Thomas, thanks for the links, will certainly come in handy!

Colin, thanks for your opinion! I agree with most of it, however I have to take into account other considerations. For example, implementing spell check just for English would be discriminative, but implementing it for many languages would be a very difficult task and would make CintaNotes grow from 10 MB to 50 MB, which many users would find frustrating.

Votes of course are not perfect, but we have nothing better I'm afraid. You think that spell check is essential, but many think that you can get by without spell check, but encryption is essential. Who is right?
All in all, IMO votes are a good enough approximation of user demands. Our goal is not to get all possible users, our goal is to find our own niche and stick to it, because for a small software company with 2 devs it is impossible to implement all features and compete head to head with giants like Evernote and Onenote. At least at this moment

And to survive, we have to do exactly the opposite: not go with essential, but go with unique features, ones that Evernote and Onenote don't have. 

Agreed. And implementation will use a custom collation. Hopefully we'll even be able to find a ready-made implementation of the natural order sort function.
But first we need to see some votes, of course ;)

Thanks for the clarifications!
Ok I'm now changing the status to "Planned". However I unfortunately can't make any promises on the dates, since there are a lot of other features planned which have more votes than this one. Hope you understand!

Thanks for your opinion and welcome to the forum! I agree that spell checking is an important feature. However for some reason it is not in the top 10 regarding user votes. If users vote for it more actively, and it gets to top 10 requested features, we will certainly consider it for one of the upcoming versions!

SQLite does support user-defined functions, but it seems that SQLite doesn't support functional indices. However it does support custom collations, and this is what should be used here. Still I think it will be much slower than the Unicode  collation that is used now.

So let's wait till this suggestion gets some votes first :)

Thanks for the clarification, Thomas!

I think there's a reason why this is an option: "natural" sorting requires extra processing and is slower than the default sorting. If you don't need natural sorting, why should you pay with performance for it?