Library clips

While all the talk is about web-based RSS Readers, especially Google Reader, Jack Vinson has returned to using the SharpReader desktop client to keep up with the lastest, why, because it bundles conversations with a threading feature…Jack has spoken highly of this feature before (which I’ve picked up on two occasions).

This threading feature is what I want out of Google Reader, I call it clustering or a memetracker feature.
Actually I’ve mentioned SharpReader’s unique threading feature in a former post (scroll to the SharpReader heading):

“From an RSS reader point of view you can view a local distributed conversation from your collection of RSS feeds.
SharpReader identifies common links from feeds in your RSS reader, organises a kind of thread…”

And from their site:

“Advanced threading support allowing you to view connected items together in a threaded fashion. SharpReader detects and shows connections between items if they have same link, if one item links to another, if two items both link to the same external webpage, or if an item has comments (for feeds supporting the standard). “

So it seems Jack and myself want RSS reading made easier when we want to follow conversations (conversations that are usually tied by links).

If you read an interesting article in your RSS Reader, you no longer have to think who else in my subscriptions has linked to this article and possibly linked to other articles in my RSS Reader…with SharpReader the conversation is threaded for you, saving loads of time.

Currently in other RSS Readers you would have to click through your 200 subscriptions to see who else is talking about that article…now that Google Reader has search it’s a bit easier to get a digest, but you can’t read it all within the same window.

Note that, SharpReader only threads items together if one links to the other, and also it will thread a page to another if they both have common outlinks.

eg.
article B links to article A therefore B is threaded under A
article B also links to article Z, and article C links to article Z (but article C and B don’t link to each other) yet article C is still threaded under article B as they both link to a common page (this common page is article Z).

I can see why they call this threading, as it’s a explicit distributed conversation…articles threaded just because they both link to a common outlink is not an explicit distributed conversation, but is surely part of a clustering or memetracking feature, it’s like grouping many conversations about the same thing in the same spot.

I guess Google Reader search achieves this by keywords, if article A, B, C, and Z all mention the same term you are searching for, then you will see all these posts in the one search result (digest).
Article X and Y may have also mentioned this term, but these 2 articles don’t link to A, B, C, and Z, and don’t have any common outlinks, in this case you have more posts about the same thing.

The keyword digest has it’s benefits, but just say the conversation doesn’t have an explicit keyword in common, you are not going to get the search result digest you require…but as long as they link to each other SharpReader will thread ‘em for you.

Conversation Digests

1. based on a common keyword
2. based on links
3. based on common outlinks

a 4th point would be offering you similar items from outside your RSS Reader…so basically what we want is megite inside our RSS Readers.

A further point is posts that attract heavily threaded items may rank on the top of the pile of your latest unread items, as it’s considered popular, this way you are reading hot conversations first, then you move on to reading posts that haven’t generated much conversation yet.

But it’s all a matter of timing, just say you open your RSS Reader and article B, C, and Z haven’t been published yet, all you have read is article A.
Then the next day you open your RSS Reader and you click article B, and you see article C, Z and A have been threaded. This is important that, even though A has been marked as read, it’s still threaded.

Another thought

I found an article in my RSS Reader that links to another article in my RSs Reader, I tried to do a search in Google Reader by using the “link:” command, but it didn’t work, this didn’t work in my Reading List Google CSE either.

Even if it did, I will only get links to article A, I will not get posts that link to posts that link to article A (did that make sense)…this is a very narrow slice of the conversation, just like Technorati link search.

Whereas, BlogPulse Conversation Tracker is exactly what we are after, only if we could limit this to our OPML…this is most of what SharpReader does, only SharpReader is also nice and tidy within your RSS Reader, you have the conversation ready for the reading.

Just the same you can use BlogRovr to see who links to a post from your OPML, but then this isn’t a complete conversational digest like threading.

Publish digests

I’d like to publish a conversation digest, why can’t a threaded post, like one of Jack’s screenshots have a unique URL, or even a widget, I mentioned this here.

If Jack were to delete a feed then one of the threads would disappear from the widget, if he added a feed it may have a post that links to a post in the widget and be threaded…so these are live meme digest widgets.