Equivalent for JSR-275 ?

299 views
Skip to first unread message

Jan Goyvaerts

unread,
Jul 23, 2012, 7:30:30 AM7/23/12
to The Java Posse
I've been asked to find something for units of measures & conversion. In the early days there was JSR-275. Which today is rejected. A pity because the implementation has merit.

Is there an equivalent for this or does everybody cooks his own library ? :-)


John May

unread,
Jul 23, 2012, 8:01:16 AM7/23/12
to java...@googlegroups.com
I'm not sure if it's "equivalent" but JScience provides an implementation of the Units of Measure interfaces.

Fabrizio Giudici

unread,
Jul 23, 2012, 1:10:57 PM7/23/12
to The Java Posse, Jan Goyvaerts
On Mon, 23 Jul 2012 13:30:30 +0200, Jan Goyvaerts <java.a...@gmail.com>
wrote:

> I've been asked to find something for units of measures & conversion. In
> the early days there was JSR-275
> <http://www.jcp.org/en/jsr/detail?id=275>.
> Which today is rejected. A pity because the implementation has merit.

Why has it been rejected? And WTF is going to be included in Java 8 at
this point, apart the first class functions?



--
Fabrizio Giudici - Java Architect, Project Manager
Tidalwave s.a.s. - "We make Java work. Everywhere."
fabrizio...@tidalwave.it
http://tidalwave.it - http://fabriziogiudici.it

Simon Ochsenreither

unread,
Jul 23, 2012, 1:27:22 PM7/23/12
to java...@googlegroups.com, Jan Goyvaerts
 
... apart the first class functions?

First-class functions aren't part of Java 8. You need an explicit or inferable nominal type on the LHS.

Steven Siebert

unread,
Jul 23, 2012, 1:39:58 PM7/23/12
to java...@googlegroups.com, Jan Goyvaerts
The JSR was twice rejected for technical reasons, including the use of the javax.measurement namespace.  The leader of the JSR (JScience author) considers the JSR dead that last I heard.

I use JScience in production, and have built extensions to it's spatial conversions.  I find it to be a very good library and would recommend it.

Regards,

S



--
You received this message because you are subscribed to the Google Groups "Java Posse" group.
To post to this group, send email to java...@googlegroups.com.
To unsubscribe from this group, send email to javaposse+unsubscribe@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/javaposse?hl=en.


Ricky Clarkson

unread,
Jul 23, 2012, 1:55:15 PM7/23/12
to java...@googlegroups.com, Jan Goyvaerts
How does that make them not first-class?

On Mon, Jul 23, 2012 at 2:27 PM, Simon Ochsenreither <simon.och...@googlemail.com> wrote:
 
... apart the first class functions?

First-class functions aren't part of Java 8. You need an explicit or inferable nominal type on the LHS.

--
You received this message because you are subscribed to the Google Groups "Java Posse" group.
To view this discussion on the web visit https://groups.google.com/d/msg/javaposse/-/l0c4206Ydg0J.

To post to this group, send email to java...@googlegroups.com.
To unsubscribe from this group, send email to javaposse+...@googlegroups.com.

clay

unread,
Jul 26, 2012, 8:01:47 PM7/26/12
to java...@googlegroups.com, Jan Goyvaerts
On Monday, July 23, 2012 12:10:57 PM UTC-5, fabrizio.giudici wrote:
And WTF is going to be included in Java 8 at   
this point, apart the first class functions? 

Java 8 is still a big release. Adding lambdas well will be a big deal, but additionally:

major collections library update. Heavily uses lambdas. Parallelism functionality. Maybe persistently immutable collections (currently you need third party libs to get this)
annotations on types: this opens up a lot of potential
jrockit integration. No PermGen.
Functions in interfaces (similar to traits or mixins in other languages)
New DateTime library.

AFAIK, Java 8 lambdas meet all the criteria of "first class functions". Java 8 purposefully keeps the "final" requirement on outer variable capture, which violates some people's definition of a closure: Dick Wall said that the "final" variable requirement makes the lambdas not true closures. IMO, I think the JDK devs are right that the "final" requirement is a good idea and avoids lots of error scenarios.


Reply all
Reply to author
Forward
0 new messages