| Re: [birt-dev] Java Event Handlers |
|
Don, I am not sure which documentation you are referring to. I am looking at BIRT 2.0 on-line help and it does not mention the adapter classes, it only references implementation of the interface. Perhaps you are referencing the BIRT book? Also, I agree with your assessment of the advantages of using the interface directly, but the direction the PMC voted on was to opt for removing upgrade issues and recommending the use of the adapters. Scott Don French wrote: There are also some advantages to implementing an interface rather than extending an adapter class. When you implement an interface, Eclipse generates the stubs for you, allowing you to see not only all the methods in the interface, but also the arguments to the methods you are about to implement. When you extend an adapter class, you have to refer to the Javadoc or other documentation to find the method signatures. Also, if you implement the interface and the interface changes, the compile errors let you know what you need to change. Whereas if you extend the adapter class and the interface has changed, you will not get compile errors, but neither will you have functioning event handlers, and the report will run but without the desired results and the reason may not be so apparent. The documentation mentions both techniques and explains the advantages and disadvantages of both. Is there some reason that we need to go further than that? Don French Developer Communications -----Original Message----- From: birt-dev-bounces@xxxxxxxxxxx [mailto:birt-dev-bounces@xxxxxxxxxxx] On Behalf Of Scott Rosenbaum Sent: Monday, February 06, 2006 1:00 PM To: For developers on the BIRT project Subject: Re: [birt-dev] Java Event Handlers Petter, I brought this up to Wenfeng at the PMC today. After discussing the alternatives, we decided that the recommended approach will be to use the adapters. Apparently this is in keeping with Eclipse standard practice. I will send an email to Janet in docs to change the BIRT documentation. In addition, I will work with Jason Weathersby to make sure that the web site highlights this method. If anyone has any additional thoughts or comments on this, please let me know. Scott Rosenbaum BIRT PMC Petter Ivmark wrote: |