Quarter Life Crisis

The second Web¶

2520 words on Software

There’s this popular contemporary phenomenon known as Web 2.0. It’s great at times and annoying at others. But should we love it or hat it? And why?

What is Web 2.0?

Ah, a good question to begin with. And frankly, pretty much everything could be Web 2.0. The silly ‘version number’ thing probably indicates that examples for Web 2.0 will be more or less geeky and the ‘.0’ part of the name warns you that they are more broken or less stable. That much is in the name alone – which possibly makes things look worse than they are.

Other aspects seem to be the actual networking of network sites (wooot, this sounds new and exciting!), audience participation (also known as people creating content for free or even paying fees to participate which companies then graciously host), the phenomenon knows as tagging and commenting, also known as social software and a general laissez-faire attitude of people vis-à-vis their valuable data.

Finally there’s the technical underbelly which includes loads of late 1990s web technologies such as JavaScript and using it to manipulate a web page’s DOM, potentially Flash for even more interactivity and a generous stack of distinct file formats trying to fullfil the very same task for the benefit of the vanity of some format author or another.

I’ll cherry-pick a few of those points and give my views on them. In some places things will look exciting in other depressing.

Technical Networking

One thing you see a lot with the whole Web 2.0 thing is that despite having many different sillily named services for your bookmarks, photos, listening habits, sound snippets and so on, you can usually extract the interesting data from them and include them elsewhere, on your web site, say.

In a way that’s not too exciting in a way as you’re essentially moving your data around and the little snippet of JavaScript to do that isn’t exactly rocket science. Yet, such data transfers used to be very hard and likely to break every now and again in the past. So while this isn’t exactly exciting new technology but just making existing technology sufficiently accessible, it’s not a technical revolution per se. But it’s still a great step forward.

It seems quite typical that what is perceived as great technical progress isn’t actually that great from a technical point of view. Frequently the technical basis for such progress has been around for a while and it just took ages for some businessperson being lucky enough to let the geek with the right ideas have their money.

Your Data

In my opinion the most questionable point about those new web services and sites is that in the end they derive their quality from your data. It’s your time and effort that makes the sites great. So you should get some benefit from that. For example the ability to export your data from the site in a way that gives both useful results, i.e. can be used to integrate the data in other services or applications, and is easy to do.

I don’t think that offering that kind of service is in the immediate interest of the Web 2 companies as it may loosen the tie-in they have on you as a user. That’s a tried and tested technique in the computer business of course. And it’s one that I’m seriously uncomfortable with. And while those recovery or backup technologies do exist for some of the Web 2 services, they usually are third party products and not part of the service itself.

And this isn’t just about jumping ship and wanting to use a different service. Young web companies can be bought or close down because they’re bankrupt. That’s not unusual and I’d just like to have a backup of my data in case that happens.

Tagging

Tagging is one of the hypes in the Web 2 world. And being the anal retentive person that I am, I surely like to have everything cleanly tagged. The problem is, however, that the situations where this is actually useful for me have been rather limited so far. Not only does the use of tagging require a certain discipline – I have to ensure that I use the very same tag if things are to be grouped together – it also has a very fragile balance to strike: that of usability vs. usefulness.

If the direct benefits of tags for me are rather low – and I’ll come back to what I suspect to be the reason for that in a moment – setting them up should better be mostly effortless. Which means that I most likely end up with sub-optimal tags. I probably won’t have a hierarchy of tags – one where I have setup a tag ‘Mando Diao’ say which automatically implies tags like ‘Music’ or ‘Sweden’ . Making the entry of tags easy seems to imply that they will end up being less expressive.

Another issue which tags need to master is their competition. Frequently when looking at the tags you see these days I have the impression that they are less useful than a simple full text search with a tweak or two would be. Remember Yahoo’s nicely filed catalogue of web sites in the old days? Google blew it away by giving better results from just looking at the text and links. Many of the Web 2 sites have to work on that aspect urgently. Some of them don’t even seem to have a meaningful or powerful find feature implemented at all. And those which do will often have a horribly slow or weak one.

The final charme of tagging comes from the ‘social’ aspect of many Web 2 affairs. Other people can tag things as well and – so the story goes – you can use tags to explore similar things done, seen or heard by other people. This sounds all great and exciting. But when using it, I only thought that the results I saw were mildly interesting at best. This may be due to the sheer mass of material on some sites which drown the bits that you may actually be interested in.

It could also be because people have vastly different tagging behaviours: Some people just use none to two tags for things where others use half a dictionary. Some people refer to black and white photography by ‘black and white’, possibly without spaces, while others go for ‘bw’ only, or perhaps ‘b/w’, then people might be German and go for ‘sw’ and so on. The ‘correct’ tag to choose for something is highly personal. And trying to build a collaborative directory of things from personal tags looks like a highly non-trivial task to me. At least if the results are supposed to be reliably good. The Web 2 people have a long way to go in that area.

Finally, I keep thinking that there ought to be some kind of standardisation in certain areas of tagging. It’ll probably be a challenge to come up with a good list for doing this, but being able to correctly indicate things like time, place, language or relation to other people would make things much better than they are these days where all we have are a bunch of locally applicable quasi-rules for doing this.

Social Networking

Wherever you hear about the second coming of the Web, you’ll quickly run into the word ‘social’. The sharing of tags described above certainly is an aspect of that. But there’s more. Most significantly the number of views of your œvre seems to be essential for the Web 2, as is the ability of people to leave comments.

To me most of those comments look like they are OMG, LOL! or Nice framing and fantastic contrast. And I find those ridiculous – except if there’s really something excruciatingly funny, which mostly there isn’t. In particular, actual discussions or substantially helpful remarks seem to be rare. It’s more about communal navel gazing, showing off, vanity and so on. If you wish, the word ‘social’ gave it away.

And then there’s the bell curve. Err, stop, not the bell curve but rather that exponential curve you see whenever people start talking about the ‘power law’. I’ve failed to get the point of it, but the essence of it seems to be that despite all the openness and freedoms that the system – or social software – gives them, people still tend to converge to the most average contributions. Which of course are labelled ‘most interesting’ as a consequence.

I yet have to figure out whether this is a bad sign for the Web 2 systems or rather for people. And I fear it to be the latter: We are in a rare situation where all the content is created by people just like yourself. And in a situation where there isn’t (is there really not?) a publisher or other entity to make decisions regarding your taste and interests for you – you can just look at everything and decide what you like best. And despite that freedom to discover things for yourself there’s still a strong urge to converge with the masses. Odd. And sad in a way as the ‘long tail’ contains so many great things.

I also see another weakness of the current state of the social software in this: With the algorithms and statistics they use, pointing people into that general direction of what people now refer to as ‘interestingness’ may be the best they can do. And that’s a bit lame as it really questions the neutrality of the services.

An interesting example in that context could be the feature introduced by delicious recently that there will be suggestions for the tags you use on a link if that link has been bookmarked before by others. In a way that’s really cool as it makes tagging even easier and leads to people using more coherent tags on the same site. But on the other side I wonder what this does to the free-form approach that used to be there and to which extent I cannot develop my own tags with this scheme in place as I’m influenced by what other people decided before getting my own turn.

Finally, being ‘social’ means that people are going to try playing social games. People want to be interesting and popular. They want their favourite websites to be known, the photos to be appreciated and their opinions to be heard. And there’s nothing wrong with that. In fact, it’s really helpful, interesting and educational in many cases. However, numerous people seem to play those systems. They add gazillions of tags to their contributions. And, if they exist, put them into many of the existing grouping schemes.

That may in part be caused by the endless duplicates you have in the tagging and grouping schemes, but I suspect it’s also caused by people realising that the way to become ‘popular’ is showing your goods to as many people as possible. In fact, there seems to be a whole industry of writers by now who write ‘reports’, ‘analyses’ or simple blog posts on how to best become ‘popular’. And the bottom line of these seems to be that you mostly should be busy making contacts and getting the word out. I’m inclined to consider this to be a modern form of spamming, but others consider it a good way to success – whatever that word means in the playful second web.

Conversations

If I recall correctly, the cluetrain manifesto got people started on the word ‘conversation’ with the claim markets are conversations. That was a great idea and a good point to make. But it killed the word ‘conversation’. Ever since I have the impression that the word ‘conversation’ can’t just be used in the second Web without that idea lurking around.

Somehow a ‘conversation’ has become a holy grail as a consequence and is always ‘good’. Someone wanting a ‘conversation’ wants to be perceived as an open minded and equal participant, even if he is a bully who just wants to voice his odd opinions and ignore all the bits of the ‘conversation’ that follow which don’t agree with him. That’s not having a conversation in my book.

And as the cluetrain book dealt with businesses to begin with, I wonder to which extent its reasonable to transfer ‘facts’ from there to the real world, i.e. the world with people in it.

Webstuff

While I’m not entirely positive about many of the Web 2 effects, I am really enjoying the user interface aspects of it. Back in the 1980s when I was playing computer games with my friend Martin on his brothers Schneider CPC or my Atari ST, he made one of the cleverest remarks I have heard regarding computing: I think I’ll invent a computer that doesn’t need to load.

It’s so spot-on! From the C-64 to pretty much any Photoshop version to web pages to the iPod, loading has always been a big annoyance. But one you had to live or at least deal with (although I consider it bad software design in the iPod but that’s another story). And whenever systems became more powerful, so did the software they had to run, leaving us in a state where the loading time always remained noticeable.

That’s particularly true on the web. Internet connections, computers and browsers became faster over years. But site designers have been keen on adding ‘features’ and many images, animations or scripts to their creations at an even faster pace. And all of those for no substantial benefits to the user but mostly to add popup windows and more annoying ads.

And that’s an area where the second Web excels. In their stream of new thinking they saw that we’ve been having the DOM for years and that it is accessible by scripting. And they used that fact for good: They removed many annoying full page loads and added the ability of pages to submit and present data changes in one really quick step. Almost like a proper GUI application. This made the web’s interactive usability a bit better. Things which used to require many clicks and reloads can now be done in a few smooth moves. And I love that every single time I use it.

I should note here that this isn’t all great and shiny as using those interactive techniques will require significant extra effort by the people designing the sites to ensure that crucial details of the web such as linkability and accessibility are preserved (or exist to begin with). As far as I can tell not enough effort has been made for those points. And it’s not clear (cue my pessimism) whether it will ever happen.

Conclusion

A significant number of weaknesses and real dangers for user data remains in the second Web and I doubt that they will be solved as doing so would be more for the benefit of the user than the Web 2 company. But as a learning experience alone, looking at the second Web will be worth it.

April 26, 2006, 0:14

Tagged as software.