It's always seemed obvious and reasonable that posts could be linked just to topic tags, even without a paper citation. And this would be easy to do.
The one problem is that when a post contains a paper ID that fails to parse (due to a typo or whatever), it would end up getting classified as such a "topic post" (assuming it does have topic tag(s)). Depending on your point of view that could be seen either as an improvement or worse than the current behavior (which simply won't index the post until the user fixes the paper ID to something recognizable). Inevitably some people will find their posts get indexed this way contrary to their intentions, so we'll have to give them an interface for fixing the error. (For us, the easiest way is for them to edit their original post, but they may not want to do that?) We could give them a way of adding a paper ID directly on a selectedpapers.net page for their post.
It's always seemed obvious and reasonable that posts could be linked just to topic tags, even without a paper citation. And this would be easy to do.
The one problem is that when a post contains a paper ID that fails to parse (due to a typo or whatever), it would end up getting classified as such a "topic post" (assuming it does have topic tag(s)). Depending on your point of view that could be seen either as an improvement or worse than the current behavior (which simply won't index the post until the user fixes the paper ID to something recognizable). Inevitably some people will find their posts get indexed this way contrary to their intentions, so we'll have to give them an interface for fixing the error. (For us, the easiest way is for them to edit their original post, but they may not want to do that?) We could give them a way of adding a paper ID directly on a selectedpapers.net page for their post.