Is Java 7 at least as stable as Java 6? [closed]
Asked Answered
B

2

19

I remember when Java 7 was initially released, there were many suggestions not to use it for anything as there were some bugs in compiler optimizations. This was apparently not merely hypothetical. I haven't followed the situation closely since then; have these issues been worked out and it is generally considered safe to use at this point?

This page from java.com makes me think perhaps it's not quite ready yet, but I don't want to read too much into it. If it is not yet safe, is it sufficient to use -XX:-UseLoopPredicate as a VM argument?

Note that I am not only referring to the specific Lucene issue but the general stability of Java 7. Is Java 7 at least as stable as Java 6?

Blasius answered 23/1, 2012 at 19:44 Comment(14)
The bug reported, which made Lucene (and Eclipse) crash, has been fixed.Oby
This means you cannot use Apache Lucene/Solr with Java 7 releases before Update 2!. Java Platform (JDK) 7u2.Intitule
How should anyone know about the "general stability" of Java 7 in a way that's suitable for SO? (I.e. what answers do you expect besides what's basically a gut feeling?)Hydrangea
@Inerdial I don't know if there are any standards, but there could be. For instance, there could be an independent test suite by a 3rd party or something. Nobody can know if something is bug-free, but it's conceivably possible to answer this question with something besides gut feeling.Blasius
@Inerdial Also the existence or non-existence of any sort of consensus among developers would seem appropriate for an answer.Blasius
A "consensus among developers" is more or less an aggregated gut feeling, as well as probably bad-subjective-question territory. As far as the hypothethical test suite goes, that would be better suited to a separate question.Hydrangea
@Inerdial: "How should anyone know about the general stability of Java 7 in a way that's suitable for SO?" Simply make it like this: "Is it as stable as the latest Java 6 is?". Sounds quite simple to me : )Hedgerow
It worth remembering that JVM version 6 and 7 is much the same except there are different options on/off by default.Borders
For those that think this is too localized, note that I am not just asking about the specific Lucene issue.Blasius
Have you stopped beating your wife yet?Tewfik
@CodyGray I see what you did there. Is there a different reason you or others think it's too localized?Blasius
For reference, here is what the CEO of Stack Overflow thinks about "too localized".Blasius
I didn't vote to close this as "too localized". It doesn't seem like a particularly bad question. I just intended my comment as a subtle prod to consider modifying the title slightly. (I secretly suspect that's what is prompting some of the close votes. They just couldn't find the "inflammatory" reason.)Tewfik
Here is a discussion on Meta.SO about whether this is too localized.Blasius
H
14

From the very article you linked to:

UPDATE OCTOBER 28, 2011: As noted on Uwe’s blog, Java 7u1 is documented to include the patches to address these issues.

The article that links to says:

Last night, Oracle updated the release notes of Java 7u1 and Java 6u29, stating that they fixed the three Lucene-relevant bugs (plus another one related to that). Based on this confirmation, it's now safe to use Java 7 Update 1 (and later) with Apache Lucene and Apache Solr.

This answer paid for by The Committee For Clicking On Links For You.

Hydrangea answered 23/1, 2012 at 19:51 Comment(2)
A subsidiary of Read Your Own Links productionsSouthwards
I wasn't referring exclusively to the Lucene issue; it is just the noteworthy example.Blasius
X
1

The major bug that Uwe Schindler from Lucene and Solr annouced was fixed in Java7u1 (read more from him here. I believe it's safe to use Java7 now, especially given that there has also been a second update released for it that "improves reliability and performance".

Xylia answered 23/1, 2012 at 19:52 Comment(3)
a second update that "improves reliability" is really not something I like to read. It reads like "first update wasn't reliable, this second update is bit more reliable". :(Hedgerow
@user988052 "improves reliability and performance" is the standard corporate boilerplate for "we fixed some stuff we can't be bothered to enumerate". Cf. every Apple release notes ever.Hydrangea
@user988052: Back in 2002, I co-authored a paper that looked at how often security patches had to be re-released because the first patch was incorrect -- out of 136 vulnerabilities, 92 had good "first patches", 20 patches had subsequent patches, and 4 of those were faulty. Perhaps the world has improved in ten years, but I'd be surprised. :)Pichardo

© 2022 - 2024 — McMap. All rights reserved.