Saturday 24 May 2008

My Content Mess

So, I was making a diagram of all the services I use online, just to help myself better understand how things are connected. To start with the fact that I need a diagram to make sense should be cause for concern. I am a programmer and that should mean I am quite good at holding things and connections in my head. Still, I needed a diagram, so here it is:Now, while making the diagram, I made a number of observations:

For one thing, I use more services than I thought I was. There are 13 services listed here, and this is not everything.

Also, their interconnections change relatively fast. I repeatedly had to update the connections in the days I was making this, so I don't expect it to stay current for a very long time.

Another observation: I have followers in many of these services. Aside from my blog readers, my Google reader friends can see my shared Items, I also publish my starred items feed, I have a lot of connections in Facebook that can see my activity there, and also some connections in Friendfeed. In fact, since all these services offer open feeds, there are people who could be following me on any of them. There may also be overlap in these sets of people, and I try to maximize it by sharing content between services as much as possible.

Although this all seems a bit complicated, it is better than having one provider do everything for me. Yes, I could do most of these things with a combination of Facebook and Google Reader but it would not be nearly as easy, effective or enjoyable as it is now that I use the best of breed tools in each category. I do have a number of problems though:

1. Too many passwords and accounts. Please someone start using OpenID so I can log in everywhere with single set of credentials. I know it's not perfect, but it's the best we have.

2. Facebook is selfish. What can I say? In the diagram, you can see the little arrows going in, but nothing going out. Why? Because Fb does not let me take my data out of their walled garden. Apparently, what happens in Facebook, stays in Facebook. That's their choice, but that's also the reason I am moving my activity elsewhere. For me, Facebook is turning into a place where stuff gets posted automatically for my old-fashioned friends to see. I avoid posting original content inside Facebook. My cool friends are moving to Friendfeed, that apparently trusts its users more and traps them less.

3. Overlap. What if I write a blog entry and then share and star it on Google Reader, and Facebook, and FriendFeed? Well since I have linked all these together, It may end up 3 times on FriendFeed and 5 times on Facebook. But it is the only way it can get to the widest possible audience, all the people who are following me on all these services. Someone want to do a smart filter that merges all the identical entries? I have heard talk about something like this on Friendfeed, let's see.

4. Not only can I not merge overlapping articles, I also cannot aggregate comments. So what if I post something here, then share it inside Facebook and Friendfeed? There is a possibility that independent comment threads will start and I will have to follow them in three places. There are a number of new ideas coming up in this area, I will look into using them and I hope they generally catch on. (Of course, keep in mind #2)

5. Friendfeed and recently Facebook, are acting like aggregators. This is cool. I like it. But they don't let me integrate with any site I want. They offer me a choice. For the moment, FriendFeed offers everything Fb offers and more, but this may change. My problem is that I have other websites where I am or could be active. Examples are Zotero, Threadless and DeviantArt. The first website that makes it easy for its users to write the adapters themselves, is the winner. Less work for them, more capabilities for us. My money is on Friendfeed, but we'll see.

6. Status messages: there are just too many! I have status on Facebook, Twitter and Google Talk. They generally offer me the same utility. Someone make it easy for me to keep them synchronized? (Of course, keep in mind #2)

7. Bookmark managers. Del.icio.us and Google bookmarks each have their own advantages. Del.icio.us has a great firefox plug in and a nice social aspect, but Google Bookmarks makes my Google searches more relevant. So I want them both to have access to my favourites. I have found a way to transfer data from the one to the other, more or less, but I would like them to be synchronized. I also would like them to be synchronized with foxmarks which I use to keep my browsers at work and at home synchronized. Any ideas?

8. And now for my favourite complaint: Everything is centralized. Waaaaah! Why doesn't anyone build something where I can keep doing my work as I like without being tied to a single entity? Or many single entities for that matter. Perhaps then all the other problems would be moot since I would be able to solve them myself. Oh well, I guess my PhD work is relevant here, in the long-long term.

Overall my little content network is complicated but works mostly Ok. I wonder how many of these will be solved by next year. Maybe there is a ray of light on the conventional horizon called SwitchABit. It's still in development but sounds interesting.

Yet another blog post that I thought would be small but turned out huge. Oh well. I need more experience blogging.

Tuesday 20 May 2008

A Spam-less e-mail system

The last two posts have been quite different to each other. While the first one was an organized, principle-establishing and mature but rather basic article, the second one was more of a braindump, thinking out loud kind of post. Seeing this pattern that is forming, today I would like to introduce yet another category of posts I want to make: Ideas. When I say ideas, I mean technical concepts that can be turned into products or services. You may ask why I do not use these ideas myself, to make these products and profit. Why publish them in the open web to be 'stolen' by anyone? The truth is, and I don't mean to be pretentious, I have far too many of them for the amount of time I can expend on developing them. So while I am working on a big one for my PhD, I might as well hand the other ones out for adoption and hope to get visiting rights once in a while. And of course there is valuable discussion that can be made, and technical points that can be raised and a great educational experience for me all around. I do not pretend that all my ideas are valuable or that they do not need any correction. So without further ado, I present to you, thespam-less e-mail system, let's call it v-mail.

To set the mood, let me talk about the social network messaging systems. Many prefer them over e-mail. Why? because they have verified everyone who can send me a message and therefore have a very high signal to noise ratio. In fact, almost all the messages that reach my inbox are valuable and directed towards me personally. So it is easy to see why people may prefer this method of messaging. What are the problems with this system? Well there are quite a few. For one thing, centralization. Everyone needs to have a facebook orwhatever-book account in order to send me a message. Another is the difficulty in adding people on the go. If I meet someone and tell them to mail me, I do not want to have to go home and add them to my social network first. Also, I may have my mail address published somewhere and people I have never met before may want to e-mail me. This is impossible if I require prior approval for someone to send me e-mail. So how do we exit this seeming catch-22?

My answer is, by applying the best of breed solutions to each problemseparately. What I mean: Let's start with the social networking messaging system. So we have a mail provider, let's call them v-mail.com, that allows you to create an account and add a list of other addresses that you accept messages from. They may give you a standard e-mail address (say, alexandros@v-mail.com) but any e-mail from a source you have not added to your list gets dropped and never reaches you. So this would be an implementation of the basic social networking messaging system. In fact it is one step further, since it allows for multiple servers to communicate, much like the current e-mail system. All you need is an account with any provider of this new mail system. It does not have to be the same as mine, in fact, nobody is stopping you from hosting your own provider and becoming completely self-sufficient.

Next problem: Adding people is hard work, especially if you have to remember to add people later. For this we can use an sms-service that you can use your mobile phone to add people with. So when you add a new person, you may send a simple text message (e.g. add someguy@x-mail.com) to your provider. This would instantly solve that problem. This person can now send a message that will get through. Of course to start your list off you could import contacts from various social networks, address books and e-mail applications.

How about anonymous people that want to send you messages? Well, this has been solved already. Go to any website that wants to get contacted by its readers. They will not publish an e-mail address for fear of spammers. What they do is expose a contact form with a CAPTCHA (these strange numbers you have to type in so that they know you are a human). This contact form could be available at my mail provider's website so that anyone can contact me as long as they provide a valid v-mail ID and prove that they are humans by typing the CAPTCHA. This could also be used to request that someone adds you to their allowed list.

So by merging social networks, e-mail, mobile phones and the web, we can create a composite system that guarantees we only get verified e-mail. Let's go over some of the benefits:
  • You could use a typical e-mail program and e-mail standards. Additional functionality such as list management would be nice but notnecessary to start with. In fact you could use this new e-mail program to communicate with traditional e-mail addresses, as long as you add the people you correspond with to your allowed list, which could be done automatically when you send an email to someone. In a sense, it's backwards compatible.
  • The user is back in control. If someone on your list starts spamming, you can simply remove them and never receive another e-mail from them.
  • Leaking your email is no longer a problem. The information of your e-mail address is useless to a spammer since they cannot send automated anonymous messages anymore.
  • This solution can work well in a decentralized environment. no need for the massive processing power required by spam filters, anyone can do it on their home computers.
  • It's good for the Internet since spam is a massive waste of bandwidth.
Ok, have to admit I enjoy creating all these scenarios and exploring the ideas in my head. So here it is, for your viewing pleasure. Let's think about it: Would such a service be useful to you? What downsides can you think of?