I’m sad to hear this as I was a happy user of RawSugar. It was the only tag based service that didn’t just copy delicious to add a new features or polished design.
Apart from a few academic papers and proof of concept applications I never saw anything new in the tagging world until Rawsugar came up with its implementation of tag clustering and auto-tagging. These features probably need some more polish but they were steps into the right direction.

It’s not hard to decide where I should move all my bookmarks: Simpy has nice search features, a friendly user base and a very likeable developer. Otis does care for Rawsugar users and implemented an importer.

Apart from that decision I’m left to ask: "what was wrong with RawSugar?". I think this is an important question as tagging is web 2.0 and web 2.0 most certainly is another bubble going to pop somewhen – the question then is who is going to stay.

what was wrong?

Knowing that I’m not in the best position to analyze, I have a go figuring out why this project failed. Some of these objections I already mentioned, some can be found reading some (sometimes very mean: has techcrunch turned into digg?) comments.

user interface

The web site has always been too cluttered. The page structure even grew with the number of features. Rawsugars biggest competitor, delicious, was always very reluctant in exposing new features to the user.
Then its whole look doesn’t really fit into the web 2.0 world (Ok, delicious isn’t that much better, before the redesign about a year ago it looked like plain xhtml without any styling). It’s not about rounded corners, it’s about freshness. RawSugar always felt a little bit heavy (although it had the best response times).

interaction with community

I always felt like RawSugar doesn’t have the early adopters in mind: For questions it did use a forum, its developers weren’t really involved in the blogosphere. But then they wanted to gain ground in the blogosphere. It very much focused on a feature called "tag search for your blog" (you can see this feature in action when you switch to the index view of this blog, it is a very good feature indeed). This just didn’t really fit. It didn’t feel they were "of us". I hope you get what I mean – I cannot really describe it

ah well..

Maybe these arguments are all in vain. Why did delicious take off? Because it was the first social bookmark service (in fact it wasn’t)? There are many things Rawsugar did right. Why didn’t the user base grow more substantially? I’m keen to hear your thoughts!

Anyway – best luck to all the guys involved in RawSugar. Especially Frank – I did have a great time with you preparing our talk for www2006!

That’s the feeling that arises when I think of all the bookmark services out there that ought to file urls I somehow find notworthy so that I can quickly recall them afterwards: There’s too much clutter and the services are just too slow. Therefore I again begin to save my bookmarks at other places: In firefox or in some text documents lying somewhere on my hard drive (probably I should tag them?).

«I want it now»

As for myself, if I don’t get my data within a fraction of a second, I don’t feel that I have good access to my data. I think one of the best things of Google Web search is that it is incredible fast. Tag services are not. That’s a pity because within tag services I search my lowly amount of bookmarks (around 2000 bookmarks at that moment) whereas with Google I search the whole net. I often find myself at the situation searching for a url on Google rather than looking it up on my bookmark service. Even if I have to do 3 Google searches – I feel at control. If I have to wait for 3 seconds for an answer, that’s simply too much.

An experiment: I ran for ((i=0;i<10;i++)) do wget -q --user-agent="Mozille Firefox 1.9" $url; done; for the three big players in tagging services: 181del.icio.us, 182Simpy and 183Rawsugar. For $url I chose a page displaying all bookmarks saved with a certain tag. I also ran a query on Google for that “tag”. The numbers are number of seconds per query. The queries were run on different hosts with different internet connection, all in Switzerland.

I’m happy to see that with Rawsugar I’m with the fastest bookmark service (2 times faster than Delicious, 3 times faster than Simpy). On the other hand: all the bookmark services are at least 4 times slower than Google. I know that Google has set a high mark – but emotionally that’s the response time I’d like to have when querying for my data.

«I just want my bookmarks»

Have a look at the “result areas” (highlighted in green) of the three bookmark services.

Delicious: Result area in green

Simpy: Result area in green

Rawsugar: Result area in green

To come back to the comparison with the closet: Even if there are 20 magic hands with documents sticking out of the closet, it is crucial that the most important document is the one closest to my face. That means, when using those bookmark services, my eye should first notice the most important link. My opinion: The service that solves this best is Delicious. The results appear on the leftmost part of my screen and the result area covers 64% of the screen. Regarding that matter Rawsugar is far worse: My eye has to search for the first result. I find it natural to start reading at the left of the screen. But at Rawsugar there are so many links in the head of the page and then the left column helps me to refine the result – a great feature but at the wrong place – furthermore: if I want just get a link I know I filed under “important”, I don’t need this at all. Even Delicious’ “saved by xxx other people” – shaped in different colorings – is too much clutter for me. Maybe I’m a puritan but – what the heck – I just want my 9 bookmarks!

Beyond the criticism

Sorry to only criticize. I like those services a lot. They are helping organizing links and they are for free. To turn this post into a constructive comment, two possible solutions to the problems mentioned.

Bookmark services offer a «Minimal mode»

Yes, tagging is about collaboration. But I fear the personal value of tagging is too small – people might leave tagging services because they don’t feel their bookmark problem solved. There’s too much network and too many features. I propose a “minimal mode”, a result page that just shows select * from bookmarks where user="phred" and one_of_its_tags="important". In firefox there are 184Smart Keywords also known as “quick search”. I set up “myr” to search for my rawsugar bookmarks. Typing “myr important” into my location bar I end up on www.rawsugar.com/links/phred/important. If there would be a site on rawsugar that displays the same results just in this minimal mode, I would bend my smart keyword to this result page and would be happy.

Someone codes a «Tag agent»

I imagine a “tag agent” that has incredible response time and no clutter. It could either get the results via rss/api from the bookmark service or it could hold all my tagged bookmarks in it’s own database. Such an agent could be installed on a password-protected part of my website so I can access it from wherever I am. The problem is: I don’t have time to write an application my own. I have thought about such an application, I’ve got plenty of ideas how it should look but I’m afraid I won’t find time to code such an application. I repeatedly get emails from people writing yet another thesis on collaborative categorisation. Instead of writing about insights on the mental actions taking place while tagging – do the tagging world a favour and write such an agent. If there’s such a tag agent already, please leave a comment.

185186187188

After the presentation at the tagging workshop at www2006 we felt the need to give our paper a more www-friendly, I-don’t-want-to-read-through-those-theoretical-equation-flooded-papers face.

So, here you go: Automated Tag Clustering: Improving search and exploration in the tag space. To read this document you should have a clue what tags are about, you should also know some tag services as delicious or flickr so you can understand the limitations these services currently have. If you don’t want to read through the whole papers, the numerous figures give you a good summary. Finally, to wet your appetite, here a few excerpts of the document:

Currently tagging services still provide a relatively marginal value for information discovery and we claim that with the use of clustering techniques this can be greatly improved [from introduction]

The whole promise of collaborative tagging is that by exploring the tag space you can discover a lot of useful information you would not find with traditional search engines. When your information need is not well defined, the idea that you can explore and see what other people tagged with certain tags is very attractive. We believe that tagging will be able to reach a very wide audience only when exploration techniques will be effective. [from limited exploration]

Although a great visualization paradigm, we believe that with today’s tagclouds it is hard to find more than one or two tags to click on. Tags are not grouped, there is too much information, so that you find lot of related tags scattered on the tag cloud. One or two popular topics and all their related tags tend to dominate the whole cloud. For example, looking at the del.icio.us tagcloud, one would mostly see tags related to web design and technologies. This is because these topics are overwhelmingly more frequent than anything else. [from limited exploration]

Tag web2.0 nowadays is so popular and is combined wildly with anything. In fact this tag is so overused that if you look at tag bookmarks in the del.icio.us dataset, the most used cotag is web2.0[...]. Basing tag similarity on these numbers often doesn’t make sense at all. The similarity measure should be chosen so the popularity of a tag doesn’t affect the set of a tags related tags. Don’t cut the long tail. The success of blogs is driven by the importance of the long tail. We all know that it is crucial to support the niches. Tagging applications should empower the long tail too. If you just sort by popularity, you’d loose all those niches. [from choosing a similarity measure]

We’d be happy to get any kind of feedback on the article. Just post a comment to this blog post.

Edit (4 years later!): A few guys asked me about the source code: Source code with syntax highlighting, download.
You need kmetis to make this run, see usage() to see how it should be used.