UserPreferences

WikiPageAsEntry


See GeneralizingEntry for context.

[JamesSnell] A careful and useful distinction should be made between Weblog entries and other things like Wiki pages. Weblog entries are a manifestation a specific authoring and publication style that is distinct from other forms of publication. The abstract model being defined here is well suited to Weblog entries, news articles, and other periodic information publication and syndication sources, but is not suitable to describe Wiki pages. While this should be obvious, it does warrant some consideration.

Maybe not obvious.

A wiki page can fit in the Entry model. How is this page you're looking at an Entry?

This fits uneasily because it allows multiple authors. We can reconcile that by observing that wikis are a public style of editing, not really that different from what a single author does privately - writes and edits until it's ready to post, then adds updates and allows comments for a while until the post goes static. The wiki just exposes the process of writing. The resulting page certainly fits in the broad model of Entry. An unknown number of anonymous authors can certainly create a valuable entry.

The wiki "entry" also fits uneasily because it suggests a lot more updates than a weblog entry. But both weblog entries and wiki pages eventually stabilize and hopefully provide a permanent resource to link to. The weblog "convention" is that an entry is written and published into the fray of short term interaction. But what about the reader who arrives a year later? That reader is less interested in author and publication date. She wants the useful content. Notice that some of the comments to Sam's initial post are simply pointers to useful wiki pages. Those pages are then entries by reference, just like a reference to yesterday's entry.

This leads to another conceptual clash: webloggers are focussed on the now, which leads to the idea that notification of an entry is based on a regularly polled feed. But the more important notification is long term - notification when the entry becomes useful, be it by link, by search, or by some method not yet invented.

[AdinaLevin] This raises an interesting question about the behavior of a wikiblog. If there is content that is edited collaboratively in wiki mode, and then published as a weblog entry on a given date, should the attributes of the weblog post be the subset of attributes that pertain to the weblog (the date posted, the author who posted it); thus discarding the data about the document's draft prehistory? Or should the date be "date posted" but all author information preserved.

Yes, that makes the point well. Do we stretch (within reason) the definition of entry so that the wiki page is posted AS an entry, with multiple authors? Or, is the the wiki page just another web page that is mentioned by an entry? And what if the wiki page refers to entries, perhaps summarizing them. Then the thread goes entry-wiki-entry-wiki instead of entry-entry-entry-entry.

[TimothyAppnel] There seems to be consensus that a system can be an author and that there is only one primary author with optional mutltiple contributors. In the above scenario the wiki system should be the author and the users that contributed to a page/entry should be included as contributors.

Nice. The wiki page "authors itself".

(Please take this opinion in good humor) "Weblog entries are a manifestation [of] a specific ... style that is distinct from other forms of publication." How unique really? "Authentic" voices have been around since there have been voices. Weblogging resembles oration, debate, and opinion through the ages. And for all the bloggarts' strutting and fretting about the now and the voice, the fact is that we are ALSO infinitesimal anonymous contributors - as economy and electorate, as coders, webloggers, and wikins. It is very good that people continue to write and read. It is very good that ego continues to drive us. The ultimate good, though, is in the accumulated knowledge, not the manifestation of a style.

Prior Art: [WWW]RDF Site Summary 1.0 Module: Wiki

[CraigEwert] "There seems to be concensus that ... there is only one primary author...". I don't see this seeming consensus at all. On the NumberOfAuthorsDiscussion page there is a clear majority in favor of single author, but majority isn't the same as concensus. On this page JamesSnell asserts that the Echo format is intended to slant towards Weblogs as we know them. But, who knows whether the single author convention is a true characteristic of the form, or an artifact of the single author capabilities of current systems? I'd hate so see single author enshrined in the spec for no better reason that an accident of history, particularly in the absense of compelling reasons to require it. And I haven't seen any reasons except "the current tools do it/expect it" and "That's what most weblogs/news articles have". Neither of those compel me yet.

In addition, the idea of a Wiki page being its own auther strikes me as backwards. The point of an author field is to provide additional information. If you are just going to repeat the link field, don't bother.

[JanneJalkanen] There are a few ways to view the relationship between WikiPages and WebLog entries:

1. RecentChanges as a weblog. This is characterized by wikis offering the RecentChanges list as an RssFeed, for example. 2. WikiPageAsEntry - it is an useful abstraction to treat each wiki page as an entry, and then aggregate them to a single page, which serves as a front page. This works well if you have your own personal wiki which doubles as a weblog. Also, it allows you to do stuff like have multiple people holding their own weblogs inside a single wiki, and include other people's entries into their own weblogs directly, instead of referring to them. The notion of categories emerges sort of automatically from this as well: each category is its own wiki page, which is used as an aggregator to collect other wiki pages; either from multiple weblogs or just wiki pages. 3. Page diffs as entries. Each diff (addition, deletion, change) can be considered a new blog entry. Unlike in regular blogs, where you just add new entries, a wiki also allows "this text deleted" entries.

I don't really know which one of these is "the right way" to think of blog-wiki interaction. I am quite successfully using #2 in my own weblog, and many Wikis abstract themselves as RSSFeeds using #1. I don't know of any wikis that do #3, but I'd wager that's only because nobody has yet found any use for it. It might be useful to keep tab of comments, for example.

Wikis often serve different purposes, and a company intranet (or a company public web page!) should probably be treated differently from a personal wiki, a project wiki, or a public wiki.

I would like to invoke the notion of ChunkizedConversation, as suggested by BenHammersley: RSS is a medium for "chunkized conversation", much like email or USENET. We have clearly defined "entries", or "posts", that follow each other in a linear or threaded fashion. The question is, what is the proper way to chunkize wikis, should we care, and is it possible there are multiple ways of doing it?

I would be completely happy if I could choose whether my necho feed from a wiki is in format #1, #2 or #3, as defined above, and I don't think the Atom standard should define what is the right mode of chunkization for wikis. Phew :-)


As things stand, does the spec handle the case where someone sends a comment to a wiki page? Some wikis have this feature, some storing comments separately from the page contents, some simply appending them at the end. --JohnAbbe


CategoryModel