JDK 1.5 or JDK 1.6

P

Pep

Hi, is JDK 1.6 considered stable and should applications be upgraded
from JDK 1.5 to 1.6 if so?

TIA,
Pep
 
A

Andrew Thompson

Pep said:
Hi, is JDK 1.6 considered stable ..

I consider it stable (and do not speak for anyone else).
..and should applications be upgraded
from JDK 1.5 to 1.6 if so?

Huhh? What do you mean 'upgraded'? If you
are currently running them under 1.5 only, test
on a few 1.6 JRE's, and if it runs as you expect -
add 'JRE 1.6' to the 'compatible with' list*.

If there is 1.6 functionality that is of interest, it
might be possible to cause the app. to fall
back gracefully to accomodate the 1.5 JRE's.
This strategy provides the 'best of all worlds'
experience for the end user.

* If using web start, this would be as easy as
changing the version attribute of the java/j2se
element of the JNLP file - from 1.5*, to 1.5+.
 
R

RVince

Not sure of this, but I think 1.6 was primarily released as the latest,
stable 1.5.x but for Vista. Any comments on that statement guys? -R. Vince
 
P

Pep

Andrew said:
I consider it stable (and do not speak for anyone else).


Huhh? What do you mean 'upgraded'? If you
are currently running them under 1.5 only, test
on a few 1.6 JRE's, and if it runs as you expect -
add 'JRE 1.6' to the 'compatible with' list*.

If there is 1.6 functionality that is of interest, it
might be possible to cause the app. to fall
back gracefully to accomodate the 1.5 JRE's.
This strategy provides the 'best of all worlds'
experience for the end user.

* If using web start, this would be as easy as
changing the version attribute of the java/j2se
element of the JNLP file - from 1.5*, to 1.5+.

--
Andrew Thompson
http://www.athompson.info/andrew/

Message posted via http://www.javakb.com

Thanks. By upgrading I meant is there much mileage in moving our code
base to JDK 1.6. As a rule we would not move up without doing a full
regression test of our code but like most companies we are extremely
busy and do not have a lot of resources available for this exercise.

So it's really a question of the benefits in doing this. If, for
instance, 1.6 doesnot resolve a whole host of bugs or similar benefits
then we probably would not do the upgrade right now. The code does not
require 1.6 per se.

Cheers,
Pep.
 
P

Pep

RVince said:
Not sure of this, but I think 1.6 was primarily released as the latest,
stable 1.5.x but for Vista. Any comments on that statement guys? -R. Vince

Thanks for the reply. By "Vista" do you mean microsoft windows?

Cheers,
Pep.
 
L

Lew

RVince said:
Not sure of this, but I think 1.6 was primarily released as the latest,
stable 1.5.x but for Vista. Any comments on that statement guys? -R. Vince

The way I parse your statement it doesn't make any sense, so I must have
misunderstood. It sounds like you're saying Java 1.6 (a.k.a. "Java 6") is the
latest release of Java 1.5, which it is not. Java 6 is the latest released
version of Java. It is stable. How it is "considered" I do not know. Java 5
is the old version. Java 7 is on the way. Java 6 is the current version of
the language.

Incidentally, Java 1.4 is already in "End of Life" from Sun.

Vista is another issue, but I sure don't see how it is relevant.

I use Java 6 exclusively unless a client orders me otherwise. (Platforms like
JME are an obvious exception.) I can think of no reason to prefer an old
version on platforms that support the current release. Even if the target
platform is an older version, I use JDK 6 and cross-build to the earlier
version. ("-target" and "-Xbootclasspath" are your friends there.)

Use Java 6.
 
M

Mike Schilling

..
I use Java 6 exclusively unless a client orders me otherwise. (Platforms
like JME are an obvious exception.)

Application and web servers are another.
I can think of no reason to prefer an old version on platforms that
support the current release. Even if the target platform is an older
version, I use JDK 6 and cross-build to the earlier version. ("-target"
and "-Xbootclasspath" are your friends there.)

Hmm. If you know that your app has to run under 1.5 (because it's going to
run in a web server that only supports 1.5), why go to the effort of
cross-building it, rather than build and test it with the version it will
run under? If there's a bug in 1.5 that's fixed in 1.6, I'd prefer to
discover it sooner.
 
A

Andrew Thompson

Pep said:
[quoted text clipped - 23 lines]

Please trim text no longer immediately relevant,
and post replies directly after what is being replied
to. Note that 'trimming' above was automatic by
my web interface to usenet. Most of what it
requoted would be what I would have chosen to
delete..
Thanks. By upgrading I meant is there much mileage in moving our code
base to JDK 1.6.

I still do not understand why you would want to
'move' to 1.6, rather than ensure the application
will run under either 1.5, *or* 1.6.

But if your application runs just fine* under 1.5,
and you only intend to support 'a version',
why 'move' it to 1.6 at all?
...
So it's really a question of the benefits in doing this.

I can see very few, if the app. is only advertised to run
under one version, and the customers have that version.
...If, for
instance, 1.6 doesnot resolve a whole host of bugs ..

* Bugs that affect you app.? What bugs?
...or similar benefits

Like what? The only things I can think of besides
bug fixes are..
- something linke a newer, better, packing mechanism
(probably thinking pack200 in Java 1.5), or
- new functionality - new classes, methods, API's.
What kind of 'benefits' are you expecting or hoping for?
 
R

Roedy Green

Hi, is JDK 1.6 considered stable and should applications be upgraded
from JDK 1.5 to 1.6 if so?

I am having no troubles with it, but then I usually use it with
-target to some lower level.

It is basically the same compiler with bugs fixed and new features. If
you don't use the new features, normally the code should be even
better.
 
R

Roedy Green

Please trim text no longer immediately relevant

If he did that, he would get accused of "censorship". There is no way
to win! I prefer trimming since the rest is irrelevant and just a
distraction from my message
..
 
K

~kurt

Andrew Thompson said:
I still do not understand why you would want to
'move' to 1.6, rather than ensure the application
will run under either 1.5, *or* 1.6.

Many developers include their own copy of the JRE with their software.
I never liked that idea much myself - seems silly to have a bunch of
JREs installed on the computer, and it kinda defeats the platform
independent advantages of Java (the developer needs to make a different
package for each OS they support).

- Kurt
 
A

Andrew Thompson

Lew said:
The way I parse your statement it doesn't make any sense, so I must have
misunderstood.

I had trouble understanding what R. was trying to get
across with that sentence..
Vista is another issue, but I sure don't see how it is relevant.

There is a Java/security problem with Vista, but AFAIU,
it affects any Java installed - it is probably just that
most Vista boxes are getting 1.6.
 
P

Pep

Andrew Thompson wrote:
<snip>

Don't have a problem trimming, I usually do but got lazy this time :)
I still do not understand why you would want to
'move' to 1.6, rather than ensure the application
will run under either 1.5, *or* 1.6.

This is a inhouse application and we do not support multiple versions
of software as there is no need. All our platforms, i.e. servers and
workstations are rolled out with standard builds, so there is no need
to support multiple versions.
But if your application runs just fine* under 1.5,
and you only intend to support 'a version',
why 'move' it to 1.6 at all?
..

I can see very few, if the app. is only advertised to run
under one version, and the customers have that version.

Like what? The only things I can think of besides
bug fixes are..
- something linke a newer, better, packing mechanism
(probably thinking pack200 in Java 1.5), or
- new functionality - new classes, methods, API's.
What kind of 'benefits' are you expecting or hoping for?

Well for instance moving from 1.4 to 1.5 provided better string
formatting. Myself I am hoping for none really. I just have a pair of
anal colleagues that will argue the cup is half empty just for the
hell of it. As senior developer I have the responsibility to decide
what our IT strategy is with this project. So if nobody is jumping up
from the community saying something like it really is a" must" due to
serious security bugs in 1.5 or similar then I have no intention of
moving up.

So I guess we'll be staying on 1.5 then, following the "if it ain't
broke don't fix it" rule :)

<snip>

Cheers,
Pep.
 
P

Pep

Roedy said:
I am having no troubles with it, but then I usually use it with
-target to some lower level.

It is basically the same compiler with bugs fixed and new features. If
you don't use the new features, normally the code should be even
better.

Seems the consensus is that there is no viable reason to move up to
1.5 if we are not planning on using any newer features, which we are
not. So we'll stay on 1.5 for the forseeable future then, unless
another part of our company requires a move in the future.

Thanks everyone for all the input :)

Pep
 
O

Oliver Wong

["Should I upgrade to 1.6?"]
Well for instance moving from 1.4 to 1.5 provided better string
formatting.

Yes, but if you never use the string formatting, then you won't
actually take advantage of that particular benefit. Most of the people
here at my company don't bother with the new "printf" thing (I'm assuming
that's what you're referring to).
Myself I am hoping for none really.

There are lots changes. Whether or not they are relevant to you, we'd
need a crystal ball (or access to your source code, and a lot of spare
time on our hands) to guess.
I just have a pair of
anal colleagues that will argue the cup is half empty just for the
hell of it. As senior developer I have the responsibility to decide
what our IT strategy is with this project. So if nobody is jumping up
from the community saying something like it really is a" must" due to
serious security bugs in 1.5 or similar then I have no intention of
moving up.

This will never happen. If there are serious security bugs in a given
version, they will update THAT version. Notice that there exists 1.6,
1.6.0_01b6, 1.6.0_02b5, 1.6.0_02b6, etc. These where the fixes to "serious
security bug" will show up. Not in 1.7.

- Oliver
 
O

Oliver Wong

Pep said:
Seems the consensus is that there is no viable reason to move up to
1.5 if we are not planning on using any newer features, which we are
not. So we'll stay on 1.5 for the forseeable future then, unless
another part of our company requires a move in the future.

Here's a viable reason:

If you're always running the latest version of Java, you very rarely
have to be worried about whether or not the snippet of code you're looking
at was intended for a newer version of Java than the one you were running.

I remember a while ago, a bunch of posts were made here saying that
their compilers were rejecting examples they copied and paste from sites
or books. The problem was the examples were using generics, and they were
stuck on a 1.4 compiler.

I think Java programmers don't ever really "plan" on using newer
features. Rather, they look at the set of features available to them, and
then pick the best one for the job. Having more features to choose from
increases the likelihood of finding a perfect fit between a feature and
the job which needs to be done.

I use methods in the API introduced in 1.6 all the time without even
realizing that it's not available in 1.5. Right now, I'm not actively
looking at the 1.7 API and drooling, rubbing my hands together, dreaming
about taking advantage of these new methods. But when 1.7 comes out, I'm
sure I'll be unknowingly using 1.7-only methods too.

- Oliver
 
?

=?ISO-8859-1?Q?Arne_Vajh=F8j?=

~kurt said:
Many developers include their own copy of the JRE with their software.
I never liked that idea much myself - seems silly to have a bunch of
JREs installed on the computer,

It makes a lot of sense to me that the installer has the capability
to install the JRE if the users does not have one.

The only little extra trick needed is that it should not
install it if a suitable version already were installed.

Arne
 
R

Roedy Green

I use methods in the API introduced in 1.6 all the time without even
realizing that it's not available in 1.5. Right now, I'm not actively
looking at the 1.7 API and drooling, rubbing my hands together, dreaming
about taking advantage of these new methods. But when 1.7 comes out, I'm
sure I'll be unknowingly using 1.7-only methods too.

For in-house stuff the corporation as a whole will decide when it is
time to flip to the new version. It causes some disruption and there
is considerable cost to getting the new version installed and the old
one removed. Java install/uninstall is very shaky even one a brand
new machine like mine. I nearly always have to do some manual touchup
to delete old files.

For preparing material for the world, try to use the oldest version to
give the widest audience. I have stopped doing that for programs
aimed at programmers. I figure they are smart enough to have a 1.6
installed even if they develop for some other version.
 
K

~kurt

Arne Vajhøj said:
It makes a lot of sense to me that the installer has the capability
to install the JRE if the users does not have one.

I would agree that is a nice feature to have.

- Kurt
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Members online

Forum statistics

Threads
473,755
Messages
2,569,537
Members
45,023
Latest member
websitedesig25

Latest Threads

Top