My views on tapestry

I have recently studied tapestry4, and tapestry5 is under development. I should be able to release it within the year!
Tapestry5 completely restructured the entire architecture. Currently, tapestry5 draws on the advantages of many other outstanding projects, including ror and Google guice.

I have seen positive comments, such as: "components-oriented web frameworks like apestry have indeed overturned all my Web Framework experience.
"
Of course, there are also negative comments, such as: "the most annoying combination of code and annotation functions, never think this is progress"; "although you do not like the page configuration file of tapestry, however, I hate annotation, and write configuration and code logic in a pot of stew. To change the verification conditions, I have to re-write the class file.
The worst part of the hwl idea-developed tapestry framework is that you have to do whatever you like template, verification, IOC, various page components, session management, upload/download, etc, others' existing excellent architectures are not easy to integrate into tapestry, which is open and interactive. I think only the ognl language can be learned in tapestry. This is not developed by hwl.
It is better than just the development idea of spring. Don't reinvent the wheel, just consider how to use it better. It seems that there is no fundamental change in T5 development ideas. I plan to switch from T4 to S2 or ww and continue to watch t5 and T6 ."

However, sometimes it is necessary to re-invent the wheel, as shown below: "Even if the existing solution is worse, it has accumulated a lot of code, development efficiency is still higher than the new solution from scratch. If we only focus on the commercial value that software can provide, we may always use off-the-shelf things instead of making another things. In the past, potian criticized sap's design as falling behind and totally out of line with object modeling. Such criticism may be correct, but think about how you can accumulate the amount of Code such as SAP? Is it easier to implement the same business based on sap or your own erptao? What if it's a business process that you never thought of before?

On the other hand, if everyone only considers the problem from this perspective, the technology will never develop. In the long run, the backward architecture provides limited commercial value. For example, the cost of OLAP-Type Analysis on SAP is very high, because the basic database design of SAP has never considered Analytic Applications. So my friend Xel said, "Sometimes it's about re-inventing the wheel !"

In short, I think tapestry may not be the smoothest path, but it is definitely a desirable path.

Let's perform web development more freely.