[rfc][icedtea-web] reproducer for pr822 (I hope)

[Jiri Vanek]

On 08/27/2012 03:13 PM, Pavel Tisnovsky wrote:
> Hi Jiri,
>
> I think that your reproducer is correct, but please try to fix following minor things:
>
> %s/Cought/Caught/g       :-)
> %s/exapmle/example/g
> %s/dicutabile/questionable/g
>
>
> String sss="Applet killing himself after 2000 ms of life";
>
> The value "2000" is based on the actual value of a Killer's public
> attribute named "n". It would be IMHO more clear to rename
> this attribute to TIMEOUT and make it probably final. Also the
> string above should read:
>
> String sss="Applet killing himself after " + Killer.TIMEOUT + " ms of life";
> (similar string is used three times in the code).

This cannot be fixed, sorry - the 2000 constant is in file, which can not be passed to testcase's 
classapth:((

Everything else fixed ad test pushed.

Thanx a lot!
J.
>
>
> With these fixes it's ok to push this reproducer to HEAD.
>
> Thank you in advance,
> Pavel T.
>
>
>
> ----- Jiri Vanek <jvanek at redhat.com> wrote:
>> ping?
>>
>>
>>
>> updated for latest states:
>>
>> 2012-05-30  Jiri Vanek  <jvanek at redhat.com>
>>
>> 	Added reproducer fro multiple signatures (and PR822)
>> 	* Makefile.am: SIGNED_REPRODUCERS growth for signed2 directory
>> 	* tests/reproducers/signed2/MultipleSignaturesTest/*: test for launching
>> 	application containing two signed - each by different certificate - jars
>> 	and different signatures are applied on different packages
>> 	* tests/reproducers/signed2/MultipleSignaturesTest/resources/MultipleSignaturesTest.html:
>> 	* tests/reproducers/signed2/MultipleSignaturesTest/resources/MultipleSignaturesTest1.jnlp:
>> 	* tests/reproducers/signed2/MultipleSignaturesTest/resources/MultipleSignaturesTest1_requesting.jnlp
>> 	* tests/reproducers/signed2/MultipleSignaturesTest/resources/MultipleSignaturesTest2.jnlp:
>> 	*
>> tests/reproducers/signed2/MultipleSignaturesTest/srcs/somecrazytestpackage/MultipleSignaturesTest.java:
>> 	* tests/reproducers/signed2/MultipleSignaturesTest/testcases/MultipleSignaturesTestTests.java:
>> 	* tests/reproducers/signed2/MultipleSignaturesTestSamePackage/*: test for launching
>> 	application containing two signed - each by different certificate - jars
>> 	but different signatures are applied on same packages
>> 	*
>> tests/reproducers/signed2/MultipleSignaturesTestSamePackage/resources/MultipleSignaturesTest1_SamePackage.jnlp:
>> 	*
>> tests/reproducers/signed2/MultipleSignaturesTestSamePackage/resources/MultipleSignaturesTest1_SamePackage_requesting.jnlp
>> 	*
>> tests/reproducers/signed2/MultipleSignaturesTestSamePackage/resources/MultipleSignaturesTest2_SamePackage.jnlp:
>> 	*
>> tests/reproducers/signed2/MultipleSignaturesTestSamePackage/resources/MultipleSignaturesTest_SamePackage.html:
>> 	*
>> tests/reproducers/signed2/MultipleSignaturesTestSamePackage/srcs/MultipleSignaturesTestSamePackage.java:
>> 	*
>> tests/reproducers/signed2/MultipleSignaturesTestSamePackage/testcases/MultipleSignaturesTestTestsSamePackage.java:
>> 	* tests/test-extensions/net/sourceforge/jnlp/ServerAccess.java:
>> 	introduced VERBOSE_OPTION with "-verbose" value
>>
>>
>> On 05/30/2012 06:20 PM, Jiri Vanek wrote:
>>> Hi!
>>>
>>> This reproducer is trying to lunch application/applet consisted from two jars, each signed by
>>> different signature.
>>>
>>> Applets are behaving correctly - are launched when packages are signed differently.
>>> jnlp application donot:
>>> a) when
>>>     <security>
>>>       <all-permissions/>
>>>     </security>
>>> is included in jnlp, then it dies with "jars are not signed exception"
>>> b) when the
>>>     <security>
>>>       <all-permissions/>
>>>     </security>
>>> is missing then restricted code is allowed (in case of same package it dies correctly with
>>> "different signatures for package" exception.
>>>
>>> I consider both a and b as incorrect, but we can live with b until a is fixed.
>>>
>>> So the last test method of both testcases should be @KnownToFail :)
>>>
>>> btw I'm not sure if this is representing pr822 but i guess it is.
>>>
>>> 8 new tests - two of them failing.
>>>
>>> Thanx in advice
>>>    J.
>>>
>>>
>>
>