As the suggestions platform guidelines describe, the intention is just to collect feedback, and we don't guarantee that any suggestions will be implemented. The reason I keep coming up with "we're a small company" is it's true, and it's by far the biggest limitation we face in our ability to get things done. Frankly given our size I'm proud of how much we actually do get done.
Only 2 are marked as "shipped" this year.
That's misleading - I figure you're looking at the "shipped" category, sorted by recent. In that case the dates are when the submission was submitted, not when we marked it shipped. So if a suggestion posted in 2020 was shipped this year, in that view you'd think we didn't do it this year, but we did. Given the suggestion platform has been up close to 4 years, and 200 ideas are shipped, that works out to about 50 a year, or one a week. And that includes ideas that were major projects that took months to complete, like the scene graph feature.
It would be really helpful if you could mark ideas that you are considering to implement.
The problem with the size of the suggestions tracker is with 1000 ideas, it's a huge amount of administrative work merely to review them. Providing a useful comment is often difficult unless we research the options and experiment with some prototyping. Doing this for every idea is just an absolute mountain of work and could tie us up for months, during which time we get nothing done at all. So, perversely, providing useful feedback on every idea means would mean we actually implement fewer features.
I don't think the "minor suggestions" category works that well, because users often don't have the technical knowledge to know if what they are asking for is actually easy or not. It can still take a fair amount of research (e.g. examining thousands of lines of code to estimate how easy a change would be to make to our existing codebase) and prototyping work to evaluate what is ostensibly a minor suggestion, which ends up being a lot of work. In fact I wrote a whole blog about this a while ago, on the unexpected complications of minor features, which covers two cases which turned in to loads of work, for example.
Every time I post a new idea, I have to revoke a vote from one of my older ideas.
Well, I think this is kind of reasonable actually. With 1000 ideas, which I estimate is probably about 10 years worth of work, how useful is it really to throw another idea on the pile? I'd rather we had fewer popular ideas, rather than a mountain of things only one or two people want. That's what the limited voting is for. It's meant to force you to pick a few things you really care about, and not so you can inundate us with hundreds of ideas amounting to an infeasibly huge amount of work.
In the feature guidelines I wrote back in 2017, I did write about some of my fears how this might end up working badly:
In particular we are worried about people making statements like "Scirra don't listen to their customers - they're ignoring this feature even though everyone wants it". If that kind of thing happens a lot, we will probably just shut down the suggestion platform.
So if the suggestions platform just ends up being a big source of disappointment and complaints, I guess we could just shut it down. But do you really think that would be better? At least people can vote on popular things as it is.