Library clips

Technorati Favourites is a public RSS Reader…read a river of news from your favourite blogs, also search.

Now your favourites has an OPML Reading List (something I called for earlier), and you can also tag the blogs in your favourites in order to see a river of news based on a topic tag. Each tag river of news has a spliced feed, along with the general spliced feed…but each tag river of news seems to lack an OPML (another thing I mentioned earlier).

Micro Persuasion asks, why not make user tags of blogs via Technorati Favourites, the same tags applied to Technorati Blog Finder…well it’s user vs. author tags…they kind of have 2 blog directories now, one defined by users, the other defined by authors.

I just posted an update on Feed Rinse, which led me to bring up ZapTXT in comparison, which got me thinking on Reading List folksonomies once again.

ZapTXT doesn’t allow you to import an external OPML, but you can import feeds and tag them.
This is aggregated into a tag cloud, when you click on a tag you will see feeds from a user-defined point of view, at this stage you can select feeds to build an OPML URL.

In the last post I mentioned that user-defined tags in a Reading List may be ineffective or detrimental…not due to users not being good indexers, but the problem of the task of re-indexing (how will users be notified that the content of a Reading List has changed, therefore their tags may be inaccurate, and will they care, or even worse will they even know, as they may have abandoned their accounts).

No solution yet

If you save another person’s Reading List it would be better to save a non-dynamic version, moreso (like feedbite), you would clone their Reading List, add any new feeds you like or delete existings feeds, and apply your own tags.
Now this has become your own Reading List, it lives at a different URL and has its own OPML URL.

But this still means that only one person is tagging a given Reading List, so I guess it still isn’t a folksonomy, it’s just a handy cloning feature.

FeedRinse and ZapTXT are similar services as you can filter a batch of feeds and then export them again (ZapTXT also is an RSS alerting system).

Of the two systems FeedRinse is a more sophisticated filtering system…here are the features:

Enter some feeds or import via an OPML file (lacks entering via an OPML URL)
Filter feeds (boolean, also fielded…term appearing various fields - title, body, author, tag)
Export feeds via an OPML file or as a OPML URL (Reading List)

This OPML URL Reading List is for your whole feeds section…if you went back and deleted your space and started again this will also reflect on people who have subscribed to your OPML URL (Reading List).

Naturally, the next step is to make channels, so you can have heaps of OPML URL’s (Reading Lists), and this is what FeedRinse have offered, great stuff!

Although on my trial the channels section seemed to have an error…but I guess the idea is to have a my channels section containing all your OPML URL’s (Reading Lists). Clicking on an OPML will list all the feeds (filtered or not), an then export the raw OPML URL to wherever you like.

It would be good to create the OPML channel first, then import the feeds in (individually or via an OPML)
…at a later date, to add new feeds to an OPML channel, just open it up, enter the new feeds or external OPML, filter if need be, then save.

Alternatively you could enter a feed or external OPML in the My feeds section and then select a feed to send to an OPML channel, or even batch select feeds to an OPML channel.

Still missing is creating a universal filter for all your feeds, or for a whole channel…I’ve been told these global filtering features are coming soon…ZapTXT is ahead on this one.

NOTE: ZapTXT doesn’t accept importing of external OPML’s at the moment, you can only create your own from existing feeds in the system. If the system doesn’t have the feeds you are after, you can add the feeds in manually, then they will be available to add to an OPML.

[ADDED: He is a screen shot. A comment on this post also asks if FeedRinse does de-duping…FeedRinse doesn’t generate a spliced feed (this may be a good idea to implement in a future release)
…looks like you have to pay for channels]

[ADDED: tag cloud…is this for feeds or channels…are these tags hyperlinked?]

On my last post about feedbite, and in fact on several posts about similar services, I have been labelling them as Reading List folksonomies.

Thanks to a comment by Thomas Vanderwal I have gleaned some more insight about the specifics of a folksonomy…in my excitement I overlooked the real meaning.

This post is extending on a comment in that post, here it is:

“Thomas,

So in essence the community is not tagging an evolving vocabulary.”

Again

Even though people that contribute Reading Lists are users, they are more considered as authors, therefore the tags they apply to their Reading Lists are author-defined tags.
At the same time these authors can also be users, they may browse the same service (eg. Feed Collectors) and decide that they would like to save a Reading List created by someone else, which they can do in the friends section….but you cannot apply user-defined tags when you save other people’s Reading Lists.

This means that a tag cloud would consist of author-defined tags only…but it still is a community cloud of some sort.
I’m not sure of the term to use, but I see it as an author-defined vocabulary, as people who create the Reading Lists get to apply the tags, the service (eg. Feed Collectors) doesn’t get involved in indexing each Reading List with a subject term or tag.
NOTE: tags don’t have to be subject terms

Question

Is there a need for others to tag other people’s Reading Lists?

If I discovered a Reading List I would save it and tag it with a term I like to use for personal reasons
…at the moment the places I could save and tag it are in an RSS Reader (eg. BlogBridge), an outliner (eg. OPML Workstation), etc…

Why not the same in the public arena, I could save someone else’s Reading List in Feed Collectors and tag it with my own tags (for personal reasons, or for public reasons - to emerge a vocabulary).
Then we could discover a Reading List by author and user tags combined…maybe there could be two tagclouds..this could confuse.

The problem is that if someone at a later date edits their Reading List, the tags they applied may also have to change…this is OK…but, if the public have also tagged this Reading List, they will not be aware the contents has changed, and their tags may no longer be appropriate.
Even if they could be notified, they have to react to this notification, if they don’t bother the tags may continue to be misleading.

This is where the potential of a Reading List folksonomy is different than a bookmark folksonomy, there is a greater chance that the aboutness of a bookmark (webpage) won’t really change that much…in contrast, a Reading List is more organic and volatile, it has a greater potential to change as the author’s interests change.

There may be two scenarios:

1. The author may change some feeds, then change the tags slightly to reflect the different feed set, but they are generally under the same topic.

eg. the author may have initially tagged this Reading List: web2.0, wiki, rss, blogs…later on they may have deleted the feeds relating to “wiki”, so they consequently removed the “wiki” tag

…it would then be appropriate for user tags for this Reading List to also delete the “wiki” tag, as it is misleading, but there is no way this can be enforced.

2. The author may change all the feeds and tags, and re-name the Reading List.

The user tags would now be misleading altogether…leading to bad browsing experiences.

In the end are user tags appropriate for webpages that may change their aboutness?

In a centralised vocabulary or author defined only vocabulary as long as the author changes the tag, the system stays clean, it only has to be done once, as the tags for that item have only been applied once.

Folksonomies may not always be the appropriate management or discovery portal for all types of objects…as much as I love them!