Why RSS?
Every time the subject of RSS comes up, someone will ask me “Why not just read their blog in your browser?” (generally in different terms, depending on how crabby they are at the time). Here’s one answer: while I think of my blogroll as being my personal blog reading list, nothing more, most people think of a blogroll as being more like a public declaration of undying love, and so they are understandably reluctant to add someone, just to follow a single thread for a week, and are just as understandably displeased when I add them to my list for a couple of weeks, and then unceremoniously jerk them back out when they stop talking about whatever caught my eye. If I just throw their RSS feed into my aggregator instead, then for all but the most obsessive stats-watchers it’s no great shakes if I later delete it, when the conversation I was following dies out.
It’s just another tool for following conversations around blogspace, just like automated referral scripts (speaking of which, mine’s a bit delayed, until my brain starts working a bit better). It’s a shame that my script to generate an RSS feed from dotcomments comments never caught on (according to my stats, almost nobody ever subscribed to my comment feeds, and a total of zero people asked for the script), since I thought that was one of the best possible uses of RSS: there are other ways, like weblogs.com and blo.gs, to keep track of when someone posts something new to their weblog, but other than the scriptygoddess subscribe by email script there’s no good way to find out when someone replies to your comment on someone else’s weblog, short of constantly going back to check. I know if I left a comment on my Blogger FAQ, I’d rather wait for a reply to show up in Amphetadesk, rather than checking back two or three dozen times, waiting for me to reply to me.
RSS on comments. I like it and i dont. I mean. Its a hassle some times to follow a thread in that you need to add it, delete it, etc.
as for the blogroll. My thinking is this. I use it as a guide for my regular browsing. So I rarely change it. Either a site keeps me interested long term or it gets lost in the fluff.
Sounds like you want to change how we browse, or make it more interactive to some degree. Remembering that for many browsing is just a time killer I can see why there has been a lag in getting the ducks in a row. I would use your referrer script if you got it done. Does that help :P
I also mean to set up an addon for blogroll in bbt but ive got a long list of things to get done. Besides, im concerned about problems. I mean I dont ever want to become the primary support person for your script and if i integrate too much thats the roll I will be in.
im rambling.
Adding and deleting could probably be easier, since there is some level of effort involved in right-click, copy shortcut, start/open aggregator, click to add, paste, ok. However, Mark’s Subscribe bookmarklet does point the way to easier adding. It shouldn’t be too tough to do an IE context menu item that, when you right-click a link, will pass the url to Amphetadesk. Off hand I don’t know how you get the link that was clicked on, but I’ll bet I’ve already got one that I can copy. Deleting? I’m usually willing to make the effort, once someone bores me often enough.
Yes, I do want to change how bloggers browse. There are way too many people who use their blogroll just for link-whoring, so that it ends up half-rotted and useless. So I want to change that, and have people use it as their main navigation device. With blo.gs sorting it for you in update order, it’s just too damn handy not to promote. Then, I think bookmarking something that you want to check back on a few times is a pain in the ass, so I want to change that too: throwing it into your aggregator, so that you’ll find out within an hour if it changes, is vastly better, and more likely to keep up the kind of inter-blog conversation that I like to see. Ideally, I’d like to see everyone have comments, and have people put the whole conversation into the comments of the first person to bring it up, but that’s more of a behavior change than I think I’m able to effect, so I’ll settle for getting my friends to actually use their blogrolls, and maybe use an aggregator as a sort of second-tier, more changeable reading list.
Maybe you’ll be using my referrer script, and maybe you won’t. One of the things slowing me down is the problems involved in making it my usual style of open, generally useful script. So just this once, I may end up making something just for me. Not sure yet. The other possibility is that I’m just not good enough to get it done.
anything i can do to help from a scripting perspective? I’m game :)
Wow, I really didn’t give the whole idea of of a general referrers script enough thought this weekend. I was so busy thinking about the choices between a single flat-file and individual files for each post and a database and how to combine referrals from the main page with referrals from the archived post, that I didn’t ever get around to thinking about the problem of monthly (or even weekly) archives.
You know how Mark’s referrals are per-day, rather than per post? Well, that’s because the server (Apache log or PHP) doesn’t know what post a request is for: the fragment/hash/anchor part is all handled within the browser. With my per-post archives, I can track what post someone linked to, but for you to have more than a ”referrers for this month’s posts” list, you need some javascript. Your archives have to have a javascript which invisibly sends document.referrer and location.href back to your referral script. I’m thinking something along the lines of Brent Ashley’s RSLite. Problems that occur to me so far: you’d get a referral reported every time someone refreshed the page, which could get ugly, you would only notice referrals from people with javascript enabled, and you would have a sort of odd delay where the referral wouldn’t appear until the next time your page was displayed (first display counts it, from the javascript, next display shows it, from PHP). You could probably work around the refresh problem by keeping track of the IP address that’s sending the remote script call, so that you only count one referral per post per IP. The only way around the delay that I can think of right now is if changing the src attribute of a script tag dynamically reloads it (can’t remember if I’ve tried that or not). If it does, you could display the referrers with javascript, rather than PHP, so that script src=”referrals.php?137” displays the referrers for post 137, and then use the DOM to change the src to ”referrals.php?137&ref=” + document.referrer to register a new referral and update the count all in one. That would be pretty slick, if it works, but I’ve really got to go to bed, rather than test it.