So, okay, I have the glimmer of an idea for a new CMS. But I am, sadly, not proficient in PHP, MySQL or programming in general. I've looked at a number of CMSes, but none of them satisfy all the requirements for the type of [largely] automated system I'd like to build.

So, first off, I guess I'm partly looking to, for lack of a better term, mooch off of someone else's skills until/unless I get proficient enough to do some of the stuff myself (though I get the feeling I'd need to go back to school for a few years to acquire the skill set necessary to do this on my own, and it's just not feasible; besides, I'd like the project to be open-source, so ideally I'd be working with others, anyway, to make some kind of best-of-breed system).

First question then would be where to even start looking for folks who might be interested in taking on such a new hobby/project. If I had any money (I don't) I'd hire a professional who knows their way around Web 2.0 technologies. Sadly, I don't. So, I can't.

Anyone know where programmers with free time hang out or go trolling for new projects to work on? Would anyone on this site potentially be interested in starting / taking on such a daunting project (building a CMS from scratch)? How would I find out? Luck of the draw?

Moving away from the specifics of who/how, my idea flows from the realization that there appear to be several niche sites that all manipulate the same kind of information, just in different ways. In an ideal world, I'd love to see a site that gives the option of manipulating that data in any way the site admin sees fit.

The data being manipulated, in its simplest form, are URLs.

The initial 3 functions are listed in the title:

1) Social News (links are submitted by users or admins and potentially made available to other users to comment, rate, or otherwise stumble across and enjoy).
2) Directory (DMOZ- or Yahoo-style heirarchical directory; folksonomy/tagging non-hierarchical categorization and browsing).
3) Knowledge-base (effectively a repository of links to information, potentially for research purposes).

The idea here is that in all of these, the main 'thing' being operated upon is a URL. The URL is then manipulated in various ways by various sites.

I think it should be possible to design a single system that addresses all methods of operating on the URLs.

The initial idea came from my own need to keep track of copious quantities of bookmarks and the necessity of being able to share them with friends or likewise the ability for friends to share them with me. Many of those links are to news articles, both timely and archived.

But, simply keeping a list of links is insufficient and inefficient. Eventually it totals hundreds or thousands of links. So, they need categorization and a method of browsing, searching or otherwise finding/navigating to them. The two approaches appear to be directories/folders (hierarchical) and tags (non-hierarchical).

See what I'm getting at? It seems like one should be able to design a system that could be used for any / all of the functions above. The functionality exists already in several disparate software packages.

The aim would be largely to integrate everything into a single deployable, customizable package. So, if a web designer wants to run a directory, they simply install the CMS and activate the "Directory" module. If they prefer a cloud-based folksonomy navigation system, they could enable that instead. If they want to run a social news site, they enable that option and whatever social features they want.

Again, in a perfect world, there would be a strong user / permissions system that would allow for administrators to apply strict or lax controls over who can add / suggest / view which content. At least partially to avoid SPAM getting through unchallenged. But it would also be to allow the administrator to keep, for instance, niche topical sites on-topic (approving only astronomy / cosmology articles for a space-news site, but weeding out links to the latest motorcycle news).

Ohh, and it would be nifty to have RSS aggregator integration for the social news site. That is to say, it would be nice if the site could aggregate RSS feeds and funnel them based maybe on certain filters to get automatically added to the approval queue, so there's always an influx of new content ready to be added to the site.

Maybe the admin could set up "official" site RSS feeds, and users could set up their own RSS feeds (perhaps they know of sites the admin doesn't) and could then funnel particular stories into the site suggestion queue.

Admins could then either add the news item to the site's feed, or decline the item permanently (plop it into a black hole list so that future suggestions are denied for that url; tinyurl awareness to check the forwarded link against the black hole list wouldn't be a bad idea either). While approving for the news feeds, maybe the same screen has options for adding the page to the directory feature and for tagging the item as well. A single admin interface for all available link-adding and value-added functions would be nifty and probably help streamline things.

The folksonomy should also have excellent administrative features such as the ability to remove tags (spam, offensive, etc.) from specific items or from the system more generally, ban tags (so the same tag can't be added again later), link tags? (That is make searches on similar spellings [plasma/plasmas, gray/grey] all bring up the same list of articles.), the ability to turn on or off case-sensitivity of tags & browsing / searches.

At some future point, it would be nice to have a local 'caching' feature (with link back to the original site) in case the linked site ever goes down or gets deleted. Perhaps with versioning (to see various previous incarnations of the site if content at a URL changes versus when the link was originally added).

Dead link checking should likely be integrated too.

Does all this sound like a reasonable project? I mean, I'd think it should be. You'd just have some base URL record with certain attributes in one table of a MySQL database and then other linked tables that store information specific to the particular functions (categories for the Directory, Directory / Social News description).

Right now it's mainly in the "gee, that would be neat" honeymoon phase. So far, I haven't found any pre-fabricated solutions. Seems like some CMSes have one feature, but not the other, or two features but they don't interact (case in point TikiWiki; seems to have RSS feeds, and have a Directory, and have tagging plugins, but they seem not to communicate one with the next or really integrate in any meaningful way). Many of the general-purpose CMSes seem to have many different modules available, each of which work in different ways and interact differently with other modules. Or things just aren't specifically designed to interact in the given ways and must be put through contortions to even approximate it.

It would be nice if there was a single solution that was designed from the ground up to work in an integrated, self-consistent manner so there were no ambiguities about how things interplay within the system.

I guess I'll "put it out there" and "hope for the best." Maybe somebody runs with it or knows of someone interested in taking on such a project or can point me in the right direction.

Hmm, that all turned out a lot longer than I anticipated. Guess I've had a few insights since I originally had the idea. Thanks for bearing with me...

Cheers all,
~Michael