VCL dead, Rad dead (6)-conclusion and Prediction

VCL is dead, Rad is dead

-- Topic that is not fully spoken in sd2c

<-- Previous section

6. Further future
(Limited liability prediction)
-----
Next, we will see this tool that is more suitable for organizing teams in the field and developing them. However, this tool is no longer expected to be integrated.
Implementation technologies in various fields (note that I am not talking about "Development Technology"), but provide Inter-Domain delivery standards. Or more
Provide deliverables directly. More specialized companies have received attention (for example, Macromedia), and large manufacturers have begun to purchase
And more companies in the exclusive field to integrate their business. Larger platform-based products will emerge, remote, distributed,
The computation theories and solutions that can be migrated have been popularized. At the same time, more specialized tools and
Dedicated talents: the overall scale of the project is expanded and implemented by multiple teams (like the engineering contractor team), while a single team
The personnel structure is more complex, but the team size is still kept below 10, with 15 ~ The upper limit of 20 people. Similar to testing and other technologies
Will appear as a field, similar to the current construction project acceptance and supervision will also appear. These exclusive fields are still
However, it has its unique standards, technologies and tools, and provides independent deliverables.

For the whole project, Rad completely died. It may be similar to the appearance of Sand Table/model making companies in the construction industry,
It has become a part of the product process, but there is no rad tool that can be integrated in multiple horizontal split layers. Prototype
The Rad process and iterative RUP are slowly exited, as no one can provide the iteration of a building.
Similar to Cheng, the complexity of the project has made it far from being able to control the cost of a single iteration. In this scale level
The control of the interface determines the stability of the system and whether the development can be continued. Therefore, the architect must complete the entire process and the architecture
For example, a precise data I/O interface may require specific architects to confirm and demonstrate the responsibilities. Same
In this case, local failures or failures will be discovered and verified earlier in the system, and rework will become the norm in the local area,
But it does not exist-or directly let the entire project fail.

Projects are increasingly dependent on experience and technical models brought about by experience. For example, the tower architecture I often mention is not
The outcome of brainstorming by an engineer or expert in a field is the experience summarized in practice. Enhanced project reproducibility,
The replication scale and environment are simpler (too complicated rules and interfaces are difficult to copy), and all of these depend on horizontal partitioning.
The conciseness of the layer interface.

The reuse and reconstruction of modules are still long-term and will never die out. The finer granularity of reuse will inevitably go from the current object to the business
Component Reuse, but it is still based on Basic Object reuse without business logic. As a component
The emergence, extinction, and reproduction of a long history. More layer interfaces will appear in the form of protocol standards.
On, the layer-based interface logic components will emerge in batches, "data" is still their only constraint and programming nature.

We will have a new attempt to compute the model, but the essence is still consistent with today's. The functional language will go to the front-end.
Become an Inter-Domain adhesive (although it is already the case in some areas), the language to be determined will be placed at a specific level
Brilliant colors (or this is also synonymous with the field language), and experts who have mastered the associated development tools at multiple levels
Program-independent experts with field expertise will become a hot topic. For the entire system, the former subject is to implement a specific business
Business logic, while the latter focuses on data definition. However, we can imagine that not many people use XML tape as these
Data Definition carrier. The exclusive data format will be the first choice for inter-layer interaction.

Finally, from the perspective of engineering practice, within the past 20 years, I don't think there will be a book called "The creation of software engineering".
. Of course, some gimmicks or paper sales do not boast.

==========================

Some of my related text links:

About "VCL is dead, Rad is dead" a reader asked http://blog.csdn.net/aimingoo/archive/2008/12/22/3583348.aspx

VCL dead, Rad dead (inserting) http://blog.csdn.net/aimingoo/archive/2008/12/07/3467787.aspx