I can't believe there's no working forum solution

can’t help but think of it in a simple but ignorant way... why can’t you just pass the same username/password from SMF to the MODx username/password fields in the modx_ database.

It’s not that simple: SMF uses SHA-1 passwords, MODx uses MD-5. These are one-way cyphers and you can’t unencrypt them from the database easily. You have to override the authentication on one side, and that’s where it gets tricky.

It should be possible with the right forum framework that supports a third-party API for authentication, or possibly Open ID. That said, the direction that would like be most useful in the long term may turn out to be to build a native forum/moderation engine native to MODx to go along with the other tools it provides. Consider the following idea-dump:

[*] xPDO for database abstraction and future running inside Revolution
[*] WebloginPE for user/profile management, though this in light of how Revo handles things should be considered
[*] a forum display engine (this is actually already done but needs to be ported to xPDO)
[*] a forum admin tool built on top of ExtJS which is what drives the entire Revo manager
[*] a forum moderation tool ... could be inspired by the comment moderation tools found in Jot
[*] chunks to template everything on the display side: this would make it incredibly easy for folks to have truly custom, semantic forums
[*] phpThumb to handle image thumbnails and upload resampling
[*] a private messaging system that works with Weblogin PE (this has been done elsewhere and should hopefully see a public reveal soon)

An important question becomes what periphery "forum bling" needs to be implemented. I’d say keep it simple and just do a great job with the simple forum functionality. When Revo comes around designed from the ground up to use external authentication, someone can tackle a true integration with a system that supports such things. Then you could have the right tool for the job in all situations.

… I would be more than happy to help people make their SMF theme match their Modx theme.

Ahem, I know this great CMS software that runs SMF and recently relaunched it’s site. Looks to me it still hasn’t updated it’s forum template yet…

Thanks for your input Ryan... yeah seems like it might be a matter of not really possible or feasible with 9.6.3...

Does the new modx site run on Rev?

I have been using webloginPE as I like the extendability of of the profile system and would make an excellent basis for a forum profile system. I like the sound of a custom forum cos in true modx nature it will be no doubt fairly easy to extend, hack and customise to your needs.

I would ask that the forum have the basics such as topic creation, edit you own posts, category creation (admins)... PM system and what not could be added in future but having the basics would be really nice. I would like it to use an external database or it’s own tables rather than Jot. My fear is the modx document limitations, would a custom forum have any bearing on this? Say if 5,000 topics were created, would that slow the modx system down? I’m a little ignorant to how this would work.

Currently I have over 1400 members on my SMF forum for example, thousands of posts... would something like that slow the whole system down or would using an external DB combat that issue?

I’m developping a native Modx forum (based on the one Ryan was talking about). The simple logic work but i want to go deeper and enhance my php skill along the way.

So i’ll probably propose a native forum script but it will take some time to achieve all my goal.

I’ll keep you informed once i can release an alpha version for testing. Don’t expect anything soon tough!

Hi all,

I’ve been referred here as I’m keen to build a forum into my site rather than bolting on and synchronising something else. I must mention that the SMF connector does work well, but for me the hassle of properly integrating any common data, even as simple as the menus, and having to template two different systems is asking for trouble in the long run.

So, I am keen to build the forum software myself. I would rather wait for Revo, due to the learning curve, options and benefits its provides, but would happily join in with any team that plans to start and put my energies there instead.

I have some good ideas, can code PHP, MySQL, HTML and Javascript, but unfortunately am awful when it comes to graphics. I have just stepped down after five years of being a senior administrator of a large public forum, with 2000 members and 70k posts. I do have a full time job, but I have energy and resources to put into this, am excited about the potential of MODx for my site and itself, and have already planned the time to dedicate to the task anyway.

If people are serious, then I am aboard.

So there is some baseline support. The first step is deciding on the baseline features, and I think rossco’s set is a great starting point for discussion. Next would be to implement the basic features on the front end, and that code already exists as a starting point, however it does need to be re-implemented in xPDO. After that an admin system for setting the basics up like categories, topics, etc. would be a good step. Finally, moderation controls could be built.

This honestly deserves to be a reference component. I’ll see what I can do about getting a SVN repo set up so we can track progress in JIRA and collaborate on development.

Quote from: rthrash at Mar 28, 2009, 03:41 PM

So there is some baseline support. The first step is deciding on the baseline features, and I think rossco’s set is a great starting point for discussion. Next would be to implement the basic features on the front end, and that code already exists as a starting point, however it does need to be re-implemented in xPDO. After that an admin system for setting the basics up like categories, topics, etc. would be a good step. Finally, moderation controls could be built.

First, IMO, the component API should be built as an xPDO model anyway, regardless if it’s for 0.9.6.3. The user/permission system will be significantly different, but we can abstract that in the forum model so the Evo/Revo specific functionality can be plugged in.

Next, I am also already working on a data model for a threaded forum application that could be used as the foundation for interfaces developed for both 0.9.6.x/Evo and Revo. This is the first step, and requires that we plan as many of the features as possible before we establish the baseline model.

I’d definitely like this to be a reference "core extension" and it will get the appropriate projects/permissions setup in our contribution systems (Jira, Subversion, Confluence, and Crucible) this weekend. Once we have the model/API is established, then we can work out plans for interfaces/integration points with Evo and/or Revo. But I want to point out, if this is going to be a reference core component, then all code contributors will need to sign a CLA and request access to this sub-project formally.

Following what OpenGeek has said, i’m totally in if a team rise up to make a native MODx Forum.

We can choose the desired features by refering us on this list: http://www.phpbb.com/about/features/.

Rossco ask a good question about the MODx Evolution cache which will be probably used for the forum:

if 5,000 topics were created, would that slow the modx system down?

I’ve some ideas for the forum and already did some research for the backend and a basic port of the PunBB default tableless theme with CChunkie (Evolution) which can be a good basis for a semantic template using chunk and phx for moderation tools.

A forum built on custom tables won’t be affected by the guidelines about the MODx document limit at all. From MODx’s persepctive it’d be just a few tables for the different views: category, board home, topic, admin, etc.

This is all sounding good. I would be keen to see this take on the MODx philosophies allowing flexibility and simplicity over prescriptive page output. Including forum threads into a page/template with calls like those for Ditto would be superb, only traversing threads/posts instead of pages. That way, feeds could be used on the homepage, overviews, RSS and the main forum thread view all from a single set of code.

Something I was looking to do, and would suggest here, is to approach categories like tags. Allow the administrator to maintain a category tree, and then discussions are attached to any number of the categories there. By doing this, instead of a monolithic list of categories being displayed a trawled through, top-levels could be shown on the forum home page, or users could subscribe to a finer view (e.g. (chemical engineering and/or management), instead of engineering and management being accessed separately and then searched manually). Searching would become intuitive and more relevant to the person viewing the content.

Language support would then also be simpler, with language being a category that is not exclusive. From the MODx forums, you could look for "Add-ons" and "Russian", instead of forcing all Russian people to manually trawl/try and search the whole Russian category for what they are looking for. Tags never take off because people are too lazy to type them, which is why I suggest a sort of hybrid "check those that apply" list instead.

I don’t know where the best place for these sorts of suggestions would be. Certainly, I find displaying and categorising threads on my forum to to be a difficult balance of easily accessible and well organised. A hundred categories is confusing, and too few categories can just mean too many grouped, unrelated threads. How do other people manage this sort of drilling for users? I like the idea that my users can quickly and obviously find where they want to be. Am I missing something?

I kind of understand what you mean coffee bean.

It would be indeed interesting to set a general Forum which can be also used it in Documents as a Comment Snippet like [[Forum? &mode=`comments`]] which will create simultaneously a thread in a category predefined in the Forum settings (working almost like a bridge to popular forum used as comment system, but better integrated).

Also, it would be neat to allow the user in the profile to set the default showed category (Russian, French, etc...) when he click on the Forum link.

I also think that jquery Markitup would be a good alternative solution than tiny MCE for posting and edit posts.

I was also wondering about WebloginPE features on Revolution. Will MODx Revolution allow by default to extend users profile?