tailieunhanh - Understanding WAP Wireless Applications, Devices, and Services phần 9

228 mà có thể hoặc có thể không được độc quyền. Quan trọng không kém, mỗi của subchannels có đặc điểm hoạt động khác nhau (ví dụ, kênh trở lại của iTV có thể là vĩnh viễn mở hoặc không), mà cần phải được xem xét khi thiết kế các ứng dụng. Một tình huống tương tự cũng tồn tại với các kênh điện thoại di động. | Page 227 Figure Typical stovepipe solution to multichannel support. Figure A shared business object multichannel delivery architecture. Page 228 which may or may not be proprietary. Equally important each of the subchannels has different operational characteristics . the back channel of iTV may be permanently open or not which needs to be taken into consideration when designing applications. A similar situation exists with the mobile channel. As we have already seen we are faced with the prospect of dealing with SMS and USSD messages voice or DTMF-based IVRs SIM toolkit WAP PDAs and laptops. While there is a good chance that WAP will become the dominant technology for mobile devices it will be several years before it is ubiquitous in the installed base of mobile phones. In the meantime we are faced with a heterogeneous and fairly diverse environment. It is therefore probable that the mobile commerce provider will have to be content with at least two of the above technologies. In this world with many subchannels to support it is desirable to eliminate as much as possible the business functionality from the channel-specific parts of system. In the current model the generation of the presentation layer from the underlying business objects is done on a channel-by-channel basis. While this code is not extensive in a thin-client environment there is nevertheless duplication and replication of logic. For larger new developments it will probably be worthwhile to introduce another tier in the architecture. Figure shows a new channel-neutral presentation layer. In this case rather than trying to postprocess the Web interface into other formats all the channels derive their presentation layers from a new channel neutral format CNF . In this model a single logical process constructs the CNF for all channels. This is then parsed and rendered for each sub channel. In this way the channel-specific code becomes more stable and does not need to be changed when