Sometimes, some reply-tree have an expected structure. I think the reply-tree structure ‘concept’ is pretty obvious. Well, I guess it’d be great if some patterns were identified and the display reacted to this.
I think that by giving some examples will be sufficient to illustrate the feature:
You’re reading a topic, and then two or three persons start talking to each other. You read some of that, but you’re not really interested in what they are talking, but you keep ‘seeing’ their conversation. It’s like a cloud above stuff you’re actually interested.
So it would be great if you could easily identify that as a ‘conversation tree’, and then skip it entirely. Think of it as if the vertical scrollbar were really a tree-like scrollbar, and the client would know what branches you’ve read and so on. Its not that you would miss any content, just reorder it according to your personal interest.
So, let’s say you enter a topic and start reading the replies. There’s the master tree, where the topic itself starts, and then lot’s of branches. Small branches could be merged into a single branch for simplicity (a ‘leaf’ reply, or just a small total number of replies [from that particular sub-tree]). This is because there’s not point in considering that a full fledged sub-tree if you would need to actually read it in order to decide if that conversation is ‘probably interesting’ or not.
So, continuing, you enter this post, and you encounter an actual sub-tree. It’s a conversation that has lots of replies in that particular sub-tree. It even has it’s own name (like, “aspect blablabla of this subject”). By having names to conversation trees, you can decide even more efficiently if that conversation, for you, have a higher or lower priority. It’s not that only that branching is named, other sub-branchings could also be named.
Finally, you could, instead of looking at each reply, look at those branchings themselves, with their name tagging accordingly. Maybe, also some counter features, that represent ‘participation’ (such as the sum of all replies given on that sub-tree) or most active participants. You don’t even need to visualize everything at once: just a 2 or 3-level depth at first.
Finally2, not only seeing branching naming, but also consider an entire ‘sub-tree’ as something that can be named, that would pretty much summarize various named branching that are in that direction.
How would those names be given? maybe anyone, and then some voting. A box could show up, like letting you name that branch (maybe it’s not considered a branch yet, but it’s about to) or vote on previously suggested names (by other members). When there’s a clear winner, just keep it and discard others. The problem would be with higher level naming of the sub-trees. If a sub-branch appears, the system could identify names scores based on the branch where the vote were actually given. Not just the vote is information, the reading position when that vote were given is also information. So based on that, the system could identify names that captures both branches, and more specific names. So maybe higher level naming can also be quite automatic.
Well, idk if this s a repeated suggestion… and I suppose it’d be hard to implement. But I’d like it a lot! XD