Bay 12 Games Forum

Please login or register.

Login with username, password and session length
Advanced search  

Poll

Does The test server work for you? Are you willing to sped time helping me test it?

It works for me.
- 6 (31.6%)
It does not work for me.
- 1 (5.3%)
I'd be willing to help test it.
- 6 (31.6%)
I'm not interested in testing it.
- 1 (5.3%)
I might be willing to help test it.
- 5 (26.3%)

Total Members Voted: 14


Pages: 1 2 3 [4] 5 6 ... 42

Author Topic: Agora, A better forum (Open Source Project): Now with Github and test site.  (Read 76188 times)

Angle

  • Bay Watcher
  • 39 Indigo Spear Questions the Poor
    • View Profile
    • Agora Forum Demo!
Re: A more effective internet forum
« Reply #45 on: October 04, 2013, 03:00:08 pm »

I've begun checking various hosting and programming services, to see how well they can meet my needs and how much they'd charge for their services. Currently, I've checked out iPage and am now checking out Quality Web Programming to see if they meet my programming needs. After this, I'll check a few different competitors for each, to see how they compare. After that, I'm going to begin putting together my kickstarter. Is there any other information you think I should put on the kickstarter? Do you have any suggestions as to whom I hire for my programming?
« Last Edit: October 04, 2013, 03:02:00 pm by Angle »
Logged

Agora: open-source platform to facilitate complicated discussions between large numbers of people. Now with test site!

The Temple of the Elements: Quirky Dungeon Crawler

tompliss

  • Bay Watcher
  • Goodbye ?
    • View Profile
Re: A more effective internet forum
« Reply #46 on: October 04, 2013, 03:55:07 pm »

Angle, if you need a mock-up or an testing prototype, tell me.
I am currently looking for a "real" job, but got my diploma last month, and searching the web and writing cover letters all day long is boring ;)
Logged

Angle

  • Bay Watcher
  • 39 Indigo Spear Questions the Poor
    • View Profile
    • Agora Forum Demo!
Re: A more effective internet forum
« Reply #47 on: October 05, 2013, 08:56:39 pm »

Angle, if you need a mock-up or an testing prototype, tell me.
I am currently looking for a "real" job, but got my diploma last month, and searching the web and writing cover letters all day long is boring ;)

Yes, that would be excellent. In other news, however, I'm having second thoughts about managing the kickstarter. I'm afraid I have little to no experience with that sort of thing. I was going to do it myself mainly because I figured there was no one else who would, but now I wonder if maybe I should seek someone else out to do it.
Logged

Agora: open-source platform to facilitate complicated discussions between large numbers of people. Now with test site!

The Temple of the Elements: Quirky Dungeon Crawler

tompliss

  • Bay Watcher
  • Goodbye ?
    • View Profile
Re: A more effective internet forum
« Reply #48 on: October 06, 2013, 04:03:08 am »

I think I read about that here, in the "Game programmers and designers enroll here. " topic :
I have a good deal of success consulting on Kickstarters as well as traditional financing.
But yeah, you shouldn't try a kickstarter without reading about it, and seeking advice first.
Logged

BoboJack

  • Bay Watcher
    • View Profile
Re: A more effective internet forum
« Reply #49 on: October 06, 2013, 03:47:15 pm »

I've had a similar idea some months ago and I'm done with the main programming stuff now I think.


Reference: https://dl.dropboxusercontent.com/u/86860124/enviromental.png
The argument of this picture is completely random and I just wrote down whatever I came up with, it doesn't reflect my opinion.
Oh yeah right, the links are directional, but I've swapped the arrows with the plus and minus signs some days ago and forgot about the directions. oops.
There is actually the possibility to make connections of the kind "only A and B together, imply C"/ A AND B => C
They are supposed to be reflected by the colors of the links, but I haven't really put time into that because I'm unsure about the design.

Right now everything is represented in a big mind map you can wander through, clicking on nodes moves the camera, you can also decide which parts of the mind map you want to currently see.

The back-end is written in Haskell and the front-end in coffee-script. PostgreSQL is the database.

I'd like to license everything under a free software license, BSD3 for example, but I'm not sure yet.

I'm struggling a lot with the design, which forced me to slow down on this project.
Everything should be easy to change and flexible, except the stupid SQL database and the queries. I plan on replacing these with combinator based sql some day. I've already tried mongoDB, it didn't fit this project, because the data is inherently relational.

I've not yet read the whole thread, but I will later.

I'd be happy to throw my stuff together with your ideas and I don't have any problem with giving away authority over my work.
I acknowledge that our ideas have very noticeable differences but I believe we can overcome these and work together.
« Last Edit: October 07, 2013, 04:41:13 am by BoboJack »
Logged

Angle

  • Bay Watcher
  • 39 Indigo Spear Questions the Poor
    • View Profile
    • Agora Forum Demo!
Re: A more effective internet forum
« Reply #50 on: October 06, 2013, 04:07:20 pm »

This is great! Do you have a site put together to test this on?
Logged

Agora: open-source platform to facilitate complicated discussions between large numbers of people. Now with test site!

The Temple of the Elements: Quirky Dungeon Crawler

BoboJack

  • Bay Watcher
    • View Profile
Re: A more effective internet forum
« Reply #51 on: October 06, 2013, 04:40:31 pm »

Its currently only running on my computer.
I don't want to pay 6€/month.  :(
But its a real webserver running on my computer.
You can look at it today, as long as I'm awake.
http://89.246.215.197:8000/
The website is very confusing since I have not spend any time on design, but you should be able to type "climate" into the search bar and hit enter. You should see the map then. (The search capabilities are similar to Wikipedia, but somehow the suggestions don't pop up correctly in firefox for me.)
You can also try this link: it'll link you to the "climate argument" http://89.246.215.197:8000/?id=39

My ip changes every day.

edit: I'm going to bed now. I'll probably start the server,after waking up, 7. October 12 CEST and let it run for the day, so you people can have a look.
89.246.207.161:8000?id=39
« Last Edit: October 07, 2013, 04:07:16 am by BoboJack »
Logged

Angle

  • Bay Watcher
  • 39 Indigo Spear Questions the Poor
    • View Profile
    • Agora Forum Demo!
Re: A more effective internet forum
« Reply #52 on: October 06, 2013, 06:53:49 pm »

You might want to change your user interface - I'd suggest adopting something like this fro the general view, and when someone clicks on a post, they get a view like this. Also, allow posts to be targeted at multiple other posts, and maybe make the general view more directional - Instead of spreading in all directions, make it travel downwards.
Logged

Agora: open-source platform to facilitate complicated discussions between large numbers of people. Now with test site!

The Temple of the Elements: Quirky Dungeon Crawler

BoboJack

  • Bay Watcher
    • View Profile
Re: A more effective internet forum
« Reply #53 on: October 07, 2013, 04:19:38 am »

All posts can already refer to all other posts.
You can not make the view more directional if all posts can refer to all other posts.
Its not a tree anymore if all the leaves are directly connected to the root.
You can take a look at some generic layouts here:
https://github.com/mbostock/d3/wiki/Layouts

Quote
and when someone clicks on a post, they get a view like this
I'll take a look at that. I'm not sure if/how one would switch back to the node overview.
There is a total lack of "how I imagine the user experience to be like" on my side.
Of course I've got a basic idea, but the actual design of which button does what and how is very unclear to me.
But the little pictures already help!
Also why exactly?

Services like iPage and "Quality Web Programming" are probably not what you search for.
iPage doesn't offer the possibility to run your own server software. Take a look at this https://www.heroku.com/
Quality Web Programming, they offer prices based on how many "pages" you have and only do php programming.
You don't want a regular website. You want a webapp, were everything could be on one page, stuff being loaded dynamically as needed.

I'd like to point out that I pretty much want it to be possible to use the website without reading any long text.
There could be additional information for every node and for the connections (I played around with that, the code is already able to)
but large texts don't work on this kinda mindmap and the whole thing seems to become senseless if the texts are so long that you can not have an overview about an opinion without reading alot, you could just read wikipedia instead.

I also started with the idea of helping with discussions, but right now I feel like there isn't even a place to even see the connection between propositions. There are a lot of ways to discuss, but there is no collective database on how people believe opinions should be linked.
Kinda totally different maybe, idk.
With the stuff I've done so far, we could also just do 2 websites. or so. you know, one being focused on disucssing(but that kinda already exists: all the stuff anvilfolk points out in his paper + http://www.discourse.org/)
the other one being focused on knowledge representation.

edit:
oh and I'm also totally in for everything Anvilfolk said.
« Last Edit: October 07, 2013, 06:44:25 am by BoboJack »
Logged

Eagleon

  • Bay Watcher
    • View Profile
    • Soundcloud
Re: A more effective internet forum
« Reply #54 on: October 07, 2013, 11:27:34 am »

Sorry I haven't been able to contribute, the shutdown has me laid off.
All posts can already refer to all other posts.
You can not make the view more directional if all posts can refer to all other posts.
Its not a tree anymore if all the leaves are directly connected to the root.
You can take a look at some generic layouts here:
What you can do is decide on a portion of the graph that is a tree. You do this by marching outwards from a root node (whatever your camera is focused on) along edges until you detect recursion (build up references to nodes in a collection, compare each to the next deeper set of connections) The problem is more that this becomes computationally expensive when you're handling multiple users in a forum with hundreds, if not hundreds of thousands of chains and loops of arguments in memory at once. Meaning we might end up needing a more complicated but optimized client-server interaction so that the user can crunch some of this himself.

On the plus side, that leads to the obvious benefits of letting the userbase improve the client interface themselves to their liking.

The more I think about this, the more I think it could resemble a true neural network in complexity. Particularly if it could be interlinked with unrelated forums, it would probably become quite complicated very quickly if people were willing to use it. It's actually a little creepy when I think "Well, what if this is just a natural extension of that underlying thought pattern into our social interactions?" I get a little zoomy, thinking about what it might build up to.
I'd like to point out that I pretty much want it to be possible to use the website without reading any long text.
There could be additional information for every node and for the connections (I played around with that, the code is already able to)
but large texts don't work on this kinda mindmap and the whole thing seems to become senseless if the texts are so long that you can not have an overview about an opinion without reading alot, you could just read wikipedia instead.

I also started with the idea of helping with discussions, but right now I feel like there isn't even a place to even see the connection between propositions. There are a lot of ways to discuss, but there is no collective database on how people believe opinions should be linked.
Kinda totally different maybe, idk.
These are both reasons why I think we should leave that up to the people responding - let them dissect opening posts any way they like, people instinctually recognize arguments that they perceive in text (or video, audio, whatever), so if we let them split a conversation off in this way with a clear representation of what's being done and drawn from in the individual post (let the forum-like conversation go normally, and allow users to decide when to send out these special branched arguments) I think things will naturally begin to head off into more specific and pointed conversation as it's used.

This kind of points at another option for regulation of discussion - a currency system, obtained through replying to posts and climbing a 'ladder' of sorts in the network, where one only gains the ability to post a root topic in the forum by participating in discussion. Other user's votes would influence this in a bell curve of relevance, where votes at the bottom would mean quite a bit less for the user's currency rating than the ones in the middle, but the same as the ones on the top. This might eliminate troll-votes, make spam nearly irrelevant by allowing the full usergroup to slam their accounts with poor ratings, and reduce forum bias from both those that have gained the ability to start a discussion and the people that only post one thing and then leave.

And yes, I am thinking microtransactions :P The idea is probably incredibly dangerous for the bias it introduces, so I'm not genuinely thrilled by the idea, but it might be inevitable if for-profit industry adopts the model.
« Last Edit: October 07, 2013, 12:10:17 pm by Eagleon »
Logged
Agora: open-source, next-gen online discussions with formal outcomes!
Music, Ballpoint
Support 100% Emigration, Everyone Walking Around Confused Forever 2044

BoboJack

  • Bay Watcher
    • View Profile
Re: A more effective internet forum
« Reply #55 on: October 07, 2013, 02:01:21 pm »

Quote
What you can do is decide on a portion of the graph that is a tree. You do this by marching outwards from a root node (whatever your camera is focused on) along edges until you detect recursion (build up references to nodes in a collection, compare each to the next deeper set of connections)
I don't understand that at all.
There doesn't even have to exist a portion of the graph that is a tree. The only way to display the graph as a tree is by leaving out connections, but why would I do that?

Quote
currency system
Because of these mechanics a lot of stack overflows legit questions get closed.
trolls/spam get downvoted and maybe flagged as spam, and bots can be kept out by captchas

Quote
root topic
Sounds complicated and restrictive.
Have you looked at the papers Anvilfolk supplied?
The idea described there is simpler and more general.
People could connect everything to everything, just like they do in real life.
And through their voting, the best assertions can be shown to the user.
No annoying long search for the user, no need to read a thread.

Quote
microtransactions
I wouldn't want to be a part of that.
Logged

Eagleon

  • Bay Watcher
    • View Profile
    • Soundcloud
Re: A more effective internet forum
« Reply #56 on: October 07, 2013, 03:15:22 pm »

Quote
What you can do is decide on a portion of the graph that is a tree. You do this by marching outwards from a root node (whatever your camera is focused on) along edges until you detect recursion (build up references to nodes in a collection, compare each to the next deeper set of connections)
I don't understand that at all.
There doesn't even have to exist a portion of the graph that is a tree. The only way to display the graph as a tree is by leaving out connections, but why would I do that?
To enable a UI that isn't a mind-map. That sort of visualization isn't exactly ideal to display larger blocks of text, as you've discovered, but it's inevitable that people are going to want to make more complicated initial posts as their ideas are not yet refined (like we're doing now - you're supplying some question about the clarity of my writing) so you do need one. You could probably do it by clustering the conversation together in box form and fitting argument lines around them, but then you have to deal with the nodes that are more distant possibly overlapping.

Human language presents multiple arguments and nested levels of meaning, because we're good at picking that sort of thing apart and focusing on one piece of perceived information, so it's strange to expect users to be perfect at refining each post they make into individual, atomic attacks, rather than letting others with different views and experiences do that for them.
Quote
currency system
Because of these mechanics a lot of stack overflows legit questions get closed.
trolls/spam get downvoted and maybe flagged as spam, and bots can be kept out by captchas
No doubt, there are potential problems. It's just brainstorming. What I think might be a problem, specifically, with a random argument graph vs a restricted opinionbase like I described, is that users that have no basis or understanding to question an argument will do so whenever they want, vs. whenever it's warranted by their level of contribution. Of course, other users will downvote what they see as destructive to the conversation, but only if they're motivated to spend the time doing so - if you have a great, devoted memberbase this might be common, but the idea of fair moderation on a network of arguments implies that moderators have to read each link to form a less biased opinion, which would become increasingly difficult as it got more complex - which means artificial limitations on the complexity of an argument are sure (in my mind) to arise.

Coming from the neuroscience angle, our brains also function off of a currency system. Each cell requires an intake of energy that biases the whole system towards useful behavior in the overall network. If they aren't useful, they eventually become starved of neurotransmitter rewards, and either repurpose towards a more useful function or atrophy.
Quote
root topic
Sounds complicated and restrictive.
Have you looked at the papers Anvilfolk supplied?
The idea described there is simpler and more general.
People could connect everything to everything, just like they do in real life.
And through their voting, the best assertions can be shown to the user.
No annoying long search for the user, no need to read a thread.
To be clear I'm not proposing any restrictions on the specific content of arguments, and off-topic ramblings would form their own hierarchial 'root' nodes. Additionally, as long as the information about the structure exists, it's fairly wasteful programmatically to discard it, as additional mathematics can be performed on hierarchial networks that are impossible on random ones. All it requires to keep this information around is a scaling variable for the entire network that increases as it grows new unique levels, and a single integer per node representing what level in the hierarchy they belong to and which levels in the hierarchy are their 'superiors'.

Actually I'm not proposing any changes to the core of this project, just branches that I'd like to explore in the future - obviously I can't tell you what to develop, I'm just sharing some excitement over this from a network/information theory nerd's point of view. But if offtopic posts were valuable and extensively explored, they would probably be upvoted to a higher level anyway, as they'd receive more traffic from users exploring and posting in that particular part of the graph.

I'll go and read Anvil's papers, but I can't promise that I'd understand them. I'm informally and incompletely trained, and I acknowledge that openly. Mostly my point is that we can consider other aspects and divisions of network theory to apply to this system, and that's the appeal it has to me.
Quote
microtransactions
I wouldn't want to be a part of that.
You don't have to be :P I'm not really interested either, or if I were, I'd want it to be in the position of a non-profit with ethical oversight determining what it got spent on. But devil's advocate: If this is developed into a codebase to be used in multiple forums, you'd need to patent the ideas here to prevent such a use, and defend it vigorously against imitators.
Logged
Agora: open-source, next-gen online discussions with formal outcomes!
Music, Ballpoint
Support 100% Emigration, Everyone Walking Around Confused Forever 2044

Anvilfolk

  • Bay Watcher
  • Love! <3
    • View Profile
    • Portuguese blacksmithing forum!
Re: A more effective internet forum
« Reply #57 on: October 07, 2013, 03:54:25 pm »

Sorry I've been out of the discussion for a little bit - have been talking to people from this thread about the possibility of pursuing this, which is great :)

That implementation is pretty kickass! Like I said through PM, it has a ton of potential! It's also really interesting that it has a backend with the data, and then a separate front-end. This should be fairly critical to the system we've been thinking of. You only implement the system once, and then implement lightweight interfaces for, say, Facebook, different forums, stand-alone applications, mobile applications, etc.


Eagleon, there's no worry in creating the Singularity using this, it eventually just becomes a system of equations :) Nodes don't really do any computation!

I guess the proposal here is, in general, give the ability to formally debate things to communities that are open to it. If the community isn't open to it, because it lets in too many trolls, or because people don't really care about it, it's not going to work. The community is its own moderator, through the means of voting.

Either way, it should be fairly straightforward theory-wise to make sense of having people with heavier weights in their votes. This allows for the system of currency, reputation, credibility, reliability, or positive/negative feedback on users. That could be an interesting feature for more closed communities, although it could put newer members of a community at a disadvantage... And it adds overhead.

Spoiler: Overhead (click to show/hide)

Moderators can simply be added application-side (as opposed to theory-side) by allowing them to add or remove nodes/attacks, rather than just vote.

I still feel that you don't need to lose the forum structure, or whatever underlying, typically hierarchical structure. It should be just as easy (or easier) to navigate as a forum like this one. That information simply isn't used in computation of debate outcomes... but it's there, as meta-data. Certain posts could be identified as being "first posts" or "debate starting posts" within a subforum of a forum. There's no problem with that. Once you get to within a thread, however, you're in graph country. You can still try to display things with a sense of time, by highlighting new posts since your last visit, or by having older posts in the centre, newer posts in the periphery, or try to maintain similarity with a thread by having newer posts nearer the bottom of the screen, if at all possible.

But again, that becomes meta-data, maybe secondary to the fact that you can now reason with other people in a more formal way, in a system that tells you who ends up being right or wrong. Heck, it goes deeper than that. Each node/argument is given a strength, so that you can have any number of participants proposing any number of arguments, and each of those arguments, not sides or people, has a degree of acceptability, so to speak, calculated from votes and from attacks (and votes on the attackers, and so on almost recursively).


The fact that we are writing huge amounts of text here is due to the fact that we only have the forum structure. We are forced to, in text, reply to everything that's been said before, and, in text, explain why it attacks or doesn't attack another point somebody else made. It makes everything more verbose. Hopefully people will get used to using smaller, to-the-point sentences. Instead of replying with one post, they'd reply with maybe half a dozen arguments.

Eagleon

  • Bay Watcher
    • View Profile
    • Soundcloud
Re: A more effective internet forum
« Reply #58 on: October 07, 2013, 04:24:56 pm »

Eagleon, there's no worry in creating the Singularity using this, it eventually just becomes a system of equations :) Nodes don't really do any computation!
Ah, but people do ;) And if they're using this network properly, they'd be extending their knowledge into a system that is readable and processable, and useable by people. I've always said that neuroprosthetics were a better solution than trying to understand our own motivations and behavior without bias to create a new intelligence, and that the internet could be considered one, albeit quite a bit less efficient than we'd like. If you made this hierarchial (which the internet and the brain are both theorized to be) they might fill that role quite nicely.
I guess the proposal here is, in general, give the ability to formally debate things to communities that are open to it. If the community isn't open to it, because it lets in too many trolls, or because people don't really care about it, it's not going to work. The community is its own moderator, through the means of voting.
Absolutely, but in the meantime you have apathy and lack of free time by users threatening the stability when trolling and spam happens, and no one is there to downvote it. I know I don't actively use 'thumbs up' or whatnot often - I'd rather post a response, because it lets me present new points of interest in full visibility to the topic at hand. Giving a motivation of some sort would improve that and get more people voting, vs spewing nonsense *coughEagleoncough*
Spoiler: Overhead (click to show/hide)
(Edited for more explanation) -This is the purpose of the hierarchy in the implementation I suggested - stagnant debates are, by the structure of this forum, going to be the ones in the middle. They get trimmed the most, so the ones that survive are going to be very strong, and pretty much universal. As you near the top, you get into novelization land, where people want to figure out what they're talking about before delving into more detailed debate. You -don't- want those to disappear, because they represent the basis for the entire discussion. The upper-most ones are going to have lively new discussion branching outwards if anyone sees something new and important to discuss on that level, and will support and reenforce the most amount of nodes, so they should be the rarest and least likely to disappear. The lower levels will be getting the more refined, pointed arguments, which is a good place to demonstrate your core knowledge of a subject and build a reputation anyway. Altogether you get rewarded most for improving the core understanding of the next higher argument. But maybe not.
I still feel that you don't need to lose the forum structure, or whatever underlying, typically hierarchical structure. It should be just as easy (or easier) to navigate as a forum like this one. That information simply isn't used in computation of debate outcomes... but it's there, as meta-data. Certain posts could be identified as being "first posts" or "debate starting posts" within a subforum of a forum. There's no problem with that. Once you get to within a thread, however, you're in graph country. You can still try to display things with a sense of time, by highlighting new posts since your last visit, or by having older posts in the centre, newer posts in the periphery, or try to maintain similarity with a thread by having newer posts nearer the bottom of the screen, if at all possible.

But again, that becomes meta-data, maybe secondary to the fact that you can now reason with other people in a more formal way, in a system that tells you who ends up being right or wrong. Heck, it goes deeper than that. Each node/argument is given a strength, so that you can have any number of participants proposing any number of arguments, and each of those arguments, not sides or people, has a degree of acceptability, so to speak, calculated from votes and from attacks (and votes on the attackers, and so on almost recursively).

The fact that we are writing huge amounts of text here is due to the fact that we only have the forum structure. We are forced to, in text, reply to everything that's been said before, and, in text, explain why it attacks or doesn't attack another point somebody else made. It makes everything more verbose. Hopefully people will get used to using smaller, to-the-point sentences. Instead of replying with one post, they'd reply with maybe half a dozen arguments.
Verbosity serves a purpose in communication. We are not naturally inclined to limit it. I appreciate your work (from my very basic understanding of it) on an informational level, I love things like that which break down a system into clearly understandable parts that fit together nicely, but I think it's unwise to ignore the nature of the beast that's using a system like this and expect people to speak in crystal, mechanical clarity (following messy tangles of networked parts that appear completely unorganized), when they have flowers for brains ;)

In any case, we're still at a level where our options for this are not limited by feature fitting towards a specific application, all I can hope to contribute is a nod towards the interrelated aspects of UI work and the gravity of discarding the raw information it could take advantage of down the line.
« Last Edit: October 07, 2013, 04:59:42 pm by Eagleon »
Logged
Agora: open-source, next-gen online discussions with formal outcomes!
Music, Ballpoint
Support 100% Emigration, Everyone Walking Around Confused Forever 2044

Antsan

  • Bay Watcher
    • View Profile
Re: A more effective internet forum
« Reply #59 on: October 07, 2013, 05:09:33 pm »

I guess this won't help, as most people don't do Common Lisp and you already have got something, but either way:
I have a graph library available and I would like the chance to see which features are missing.
If you use quicklisp, it is available there as "fgraph".
If you just wanna take a look, it is on bitbucket.
https://bitbucket.org/thomas_bartscher/cl-fgraph

Quote
What you can do is decide on a portion of the graph that is a tree. You do this by marching outwards from a root node (whatever your camera is focused on) along edges until you detect recursion (build up references to nodes in a collection, compare each to the next deeper set of connections)
I don't understand that at all.
There doesn't even have to exist a portion of the graph that is a tree. The only way to display the graph as a tree is by leaving out connections, but why would I do that?
Do a breadth first traversal of the graph and collect the edges and nodes you traverse into a new graph. Viola, you have a portion of the graph that is a tree. Of course it does not include all connections between the found nodes, but for displaying purposes you don't need those available all the time. Display all the edges the node your mouse is over are incident to.
If you want to cut down on computation down, stop the traversal at a specified depth or maybe after you collected a certain number of nodes.

If you wanna get fancy, try to come up with some measures that describe the distance between two nodes and don't do a breadth first traversal but replace the queue in the bft with a priority queue.
Maybe provide different measures so people can have different views on the structure of the discussion.

Either way, it should be fairly straightforward theory-wise to make sense of having people with heavier weights in their votes.
Thought to that: Use a network of trust to determine the weight of votes. Not that I know how to do that, and I guess it needs a lot more space and time.
Quote
And it adds overhead.

Quote
I still feel that you don't need to lose the forum structure, or whatever underlying, typically hierarchical structure. It should be just as easy (or easier) to navigate as a forum like this one. That information simply isn't used in computation of debate outcomes... but it's there, as meta-data. Certain posts could be identified as being "first posts" or "debate starting posts" within a subforum of a forum. There's no problem with that.
Uhm, I am not quite sure what you are proposing, but it seems backwards in relation to what I thought one would do.
Any post has a time-stamp/version history (yeah, let's get fancy). If you want to see how a certain thread has evolved, choose a starting point, an end point, a heuristic to measure distance between nodes and then run a path-finding algorithm. The output of that algorithm is a thread that can be viewed by the user. It is implicit in the graph and the heuristic, but it needn't be anywhere explicitly.
Logged
Taste my Paci-Fist
Pages: 1 2 3 [4] 5 6 ... 42