Library clips

Glenn Reid has a great post from earlier this year on threaded blog conversations…for a way to augment the distributed conversation concept he is suggesting threading blog posts from different blogs as a linear conversation by pinging a threadID to an RSS engine such as Technorati.

This seems like it’s worth a trial, I mean distributed conversations are always going to be there as back up by trackbacking the post you are referencing or viewing the incoming links at Technorati or BlogPulse Conversation Tracker.

Tagback

Another way of doing this is using Technorati Tags and just giving the tag a unique name, this was given the term, Tagback.
When you make a post leave your unique Technorati Tag at the end of your post, leaving a message that if you reference this post please use this Technorati Tag…this way the distributed conversation is stored in the one place…and because the tag is unique you are reducing the noise, if any.

But when searching in Technorati Tags how do you browse and discover Tagbacks or conversation tags, these tags lack a prefix (or namespace) that would define them as conversation tags amongst the normal tags…so it is accomplishing the first step of collating all posts on a conversation, but what about findability and browsing for conversations.

Suggestion

Instead of pinging Technorati with a ThreadID system or a unique Tagback you could set up a “blog for threads” that you can ping, or why not just set up a service similar to Topic Exchange, but it would be Conversation Exchange.
You could have your own account space, like in del.icio.us, with an item list of all the conversations you have initiated, and also a list of conversations you have contributed in…maybe you could add tags to these linear conversations (or distributed gatherings) so you could organise, share and browse the portal by a conversation topic (a conversonomy even…he!..he!)

…this idea is simply leveraging on the Topic Exchange system by allowing a user space, and folksonomy features.

The way this would happen is similar to Tagback but more like Topic Exchange (mentioned above), at the bottom of your initial post (the conversation starter) you leave a “conversation link” that points to a permalink from your space at the Conversation Exchange. Then someone responding would ping this link in their post and also leave the same “conversation link” at the end of their post, and so on…(you could also trackback the blog as per usual anyway as a back up).
This way every post in the distibuted conversation is pinging a site (the Conversation Exchange) where the discussion can be stored as a discussion thread, and can be tagged for subject browsing, or searched by full-text.

So there is no bookmarking taking place, all posts including the initial post are populating the spot in the Conversation Exchange by pinging.

See an example of how this looks at the Topic Exchange, here is the blogwalk channel
…the difference with the Conversation Exchange would be that they are more specific one off discussions, not sure how you would close the conversation (stop people pinging), maybe there is no need for this actually as conversations re-emerge or people come across the conversation a while later and may want to add to it.

NOTE: the discussion starter needs to be an indepth post…not a one liner…it needs plenty of ideas and content to generate discussion.

Why do this?

Well if you want to view a conversation, go to the Conversation Exchange and browse by subject, just like a forum or a folksonomy…so it’s a place to aggregate all the conversations that are happening

…at the moment when a conversation has taken place how would you firstly know about it (discovery), and how would you find it, there is no central place where they live like a forum or a folksonomy, or even a wiki, as they are distributed in the blogosphere.

How is it different to the Topic Exchange

It would be a different sharing and browsing experience as it would be a folksonomy, where all users would have a space, in contrast to browsing the system categories or via the ODP based directory at the Topic Exchange.

If you browse the channels/topics at Topic Exchange you will see a topic called “ipod”, now this type of topic would be too broad for the Conversation Exchange…the idea of the Conversation Exchange is to capture specific discussions, an example of this is the current screen problems with the new ipod nano.
Capturing these specific conversations wouldn’t be effective in the current Topic Exchange as there would be hundreds and hundreds of similar topics about ipod nano’s screen problems, instead you would need a tagging system to organise this instead of the topic directory.

Another example of a specific conversation was the discussion that was taking place a while back on Full-text vs. Excerpt RSS feeds…now this is too specific for the Topic Exchange, it would be more suitable for the Conversation Exchange.
By saying this I mean it could belong within the “RSS” topic at the Topic Exchange, but then how would you find the specific dialogue amongst all the other dialogue taking place at the “RSS” topic.
Conversation Exchange would plan to go a step further, to a more micro level, in preserving/indexing (tagging) little packets of specific conversation within a topic.

The other difference is that the topics in Topic Exchange are neutral, whereas in Conversation Exchange someone owns the topic (conversation) as a permalink in their user-space, but it can be shared and tagged as it would be in a folksonomy.

RSS

Then you could implement RSS features

…subscribe to the RSS of a tag to see the latest (initial posts) conversations about to take place that have been tagged with a specific tag by the conversation starters..then others could tag the conversation anyway they see fit.
This would be great because if you are interested in discovering conversations about “Blog Ranking”, you could view this tag and see the initial posts of all the parallel discussions taking place in the blogosphere about “Blog Ranking”.

So firstly you would subscribe to the RSS of a tag (find new conversation starters about that tag/topic), then once you have found a conversation you like, you would subscribe to the RSS of it…then wait to be notified of new content as the discussion unfolds, it’s all done for you, you would no longer have to manually follow a conversation by using the RSS engines or by having the right feeds in your RSS reader or by surfing from post to post.

Issues

This system is reliant on the bloggers doing all the work, they have to make sure they don’t forget to ping the permalink at the Conversation Exchange, just like Trackbacks (also the spam problem)…these are the current problems at the Topic Exchange.
Although someone has tried to automate pinging the Topic Exchange…see here.

Another problem is that when the conversation is ending you could do an incoming link search of the initial post and check what’s missing and maybe manually insert those posts into the conversation, and also exclude non-relevant references …this is also reliant on the expertise of the initial blog poster…but how would you do this if there is no bookmarking taking place as it works by pinging (is there a function for post-editing, as it’s basically an automated system).

Also, although a post is linking to your post, it doesn’t mean it is talking about the same thing as your post, this is something the Conversation Exchange can not exclude.

Also another problem is what defines a discussion

…in the (fictional) Conversation Exchange it would be an arbitrary choice, the person who is publishing the initial post and pinging the Conversation Exchange, is saying “I’m starting a discussion”…just like with Tagback.

In comparison, with the current blogosphere, the start/end of a discussion is in the interest of the beholder
…if they were to enter a URL in the Blogpulse Conversation Tracker, the conversation tree shows the discussion (although it doesn’t show the back discussion of the URL you have entered), but how do you know if the conversation started at the URL you entered, and how many conversations are splintering into other focused conversations…you really have to analyse all posts to make a human based decision on the real conversations, all the RSS engine is really doing is showing reference links, not explicit conversations.

Once you have done this you could compile a linear list of URL’s representing a conversation and publish it for others to see…so the RSS engine is just doing the ground work (a good job at that), but it needs a human touch to really impose the start and tail, and also excude irrelevant posts.

But this is manually compiling a list, the idea is for this to happen automatically.

SharpReader

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…what if this were to include to feeds outside your RSS reader.

From the website:

“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). “

…here’s a screen shot.

…also on RSS Bandit.

BlogPulse Conversation Tracker

You could say that the BlogPulse Conversation Tracker already does what this post is suggesting, automatically, without the conversationalists having to use trackbacks or tags (there is no effort by the bloggers which is a great thing), but the limitations are:

- Time limit

Will it keep old conversations

- Locating start of a conversation

If you enter a URL of Blog C, you will see all referenced posts from Blog D,E,F, etc.., but it won’t tell you that Blog C is referencing an earlier post in the conversation which is Blog B, and that Blog B is referencing Blog A (the initial post, the conversation starter).

So the only way to find the inital conversation post is to view Blog C and follow the links back manually to Blog B, then to Blog A.
Then you would enter the URL of the initial post from Blog A and see the converation thread in it’s entirety.

Example of a conversation explained above:

Blog C Feester 500
Blog B We made a list!
Blog A Link Love Lost or How Social Gestures within Topic Groups are More Interesting Than Link Counts

Now the other problem is how do you define where a conversation starts, as Blog A is also referencing other links, we can’t keeping going back to find a post that doesn’t link to anyone, this does not define the start of a conversation.
So it’s up to the user to browse and decide whether the conversation they are reading is starting at the post they are on, or another post, so forth…wherever their interest lays is as far back as they need to go for their notion of the initial post of a conversation, and also the end post of a conversation.
That is, the related topic or more specific topic of Blog C may be the start of a conversation for someone else, their idea of the conversation from Blog A may not be of interest…so we have posts that reference each other and spike off into different conversations that all make up a loosely related conversation by a common link.

…and (as mentioned before) linking alone can’t declare that a post is part of a conversation, it could link to a post but be talking about something totally different…something that can’t be helped, unless you examine and edit the conversation and re-publish it (but then the original idea is for automated conversations - both the process [via pinging] and the presentation [all the pings collected on one page forming a linear list]).

As mentioned before, for the Conversation Exchange, it’s up to the initial blog poster (conversation starter) to explicitly decide they are going to start a topical conversation…even if they are referencing earlier links that motivated their post, they are still able to declare that a conversation is starting with this post (and tag it with a subject term, which others can re-tag with their own tags)
…no-one at all can change the contents of the permalink at the Conversation Exchange as it’s populated by pinging (maybe the owner [initial blog poster] can have editing rights as the permalink is in their account space).

- Discovery

At the moment in BlogPulse Conversation Tracker you can’t browse conversations, you can only see conversations by manually typing in a URL…there is no discovery (well you can discover general conversational topics at Topic Exchange, but not in the sharing and browsing environment of a folksonomy)
…and these aren’t exactly conversations, they are topics that contain many conversations.

We need more full fledged discovery, where you can search/browse for conversations by subject, maybe BlogPulse Conversation Tracker could incorporate tagging of conversations (share and discover blog conversations by user-defined tags)

…this seems an easier and more probable solution, as BlogPulse is widely adopted and there is no effort by bloggers (they can blog as per usual - no need for pinging, except for BlogPulse)

But then is there a need for this, you could find a post you like, enter it in BlogPulse Conversation Tracker, and then bookmark that URL in a social bookmark service…although, a specialised folksonomy for only BlogPulse Conversation Tracker URL’s would be ideal (Is it possible to restrict URL’s to be included in a folksonomy to only one domain?)
Similar to del.icio.us, if you browsed the folksonomy and click on the title of a bookmark it will go to the BlogPulse Conversation Tracker URL page for that link where you can view the conversation tree.

The problem with BlogPulse Conversation Tracker is that all you are really doing is bookmarking a URL that marks the start of a conversation, there are no customised parameters, you see the whole conversation tree from that URL onwards.
That is, you can’t pick from the conversation tree to see which links belong in the conversation you are tagging, you can’t even pick a link for the end of the conversation, all you can do is bookmark what you choose as the seed of the conversation, and it defaults to show all links to the URL.

The only possible way is to enter a URL into BlogPulse Conversation Tracker, and be able to select or tick all the posts you think are part of the conversation you are tagging.
But then where would your list live, you can’t just point to that page in BlogPulse Conversation Tracker anymore, as you have to now point to a permalink, you’re customised version of that BlogPulse Conversation Tracker page.
So in effect, everyone would need permalinks in this BlogPulse Conversation Tracker folksonomy.

This is the difference the Conversation Exchange folksonomy would offer, in that if you click on the title of a bookmark it goes to a permalink in your userspace, showing the linear conversation list (so the permalink in the userspace is the native site)…if you want to see the extended conversation or conversation spikes you could include the BlogPulse Conversation Tracker URL of each entry, and place it next to each entry within the item (linear list).

Conclusion on the Conversation Exchange

Good idea but:

- too hard to rely on bloggers to ping a portal (also spam issue)

- the head and tail of a conversation is different to everyone (ie. conversations within conversation, spiking off, tangents, etc…this is an issue regardless of the system anyway)

…but the purpose of this system is to take the plunge and make a decision of where a conversation lays.

- does the simple act of linking declare a post is part of a discussion (how do you define this)…again this is the case with all RSS engines anyway.

BlogPulse Conversation Tracker does a good enough job, if only it implemented some of the features discussed…or hopefully this post has triggered an idea.

I’m beginning to think that there isn’t an effective or moreso easy way to try and convert a distributed conversation (indirect discussion) into a linear conversation, for the purposes of sharing and discovery (posterity).

Manual Alternative (After the fact, or as it happens)

The only way it seems possible is by manually compiling a linear list of links resembling a conversation as a summary (or review) blog post or wiki entry that you can then bookmark, tag, and share in your favourite social bookmark folksonomy (or even a social bookmark folksonomy just for summaries or reviews of blog conversations).

The current problem (for our purposes according to this post) with most social bookmark managers (eg. del.icio.us) is that each item or bookmark lacks its own permalink (although Furl has permalinks), so clicking on the item title launches to the native site and not to your own permalink within del.icio.us…also each item as mentioned refers to only one link, not a collection of links.

The answers to this are available in 2 new services; firstly Jeteye enables you to make Jetpacks which live as permalinks in your user space…a Jetpack is a webpage consisting of a compiled list of links, the heading of the webpage can be a topic title, clicking this title is the permalink for this page (only Jeteye is not a fully fledged folksonomy).
The real answer is at Wink, part of their service is where you can organise, share, and discover topic lists, tagging them as you wish (again del.icio.us only shares single links whereas Wink enables you to share a lot of links listed on one page that lives in a permalink in your user space).

…off on a tangent

Maybe we need a blog collective doing summary posts of distributed conversations in the blogosphere…then these posts could be re-syndicated in a master blog (what we call a Public RSS Aggregtator)…also allowing users to tag entries.
A choice to browse the portal by the cumulative blog categories (author-defined) tag cloud or the user-defined tag cloud.

Wow, a Public RSS Aggregator folksonomy, how did I get here!

Also see some related posts:

Preserving blog conversations
Blog Clouds