On several reasons for the success of Hibernate framework

The fast release of the release

Maintain an active development speed, often in release, or even within days, to the next version of the previous version. This is the best way to keep the software away from bugs, and to reassure users that Hibernate's development is very active, and that it also has the benefit of discovering which features users really need.

Regression test

I think the entire Java community is now very focused on automated regression testing. If the functionality and design of the software are significantly modified, a comprehensive test suite is too important for software maintainability and stability. We should be aware that if a new feature of the software is not tested for regression, we should not be able to do it.

Make the best of a function

If you don't do it, do it, you must do it best. We do not do the best function, we do not do it, throw it to other software to do it.

Avoid excessive design

Waste a lot of time and energy to abstract software functions and expand the flexibility of software, but also spend more time to solve your users face the actual problem! Simply put, the software is most important to run, do not try to solve your users do not care about the problem. Even if your software design is not elegant and not related, anyway, or initial stage. In the future can be refactor, you should pay attention to the problem is the timely use of useful functions to do.

Centralization

Before you have to decide on a democratic vote, at least you've got the outline of the software ready. Software development needs to be led by one or two enlightened people to keep software development consistent without creating too much disagreement and to ensure that the development team is focused on the best possible functionality. I think that the biggest risk of OSS software is that opinions are not unified, the stall is too big, the result is not doing anything.

Document

Nothing is more important than a document. If your users do not know that your software has such a function, it is tantamount to not this function, simply remove it, save the source code to increase the complexity.

Avoid standardization

Good standards can bring software interoperability and portability, and bad standards can stifle software innovation. The best software is in the continuous attempt, continuous error, continuous accumulation of experience in the process of production. In fact, the standard is often more close to user needs.

Run the hibernate in 10 minutes.

Potential hibernate users have lost their interest in hibernate for the first time when they downloaded hibernate, they could not spend half an hour to install, configure and troubleshooting.

Our slogan is that new users (assuming they have enough JDBC knowledge) will run the Hibernate demo within 5 minutes, and they can write the simplest hibernate program of "Hello World" in 1 hours and run normally.

Responsibility of developers

Users are invariably confronted with problems, and the development team has a responsibility to provide help. The user lets us know the vulnerability of the document, the user let us know the test case of small bugs. In addition, no user to use our hibernate, we also develop it to do, is not a waste of time!

There's a joke about bugs: users don't mind finding bugs in new features (the translator: Windows users seem to be so), as long as you get rid of bugs quickly. "Responsibility" means that the bug fix should be within 1 weeks. It is an ideal goal to have an average of 24 hours or so to get the bug report to the bug fix code and submit it to CVS.