You know, I stopped caring about JSR-310. Provisionally. Here's the problem of jodatime: It sufferes from -just-this-once- syndrome.
If I need to format a date, I know jodatime is easier, but, for just this once, It's not worth it. I'll just stick with SimpleDateFormat. Same applies for parsing dates, and even for limited arithmetic (though the horrors of java.util.Calendar means I'll go for jodatime real fast when arithmetic gets involved). Of course, the next time, it's still just-this-once. Adding jodatime to a project adds a dependency, requires me to do some setup work on the project and across various build tools, etc, etc. Often, it's one of the first things you run into when building a prototype, often before you've got your dependency system all sorted, and you just couldn't be bothered. Then, later, when you've got everything set up nicely to handle dependencies (you've got maven rolling or some such), you still feel it isn't worth switching as you then feel that for consistency's sake you would have to go through all existing uses of Date, SDF, and Calendar, and fix them too. Thus: just-this-once syndrome. Making JSR310 vitally important to solve date handling in java; a third party library is no good due to just-this-once syndrome. But here's why I provisionally think it's no longer important: jigsaw. *IF* jigsaw works right, then just-this-once syndrome should be a thing of the past. You install joda-time once, possibly configure your IDE that you'd always like to see certain classes in it in auto- complete dialogs, and that's that. To add joda-time, you just add a module reference in module-info.java. I bet IDEs will quickly extend their already ubiquitous automatically-add-the-import-statement functionality to also automatically adding the module dependency. If it's that easy, then the just-this-once problem goes away. Then the only 2 remaining benefits of JSR-310 are: (1) Improving joda-time itself. In other words: the things that make JSR-310 different from joda-time. Given that Stephen himself said that joda is a scratched itch, these things (timezones and obscure calendar systems support) are a non-issue for almost everyone, me included. So, non-issue. (2) It gets the official sun stamp of approval by being in the java.* package hierarchy. Well-meaning idiots who ascribe some sort of magic power to the java.* package will need less smacks with the cluebat before they'll accept that java.util.Calendar sucks, and they should be using something else. Useful for the ecosystem, but, personally? Not at all relevant for me. so, for pretty much everyone capable of helping out, with jigsaw, the itch will get scratched when JDK7 comes out. Hopefully. And to be honest, this is how it should be. Using a simple no-frills entirely stand alone third party library like joda-time should be exactly as easy to use as anything in rt.jar. If jigsaw makes that happen, java is going to be so much nicer. (both the language and the platform, so for once this is great news all around, even for scalanistas, groovians, and jythoners). On Dec 1, 4:21 pm, Casper Bang <[email protected]> wrote: > Actually Stephen answered that question recently via his > blog:http://www.jroller.com/scolebourne/date/20091120 > > But it sure would be nice with a fixed date/calendar API. Sadly it's > probably the wrong time to ask for Sun to sponsor JSR-310 work. > > /Casper > > On Dec 1, 3:16 pm, Brian Leathem <[email protected]> wrote: > > > > > Discussions about JSR 310 seem to have started > > again:http://markmail.org/message/pp5pozntqmj5li2d > > > Why can't the JCP just rubber stamp Joda-time? > > > Brian -- You received this message because you are subscribed to the Google Groups "The Java Posse" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/javaposse?hl=en.
