A Commentary on Comments

For those of you who do a lot of weblog surfing, and who frequently participate in discussions at those sites by posting comments, I think there’s a need for a centralized system to manage that content. I’m talking about a method of aggregating those contributions in a single location, ideally on one’s own Web site but perhaps also on a page hosted by a remote application, combined with some pinging intelligence and a facility for management by their original author — you.

Think about it like this; taken altogether, you can look at everything you’ve written on other people’s weblogs as a body of content that you’ve generated for free — it’s only fair that you should be able to maintain a centralized archive of it, and to be able to display the fruits of your labors. Of course, the archive would include abstracts or excerpts from the original weblog post, as well as a URL directly back to it. That way, everybody wins.

Everything You Ever Wrote

And why stop at weblog comments — why not aggregate comments posted to discussion lists and bulletin boards, too? Oftentimes, those contributions are at least as valuable as weblog posts themselves. If you’ve contributed several excellent passages’ worth of commentary to a particularly robust mailing list, they shouldn’t be lost to that list’s archives (no matter how accessible) for all time… they should be made available directly on your own Web site.

This Is a Job for…

There’s probably an idea for a revenue-generating line of business in here somewhere, and I’d do it myself if I could see it clearly. But right now, I only imagine this as an adjunct product for a company like Six Apart — it’s no accident that there are shades of TrackBack and TypeKey in this concept, but hopefully the final execution would be a bit more universal.

Ideally, it would be completely open source, and championed by one of those brilliant, industrious and frequently very young lone wolves that commonly turn the Internet on its head. Given my own relatively advanced age (!), limited time and insufficient technological capacity to overcome the nontrivial challenges of getting all of these various content platforms to talk to one another, I’m clearly not the guy to do it. But if you think you happen to be, and you can pull it off, I get dibs on designing the user interface!

+
  1. Awesome idea. I know that Kottke does something like this, but I’m pretty sure it’s manual labor on his part. However, it seems to me to be a logical extension of TrackBack… maybe TrackForwardЎ? Heh heh. Well, I have been messing around with the PHP in WordPress lately, so I’ll take a look, but no promises: I’m more of a design guy, like you. Keep up the good work.

  2. It seems like copy and paste would do a lot of what you’re talking about. In fact, I wish I had copied more of the posts and remarks I made on other peoples sites and kept them for when I make my own site.

  3. Copy and paste will work, I guess, if you want to quickly grow tired of maintaining the aggregate, and if you want to forgo the option of nifty features like being able to automatically keep track of how many other comments were contributed to the post you commented on. I just thought of that! Seriously, once you get weblogs to talk to one another, some powerful things can happen.

  4. if truly automatic is seems like a system that would be either clumsier than simple copy/paste or extremely easy to spam.

    an easy temporary method would be to send a small packet of post data to /tf/ of whatever website people use in their comments, that way no forms have to be changed and it would usually only create a 404 error if the person isn’t using it….or you could just send it to a central site based on e-mail addresses. both ways are easy to spam, but require no extra work or confusion for people not using the system.

  5. I thought Typekey was going to be the beginning of all this, but unfortunately its execution is still a little clumsy. It would be much more attractive if there was more of a POINT to it (like the archiving of all the comments you’ve ever made on anyone else’s blog… and the syndication thereof… whoa.)

  6. P Winchell: I’m sure something in your remark there made sense, but I couldn’t parse it. I’m also not sure how much benefit there would be to spamming this system, either — it would be a fairly simple matter to secure your own aggregate to avoid that happening, and why would any spammer want to create their own aggregate of all their junk comments?

    Anyway, any new community/discussion technology one could dream up can easily be shot down in the first five minutes by uttering the words “easily spammable” followed by some technospeak. It’s true, spam is a concern and technology is spooky, but why let that stop us?

    Virginia: Yeah, TypeKey is a bit of a mystery to me. I don’t really object to its centralized user authentication as much as some might, but I can’t really be bothered to implement it on my site. What you’re talking about is a great idea and one route to centralizing all of your own comments written on any weblog. Unfortunately, it suggests that everyone would need to run TypeKey, too, something which I doubt is going to happen any time soon.

  7. My, what a fascinating idea. You’ve got me interested.

    We could do registration by email address (think Gravatar). Create some simple plugins for WP, TxP, and MT (and others — phpBB? that would be cool) so that, upon posting a comment, it pings the remote (TrackForward? Hmm..) server.

    You could then do all sorts of fun stuff on the remote site with archives of all your comments, generate an RSS feed to use on your site (Recent Comments). Yes…very cool.

    I suppose I could, after stealing your idea, let you design the user interface 😉

    We should talk more about this…

    B.J. Schaefer
    The Brainstorming (and Bored) Web Developer, Who Has Been Searching For Something To Build For Months

  8. This is too funny. Just today someone sent me an email talking about this idea he had just thought of — which, of course, was that he always forgot about comments he made on other sites and wanted a way to aggregate them somehow.

    I then preceded to tell him that it was really funny that he mentioned it to me because about a month ago I looked into making this idea into a business, but couldn’t think of a way to make honest money off it so I trashed it.

    I then said ( exact quote, I promise 🙂 “If you and me have thought about it, I’m assuming several others have too :)” Isn’t it amazing how people repeatedly think up the same solutions at the same time? Happened with the telephone, tv, radio, etc.

    I would love to see this actualized. I imagine something like deli.cio.us with a straight-forward API that would allow bloggers to add a simple Javascript function to their sites and show all the conversations they’re currently having around the web. Maybe if we put our heads together …

  9. sorry for my comment having much technobabble and being all about spam.

    I do think it would be a wonderful system to use. It would be relatively easy to make said system so that it could use either central servers(trackforward.com is open) or personal systems(tekgo.org/tf/ would work for me). The easiest way would probably be to just add a “trackforward” address in the comment form that MT or WP or whatever would deal with in the background.

    of course the biggest hurdle would be getting enough people using this that it would be usable, which would probably mean that it would have to come enabled by default in blogging software.

  10. I am personally appalled by most of the comments I make on web sites if I re-read them more than a few days after I write them, BUT, if I wanted to pull off what you’re looking for, I’d probably just start signing all my comments with something like “Khoi Vihn Diesel” and then I could run an offline Google search on that term, follow the results, and parse the relevant comment blocks out of the resulting pages. The results could then be cleaned by some regular expressions, or an unpaid intern, and then stored in a database for future use.

    That’s how we do it in… the ghe-tto.

  11. What an intriguing idea! The problem is always going to be CMS’s – there are so many now, long gone are the days of most people using either MT or Blosxom. The Gravatar system has managed to work across a good range of CMS’s though, so I guess it could be possible – it is just text after all.

  12. I think that a centralised system is not really what you want. Look at weblogs.com — to me, it feels “overwhelmed”, slow, and sometimes even unreachable. And that’s just blog *posts*. Don’t know about you, but on a lot of site, the number of comments far outweigh the number of posts. Unless this idea is executed by Google (or another party with similar horsepower / bandwidth to burn), it’s going to collapse, probably sooner than later.

    No, decentralized is probably the way to go. Make it an extension to the various CMS’es: at least the interesting / most prevalent (sp?) ones have an XMLRPC interface which they announce, so you could base this on the web site URL people fill in at the comment form. Your software then pings my software with the comment, and maybe whether it was accepted or bounced, and then my CMS sorts things out. Reverse trackback, if you will =]
    No, not all CMS’es would be able to partake in this, but as I said, the bigger ones probably would.

    Alternative, centralised: people create an account at your trackforward site, and the system gives them a unique URL which they then use as their web site URL in comment forms. Some form of link back magic happens (either through the reverse trackback method I outlined above, or by some other means, ie. RSS), and at the trackforward site, people can see where they reacted (complete with link to that location). Hell, why not make it public, so that you (the author) can see where that _really_ insightful comment by X came from, and whether X has more insightful things to say. Oh, and a link to that persons web site, if available, of course =]

    However, that’s the centralised route, and I don’t even want to think at the kind of traffic you’d get if a lot of people would start to use this. =]

    @Jon: no, you’re never going to “get them all”, but that goes for most interesting technology. Not all blogs are trackback-enabled, either, and that didn’t stop it from becoming quite popular. As soon as some big names start to use it, more and more “little guys” will want it too.

    (apologies for any funny spelling glitches — I got carried away, which seems to disable my internal spell checker =] )

  13. Hm, that’s odd: I see no trackback URI mentioned, nor a link to the comments feed. That last thing would solve part of your question, of course (so switch to WordPress, then! =]), but this Trackback thing…
    Any specific reason as to why you’ve disabled that?

  14. Flickr seems to be moving this way as well. You can of course, look at recent comments on all your photos from your updates page, but also track any comments you’ve made on any other photostream, which is very handy.

  15. Max: Sheer laziness accounts for the fact that I have not enabled TrackBack (and to a lesser extent, TypeKey too). Also, I’m concerned about how long (i.e., tall) these article pages will get if I add TrackBack data, though I do have some ideas on how to solve that. Hopefully I’ll find some time to tackle that soon.

  16. I think you could probably hook this into the URL (in your case, “Web Site”) box most of us fill in when we leave a comment. So, for example, when I press “post” on this comment, your software (MovableType, TypePad, WordPress, etc.) would send a ping to whatever address I listed, which my weblog would then pickup and use however I’ve configured it.

    I’d like to see two things from this:

    1. A space on my blog that lists all comments I’ve posted on other people’s pages.
    2. A way to track comments that have been left on discussions since I’ve left mine. All you’d really need would be to ping each commenter’s page with a date, and then my page would list my comments, plus something like “There have been 4 comments left since this one. The most recent was at 10:30.” Then I’d be able to check my comment page and follow discussions back to their source.

    I really think that lots of weblog discussion fizzles out before it’s had its chance because people forget to check back the next day. anything that helps people stay involved in a converstation would help.

    Of course, this would also open up another avenue for TrackBack spam…

  17. I’ve just started bookmarking sites I’ve commented on at delicious, with the tag “commented”. This way I can keep at least some track of posts where I’ve left my trace. But still, it should be automated, and I want to be notified when someone posts a new comment or a response to mine.

  18. It’s not completely what I’m talking about here, but for automated subscriptions to comment threads, add a comment to a post at Narayan Nayar’s excellent Etherfarm weblog (powered by Expression Engine). You’ll then receive notices when someone posts further comments. It’s pretty slick.

  19. Khoi: I know what you mean with the “pages will grow, fast” thing — in combination with the odd-looking snippets that some blog apps produce, it has prompted me to discard the snippet and just do a list of sites that trackbacked. Binary Bonsai did that too, IIRC, and so did Kottke. Saves heaps of space, and since trackbacks are meant to generate traffic for the trackbacker, why not facilitate that? =]

    David Ely: that’s about exactly what I meant! Did I word it badly or did you misread my comment? (the former might very well be an option…)
    As for trackback spam: I’d really hate to see nifty tricks being shelved out of fear for spammers. Not that I think some folks won’t use this for exactly that reason, it’d suck. Damn spammers.

    Jacob: have a folder for that in your bookmarks bar, and drag the URL of the site you’ve just commented in there. Added bonus: it’s faster than del.icio.us (no server request), and if the blog app you’ve just commented to gives you a segment (like #comment123) it also takes you to your comment, ready to see what follows it.

    Khoi (again?): The mail-me-if-someone-comments-after-me is available as a plugin for WordPress and Pivot — and I’m sure someone ported it to MT as well (if it didn’t originate there). Yes, it’s nifty. Make sure you get one that allows you to “unsubscribe” from those notifications, though, because it sure is irritating to get the occasional mail informing you of updates to some topic you replied to half a year ago. =]

  20. I wonder if this isn’t something that would best be handled client-side. In the interim, without the development of some robust plugin architecture, I can see a Firefox extension that would suck up all the relevant details of a given comment you’re making, or provide a pop-up window through which to comment, which then stores the information client-side in a cookie or something which, at some point, is transmitted to a server.

    Of course, to get all the interconnected wizardry, this client-side interface would need to be only part of the solution: whichever server receives the data would need to be set up for handling it properly. That would mean that the extension would need login and authentication capabilities as well as an ability to pull from the server.

  21. I created bbBuddy (bulletin blog buddy) exactly for this purpose.

    I noticed people were losing alot of the content that they were creating.
    Posts on message boards would get pruned, comments would get lost…

    I also thought that people may want to look back on their content just for nostalgia sake.

    I know people are generally lazy so I created a bookmarklet that would save any text highlighted to bbBuddy. Nothing last forever and I didnt want people’s content to be locked in at bbBuddy, so bbBuddy also allows people to download all of their posts. You can also add labels to your posts to organize them.
    I also made it clear that bbBuddy would be permanently free.

    The problem with bbBuddy is that it is not automated. You have to highlight the text you want to archive and click the bookmarklet. However, I figure if people dont want to do that then their content is probably not worth archiving anyway.

    Please check bbBuddy.com out and give me your feedback. Make sure to try out the bookmarklet.

  22. Khoi (and others): I assume that lots of folks already read this, but for those that missed it: check out the interview with Tom Werner by John Oxton at http://joshuaink.com/blog/240/seven-and-a-half-questions-for-tom-werner . Tom Werner, for those who don’t know, is the guy behind gravatar.com, and he plans to fulfill your request =]

    (Sorry for the link not being a link, but I don’t know where I can use HTML, where I should Textile or something simiar, and where I just plain can’t use any kind of markup. Khoi, if you can / want to make it a proper link, go ahead…)

  23. This is very interesting.
    What would be *really* useful, though, would be an archive of my real-world comments. Never again be confronted with a “but the other day, you said completely the opposite”. With all your comments neatly archived and indexed, you can check your position on important issues whenever you like. Never contradict yourself again!
    If only…

  24. Kottke does that mannually, I asked him. I’m doing manually too, using custom fields (WP 1.5) and a special category. It’s easier and more interesting than I thought It would be. All I have to do is create a normal post copying the text and the permalink of my comment, and I can customize the name of the blog and the title of the post I’m commenting (some blogs don’t have titles for posts and some blogs have names too long). So I have just 4 fields to fill in and I guess it’s better than if it was automatically – what I find VERY hard to build because that very intelligent script would have to get all the specific data of a single comment and post it in a single or double click, and not all blogs or comments systems are prepared for that. I guess like the tags, the standard is yourself. Or not. 😉

Thank you! Your remarks have been sent to Khoi.