Adding to the TypeKey fray

This entry was published at least two years ago (originally posted on March 23, 2004). Since that time the information may have become outdated or my beliefs may have changed (in general, assume a more open and liberal current viewpoint). A fuller disclaimer is available.

Okay, so while I wasn’t paying much attention, TypeKey became the most recent firestorm to sweep through the weblog world. By now, you probably already know at least the basics of TypeKey — and if not, then you’re not likely to be interested in the rest of this rambling (in brief, it’s an identity verification system for weblog commenting…think Microsoft Passport for weblogs).

There’s been a lot of interesting discussion of TypeKey since the first announcement. I haven’t been able to wade through all of it, but I’ve found Shelley Powers‘ three TypeKey-related posts (TypeKey: The Patriot Act of Weblogging, TypeKey Scavenger Hunt, TypeKey: Final Act), the related discussion in her comments, and the comments in response to Jeff Jarvis’ post Comment to be extremely worthwhile.

Looking over the TypeKey FAQ recently posted by SixApart, my first impression is that it sounds like a good system. Identity verification combined with comment moderation (to come in MovableType 3.0, and I’m assuming also in a future TypePad update) can go a long way to combatting both spam and unwanted comments. It likely won’t be a 100% solution — but then, chances are there just isn’t a 100% solution (there are always smarter mice to avoid the better mousetraps).

However, Shelley brings up two very important points (actually, she brings up a few more, but these two spoke more to me) — points that are prompting her to avoid TypeKey, neither implementing it on her weblog when it is released, nor signing into the service as an authenticated commenter. While I don’t feel as strongly about these as she does, they’re certainly worth considering.

The first is simply scalability and performance — if every comment on every weblog needs to go through some verification process before it appears on a post, how quickly will the system be able to respond as more and more people sign into the service?

We who went to Movable Type or other product that we host on our own servers did so specifically because we did NOT want to have any form of dependency on a centralized system. We did so, for the most part, because we have been burned on either performance or access because of the centralization and scaling problems.

…[Mark Pilgrim] lists several centralized systems that he believes do scale well and serve the community, and it’s true these have managed to scale and are useful, but each and every one has failed when I’ve tried to access it at least once a week.

Blogdex was inaccessible off and on this weekend, and Technorati was hard to access last night, and I couldn’t access Bloglines two or thee times last week, and I got some kind of odd error with Radio comments a couple of weeks ago, too, and, well, the list goes on. The problem with centralized systems is not that they fail completely and breakdown permanently; it’s that they behave oddly or inconsistently, or poorly under load.

…the thing with Technorati or Blogdex or Bloglines (I haven’t used Feedster) is that I’m not dependent on them to write to my weblog, or for my commenters to respond, or for my pages to be accessed. Only my own system resources, or the Internet in general between my server and each of us can impact on this. With TypeKey, though, that’s changed.

Unfortunately, at this point, there’s really no way of knowing how Six Apart plans on handling this. TypeKey’s announcement is only a few days old, and details of the underlying systems (both software and hardware) are still forthcoming. Until TypeKey moves out of the testing stage, goes live, and starts getting hammered by everyone who signs on, we won’t know the impact that the system might or might not have on our sites.

Historically speaking, as Shelley points out, things aren’t looking too rosy. I’d like to think that Six Apart realizes this, and will have done everything possible to ensure that these issues aren’t a major factor — but then, there’s a lot of things that I’d like to think that aren’t borne out by real-world evidence. At the moment, “wait and see” is the only real approach we can take.

The second issue that caught my interest was one of conversations, who we allow to participate in them, and what we allow them to say.

Odd thing, weblogs and comments. We say to each other, “Our weblogs are our homes and we should be able to control what’s said in them”. Yet, they aren’t our homes, are they? You don’t keep your door open for anyone to just walk in to your home, do you? Weblogs are published online supposedly because we want a broader audience for our thoughts and writing then just our friends and family.

They aren’t really our ‘homes’, and the analogy fails in so many ways, but they are our spaces, so we have a right to control them and hold people who comment accountable, don’t we?

But who holds us accountable? I’ve seen again and again, the weblogger write the most inflammatory material in an essay, and when you respond to the tone they set in their writing, or to their responses to your earlier comments, you’re told to be nice, or be gone.

We say, commenters should be held accountable for what they say. I say, but then, who holds the weblogger accountable?

So far, I’ve kept a very open comment policy on my weblog. Generally, outside of removing comment spam and deleting duplicate comments, I do very little editing of what people contribute to my site. That’s not to say that I’m not tempted at times — I have one particular post that has picked up some extremely disturbing racist comments — but to date, the only major comment deletion or editing I’ve done has been at the request of the person who left the comments (and that was under admittedly unusual circumstances).

Truth to tell, I’ve never entirely understood the impulse to delete comments that don’t agree with something I’ve posted. I’ve had some very interesting discussions with people who didn’t agree with something I’ve written — if part of why we bother to blog is to invite discussion, why would we want to limit that discussion to the proverbial “echo chamber” of nodding sycophants? Seems to me that that approach makes for some dreadfully boring “discussions”.

At other times, I’ve seen people get very aggravated about a comment that seemed to be overly rude, aggressive, impolite, or offensive in some way. Well, perhaps…but while it would be nice if all debate could be structured perfectly politely, is the occasional jibe, jab, or verbal tweak really worth deleting the comment (or even banning the commenter) and in the process removing the actual content of what was said? Myself, I certainly don’t think so — in fact, I often get a certain perverse pleasure out of responding to those posts. While I’ve never quite mastered the stereotypical British practice of being able to say the most brutally vicious things in an impeccably polite manner, I’m always willing to give it a shot. ;)

So, with all of that…will TypeKey work for me (if and when it is integrated into TypePad)? Well, at the moment, keeping an open comment system has been working fairly well — the comment spam hasn’t been hitting me hard enough to make it terribly difficult to deal with, and I do enjoy getting feedback when someone feels moved to comment. I’m also not a big fan of comment moderation — as it would require me to approve every comment before it appeared on the site, it could lead to some oddly dis-linear conversations as I’m not at my keyboard at all times, anxiously awaiting the next comment to land in my inbox.

At the moment, it looks like TypeKey-enabled weblogs will have a few options for how they handle comments.

  1. Only accept TypeKey-authenticated comments where the commenter sends an email address
  2. Only accept TypeKey-authenticated comments
  3. Accept TypeKey-authenticated and moderated comments
  4. Accept TypeKey-authenticated and regular comments
  5. Accept moderated comments
  6. Accept unmoderated comments
  7. Accept anonymous comments

My current plan is to go for option four. Those people who sign up for TypeKey will be able to use their TypeKey logon to verify their identity. Those people who don’t use TypeKey will still be able to participate normally. Ideally, I’d like to find a way to signify TypeKey validated commenters (possibly with a special icon by their name) — I’ll have to look into the viability of doing that once TypeKey is available to me.

Given that option four will perform essentially identically to my current open, unmoderated system, why bother? Well — first off, curiosity. I’d like to be able to play with the TypeKey system from both a commenting and administration standpoint, and this will allow me to do so. Secondly, it will allow those people who do use TypeKey to use that logon to comment on my site along with any other TypeKey-enabled sites they visit, without having to remember separate information for my specific weblog (sure, it’s easy enough to do, we probably all have a good number of username/password combinations rattling around in our heads…but why not make it that much easier to keep track of them all?). And lastly, should I get to the point where I feel the need to institute a more draconian comment policy (though I hope that doesn’t happen), having TypeKey already enabled will make it that much easier.

In the end, then, the questions and concerns that Shelley has raised are very worth keeping in mind, but they’re not enough to keep me away from TypeKey, as I think that there could be some very good benefits to the system. Here’s hoping that once TypeKey goes live it’s the benefits that play out under real-world conditions, and not the potential downsides.

iTunes: “On the Run (Hot Tracks)” by Bigod 20 from the album Roadkill 1.04 (1992, 6:21).