So count me as +0 for renaming *Data if it irritates people. My tolerance level for misspelling is pretty low, but having *Data in the class name doesn't bother me.
Craig On Sep 12, 2006, at 10:42 AM, Allen Gilliland wrote:
sorry, late reply.actually, this is probably as much a pet peeve as anything, but the *Data convention is one of the worst offenders in my mind. i guess i just don't like the idea that a build tool is telling me how is should name and/or structure my code. that just seems backwards to me.i would prefer it if xdoclet had a better way of being told what pojos we wanted to be mapped. for example, it would only map classes that had a special annotation @persistent-class at the class level comment.-- Allen Anil Gangolli wrote:Agree with most of these, but the "*Data" convention is used by the build to determine the set of classes over which to run Xdoclet. We'd need an alternative mechanism. We could name them explicitly, but I think this convention is useful to easily identify the pojos that have associated persistence behavior.--a.----- Original Message ----- From: "Allen Gilliland" <[EMAIL PROTECTED]>To: <[email protected]> Sent: Thursday, September 07, 2006 10:33 AM Subject: Re: Misspelled ReferrerManagerI agree that we should rename it, but I think it's too late to do this for 3.0. We've already spent way too much time testing/ refining the release to go in and do a sweeping change that touches lots of classes.Also, I don't think there is any reason that change can't be made for the next release. I would actually prefer that we do a single rename for all of our classes rather than keep picking at them one at a time. For example I would be a strong supporter of a few other renamings ...1. Website -> Weblog 2. Remove "Data" from all the pojos. 3. Referer -> Referrer 4. Index -> Searchand probably a few others if I looked harder. so rather than try to squeeze this in for 3.0, why don't we plan to make it a feature for 3.1 and do a more complete renaming.-- Allen Craig L Russell wrote:I noticed (again with 3.0) that the ReferrerManager interface name is misspelled. According to my IDE, there are 35 occurrences of this interface name in the entire source code.It seems that 3.0 might be the right time to fix this.If everyone agrees, I can provide a patch. I don't think I should check it in, because I'm not comfortable enough with making sure that I didn't cause a regression because of a missed update.Craig Craig Russell [EMAIL PROTECTED] http://db.apache.org/jdo
Craig Russell [EMAIL PROTECTED] http://db.apache.org/jdo
smime.p7s
Description: S/MIME cryptographic signature
