Yea ... going to say no to this one mostly because auto-translation software isn't really the best right now. And I don't think we want to dump a lot of money and time into building out a feature that Google hasn't even gotten right yet.

To best explain what I'm talking about: http://xkcd.com/1425/

Hey, sorry if the communication is a bit late or in some cases non-existent. We're trying to keep up with it but often times we're just too busy with our faces glued to the code. As far as the suggestions are concerned. We are keeping track of them and as others have said, theymos has final say. But in the mean time we track what we can and offer our suggestions or feed back as well.

As the forum software starts to firm up more in terms of features, we'll probably be more vocal but as of right now, we're trying to go through all the things mentioned in the stickied thread of this board. Those have top priority.

As for the beta launch, it's coming! The date was self imposed and code wise we're there already. We are just working out the kinks in getting it up on a public server. So sit tight. It should be here soon.

1.) Flarum seems to be a student's project that hasn't reached a fully useable state yet. Although we're not fully done with our current forum software, we have gotten a bit farther than their current code base.

2.) Jeff Atwood is one of my heros so I can't really complain about the quality of discourse but I do believe it was written with the idea of replacing the comment system for blogs first. The idea of being a stand alone forum came second so as a forum user, you tend to lose a lot of the features and the UI isn't made with that in mind.

Honestly, the closest forum that can compete with what we're going for is probably nodeBB. We have a lot of catching up to do to be as feature complete as they are. We're already ahead in some aspects.

I need to update those images. I re-hauled the editor a few times since then. There's image support and a cleaner look now. I doubt this will be the last time I update the editor though. The designer hasn't gotten to the editor yet. Also, my next goal is to break out the editor into something that floats along the screen instead of being in a fixed area.

I'm not sure what you're trying to say? Supporting hotlinking through a cdn and building in support for image upload is more work than just using imgur. Also, it creates a third party dependency that we can't control. So if imgur happens to go down (which it has before). Then all images on the server would be down to. If you want to use imgur, you still can.


Also, I need to make a modification on the last post I made. The hotlinking support will no longer keep a copy on the cdn. It'll just stream the image from the original source. I've got that working so far. I just need to move on to image uploading.

The plans are to support both image upload and to support hotlinking. The hotlinking is going to work the same way most other pages handle it (github, facebook). Basically if you upload an image, it'll be stored on our cdn and the BBCode to that link will be supplied to the writer in the post editor. If hotlinking is used, the url will be preserved for the user but the hotlinked image will be moved to our cdn and that cdn url will be displayed to everyone else.

The main reason for this system to provide a secure way to handle images. All images will be validated for exploits along the way to our cdn. This way someone can't just hotlink an image and when the post becomes successful, bait and switch it for someone else. In the end, all images will reside on our cdn.

I hope it goes without saying that the cdn will be on another domain to prevent XSS attacks as well.

The other reason for this is to allow for a better user experience for the user. We don't want the user to go to another site just to link or upload an image.

Tangentially working on this right now. All signs point to yes.

BTW, sorry for the lack of interaction in the past couple of weeks. Been heads down working on a lot of moving parts.

I'm still hesitant mainly because even in signature campaigns, the OP is quite often edited with new information. Also, there are circumstances (although rare) where a later post might need to be edited. In the event that the first version of a post needs priority above any edits, we can file an issue for a specific view of a thread where the posts shown are the original post and you'll have to click to expand for any newer versions. (the reserve of what will be the default layout)


I think a concrete example would help more in this area.


So just thinking off the top of my head, I feel like signature campaigns should be allowed to have users edit their post, but the user may only post once in that thread. Also, when the campaign is over, the OP can lock the thread at that point, keeping others from posting in it. I think that solves the majority of issues with signature campaigns. If we want to take it a step further, we could also allow OP to tag a certain version of a post and that version will become the 'default' post.

Contest may have a different structure where posts aren't ordered by time as like most forums but may be sorted by OP's desire. (Maybe winning posts show up right under OP, I'm really just freebasing at this point so don't take any of this as truth.)

Group Buys might follow the same structure as signature campaigns but OP could tag posts where the orders have been accepted (versions be damned). Maybe those tagged posts could get hoisted to the start of the thread.


The great thing with what we're building is that we have the ability to really go wild with the features since we're building from the ground up and have complete control of the stack. But in order to build something that the COMMUNITY needs, we need concrete examples. I don't mean to preach but solid examples is always better than notional ideas because we can point to some use case when we actually try to build the damn thing. Again, sorry for the rant.

The new forum will have versioning for all posts. This means that every change to a post is kept in the database and all users will be able to see all versions of a post. This makes the point of this thread rather moot since it will be on by default for all posts.

That's just one of the repos =). We've split the code apart over repos to better modularize the entire infrastructure.

Also, the slickage one is deprecated so don't follow that one.

Lastly, the dependencies have changed since first writing that README. Couch and redis are gone for now.

All boards, posts, threads, users, etc will all transfer over.

And no pic yet. =(

The alpha release is just a code release at this point. But you can try and run it yourselves, just don't expect the world with it.

Since the github pages for the new forum are public, anyone can install and run it but don't expect any support yet. It is still in alpha. So feel free to browse the code and if you're brave enough, install it. (It really just one command to install anyway)

We won't have a public staging server up. So if you guys don't have node installed or aren't comfortable setting that up, please wait till a later release. But feel free to keep watching our github repos. Those are all public.

Alpha-Preview will be done by end of August. Not the whole forum.

Being that it's an Alpha release, don't expect much other than a working forum. None of the bells and whistles should be expected.

I like this idea. We were just looking at slashdot and their implementation of this. Again the metrics that we'll be using may change soon since we want to give maximum flexibility.

For the forum, yes.
For hackernews, that's how having only likes changed content. I was just trying to illustrate the point that having just likes also has it's own drawbacks.

A concrete but fictional example would be if post likes were tied to user rating. Many users with malicious intent could be pushed into higher and higher user rating tiers (if such a thing existed.) Basically any system that only has a up button could easily have everything pushed "up." ("up" could be anything too.)

Also please consider the psychological aspect of a like and dislike system. Reddit has become famous for karma whoring. Some people go to extreme lengths for a virtual number that means absolutely nothing.


I would like to try and figure out a way to tie likes into a meaningful metric to try and fight spam posts and malicious users but I haven't figured out a viable route yet. I may switch to something entirely different metric at any point in the future too. As of right now, nothing is set in stone on that subject.

If you go on hackernews, there's only a like button but even that has shown that it was lead to unwanted side effects. In their case, many of the same articles get pushed to the top because there's nothing to say that the same content has already been posted. Think of the like and unlike as way to move content up or down. All three scenarios (likes, dislikes, likes and dislikes) come with their down sides.

Sadly, the code that you guys have been using has been modified so it no longer conforms to that page (CSS wise). There's also a lot more BBCodes then there are listed there. The goal was to support as much BBCode as the users already use. I've already moved on from this module already though.

The easiest way is to just use the @username way of doing things, IMHO.
It gives a simple key for the JS to watch and bring up a dialog that allows you to choose the user you wish to tag. Rather than keying on every single keystroke and trying to match that against a username like how facebook does. As for a BBCode option, this may fall into the backlog because it's more error prone with misspellings while the dialog, the user can simply just click on the user they mean to tag and inject the right username (spaces and all). 

Also disabling notifications for username tagging should be simple enough.

I actually added on to Vod's post because I wanted to let the community know that I was tapping into the likes functionality as a metric for reputation.

Although the reputation system I have in mind is not like the current system in any way. I actually want to continue this discussion. Maybe new thread?

ranlo

So if activity were to be removed from the equation all together so that the only way a user can improved their ranking (aside from buying an account) is to generate content that others would like/upvote. Would that alleviate some of your concerns?