Subscribeco.mments : What’s missing
The short answer to that question is : a lot. I mean that in a positive way. co.mments has such a huge potential that much of that is as yet untapped. Growth potential from this point is huge and impossible to properly define. Here’s are a few things which I think would be pointed in the right direction:
- Harvest comment metadata. You’re pulling in data from the blog page already — why display the metadata as just part of the comment body? At very least the permalink to the comment should be harvested — this would be a huge improvement by itself because we could click from co.mments to a specific comment and not just the whole page. Author data comes next, and then timestamp. It’s all there, you just gotta get it out.
- Track Main/Archive/Category pages. This works, but it’s really ugly. If I enter a non-post-page URL on which there are comments (such as http://singpolyma-tech.blogspot.com/) I expect a similar result to having entered all of the posts on that page. Instead I tend to get each comment section as a comment. Not that cool.
- I’ve said this already, and I know it’s in the works, but a RESTful API. XML or JSON(P) (or both!). You’ve got my data! Give it back!
co.mments is one awesome service with huge potential. All you’ve gotta do is tap it. Incidentally, one of the reasons an API would be cool is that then you can harness the larger sphere of programmers who will be writing code working with it. They will write code so that you don’t have to. The more you open your site to us, the more we can do for you and save you the work!
Posted by singpolyma in General on Monday, June 19th, 2006. You can follow any responses using co.mments, leave a response, or trackback from your own site.
Assaf
June 19th, 2006 at 11:56 pm
Oops. I just noticed this in the moderation queue.
Assaf
June 20th, 2006 at 12:08 am
Thanks for the great feedback.
Harvest metadata. I’m in the midst of updating the tracking engine. First, so I can improve the hit rate and capture more blogs. And second, to harvest more metadata. Once that’s done and stable, I’m going to add more interesting stuff to the tracking page/feeds.
Main/Archive/Category. I know I don’t mean to track an entire blog, only specific posts. I’m going to keep the bookmarklet working as it is. But, there’s an easy solution to that …
API. XML and JSON(P). The reason I upgraded to Rails 1.1, it makes APIs a snap and you it does both XML and JSON in two lines of code. I got a lot of ideas from users (feel free to post more) and I’m going to do them one by one.
One thing the API will let you do is write your own bookmarklet. Or a Firefox extension if you want.
Only thing holding me back: I want to get the new tracking engine up and running first, and build the API around more metadata.
Singpolyma
June 20th, 2006 at 1:19 am
Thanks for the response! I suppose since the stated purpose of co.mments.com is to track specific posts the glitching on non-post pages does make sense. Tracking the main/archive page of a blog doesn’t actually make loads of sense, now that I think more on it, because the posts there present are subject to change. Very glad to hear that metadata+API are both being very looked at. I agree that the harvesting engine should definately take priority ove any other changes
splintor
June 20th, 2006 at 3:06 pm
Another idea that comes to mind is to identify messages written by the orignal poster (usually the blog owner) and mark them differently.
So an example in http://www.dustindiaz.com/its-ma-burthday/
Assaf
June 22nd, 2006 at 12:11 pm
@splintor
great idea.
Brad
July 17th, 2006 at 9:54 am
One other thing:
An option to stop following comment threads after X days of inactivity.
Assaf
July 17th, 2006 at 11:07 am
Brad,
At least two posts I wrote a while back are still getting comments, sometimes weeks apart. So I’m not sure about that option, there’s a tradeoff.
What do you think, would it be a good idea to just stop tracking at some point?