Am 23. September 2017 14:36:38 MESZ schrieb sebb <[email protected]>:
>On 23 September 2017 at 13:10, Felix Schumacher
><[email protected]> wrote:
>>
>>
>> Am 23. September 2017 13:39:44 MESZ schrieb sebb <[email protected]>:
>>>On 23 September 2017 at 12:28, Felix Schumacher
>>><[email protected]> wrote:
>>>>
>>>>
>>>> Am 23. September 2017 01:27:13 MESZ schrieb sebb
><[email protected]>:
>>>>>On 22 September 2017 at 19:53, Milamber <[email protected]>
>wrote:
>>>>>> Hello,
>>>>>>
>>>>>> I thinks we need to remove the "or later" words from these pages
>>>>>because
>>>>>> JMeter (out of the box) won't start with Java 9
>>>>>>
>>>>>> Note: of course you can start JMeter 3.3 with Java 9 with some
>>>tweaks
>>>>>to
>>>>>> disable the check of Java version from launch scripts.
>>>>>
>>>>>Does the Java check provide any benefit?
>>>>>
>>>>>Why not drop the code?
>>>>
>>>> I think we could reduce the code to log a warning, rather than
>>>exiting.
>>>
>>>But AFAICT this will still not work for Java 9 (or the code would not
>>>need to be fixed to work with Java 9!)
>>>
>>>What benefit does the check provide?
>>>
>>>AFAICT it only avoids the java.lang.UnsupportedClassVersionError.
>>>
>>>Do we really need to prevent people seeing that?
>>
>> Java 9 decided to change the format of the version string. That is
>why it doesn't work at the moment with our check.
>
>Yes, I know.
>
>But if we fix that then there is no need to change the code to log a
>warning.
>
>In the meantime the behaviour with Java 9 is much worse than it would
>be without the check.

Well, I wanted to express, that we just emit a warning and continue. 

>
>> I think it is nicer to display a warning that the Java version is not
>supported and hope it will work, than giving the user a rather cryptic
>message and hope they search online, what that message might mean.
>
>java.lang.UnsupportedClassVersionError is a basic error that anyone
>who has run Java code will have encountered at some point.
>
>The Java version checking code has given quite a bit of trouble since
>it was introduced.
>And is likely to continue to do so.
>I just don't think it's worth it for the minor benefit that it gives.
>
>If you really want to log a warning then I suggest adding a check at
>the end.
>If jmeter exits with a non-zero status, then show a message such as:
>
>"If jmeter failed with java.lang.UnsupportedClassVersionError, please
>check you are running with the minimum required Java version"
>
>Such code should not need any modifications when the Java version
>display changes.

That also crossed my mind just after sending my last message. 

I like it. 

Felix

>
>> Felix
>>
>>>
>>>> Felix
>>>>
>>>>>
>>>>>> Are you ok for I updated these pages?
>>>>>>
>>>>>> http://jmeter.apache.org/download_jmeter.cgi
>>>>>> https://www.apache.org/dist/jmeter/
>>>>>>
>>>>>> Milamber

Reply via email to