Dragging Atlassian is very much in vogue on HN these days, and yes, they deserve some of it. That said, the spirit of HN calls its users towards curiosity rather than division, and invoking “hate” is without question a divisive tactic.
(Edit) adding a link to the guidelines: https://news.ycombinator.com/newsguidelines.html
(Edit 2): It would also be prudent of you to mention you’re building a product that is Confluence-adjacent: https://news.ycombinator.com/item?id=30295553
Lots of supporting comments over the years https://hn.algolia.com/?dateRange=all&page=1&prefix=false&qu...
The main thing I hate about confluence isn't really a confluence problem per se, but with the way organizations use wikis in general. Most organizations record jewels of wisdom among a boat-load of crap (looking at you, internal amazon wiki.) Finding the useful information in the wiki is a job in and of itself. Maybe if there were some way to enter in a term you were interested in and have the wiki software search for pages with that term. Or even better, maybe this "search feature" could rank pages based on meta-data like how frequently pages were accessed or how many links there were to a page (or how fresh those links were.) Heck, you could probably make a business out of doing this across the internet. But that is, of course, crazy talk. Everyone knows you can't add a usable search feature to a wiki.
So... confluence search is the first thing I would fix. And in a more general sense, maybe adding hashtags could help. That way I could tag various pages and use that in subsequent searches.
The medium really is the message here and the message wiki's present is "all information is unstructured and is certainly not in a hierarchy." And when you have info like that, wikis really shine. But most of the info in a hierarchical organization is itself hierarchical. It would be cool if the "spaces" concept was hierarchical. We could have an engineering space and below that a shared services space and below that documentation for various teams. If you knew the structure of the organization, and you knew you were looking for info about a service or feature of a particular team, you would have a chance of finding the info you were looking for.
And since we're talking about confluence... damn I wish it's editor would die in a fire. I usually write docs in an editor on my desktop (usually emacs if you must know) and then copy and paste it into confluence. And then spend an hour trying to figure out why I got the formatting wrong. Yes, I understand there's a WYSIWYG editor on confluence, but I use more features than bolding or italicizing a word or two. I actually try to use both "normal" and "monospace / code" typefaces and that's a major horror-show in one of the current confluence editors.
I say "one of the current confluence editors" because there's really no way to figure out what version of confluence you're using. Am I using a cloud instance? Is it an instance we're running locally? No idea. There's no way to figure it out. I tried to search through the Atlassian docs site and found an article entitles "How to find out what version of jira / confluence I'm running." The first thing was "If you're running version 1.x or 2.x then go here... If you're running version 3.1.x through 3.5.x go here..." and so on for twelve different versions of cloud and server confluence, each locating it's version info somewhere slightly different. So... if you have to know what version of confluence you're running so you can figure out what version of confluence you're running... uh... that seems a little weird. I eventually found our ops guy who installed it and he mentioned "oh yeah. I flipped the bit that says 'publish version information' because of security." So apparently you can't even figure out what version of confluence we're running, even if you knew which version we were running. Since editor features are dependent on what version of confluence you're running, you just have to try a feature and see if it works. Is a particular feature not working because you misunderstood it or because you're running a version that doesn't implement it? who knows.
Also, the auto-timeout is something like one hour by default. So... you want to finish editing things on the wiki within that hour (or two hours or eight hours or whatever your auto-logout is set for.) If you don't and you hit submit, all your content is now gone. What would be cool is if the editor could save your contents locally (HTML5 does local storage) and THEN tries to post it to the HTTP API and if it doesn't succeed, it tells you "oh you have to log back in again" but here's the main point... it recovers your edits from local storage and then re-posts them after you've logged back in. Just an idea.