Brian Zhang Larsen’s Blog

In pursuit of distraction from my scheduled choirs this Sunday, I stumbled across a celebrity match website. The idea is to upload a picture of you, and the site analyzes the picture in order to come up with a possible genetic match, or at least an answer to which celebrity you look like the most. It turns out I possibly could make money being a Vincent D’Onofrio look a like at parties or other social events. Ok, the name didn’t say me anything either, but you have probably seen him in various supporting movie roles. But after I tried a couple of times, I often resembled Christopher Reeves the most. Especially on the Pictures we had taken in china, were we were dressed up in traditional Clothing. It also turns out that both my mother and Vivian’s grandmother has a remarkable resemblance with Meg Ryan, and that if I was black, I would look like Pharrell Williams.

After reading Jesse James Garrets: The elements of user experience, something has been annoying me slightly. Why is it that he, just will not stick to standards when diagramming. He insists on using his own modified standard, which in some circumstances can be quite disturbing.

Maybe it is because he simply never heard about UML, but I doubt it. Maybe it is because he doesn’t feel like he can diagram everything using UML, which there may be some truth to. But why does he use misuse UML notation then, instead of extending it? It is the de facto standard for modeling most software related problems, which in my opinion would include websites and applications. In fact I have stumbled about this lack of standards several times in my job as a web developer, and I think that is a serious mistake. I am not a UML or standard fanatic in any way, I just like the way that it eases my work to be able to immediately understand another developer/architect or other diagrammer’s work, without an explanation, or without me using time on understand the person’s intentions.

I am not solely against the way Garrets diagram, in fact I think there are many clever ways to diagram problems, that I would agree could be hard with UML as it is, like the differentiation between decision points from the system or the user. This could maybe be done with swim lanes, but this is not always a nice and clear way to show an otherwise simple flow. There are just some fundamental problems that I have a hard time coping with. Take for instance the way he diagram decision points below.

In my head, and according to UML, the flow split should be AT the decision diamond, and not at some point afterwards. It confuses me, to see a split like that, and this repeats itself at different kinds of split points in his visual vocabulary. Another disturbing thing is the way of grouping flows.

Why would anyone illustrate a flow merging with another like that? I would strongly prefer that the flow arrow from the left entity directly went to the destination entity, instead of this merge kind of thing. These flow has nothing else than their destination in common, why not keep it at that on the illustration.

Other than these more or less minor mishaps, I think the idea of trying to standardize web flow diagramming is solely a good thing. I recommend you check out JJG’s suggestion and make your own judgment. You can see his visual vocabulary here, or check the quick reference (PDF).

Last year I talked to a Chinese friend of mine about the internet growth and the need for new Chinese marketing channels in china. The thing is that a lot of parameters suggest that the Chinese online market will explode, which it in fact already has. Later I talked to my father, and independently he had also done some slight research of this interesting and potential very lucrative market for the purpose of investing. His main interest was the large search engines and portals as baidu.com, which has seen immense growth lately.

The thing is that now a lot of foreign companies have found that China is not just an interesting place to look for products to export, or establish production facilities. It is a very interesting market to sell to too. The amount of strong buyers is rising steadily, which calls for stronger, better and more modern marketing channels.

I am unaware of how precise the “online marketing” market is in China, but I would expect it to grow very dramatic at the moment. The service sector “only” amounts to about one third of the Chinese economy. Due to increasing trade, not just exports, and increased privatization of the sector, I would expect a higher demand for domestic services.

The amount of internet users in china is growing like no other place on earth. That fact would itself support the thought that online marketing is growing too. I think that if there is ever going to be another klondyke.com era it would be starting right now, and in China. There are so many new businesses appearing daily in China, and they all more or less need to deliver their message. The online Channel is perfect for that, and is a relatively cheap instrument compared to TV commercials. It also gives the company the opportunity to delivery whatever message they want, when they want it. Once the channel (website) is established, the marginal cost of that channel will decline, because what is costly, is the implementation. The maintenance of content has relatively low cost, because neither expert knowledge nor production is needed. And unlike print ads, or TV commercials, the message can be changes from minute to minute.

According to Alexa, 3 of the top 10 sites are Chinese at the moment. All of them are targeted at the middle country, and none has a wider international target. But maybe that will be the newt step?