Jonathan, I don't see that your argumentation and attitude would lead to any positive reactions or development either for Velocity or for FreeMarker. There is room for more than one template engine, at least as long as JSP is what it is today.
We've found Velocity exactly the right tool in projects where we need to integrate a clean and easy to learn, yet effecient, content processing tool to a framework where content is developed and maintained by people who are not programmers, but talented graphic artists and document writers familiar with XML and HTML. The framework developed by Java programmers provides a full set of additional tools to access business logic of the system in a controlled way through Velocity's tool API. The final product must be stable and maintainable several years after its release and training of content producers is one of the most expensive cost items in such a project. The number of new versions and additional features not directly needed in template processing is not an issue, but stability, reliability, backward compatibility and, most of all, a clean and well-thought architecture.
Well, I see the political content of what you're saying. Or the diplomatic content, as in... "Why can't we all just be nice??"
But on a purely technical and logical level, what you're saying seems somewhat ridiculous. We obviously can't compete with Velocity in the backward compatibility department simply because the most perfect backward compatibility is surely just to have no further development. Then you always have perfect backward compatibility. (Even with the bugs in previous versions...)
Moreover, nobody ever forces anybody to upgrade to the latest version of FreeMarker. Anybody can achieve the same level of backward compatibility by simply not upgrading whatever they are using. So, to tout this as some kind of domain in which Velocity has some kind of advantage strikes as a bit absurd. After all, the issue that has been rekindled as a result of my original response in this thread is that Velocity development has come to a complete stop.
FreeMarker is too complicated, illogical and changing too rapidly that we could apply it in such projects. But it's a useful tool in projects with a tight schedule, smaller amount of resources and shorter lifecycle, where we don't have time to build a full framework around the business logic of the system, but can apply FreeMarker's extended features directly instead.
I consider that high praise. You use Velocity when you have the time and resources to do things in the properly pure MVC way. You use FreeMarker when you actually need to get things done in some reasonable amount of time. I think that's a high recommendation for FreeMarker.
Also content producers must be more experienced in programming to understand namespaces, customized directives, wrapped objects, etc.
That's true, but as far as I can see, this argument is also flawed. You see, nobody has to use those features. They are available for people who want/need them. Specifically, the presence of namespaces, i.e. the ability to define different sets of macros in separate namespaces so that variables can be defined without any risk of namespace collision -- the presence of the feature adds no complexity burden on people who do *not* use it!
In general, if you use the subset of FreeMarker functionality that is available in Velocity, it is no more or less complex than Velocity. And, of course, if you use a given version of FreeMarker and never upgrade to a later version, you achieve the exact same level of wonderful backward compatibility that Velocity users enjoy.
So why not let Velocity and FreeMarker to follow their own design philosophies and people will have alternatives to choose from based on their different demands?
This is all nice and politically correct, Ilkka, but the fact remains that, to simply stop all development does not constitute a "design philosophy". The fact that no code has been committed to the Velocity CVS for about a year is simply not justifiable in those terms.
It is up to the Velocity community to figure out what they're going to do about this, but I don't think you are doing them any favors by encouraging this nonsense about how the complete lack of any ongoing development can be justified in terms of "design philosophy".
It obviously cannot.
Best Regards,
Jonathan Revusky -- lead developer, FreeMarker project, http://freemarker.org/ FreeMarker-Velocity comparison page: http://freemarker.org/fmVsVel.html FreeMarker 2.3pre4 is out!
-- Ilkka
Jonathan Revusky wrote:
Nathan Bubna wrote:
Jonathan Revusky said:
>> ... cut cut cut ...
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
