|
|
|
|
|
by esm23
5240 days ago
|
|
I'm currently building a social media news aggregation site, and was considering the merits of using a Facebook/Twitter only login/comment services. Do you believe most potential users would be opposed to this? If using these services we would be very clear about privacy issues and options,giving our users the option of restricting their activities to our network, and not sharing the information with facebook/twitter if the user would choose not to. Also in our service agreement and in practice we would restrict our use of our users information to strictly what they share with us, not datamining them, which in my opinion violates peoples privacy. There is a lot of upside in my opinion with using the Facebook/twitter login. They are the 2 dominant social media platforms, in which our potential users are bound to have their strongest ties. If they choose to share their activities on our site with their social media network of choice, it would be their way of advertising a service they are using (and presumably enjoying if they are participating on the site), which would help our service attract users. Would appreciate any feedback and advice, especially pointing out where im wrong :) |
|
Everyone uses the web differently. I understand that other (much larger) demographics have no problem with single sign on and that you can perhaps get to market faster without implementing a dedicated sign up system. I respect that.
I do like "social" (I hate that word) news sites; a tool to help cull non-interest items, prioritize on key interests, and introduce novel information from users who share similar interest graphs would help me out enormously.
I don't want to derail the topic, but has anyone given a thought as to whether aggregation (and distribution and reader functionality) could work in a decentralized, p2p manner? Without relying on a central authority of any kind?
I would love a scaled-down, low-bandwidth bittorrent-type app that delivers news items (with highlighted commentary) to my devices, already wrapped in an readability like interface, collected from among my "interest graph peers". That way, the mining/ML algorithms and processing overhead can be implemented and tweaked by me.