MWS:Challenges
From OpenSource
Contents |
[edit] Cost of Data Transfer
There are two types of cost; monetary and battery.
The monetary cost may go away by itself over time. There are already operators that offer flatrates for data-transfer and in general the cost/byte is likely to go down and not up. Of course, just like many ISPs require some specific agreement before you can host a global website, operators might also introduce some specific requirements for mobile webservers. Not that the situation is particularily similar; for a huge audience you want a regular website, not a mobsite. Further, if WLAN is used instead of cellular, then the monetary cost issue disappears completely. The battery cost issue will not go away by itself but may actually get worse as mobile devices evolve and are equipped with more functionality that all want their share of the available energy.
In both cases we need to distinguish between the case of the web server just running and the case of the web server actually being accessed. In the former case, the cost issue can basically be ignored. There is some activity going on all the time in order to keep the terminal connected to the gateway, but the cost that causes is negligible.
When the mobsite is accessed, it does affect the battery lifetime. But so does browsing the web and talking with someone. So the real question is whether the functionality provided by a mobsite outweighs the cost in terms of increased battery consumption and phone bill? Sharing large amounts of data to a large number of people from a mobsite will not be feasible, but that is not the purpose of a mobsite anyway. For that purpose a regular website is a better alternative.
The tricky issue is that the moment traffic reaches the terminal there is cost, even if the access subsequently is rejected by the web server. In order to deal with this, you need an intelligent entity at the border between the wireline and wireless network that is capable of controlling the traffic.
[edit] Access Control
To be added.
[edit] Security and Privacy
To be added.
[edit] Availability
To be added.
[edit] Performance
To be added.
[edit] Community Aspect
To be added.
[edit] Litmus Test
Ok, so on the one hand you can do a lot of things with a webserver on a mobile personal device, but on the other hand there are all sorts of issues that need to be taken into account. A good litmus test for whether some concept makes sense for a mobsite is to ask:
So why don't you simply upload the stuff to a regular website and share it from there?
Without a good answer, perhaps it is better to use a regular website. Indeed, the best mobsite concepts are likely to be those for which the question is meaningless. But then again, even when the question is relevant, a perfectly valid answer might in some context be I don't have one and I don't trust any public sites.
An example of such a concept is a web gallery. Given HTTP connectivity, it is trivial to create one on a camera phone. In that case the question could very concretely be So why don't you simply upload the picture to Flickr? Yes, if the intention is to share the pictures with the world, it does not make much sense to share the pictures directly from a web-gallery on the phone. The bandwidth is likely to be insufficient and the costs prohibitive.
However, if the intention is to share pictures only with members of your family and/or a couple good friends, a web-gallery on the phone might be the simplest way to achieve that. At least it is likely to be significantly cheaper than sending MMSs.
