[simonsarris]

(I make GoJS)

It does compete with "free", so in one sense it's expensive. In another sense, if it saves you X programmer days, or months, it becomes cheap very quickly.

What you are buying is thousands of hours of thought about implementing interactivity across browsers, data-binding, an undo manager, etc. Considerations large and small. Put another way, what you are buying is time, which many projects find crucial. (And with support, you are buying "can you show me a proof of concept for...", "how best might we...", etc)

[jmchuster]

It's not so much that you're competing with free, it's that you're competing with open source (which is also often free). The nature of open source is that it's very often better in terms of both quality and community. Add in the overhead of acquisition, and it makes it a no-brainer.

It's very hard to tell ahead of time that a paid library is going to save you X programmer days. You'd really need some way to convince someone that you are best-in-class in documentation, in community, in stackoverflow answers, in plugins, in features, in customizability, in debuggability, before they even start the painful and lengthy requisition process. And dedicated support is nice, but is it better than being able to google the answer to every question?

Now of course if you pick a small niche and no strong competitor exists, great, it's pretty easy to become best-in-class with your thousands of hours. But pick anything that is used by FAANG, and now on top of the community contributors you've got dedicated engineers they've assigned to invest in their investment.

[true_religion]

Well, in my experience, you start with open source libraries first, and try to build functionality atop that. If becomes too troublesome then and only then do you look for a paid library.

At that point you can justify the rate by looking at how much work was done previously.

For an example, about ten years ago I was tasked with doing JavaScript charting for stock market data and used open source libs at first. After 8 weeks of working with this, it became clear that this was a tar pit that was dragging time away from our startups core problems. We then bought HighCharts.js for about 3k per year, which saved us over 10k in engineering time instantly based on the idea that we would needed at least two more months to figure things out ourselves.

[ygra]

As one of the developers working on yFiles, another commercial solution in the same space I feel the problem of competing against "free". There's also relatively few useful feature subsets that can be licensed separately for a smaller price. But we also regularly get questions what days is apart from D3 which is a lot cheaper (in visible monetary terms at least).