----- Original Message ----- From: "robert burrell donkin" <[EMAIL PROTECTED]> To: "Velocity Developers List" <[EMAIL PROTECTED]> Sent: 2001. december 19. 20:15 Subject: Re: VelocityVaue Patch
> it's sometimes hard for developers to understand that rejection isn't > usually a statement about the quality of their code or their design ideas. > social pressures combined with patch review and fixing by committers > mean that weak patches are rare and are as likely as any to be accepted. > the most common reason for rejection is that a patch would compromise a > design line-in-the-sand understood by the committers but which appears odd > to outsiders. I once did a fairly complete overhaul of Introspector so that it uses the Java's own built-in JavaBeans-compliant introspector engine. It was rejected, because Velocity as it is now does not conform to JavaBeans spec (the isXxx property accessors were added since, but there were other issues as well) and there was a chance this will break existing templates. I did the only reasonable thing to do: weighed the presented "contra" arguments myself, and as a result accepted the rejection. I'm luckier than Will because I didn't really depend on the change, since 'm hacking Velocity for fun only -- I haven't yet used it in production anywhere being accustomed too much to Freemarker... I just wanted to say, rejects happen. Whether you accept one or fight depends on whether you see the same big picture (including the technical, philosophical and social issues) for the project as those that steer the project. Fortunately Apache license lets you keep a separate version of Velocity, even if it can get tedious keeping up with patches. Attila.
smime.p7s
Description: application/pkcs7-signature
