Java 7 screwup

[L Mihalkovic]

Wouldn't it be nice if OSX Mail had the notion of voiding old outbox contents?!  This email stayed dormant for several months on a broken hard-drive until yesterday when the machine was powered again and decided to work... 

I stil stand in true admiration of the work you are doing, still annoyed by some of Mail's quirks... and having a grand time with jsr292 (I ran into some spurious crashes of the mac port for code that worked on windows, but considering the stage I am at, I can't say if this is macport bug or a lot of forgiveness from the windows version).

sorry for the noise
cheers


On Aug 6, 2011, at 12:29 PM, L Mihalkovic wrote:

> As many people pointed out this is not a new issue, but a rather old one that has been pressed forward by the decision to turn some of the dormant optimizations on by default. If you look at the number of major issues Apple has with any new releases of their hardware/software and compare the reaction they trigger in the general public versus what happened in this case, then it is difficult not to draw the conclusion that there is way more going on here than simply technology… JDK 7 is not in production anywhere, and anyone with production responsibilities should be fired on the spot if they took the download and rolled it in prod without further validation in their environment. If there was not an established steady release train of future updates, then yes, I would share the opinion that releasing the JDK as-is was a major issue. But it is not the case, and there will be a number of regular releases that will address this and other issues that will undoubtedly surface… some will be new, and some will again be old issues pressed forward by recent decisions. I would rather have a somewhat work-in-progress released JDK with the a strong commitment to back it, rather than have to wait for another full year until all the kinks are hammered out. I don't like or dislike large corporations the like of Oracle or Apple, but I do have a tremendous amount of respect for the work the engineers at both companies are doing or a daily basis to work on the huge software project they work-on and regularly deliver.. bugs and all.
> 
> Cheers
> LM/
> 
> PS: speaking of bugs that you think might delay releases…. dragging this in-progress email on Lion causes parts of the window to be blurred whenever it touched the side of the screen… really annoying.
> 
> On Aug 5, 2011, at 3:47 PM, Mario Torre wrote:
> 
>> 2011/8/5 John Yeary <johnyeary at gmail.com>:
>>> As with any major release, there are bound to be bugs. However, the OpenJDK
>>> project announced the release and freeze on the code for what would become
>>> JDK 7. No one reported any major issues prior to launch. The folks on Apache
>>> Lucene/Solr projects could have tested in advance of release and found these
>>> issues. This may have prevented the on-time release.
>>> 
>>> I am not blaming the Apache folks, or Oracle. As with any new software, you
>>> need to test it. I have not been affected by the bug(s) reported, and can
>>> happily continue on my computing way. The workarounds have been released,
>>> use them until a fix is out, or alternatively wait and use JDK 6. As noted
>>> in the article, these JVM options are disabled in 6 and the bug was present.
>>> 
>>> 
>>> No one noticed this previously so I surmise that it is a mountain-mole hill
>>> issue.
>>> 
>>> John
>> 
>> I also think this is "mountain-mole hill" and in general agree with you.
>> 
>> I don't think that Apache should be responsible for testing OpenJDK bugs though.
>> 
>> This specific issue seems to be quite tricky to get but I wonder if
>> this should not have been spotted by the TCK?
>> 
>> Cheers,
>> Mario
>> -- 
>> pgp key: http://subkeys.pgp.net/ PGP Key ID: 80F240CF
>> Fingerprint: BA39 9666 94EC 8B73 27FA  FC7C 4086 63E3 80F2 40CF
>> 
>> IcedRobot: www.icedrobot.org
>> Proud GNU Classpath developer: http://www.classpath.org/
>> Read About us at: http://planet.classpath.org
>> OpenJDK: http://openjdk.java.net/projects/caciocavallo/
>> 
>> Please, support open standards:
>> http://endsoftpatents.org/
>