I have a general question for contemplation: when does anyone consider
converting to the new Mustang version of Java, which is now in beta?
obviously you have to consider this case by case, but as a general
principle, can you stay with the current version if you don't need the
full compiler control, scripting options from version 6? or are you
supposed to switch? the conversion from 1.4 to 1.5 is still fresh is
memory, and now Sun is moving to Open Source, what effects will this
have on the standard?
Jaap
steve - 09 Jul 2006 23:07 GMT
> I have a general question for contemplation: when does anyone consider
> converting to the new Mustang version of Java, which is now in beta?
[quoted text clipped - 7 lines]
>
> Jaap
Why not wait till version 7 which will be next year, or V8 6 months later.
Seriously sun should be shot , for the millions of dollars they waste, and
force people to waste.
Instead of bug fixing a stable language they just keep adding bloat.
Follow if you want to, or instead you can just code to 1.4 standard and wait
a few years.
Steve
Chris Smith - 09 Jul 2006 23:36 GMT
> I have a general question for contemplation: when does anyone consider
> converting to the new Mustang version of Java, which is now in beta?
[quoted text clipped - 5 lines]
> memory, and now Sun is moving to Open Source, what effects will this
> have on the standard?
There is no need to switch if you don't have a reason to do so.
Generally speaking, bux fix releases are made against the past two major
release cycles (this is empirical; I don't know if there's any kind of
committment to this effect by Sun). In any case, the transition to Java
1.6 is unlikely to be nearly as tough as the one from 1.4 to 1.5. I
know companies that are still contemplating that change. (If something
happens in 1.7 to fix generics, it may turn out they were the smart
ones...)

Signature
Chris Smith - Lead Software Developer / Technical Trainer
MindIQ Corporation
Thomas Hawtin - 10 Jul 2006 07:48 GMT
> There is no need to switch if you don't have a reason to do so.
> Generally speaking, bux fix releases are made against the past two major
> release cycles (this is empirical; I don't know if there's any kind of
> committment to this effect by Sun).
The version before the one which is superseded goes into a 1.5 year end
of life period. So at the moment 1.5 is current, 1.3 has been in end of
life for over 1.5 years (last time I looked it had not been moved into
the archive section) and in the autumn 1.4 should start its end of life
period.
Not many bug fixes get backported. Just as well as allegedly in general
each bug fix introduces 0.7 new bugs on average. Funnily enough they do
seem quite keen on updating the core virtual machine. Presumably it is
sufficiently out of the way that few of the people likely to complain
notice.
> In any case, the transition to Java
> 1.6 is unlikely to be nearly as tough as the one from 1.4 to 1.5. I
> know companies that are still contemplating that change. (If something
> happens in 1.7 to fix generics, it may turn out they were the smart
> ones...)
It's changing the source from 1.4 to 1.5 clean which is disruptive. That
and switching to new classes, such as StringBuilder. If you use -source
1.4 -target 1.4 or ignore the warnings, it shouldn't be too difficult.
Two years doesn't seem too fast an update cycle.
Tom Hawtin

Signature
Unemployed English Java programmer
http://jroller.com/page/tackline/
John O'Conner - 10 Jul 2006 10:03 GMT
> I have a general question for contemplation: when does anyone consider
> converting to the new Mustang version of Java, which is now in beta?
You can start looking at the new features now. However, there is no
hurry to update, especially since Java SE 6 is still beta software. Most
tools take several months to fully support the new version, so you might
not seriously consider moving to Java SE 6 until early next year.
Early adopters and engineers starting completely new projects might
consider using version 6 right now... However, you would not want to
deploy existing apps onto version 6. That probably isn't the wisest
option at the moment.
Regards,
John O'Conner