HN Rate My Idea: Socratic Linking
whattofix.com
whattofix.com
You also will face extraordinary resistance if you become large enough to matter; many people will consider it a violation of their site, and will claim it's illegal. Eventually some will try that out in court, if you make money. You'll get people like me claiming it's unethical, and pointing out that if you do succeed and legitimize the idea that it's ok to lay things over other people's sites regardless of how they feel about it that you can count on Comcast and AT&T jumping on that capability to use it in ways you won't like, and that in the end we're all better off leaving web pages alone.
But, really, my main point to you personally is that this has been tried repeatedly and if you want to build something successful you really ought to examine the history and figure out what it is that would be so different about your approach that it might work.
The best comment so far that I've heard is to begin implementation in a walled garden in niche subject area. See if that works, then expand a bit at time into larger (and foreign) sites.
There's a fascinating question here about who owns what part of the content. If you write something, you own it, but if I say something to my friend like "check out jerf's last blog beginning about half way down -- that really pissed me off" obviously that content is mine. Now the interesting part is: who owns the connection between those two concepts? I don't think it can be you. ???
Why don't you start experimenting with your own blog?
Perhaps a product review site.
Not my intention. I do seriously consider it unethical, but it's a very intellectual sort of thing; I'm not emotional about it, and more data may yet prove me wrong.
"begin implementation in a walled garden in niche subject area"
Probably your only hope of success.
"There's a fascinating question here...???"
Only because you asked: http://www.jerf.org/iri/blogbook/communication_ethics . It's pretty easy for me to go on... and on... and on... about this topic, so I try not to without a lot of provocation now. :)
Two problems:
1) A lot of these sites that you refer to are small and the chance of multiple users of a FF plugin going to that site might be unlikely.
2) Unless you have a lot of experience with FF plugins, it sounds like a lot of work for something you may just want to experiment with.
So,
1) Make it a destination site, with a simple bookmarklet for grabbing the text and adding questions, etc. It takes 2 minutes to make a bookmarklet that can grab text and will work with more browsers than FF.
2) This way, you also expose neat questions/ideas that are ripe for discussion, from a series of small sites that might otherwise go undiscovered.
I understand it's not quite the same, but for a small project, proof of concept, it's a day or two of coding instead of a huge commitment, and it works well even if you have a tiny number of users.
There are services (like 'fleck' or 'diigo' not sure if they are still working) that allow you to stick post-it like notes on top of webpages, that has quite a bit of the functionality you describe.
It sounds like a firefox plugin would do the job.
I knew there were basic "post-it" kinds of apps out there, but I didn't think any of them combined the wiki and emotional response slants. For instance, Discus does site commenting and tracking (along with CoComment and others). What's missing to me is the cross-linking between sites based on topic area.
The HN question I was expecting was "Sounds cool. How are you going to monetize this? To which I have no answer.
Just seemed like a neat concept, so I thought I'd throw it out there so you guys could throw rocks at it.
The reason I mentioned the post-its is because your idea seems to be a special case of those, you could easily extend such a post-it like application to become what you describe.
One really sad thing is that with all the cross-site scripting jerks out there some of the nicest things that you could do with javascript (layers on top of other websites) are no longer possible.
I've been playing around with something like that on cameria.com (now absolutely defunct, but I do still have all the code in a repository somewhere) which basically dropped a chatroom on top of every website where you wanted one, using little avatars and balloons.
The idea was to recreate the feeling that you have when you visit a mall or a store but then to apply that to a website, you'd basically 'see the crowd' of the other people visiting that site (limited to some geographical filter and a maximum number of users).
This also allowed all kinds of interesting bits like crowdsurfing (before it had a name) and 'guided tours' across the internet.
Alas, the 'holes' in javascript (that I thought were neat features) got closed to stop cross-site scripting attacks and the project died.
So now everybody that wants to do any kind of layering has to either have a friendly relationship with the website that they want to display the layer on (for instance analytics makes use of this through their tag) or you have to cache the data and spit it back out.
Back then we just loaded JSON in javascript from a foreign server. I'd be surprised if that loophole was closed because if you can't load JS from a foreign server lots of stuff is not going to work anymore (like Google analytics)
The reason is that as soon as your project requires a plug in or a download the size of the 'universe' is only a fraction of what it would be otherwise.
Second (and hopefully more constructively) there have been many web page annotation startups, which is not to say your idea is the same (it's not, as far as I can tell). But it might be useful to think about your idea in terms of features and versions, and ask yourself: if my version 1.0 release looks like what some other startup has had at some point, then why did that other startup not ultimately evolve to my final idea? Is it just that they were not smart, or were there good reasons the idea wouldn't fly?
Or maybe your thinking is you would just jump to full features on version 1.0... however, in that case, thinking about the above question might still reveal some insights.
Aside from that question, what you describe sounds a bit like the online service SpinSpotter (which I think is a good idea, but severely underused).
Instead of a single dimension, my idea is that internet content consists of two things: keywords and emotional impact on the reader. So you tag the content, giving it semantic value, and then click on various indicators showing how it affected you emotionally.
So now I can provide you with associated material that is relevant and can predicatively adapt to your moods, i.e. responses.
It's cool stuff, but basically it boils down to a souped-up version of PageRank-like concepts.
I have a screen shot of a UI prototype but have to catch a plane right now. Glad to put it up if anybody would like to see it.
I was having visions of trying to parse sentence breaks, and then having users shove the good sentences in the article to the top, and bad ones down. None of which made any sense to me...
There is a bit of arm-waving going on here, I admit. But it's just a blue-sky concept right now, so a little arm-waving can be forgiven (hopefullly).
super cool project for enabling an internet-wide dialogue/communication/modification-layer