-
I applaud your plea for people to take control of their data and strongly agree in principle. However, there are a couple of problems:
- In some cases, the available open source, self-hosted solutions simply suck incredibly badly compared to the commercial hosted solutions. Examples include rnews, which is awful compared to Google reader and I can’t seriously believe that you would recommend Posh over Netvibes, iGoogle, etc. at this point. Much as I really want to like this software, it is a bit like asking a windows user to convert to linux in 1992. A good idea in principle, but it’s not going to happen for the average user until the software has comparable features and is as easy to use.
- In the case of social networking sites there are network effects to consider. For example, bookmarking sites like del.icio.us and digg get better the more people that use them and the more friends that you have on them. Increasingly, almost all web 2.0 services have some sort of friends or recommendation system (of varying utility) that relies on a large number of users for it to work well. At the moment, you cannot reproduce this by installing something on your own webhost.
What I would recommend doing at the moment is to find a way to backup as much of your web presence as possible, just as you would with files on your own computer. With RSS readers this simply means keeping a copy of your opml file, but for other services it is not so simple. I believe that installing sweetcron might do some of the work for you, since I think it works by downloading the data from your feeds to a local mysql database on your webhost. However, I haven’t looked into the code and the documentation is not good enough for me to be able to tell if it really stores everything permanently. Definitely a good idea in principle though.
In the long run, I think we need to push for the idea that social services should work more like email than like websites so that no single company is in control of our data. By this I mean that you would have a “social server” based at your work, ISP, webhost or even your home, just like you have with an email server. These servers would talk to each other in a common language, pushing friend statuses, articles, bookmarks, ‘tweets’, photos, videos, etc. to each other based on who you are following and various recommendation algorithms. Initially, this would have to work with the APIs of existing services so that you don’t lose your network by switching to the new system. Some of the popular services that are quite closed (e.g. Facebook) are going to cause problems with this, so we need to push for services to have APIs that are as open as possible as a first step.
-
I love rnews. I didn’t refer to it in this post, but you know me well enough. Seriously–I really think it it works fine as an RSS aggregator. Perhaps I am not an RSS power user, but I believe that most people would be well served by its simple, clean, and easy to understand interface. Posh–well admittedly that is more marginal. It is rough around the edges. I believe Pligg is an outstanding alternative to other social bookmarks from a software perspective. Your point that you lose the utility of sharing bookmarks with a self installed social bookmarking is valid, yet so is my point that it can all disappear rather abruptly. It would be great if one could easily export bookmark collections from del.icio.us and other such sites into Pligg for safekeeping.
I come at this as a K12 educator and, to be honest I thought about qualifying the recommendations, but I didn’t. That being said, I do believe that these are viable tools for the K12. Posh actually has features that I don’t believe can be found with NetVibes—the ability to offer a restricted menu of widget options and feeds for users.
I agree with your call for greater data portability and an open common language. I envisage a decentralized distributed network. Sometimes I have wondered if this is what the Elgg developers have been thinking about.
I’ll have to look into sweetcron.
-
You are obviously not using Google Reader. I guess I am a bit of an RSS power user, but it has so many good features I didn’t know I was even using until I tried to switch to something else. The experience is just so smooth and unobtrusive compared to anything else.
I agree with you that in teaching the ability to create a controlled environment outweighs the possible benefits of using hosted services. However, I’m mainly using these tools in the context of research, for which I believe network effects are much more important.
The Elgg developers do seem to be interested in portability and decentralization. OpenDD is part of their solution to this. Unfortunately, I don’t think they have the necessary clout to get it accepted as a standard, since the Googles and Facebooks of the world are working on their own solutions. If it were possible for Elgg users on one installation to follow users on a different installation and to join their groups etc. then things would begin to get interesting. However, I think that the architecture of the system is the wrong model for this. To be viable it would need to work on a “push” model, where content you have access to is periodically sent to your local server regardless of whether you are online, rather than the current “pull” model where you make queries to a remote database. The latter is a very costly model for large networks, and is the main reason why Twitter has had so many stability problems. If every Elgg installation potentially had to serve content on demand to users of every other Elgg installation in the world then I think that it would quickly break on a typical webhost account.
-
-
-
Pligg? IMHO, I’d choose it’s fork http://socialwebcms.com/
Worth are read Social CMS Buzz:
http://tinyurl.com/bvzy7g-
Tradenet–Thanks for the heads up on socialwebcms. I hadn’t worked on Pligg in awhile and I wasn’t aware of the fork. In house social bookmarking for our K12 school is a bit down the list as of yet, but I want to have my ducks in a row when we get to that point.
I’ll download it an give it a run. I imagine if I could lock it down as I could Pligg.
Drupal–I’ll get to it one of these days when I have the time to invest.
-
Frankly if I wanted a serious enterprise “Digg” like clone I’d choose:
http://blog.reddit.com/2008/06/reddit-goes-open-source.htmlI personally install and run a lot of Drupal sites. Why? Security. I don’t have to hack core. Modules for the most part are maintained well. Plus all the caching features as well.
-
Reddit eh? Much has happened since I first found Pligg! Thanks again.
Maybe I will do a Drupal installation and have a wack at it when time allows.
-
-
-
Comments are now closed.
Creative Commons License
Openedweb by Stephen J. O’Connor is licensed under a Creative Commons Attribution 3.0 Unported License.
7 comments