This entry is intended to provide a more detailed and structured explanation of the community development idea posted at Your job just got more difficult 1: Directory on the Feature Suggestions page. This is intended to both document and explain that suggestion in detail with the aim of creating a proposal which will outline all the necessary bits and pieces visible from a non-code perspective, in order to allow potential development of such a feature.

The idea that this could be possible is cited in a page on The Hub, entitled Categorisation of Content.

'Elevator Story'

A community edited dirctory system for h2g2 would provide better searchability than the existing Search page and would provide a way of gaining status for non-Edited Entries to include fiction and opinion pieces. It would provide another aspect to volunteering for h2g2 and might be fun, too.

The Longer Version

At present, there exists on h2g2 a system whereby Edited Entries and Help Pages are categorised within a system which is rarely talked about, and through experience I have noticed is not often used. Given the nature of the Search system, which works in a complicated way and is by definition automated, a human-edited directory of all substantial pages in the Guide would be useful for the development of the Guide as it becomes inevitably more difficult to see the 'holes' in the Guide.

Advantages

• A logically ordered and categorised directory would be quite easy to navigate and would allow people to home in on the page that they are looking for by choosing the most appropriate area. It would be more intuitive than the Search system.
• Many people are frustrated by the fact that their opinion or fictional piece is not recognised by the Edited Guide, although they are clearly important to the Community. A directory listing would be an 'official way' of recognising worthwhile pages whether they are naturally suited to the Edited Guide or not.
• By categorising the Guide, it would be easier to find the holes that need plugging, enabling authors to more precisely target gaps in the Guide. Also, seeing categories that need filling out might give people the inspiration to get writing.
• A directory system would give the opportunity to easily locate closely related material and to link to it as an Author, or to do more reading up on a subject - for example you could easily zoom in on an area of interest within a topic, or zoom out and read the basic information on a topic before reading the detail. This would ease the burden on the Author as far as explaining to confused Researchers goes.
• The many Researcher-based groups and communities on h2g2 would benefit from having their page listed and a subcategory available where members could be listed - linked to their Personal Spaces. This would not only provide a neat, easily maintainable membership system, but the group owner could be responsible for the subcategory, allowing people to 'submit' themselves to that subcategory in order to join the group, which would prevent the inevitable problem of having to root through reams of chatter looking for new applications to the 'Join' thread.
• Entries aren't the only things that could benefit - the h2g2 Picture Library is already in the directory - and Community Art could also be added, making the most of the current project to find it all!
• Scouts could pick listed Entries from the directory and recommend them if suitable, and the process of bumping them up from their existing status in the directory would be quick and easy, possibly even automated to reduce the workload on the Italics.
• h2g2 already has an index, but with over 4500 Edited Entries in it, an alphabetical list is becoming insufficient and slow to navigate.
• New Researchers could be encouraged to be enthusiastic to get their stuff listed at first, which might detract their attentions from Peer Review until they've read some Edited Guide.

Notable Features

Hierarchy

The directory would be rather like a good web design menu - to get to a page about C₆₀, you might go to:

Listings

Once you'd got this far, you'd discover a list of Entries. The listing you were looking for would probably look something like this, and would exist under Edited Entries:

A733583 by U175309
Edited by U284
Discovery, uses and properties of C₆₀. Allotrope discovered circa 1985 by Harry Kroto and Rick Smalley by accident.
h2g2 Entry: A733583, 01 May 2002.

Note that the details are not primarily intended to give a précis of the Entry, but rather to give an impression of what information is given in the Entry. Information about the subject of the Entry comes second to details of the Entry itself. In the current directory, only the title is given. This is OK for Edited Entries only, but the description and details given in the entry described above would allow for differentiation between multiple Entries of the same title. In this entry, the workload is lightened by automatically inserting the Author and Entry titles at the top, using the tried-and-tested 'self-terminating link' method where the link end tag is combined with the start tag.

The description should run to no more than a couple of lines, and should feel like it is composed of coherent sentences despite being in note form. The date from the Entry Details could also be added automatically - the only thing that would have to be typed manually is the description. Collaborative entries, where there are large numbers of authors that would fill several lines, could simply be denoted by 'Collaborative Entry'. This could be tidied by a category editor using a simple check box.

Page Structure

A standard format would have to be devised for each page. There is a suggested layout at A743357. This layout isn't as compact as the dmoz layout, but there's more information in the proposed layout than dmoz provides. They're automatic bells and whistles, so they're not too labour intensive - well, once they've been implemented, anyway.

Not Covered by the Proposed Layout

• Where several entries exist, as they should in most categories, one may stand out as the most useful on the page. If this is the case, this could be recognised by italicisation of the description and a little smiley before the title. It would sit as the first Entry in it's appropriate list (Edited/non-Edited) regardless of alphabetical priority.
• Those with editing permissions could also be able to see the Researcher Name of the person submitting, if it was different to the author.
• For the benefit of editors, the listings as they saw them in categories for which they had permission could include a status icon beside each Entry - usually . could indicate moderated Entries and those hidden pending moderation decisions, and could be used for deleted (where an Entry was particularly good the Editor might ask nicely for reinstatement, so automatic removal from listings might not be a good idea).
• Instead of having a specific 'back-end' for editors, the page could simply be added to for those areas that they have permission to edit. As with the above point, a link such as edit could be added. This is covered in more detail below.

What tools would be needed?

Well, as I commented above, there would have to be systems to perform the following tasks:

• Add pages to the category directly.
• Approve submissions - via an edit page.
• Edit descriptions.
• Delete broken listings with a confirmation splash.
• Move submissions and listings to other categories' submissions list.
• Create, edit and remove category stuctures.
• Amend category descriptions and guidelines

This would allow for total control. Category editors should not automatically be credited with the editing of subcategories as a way of encouraging more people to become category editors - although they would still get privileges. The links to these pages, as hinted at above, could simply be tacked onto the end of listings and in pages where privileges were granted - in much the same way as Scouts see a 'recommend Entry' button, and Subs see 'Return to Editors' where appropriate on Entry pages.

A fair amount could be automated beyond actually locating an Entry. Titles, authors, status¹, URL display, and other things could simply be called up from the database.

Other Bits and Pieces

• Editing Notes

  Editor Notes should be kept for consistency. They would ideally be called up and laid out on the page along with their contributor's names when a category editor was adding, editing, moving or removing a listing. This would guard against 'spam' submissions, and allows for notes to be made to aid other category editors.

• Submissions

  The submitter should provide a description hirself to aid the workload for the Category Editor. No anonymous submissions should be allowed since all postings come from registered Researchers anyway.

• Unreviewed

  Unreviewed categories would be required to allow a submissions queue to build up. These pages should only be visible to those with editing privileges over the category concerned - category editors and Italics.

• Permissions

  Category editors should have full editing permissions in their category and its direct subcategories, regardless of whether they are the named editor of those subcategories.

• Private Forums

  A limited number of private forums allows matters to be discussed within the categorising community - like the Aces or the Scouts already do.

• Moderation

  The category tree would need moderating as with any other page, if volunteers are to be let loose on it.

Examples to Learn From

Currently, only one example of an existing system springs to mind, and that's A712379 at dmoz.org. The aim of the ODP is to categorise the entire of the useful Web. That's a mammoth task, and the team is massive. A massive team would not be so necessary on h2g2 as if we scale down 3.8m listings for 50,000 editors to a team of 50 volunteers, that still makes for 3,800 listings - not far off the Edited total of 4,500 here at h2g2². This isn't as big as it seems - if you divide that 3,800 listings by the length of time the ODP has been running - that's longer than h2g2's 3 years, then it works out at a couple per day. Also, take into account that spam submissions aren't likely to take off on h2g2 due to the nature of the system, that there's a really sticky community around it and that the growth rate at h2g2 is far, far smaller than of the web as a whole, thank goodness, and it looks manageable. Also, the scheme would naturally be scaleable as the task expanded. These figures are by no means a true indicator, or realistic enough to make a decision with, but they do give a really vague idea. Also, there are a lot of inactive signups to dmoz as editors despite a rigorous acceptance procedure, whereas most h2g2 volunteers are already addicted to the site, and are likely to be a bit more proactive.

Conclusion

This scheme could potentially add value to an ever-expanding Guide and would provide an ample solution to some current inadequacies in the Search system. It would possibly make the gaps that Authors need to fill more evident, and would bring together the Edited and standard Guides in order to better find Editing candidates and to allow easire comprehensive reading of a subject area.

The scheme would require only reasonably small extensions to the current system in order to provide extremely worthwhile functionality, and might attract another group of people to h2g2. Essentially, there are considerable advantages for almost all types of Researcher, especially those involved in fiction and opinion pieces that are not suited to the Edited Guide and so currently escape any categorisation.

Revised 6th September 2002