[squeak-dev] 4.1 release tasks

Levente Uzonyi leves at elte.hu
Sun Mar 21 11:45:49 UTC 2010


Note that the new VMs for 4.1 are on the way, so we will have to revert 
the vm related expected failures, or make them VM(Maker) version 
dependent.


Levente

On Sun, 21 Mar 2010, Alexander Lazarević wrote:

>          class           |            selector            | result
> --------------------------+--------------------------------+---------
> BitBltTest               | testAllAlphasRgbAdd            | failure
> BitBltTest               | testAllAlphasRgbMax            | failure
> BitBltTest               | testAllAlphasRgbMin            | failure
> BitBltTest               | testAllAlphasRgbMinInvert      | failure
> BitBltTest               | testAllAlphasRgbMul            | failure
> BitBltTest               | testAllAlphasRgbSub            | failure
>
>> The above 6 are vm bugs, fixed in newer vms.
>
> So I guess it is ok to make them expected failures until the new VMs are
> out.
>
> ClosureCompilerTest      | testDebuggerTempAccess         | failure
> ClosureCompilerTest      | testInjectIntoDecompilations   | failure
> ClosureCompilerTest      | testInjectIntoDecompiledDebugs | failure
> DebuggerUnwindBug        | testUnwindDebuggerWithStep     | failure
>
>> AFAIK the above four are known bugs, Eliot has the fixes, the have to be
> integrated sometime.
>
> LocaleTest               | testIsFontAvailable            | failure
>
>> This is font dependent. We should probably add it to expected failures if
> the default font is not Accuny.
>
> MCChangeNotificationTest | testCoreMethodModified         | failure
>
>> That one is related to system change notifications. If you run this test
> alone, it passes.
>
> True. How do we handle this?
>
> PackageDependencyTest    | testST80                       | failure
> PackageDependencyTest    | testSUnit                      | failure
>
>> These were fixed earlier.
>
> I still get the above
>
> SMDependencyTest         | test2                          | error
>
>> VM bug, will be fixed in the next vm release.
>
> So it's also ok to make it an expected error
>
> WorldStateTest           | testDeferredUIQueueTimeout     | failure
>
>> Never seen this failing on windows.
>
> This only fails for me when doing all tests ...
>
> Alex
>


More information about the Squeak-dev mailing list