Realistically Jigsaw in no way facilitates removal of long-deprecated code.
At *best* it would allow for deprecated *classes* to be moved to a
separate, optional module. It can't do a thing about deprecated
methods. Removing those would still utterly fragment the Java community.
Personally I'd like to just see the basic "app developer level"
modularity from Jigsaw and leave the native-package-system integration
and core class library modularization to Java 9 or 10 or whatever.
Having a simple module system that is understood by both the compiler
and runtime would help ordinary developers who don't need something as
complex as OSGi (e.g. don't need to juggle multiple versions of 1 jar
within one JVM or web/enterprise app as they just manage to figure out a
combination of versions of different libraries that works and move
forward using those -- it's usually pretty easily done).
How long would I hold Java 8 for this? Not terribly, but I don't think
this is really being considered. The Jigsaw folk seemed really hung up
on the native-package-system integration and core class library
modularization in previous talks.
On 8/30/2012 10:16 AM, Kevin Wright wrote:
I always figured that anyone who wanted lambdas that much would be
sailing the choppy JVM waters on a differently lingual boat by now.
It's not as though you even have to change your ops infrastructure or
much of your tooling to do so.
(hint: rename scala.jar/clojure.jar/groovy.jar/whatever.jar to
apache-closures.jar, release your project binaries as you always did.
Don't make a song and dance about it, it's easier to ask forgiveness
than to request permission)
Jigsaw on the other hand... That means distinctions like SE/ME can be
done away with, long-deprecated code can finally be removed, startup
time, download time, and memory foot print can be reduced, etc. It
makes Java far more suitable for running on something like the
Raspberry Pi. These are cross-cutting concerns that benefit all
languages on the Java platform.
Defender methods and method handles are also just plain awesome. Even
without lambdas, I'd fully expect some powerful optimisations to
be realised on top of those two.
On 30 August 2012 15:57, Thomas Matthijs <[email protected]
<mailto:[email protected]>> wrote:
On Thu, Aug 30, 2012 at 4:03 PM, Ricky Clarkson
<[email protected] <mailto:[email protected]>> wrote:
> I want my lambdas now and I'm in a job where using non-Java
languages
> will be a difficult sell. The earlier the release the better
for me.
>
> I've seen classpath hell exactly once, actually in a current
project,
> and plan to deal with it in a different way - attempting to
> find/create a combination of libraries that don't have version
> conflicts, and where that is not possible, moving tasks out of the
> same JVM process.
>
> The other benefit would be JVM startup time, which is less and
less an
> issue each year as machines get faster and Java doesn't get bigger.
> I'd like to see the startup time be improved further, but
lambdas will
> affect me more than cutting down startup from 5 seconds to 1.
>
I agree fully, jigsaw won't fix any problem i currently have, lambdas
on the other hand would be very beneficial
--
You received this message because you are subscribed to the Google
Groups "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.
--
You received this message because you are subscribed to the Google Groups "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.
