my house and doing mockups to decide what furniture and TVs to buy, to making easy-to-understand political comic books to influence how my friends vote (old example: https://masonmark.com/stuff/2008/America.pdf), and I even used it to create the CV which helped land me the awesome job I have now (http://www.fivespeed.com/mason/resume/mason_resume.pdf).
According to rumor, the principal guy behind OmniGraffle had a fairly unusual employment arrangement and wasn't typically seen that often in the halls of Omni — I don't know the details, or even his name, but as an OmniGraffle-dependent user I'm wondering with some trepidation if he was part of these layoffs.
Because free-to-use online tools like Draw.io are "good enough" for many people who would have previously stumped up for Omnigraffle for more basic needs.
I also wonder if being MacOS/iOS only has hurt them, in the long run, more than they'd like to admit?
This story is more Omni slipping rather than the model failing universally.
Personally, I own most of Omni's software, but I won't be upgrading to new versions when they come out.
OmniFocus was once the cornerstone of my productivity. It was expensive and I didn't use the scripting aspects of it, but it provided value. The Apple ecosystem didn't bother me too much, but it was a pain, and I was doing more and more work on my Windows machine and going to my phone to add/update tasks was creating friction.
They came out with OmniFocus web, which would allow me to solve the issue... but it's too damn expensive. I've already spent ~$175 for licenses to OmniFocus for Mac and iOS, now you're asking for $50/year on top of that.
The competitors are hungry in this space, and Todoist offered more for less. I switched to their premium plan, which is cheaper than OmniFocus web itself, let alone the full subscription and includes clients for everything, plus a web-only client as a last resort.
The same pattern holds true for the rest of their software lineup. I own most of it except OmniPlan.
TL;DR. Omni isn't providing value compared to its (paid) competitors, to say nothing of FOSS.
I would have to make compromises to move my work over to Todoist from OmniFocus. It sounds like you were a bit of a light user whose situation got changed on him (Windows.) That’s fine but it doesn’t mean OmniFocus is behind its competitors.
Dealing with paid software in corporate america is a terrible experience for all involved.
As a user:
I see some software I might be interested in trying. I have to go to generally some nameless, faceless group to get them to buy it. I open a ticket, or get an inquiry started, and they start asking boilerplate- do we have similar tools already, what is the license, etc. If we get past this phase, there is typically some kind of legal review. Maybe a few weeks later we can actually have this on our machine in some form.
From a company perspective:
A user wants to buy yet another piece of software that we have to manage and keep track of. We will have yearly contract renewals to deal with most likely, understanding and holding them to their support model, ensuring they don't phone home, ensuring we are in compliance with our number of licenses, etc- this is all quite a bit of continued work. Also- how do we package this software and get it out to our users- there is a process for that as well.
With FOSS, we have none of these problems aside from the potential phone home problem, which isn't really a problem, because we restrict by default. It's just a much more pain-free process.
I don't doubt that any of this is true. But commercial software is the best way to ensure that everyone involved in developing the software is reliably compensated for the software itself (and not for some related service). The freeloader problem in open-source software is very real. If we want to make sure developers get paid, it has to be mandatory, and unfortunately that includes license enforcement and other things that we find unpleasant about commercial software.
OK, I'll bite. I have a fundamental aversion to paying for something that has zero marginal cost. I feel I'm taken for a ride if I do so.
If I pay for a book, I feel it cost some effort to produce it physically, print it, bind it, ship it. I cannot force myself to pay for an e-book, which is a pdf copied at no extra cost. Same for digital music and software or any other digital "good" or subscription. I sympathise with the creators, if they will produce something that does not scale, I will pay them for it. Otherwise, sorry, but not my job to figure out their business models.
I also have that aversion. But regard, zero marginal cost doesn't mean 'zero cost'. There's overhead to amortize. Like, the effort to have a company that hosts that stuff. And the desire to run a for-profit company, else we'd all be at the mercy of volunteer organizations.
They seem to have 'figured out' their business model already, and it involves each person copying the PDF, paying a nominal fee.
You make a good point of course -- my issue is that this amortization has become totally divorced from use.
For example, a taxi driver also needs to amortize the upfront cost of their car, but the more they drive, the more the car wears down, so it is all predictable, and somehow feels "fair".
With digital goods, you may sell it twice, or a billion times, what you earn is independent of how much cost you sunk into it (apart from hosting and other infrastructure as you mention) -- which explains the runaway successes of our times, but it's just something I am unwilling to participate in :)
Yes it seems like new rules are needed for synthetic businesses of this sort.
I've argued elsewhere on HN, that the reward for building an automated system should be, pay for building the system. In America at least, we consider the entire lifetime output of said system to belong to the builder. That is unsustainable. Imagine a future where everything is automated. There will be no need for workers, and not much further development. The economy becomes a static owner-vs-unemployed deal, and how can that survive?
Referring to reassignment of automated machine output? I don't see this slowing automation much. It makes for a simpler, more reliable process across the business. Why deal with hiring people when automation solves it forever?
And even if some folks do decide to go the 'boutique' route and hand-make something, there is always another person who will go the automated route and cut their prices by 90%. Free market after all.
It will be upsetting for sure. Lots of entrenched interests in renting automation at high rates. Maybe have to do it in a phased approach - VAT or some such?
This is the wrong way to look at this. You are not paying $12.95 for a bound copy of papers with ink on it, nor for a printed disk with magnetized bits on it, nor the electricity that powers the servers that host the transfer of those bits. This is so obvious I didn't think it needs to be said. It's advantageous for you to think of it this way, though, right? In your way of thinking, no software should ever cost anything if delivered electronically, and no more than $3.00 if delivered on a physical medium.
What about physical things with near zero marginal costs? Do you feel like your logic applies to an album on a CD when everything was on CDs?
It’s absolutely immoral to me that you see a long chain of work (research, writing, editing, marketing, printing) and judge it solely on its last component (printing). There are so many physical goods where the final piece is just a small portion of the final cost.
It's a continuum of course more applicable to some, less to some other things. I find books mostly fine, and I own a huge library, to a large extent inherited. CD albums already seemed dubious to me and hardly bought any, with software products it is entirely on one extreme of the spectrum.
Why are books fine? Publishers don't print to order. The publisher has already printed thousands before you'd even heard of the book. The only marginal cost here is delivery, and that delivery person is already likely to be going down your street with a van full of goods whether you order or not. The marginal cost is just them adding the extra stop and unload for you, which is probably pennies. Or if you go to a bookstore that already had it on the shelf, it's practically zero marginal cost.
Everything is an exercise in amortization. You seem happy to amortize the obvious parts (printing, binding, stocking, delivering), but not the behind the behind the scenes parts (writing, editing, marketing). That's a crappy distinction.
I never fully understood why we do not have the "a good tool saves me x hours, and makes me happy while doing my work so I should pay for it" culture. SW engineers are pretty well paid and companies themselves have decent margins so this penny pinching tendency is upsetting.