In a recent post about the development towards a syndication bus using WordPress Multi-User, Steven Egan brought up an excellent issue that for me remains the Achilles’ Heel for this setup: how can we syndicate comments along with posts?
Thing is when you use a tool like FeedWordPress to syndicate posts, what you cannot syndicate is comments, and while the permalink points back to the original post where the comments should be seen, there is no way to indicate on the syndicated post just how many comments there are, or reflect the recent ones in the sidebar somehow.
Steven says this in his comment on the issue:
It would be interesting to take the republishing thing and include the comments at a single location. Basically the comments happening at the original blog, rather than having it happen on multiple blogs. Even if it is just pointing to the place to join in the conversation.
What I would like is the ability for the comments to be reflected in both spaces, so that the permalink can point to the post on the course aggregation blog, for example, but once someone leave a comment there, that same comment is automatically re-published on the original post. Now this, for me, would make the whole thing perfect.
But at this point I would even settle for a little number on each post in the aggregation blog that would dynamically reflect how many comments are on the original post. Yet, this wouldn’t answer the questions of how to show recent comments in the sidebar of the aggregation blog. I’m using a workaround for this currently on UMW Blogs which basically takes all the comment feeds for the individual student blogs for a course and creates one aggregated feed of them in the sidebar (I do this BDP RSS). But this is not only laborious, but it is also far from perfect because if a student uses their blog for a variety of different classes or reasons, then comments that are not relevant to course will show up.
So, anyway, I was just recognizing a weakness in this model and wondering if there is a better way to deal with the aggregation of comments for a syndication bus like the one we are thinking through.
Image credit: Otbayley’s “Left Leg: Broken Heel Bone and Detached Achilles’ Tendon”
I like the convenience of being able to type a comment right on the post I’m reacting to. Where it ends up is no real concern of mine, and if it were just a blog post on my own site, with a link back to the item I’m commenting on, that would be fine with me, would make RSS work, and might even lead me to write better comments.
This is why I create mIDm a few years ago – an initiative that was later superceded by OpenID. Your website needs to know who I am to know where to send my comment, which means I needsome sort of persistent identity. I designed my ID to be housed in my browser header, a convenience OpenID never did undertake.
I still think this is the approach to address your need.
As someone who uses the same solution you do (for obvious reasons–you taught it to me!), I have the same question. I see that some of the blogs I subscribe to on Bloglines have a numeral indicating the number of comments on each post. How do they do that? Is there a way to hack an RSS reader for republishing/aggregating posts in the middle of a WP blog-portal?
I have a similar problem and had to hack a template to take out the comments field for a project I just did, since the field reflects “no comments” all the time no matter if there are comments or not. We originally wanted comments to appear on the aggregator blog but then there would be two places for comments, which wasn’t optimal.
What I did notice during my experimentation is that blogs from wordpress.com do show comment numbers, just in a different place. I just set a up a test to show you what I mean – I’m syndicating the WordPress.com news blog. You’ll see that the comments show up on the bottom of the post. Feedwordpress Test Comments Blog
What I was wondering is why our version of wpmu doesn’t do this for blogs syndicated from our system (does yours?). If I could get ours to show up at least the was they do for wordpress.com blogs, I was thinking about hacking the template code to have them show up in the appropriate comments section. Even though the permalink would go back to the original source, I figure it’s better than nothing. At least a reader could easily see that a post has comments.
I know this is only half a solution at best (if there’s even a way to do it on wpmu) and doesn’t create the cross-posting of comments you were talking about. But I’m interested in looking into this more as well.
I’ve always wondered why a syndication format that allowed for posts and comments to be delivered via a single feed was never developed. I’ve thought about using “after-market” tools to glue the pieces together, so that I can get both types of content delivered in a single feed to my reader, but it’s a bit of a hassle and I’m not quite sure how that experience would work in the reader. As Gardner suggests, readers (or any tools that are capable of reading/re-publishing feeds) that could consume such a format would be interesting. Let the posts and comments reside in their original location — just develop the tools that know how to consume such a feed.
I find this really interesting. I’ve been working on a post that includes this very set of comments and concepts, though from a different point of view. Hopefully I’ll get it posted tonight.
Something of interest though is the comment feeds themselves. It might be possible to embed a feed reader/display into a web page, as I’ve seen this kind of thing before. By displaying the feed through such a module/widget, you might be able to get the comments to work. How to get all that to work in an automated fashion is beyond my knowledge right now. Blogger does have post based comment feeds as seen here: http://help.blogger.com/bin/answer.py?hl=en&answer=42662 , but all I’ve done with them is enable them.
I think it would be possible to create some very interesting sites and systems using the syndication-based architecture. It might be possible to wire it together to mimic a forum with a tag-based subforum. Add that to the wiki functionality options for some blog platforms and it could be the means to implementing an Open Distributed Blog/Wiki/Forum. That would make me very happy.
Something else to note is that there probably needs to be an “opt-out” or “opt-in” feature involved. At least that’s what my estimations for eventualities predicts.
@Stephen: I don’t think that is quite right for this application, but I definitely prefer that direction over OpenID. It’s more about connecting the posts and discussions than about who’s comment it is, at least in my point of view. That means the posts could be re-posted by others and the discussion could continue through both venues.
@Gardner: I think there is a number of comments variable attached to blog posts most of the time. Using that on another site might be a lot harder than displaying it with the post.
The idea of an identity tool that allows all your work to be tracked and re-posted as necessary is ultimately the key beyond my about the the details of comment syndication. And when I think enough about them, the ability to syndicate comments from all the different blogging services out there my head begins to spin. So, a service like OpenID that would allow you to post your comments anywhere and also track and maintain them would make sense, but I wonder how those comments might be visible to others in an aggregated space. So if I commented on your Half an Hour blog, and someone syndicated that post into an edtech aggregator, how might they see how many people have commented, even before they follow the permalink back to the orginal post? I understand it isn’t necessarily crucial, but when thinking about aggregating a group of posts together from a series of individual blogs, some way to visualize the comments on those posts and that space may prove important. I also think an identity management tool that allows you to still control and track your comments, as well as other writings and activity online, is something that needs to evolve, but from what little I know of OpenID it does not seem to be going in that direction–or if it is, it seems quite limited in its promise. It does allow distributed login from a an ID namepsace you control, but not much beyons that I can see–am I wrong?
Yeah, we have been dealing with this one for a while haven’t we? 🙂 I know that there is a plugin that allows your comments to be viewed in Google Reader, but I can;t find it. Clay Burrell was talking about it a while back and I still can’t find it. What is even more interesting to me is that in the latest version of WP, comments are obviously linked to posts. If you edit a post, you can actually see all the comments associated with a specific post in the discussion field. This is new, and suggest the link between the two has been made, at least in WordPress, I just have to figure out how we could include that feild as part of the syndication, and whether it could ping back and update from the original post regularly. That would be key, and it has to be what the RSS readers do. I want to stay hot on the trail of this one, because I always think about it but then let it go, and it vanishes.
I think what WordPress.com is the very plugin I suggested in my response to Gardner. You’re right, that would be key, and it is doable. They have something happening there it is probably a quick fix. I pulled in bgblogging.com in a wordpress syndication blog and it did show the comments correctly. Do you know if they update Dynamically? Must follow up here because this would be a nice temporary fix to we can figure out who to do a more sophisticated link of comments to an OpenID or even syndicate comments back and forth between blog. BTW I have been doing the same hacks to the themes as you 😉
WordPress now links to the two as I mentioned above, and I think there has got be some way to get in there an pull that field into the RSS feed. Not sure, but there has to be a way. The only issue here is that is great for WP blogs, but what about blogger, typepad, etc. It seems like a much bigger issue on the aggregate.
You;re the inspiration for this one, and as usual you don;t disappoint with an awesome comment. I look forward to the post. The issue with feed for post comments is that it’s separate, and hence not pulling the two together. It would require two syndications, rather than one dynamic one that is pinged about comments. So, what we need is a call to the comment feed for a specific post within the syndicated posts so that it can dynamically updated comments from the original post—does this make sense?
There are two separate but related issues here. One is that comments are things I author and publish, so they’re things I’d like to track, aggregate, store, etc. The alternative is not to comment on a post but to comment in another blog post and link to the original post–not a very satisfactory solution.
The other is that for comments to work well in a teaching and learning situation, they have to be aggregated for maximum recursive effect as a course dynamic visualization. If every teacher used BDP-RSS to aggregate all the course’s comments (a pain), then could the motherblog could aggregate all those feeds in a way that tied them to the specific courses? Or is there a way to let tagging do this work? I’m grasping at straws here. The fact is that I care very much what happens to my comments, and I care even more about student comments in course blogs that have the potential for more dynamic catalysis of course activity–if they’re usefully visible.
Makes perfect sense on the drawing board, but I don’t see a nice simple way to accomplish it. So, Instead I looked for a promising direction for the present. I find that the answer usually comes if one continues creating working solutions.
I don’t see an easy way to implement the information into a single feed. A sitewide comment feed is available in both WordPress and Blogger from what I’ve seen that might be combined into a post and comment feed. That feed would then have to be checked for schema differences and such to create a feed reader that can handle the different behaviors. Theoretically that would give you the feed described, I think.
They are updating when the page refreshes. I just pulled in posts and comments from a test blog I have on wordpress dot com. Here’s the link again to the test:
The three comments I left on today’s test post updated on the aggregate blog when I reloaded the page, so that’s good.
I would think there is also a way to pull the custom field that is controlling this display into the appropriate section of the post without too much trouble (let me eat my words now).
We’re still using 2.6.3 and these comments don’t populate where they should. Do they at UMWblogs?
Also, I have tested this with Blogger and couldn’t get them to show up at all, which is a shame because we have a lot of student bloggers on that platform. So our latest project http://blogs.cas.suffolk.edu/undergradblogs is just serving as an aggregator and organizer since nothing too interactive can happen there. We really wanted to pull in comments and allow commenting, but didn’t want the students to respond to comments from both places.
Could a zombie help?
Gardner, How sick are you pulling out the zombie tunes, of course this helps, it gives me a sense of urgency I have never had before 🙂
Michelle, Maybe wordpress.com are using FeedBurner, which allows you to show the comment count at the bottom of the feed? See my example, using FeedWordPress and FeedBurner, here: http://animation.blogs.lincoln.ac.uk/2009/02/17/buddypress-a-university%E2%80%99s-social-network/
On a related note, if you use the FeedBurner plugin for WordPress, I recommend you modify it so that FeedBurner doesn’t mess with your tag feeds: http://blog.bradgrier.com/2008/06/23/tweaking-your-feedburner-feedsmith-plugin-to-support-wordpress-25-tag-feeds-easy/
Jim, don’t forget to check the WordPress codex. You can see what feed capabilities are available in WordPress by looking at the functions reference and the template tags: http://codex.wordpress.org/Function_Reference http://codex.wordpress.org/Template_Tags
see also: http://codex.wordpress.org/WordPress_Feeds and http://codex.wordpress.org/Customizing_Feeds
Reading your posts and the comments, I wonder whether we should all think about what RSS as a standard can and can’t do. There is an attribute that can display a link to comments about the syndicated item: http://cyber.law.harvard.edu/rss/weblogComments.html and http://cyber.law.harvard.edu/rss/rss.html#ltcommentsgtSubelementOfLtitemgt
I don’t have time right now to pursue this unfortunately, but will give it more thought over the next few days.
(Personally, I don’t care what happens to my comments after I’ve posted them, nor do I want to accumulate a comment profile over time. What’s the next step, recording every conversation we have with friends? 🙂
“What’s the next step, recording every conversation we have with friends?”
I’ve considered this. Too many great ideas vanish into the air. I also urge my students to take their “random thoughts” seriously.
Chit-chat, no. Conversations, yes.
That’s one thing I really like about Google Chat and Gmail, it keeps a log of the chats right there with your e-mail and associates e-mails and their responses into thread like conversations.
It’s also why I like the idea of a feed reader that has a Forum/Wiki organization. If it uses local files, it gives you a stand-alone copy of the conversations. As Gardner said, “Too many great ideas vanish into the air.” Something to add to that is too many great links/connections are lost and forgotten.
Along the same lines, it might be a good place to put in things like wiki links and aliases. … That could make for a really cool addition to Firefox. Maybe I should bring it up in the discussions as part of the project design in the Mozilla Open Education Course.
@Joss “nor do I want to accumulate a comment profile over time.”
I certainly would want to cast a lot of my comments aside as temporal chaff, but I was thinking a lot about how a packaged-together finished course might have a lot of valuable content for other teachers trying to offer the course to their own students (as well as to other people.. time-shifted students who want to learn this material a month later maybe). If the course site could aggregate the course’s curriculum, student work and selected comment threads (with teacher&student content), others could “take” or re-offer the course in their own contexts, provided open licensing exists for the content.
You might want to have a comment profile for a specific purpose, like for the #mozopened course that should be packaged with your contributions to that course… and you should be able to choose what comments get published on that list….and I don’t think this capability is present with existing tools. Blogging software lets post authors ping urls, but not comment authors..
Brainstorming a workaround: Maybe what we need is something like a firefox plugin that allows a user to ping their own comments to an aggregator on their own blog (with tags) so that you can build up a “curriculum vitae” of the posts you think are relevant to a particular class. Maybe if the trackback form can be written, a bookmarklet could work? But is this the right angle to attack this from if what we’re trying to grab is the group discussion around course materials, not individual comments?
Nevertheless, I’m mainly playing with the idea of packaging a course together so it can be viewed by new audiences and extended by new teachers… I believe in the individual blogs aggregating into a central resource format that emerged in the #mozopened week 2 seminar (And from Wiley’s Openness/Disaggregated Future of Higher Ed week1 reading http://www.slideshare.net/opencontent/openness-and-the-disaggregated-future-of-higher-education-presentation )
(I’m not an RSS code pro but just thinking generally about how syndication works…)
For the main issue of this thread, how to syndicate comment threads with posts that may be aggregated into group formats, each individual posting should have its own comment thread using one of those script images.. Maybe the post should contain a link to a comments feed as well as a permalink to the post location with comment thread?)
I’ve just started to write up what I know about WordPress feeds here:
Hope it is of use to someone.
I’ve just noticed that the latest version of BuddyPress has a setting:
‘Blog Tracking: Tracks blogs, blog posts and blogs comments for a user across a WPMU installation.’
Which seems like it’s the start of something that we’re looking for. I wonder whether that code could be turned into something more generic for tracking comments across WPMU?
Casting an eye over the bp-blogs.php file, I can see that it’s only recording comments by registered users but that seems like a deliberate decision in the code that could be changed.
Time to speak to Andy Peatling, the BP developer, maybe?
I was talking to Eddie Tejeda, the developer of CommentPress about comment author feeds – something that WordPress doesn’t currently support. You might like to know that he’s just started work on a plugin (not dependent on CommentPress): http://code.google.com/p/commentpress/source/detail?r=100
That’s just a placeholder for the code but he reckons it’ll be in place pretty soon. Feeds will be based on the comment author’s email address.
If you’ve got time, we’d also really appreciate some testing of the new CommentPress. You’ll find a lot of changes from the earlier version.
It should be packaged as a zip soon, but all you have to do is type:
svn checkout http://commentpress.googlecode.com/svn/trunk/ commentpress-read-only
into your terminal and then copy the folders into /plugins/ and /themes/
there’s a mailing list here for feedback:
thanks and please spread the word!
and if you would rather just feedback by using my test site, it’s here: http://writetoreply.org/dev2