Hi guys, I have started working on this issue by myself. My first (and simple) approach is already working with the code packaged with 0.5.2 version. Now I'm trying to apply it to the current code in SVN (BTW, the developer's doc still points out to CVS) but I'm having some problems with the Maven 2 build (I guess nothing serious :)
This first version doesn't reduce the total code of the class, just refactors the code of the startElement method moving out some parts to private methods (one for each state). I don't like this solution very much because the generated code is still huge, and now I want to optimize it, but I need some more time because it affects not only to the code generation module but also the way the XSD is parsed and validated. I hope that in a couple of days I will be able to have something working. If somebody has any advice or idea I'm glad to hear it! Best regards, Jose Luis. 2007/7/31, Changshin Lee <[EMAIL PROTECTED]>: > > OK, I see. Did you guys start to work on JaxMe to solve the issue > then? Have you ever heard from Jochen about it? > > Cheers, > > ias > > On 7/30/07, Robert Reeves <[EMAIL PROTECTED]> wrote: > > Some of our customers have a problem with CDDL. Also, we must be Java > 1.4 compliant due to other customer requirements. > > > > Robert > > > > -----Original Message----- > > From: Changshin Lee [mailto:[EMAIL PROTECTED] > > Sent: Sunday, July 29, 2007 3:32 AM > > To: Robert Reeves > > Cc: Jose Luis Huertas Fernández > > Subject: Re: Refactor startElement and get paid > > > > On 7/29/07, Robert Reeves <[EMAIL PROTECTED]> wrote: > > > For our purposes, the Sun JAXB implementation has two problems: > > > 1. It uses a Sun license, which is a problem for some of our > customers > > > > Sun's JAXB 2 RI is licensed under CDDL, which isn't compatible with > > GPL. In particular, if you just use it as a component of your > > software, it means that your software is considered as "Larger Work" > > and you can license your work regardless of CDDL. If you can specify > > such a problem from your customers, I think I can forward the problem > > to someone from Sun who you can consult further. > > > > FYI, TmaxSoft (where I worked) adopted Sun's JAXB 2 RI in their > > commercial product (Java EE 5 server). > > > > > 2. I believe it requires Java 5, right? > > > > Yes, it does. > > > > Cheers, > > > > ias > > > > > > > > > > > > > > > -----Original Message----- > > > From: Changshin Lee [mailto:[EMAIL PROTECTED] > > > Sent: Saturday, July 28, 2007 4:46 AM > > > To: Jose Luis Huertas Fernández > > > Cc: Robert Reeves > > > Subject: Re: Refactor startElement and get paid > > > > > > Hi all, > > > > > > Have you guys tried JAXB 2.1.4 available at https://jaxb.dev.java.net/ > > > ? I tested the "test.xsd" and found that generated artifacts from JAXB > > > 2.1.4 got compiled without problem. I guess your wish limit (250) > > > might not cause a problem as well. > > > > > > You can ask me further about this issue as I'm familiar with both > > > JaxMe (actually I'm a committer) and Sun's RI. > > > > > > Cheers, > > > > > > ias > > > > > > On 7/28/07, Jose Luis Huertas Fernández > > > <[EMAIL PROTECTED]> wrote: > > > > Perfect, let's see then if Jochen or any other Jaxme developers > (although I > > > > also think he's a bit alone) want to make this contribution. > Otherwise I > > > > would be happy to work on it! > > > > > > > > As far as I know, my company was thinking in Sun JAXB RI as the > replacement > > > > implementation, so I don't think they are aware that it suffers the > same > > > > problem (I wasn't aware too). Maybe it's possible to split the > costs, but we > > > > will have to wait until monday to know it. Anyway, I don't think > it's worth > > > > to discuss the details now until we know if Jochen is interested. > > > > > > > > Meanwhile I will be taking a look at the code to be able to estimate > the > > > > effort more accurately. Even if I'm not chosen, I will try to do my > bit! :) > > > > > > > > Best regards, > > > > > > > > José Luis. > > > > > > > > 2007/7/27, Robert Eric Reeves <[EMAIL PROTECTED]>: > > > > > > > > > > > > > > > > > > > > > > > > > Cool. We'll wait to see if anybody responds, if you wish. > > > > > > > > > > > > > > > > > > > > How much time would you need to complete this refactoring project? > > > > > > > > > > > > > > > > > > > > As far as your company goes, keep in mind that the Sun JAXB > implementation > > > > (both 1 and 2) has the same problem. Do you know of any other > > > > implementation? > > > > > > > > > > > > > > > > > > > > If not, then maybe your company and my company can split the cost. > You > > > > could work on it at work, and we would subsidize some of the cost > for your > > > > company. Then we both benefit! > > > > > > > > > > > > > > > > > > > > Thanks! > > > > > > > > > > > > > > > > > > > > Robert > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > From: Jose Luis Huertas Fernández > > > > [mailto:[EMAIL PROTECTED] > > > > > Sent: Friday, July 27, 2007 11:39 AM > > > > > To: jaxme-dev@ws.apache.org > > > > > Subject: Re: Refactor startElement and get paid > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Hello Robert, > > > > > > > > > > we are facing the same problems here. These days we are thinking > if it's > > > > worth for us to fix this "bug" or it's better changing to another > JAXB > > > > implementation. > > > > > > > > > > I think our company is not willing to invest our time refactoring > Jaxme > > > > and probably they will move to another implementation. Anyway I was > thinking > > > > to make this changes by myself at home and this could be a good > incentive. > > > > > > > > > > But, to be fair and honest, people who has contributed directly to > the > > > > Jaxme development deserve this opportunity much more than me. Just > in case > > > > that no Jaxme developers wanted to make these modifications contact > me if > > > > you are still interested. > > > > > > > > > > Best regards, > > > > > > > > > > Jose Luis Huertas. > > > > > > > > > > > > > > > 2007/7/27, Robert Eric Reeves < [EMAIL PROTECTED]>: > > > > > > > > > > > > > > > > > > > > > > > > > All, > > > > > > > > > > > > > > > > > > > > There is an upper limit to the number of sub-elements a > complexType can > > > > have using JaxMe. Using the attached XSD, test.xsd, you will notice > that > > > > JaxMe generates a startElement method in TesttypeHandler.java that > is larger > > > > than the java compiler can handle. > > > > > > > > > > > > > > > > > > > > Phurnace Software is willing to pay you to refactor JaxMe so that > the > > > > startElement is significantly smaller. Of course, we would hope this > code > > > > would be committed to the JaxMe mainline development so that all can > > > > benefit. I think an upper bound of 250 elements should suffice, but > more > > > > would be welcome. > > > > > > > > > > > > > > > > > > > > This is a chance to get paid to work on JaxMe. Please let me know > if you > > > > are interested. Compensation details can be worked out privately. I > can > > > > assure that this project will be well worth your time. > > > > > > > > > > > > > > > > > > > > Thanks! > > > > > > > > > > > > > > > > > > > > Robert > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > --------------------------------------------------------------------- > > > > > To unsubscribe, e-mail: > > > > [EMAIL PROTECTED] > > > > > For additional commands, e-mail: [EMAIL PROTECTED] > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > >
