I would argue against anything that would increase the possibility of subtle errors. If a solution like the one you suggest could be done while "fireproofing" against problems when classes change packages, then I wouldn't be so against it. I just hate having to track down subtle bugs in a system that were caused by attempts to make things easier (e.g., most of Windows).
As far as the searching algorithm you suggest, I'd have to say that I don't have anything against its usefulness but I would have an issue if it forced a lot of computation to handle the searching. As it is now in WW2, I understand that a lot of the getText() functionality is maybe a little slower than it could be because it searches from the current class on up through the package hierarchy (e.g., if no text resource is found for "com.package.sub.MyClass", it searches in "com.package.sub", then "com.package", etc.). I appreciate these features on the one hand but since I find the alternatives the avoid to be not that much more difficult, I'd rather avoid the performance hit if possible. I know we're talking ms here of difference but it adds up. WW1 was pretty fast, especially with velocity, and I've seen disappointing results with ww2 thus far. I would think it would be best to concentrate on performance tuning at this point, now that it's in Beta, and save convenience enhancements for later. Of course, it's not up to me. Oh, and for the record, even though I'm disagreeing with you on these points, I do agree that efforts should be taken to make ww2 easy to adopt. It's just that I'm interested in making sure that those efforts don't interfere with how more advanced users do things, by introducing the chance for subtle errors or hurting performance. -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of boxed Sent: Sunday, August 17, 2003 4:30 PM To: [EMAIL PROTECTED] Subject: Re: [OS-webwork] Simplicity of WW2 - Practical ideas Mike Cannon-Brookes wrote: >You can already test actions to setup xwork.xml - just instantiate the >object, call your setter methods and run! > > Are you trying to scare users away now? I was talking WW2, not XW, so a web-based interface where you can get immediate feedback in the environment you're gonna use seems reasonable. Writing even more code is hardly user friendly. >People doing J2EE understand XML, they have to. All descriptors are >XML. Xwork.xml is not _that_ complex for a hello world example, most of >the elements are optional. > > I think you overestimate "people doing j2ee". Maybe you need to come and spend a day or two in #java and listen to the questions people who are trying to understand WW (1 and 2) are asking. Many are new to java, people just don't learn programming the logical way by working upwards, they start in the middle and work outwards. Sure, it's not "_that_" complex for a hello world app. It is however an order of magnitude more complex than it was in WW1 and _already that level_ confused a lot of newbies. If we want WW2 to be used and to be able to compete with things like Struts, you can't have this kind of elitist thinking. If you have a competent person who knows XML and all that, the odds are he's already used to Struts and we do NOT want to make him give up on his evaluation of WW2 because it's a big hassle. We want to give the view (illusion if you will) that it's easier than Struts, and not just a little easier, a _lot_ easier. No one will change if they don't see their time invested will be worth it down the line. >However, there _is_ a problem with WW2 at the moment that if a view is >not found, no debug page is shown. I think it should be ("action >returned "input" but not "input" view found). > > So you agree with the full debug output? Fine, I still hold to my view that if we give the illusion of ease of use, people will believe it is easy to use. Call me crazy, but I think people believe what they see. I'd also like some arguments against my ideas other than "it's not so complex" and "I don't think so". You began the email with "I have to say that this is a _bad_ idea." and then proceeded to give zero technical arguments against the idea. I am fully willing to throw my idea in the garbage bin, given actual problems with it. You have not shown any. Anders Hovmöller ------------------------------------------------------- This SF.Net email sponsored by: Free pre-built ASP.NET sites including Data Reports, E-commerce, Portals, and Forums are available now. Download today and enter to win an XBOX or Visual Studio .NET. http://aspnet.click-url.com/go/psa00100003ave/direct;at.aspnet_072303_01/01 _______________________________________________ Opensymphony-webwork mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/opensymphony-webwork
