New Trunk Image: 10548

Grab it here: http://ftp.squeak.org/trunk/Squeak4.2-10548-alpha.zip

On 2010/09/22 05:31, Casey Ransberger wrote:
This brave and noble soul tested it on Windows XP.

First comment is that the "What are your initials?" popup has got to go
if we want to automatically run tests. (But I guess this is easy to fix
in a script that runs the tests?)

"2805 run, 2760 passes, 9 expected failures, 35 failures, 1 errors, 0
unexpected passes"

Error:
WeakFinalizers>>#testNewFinalizationSupported (because there's no
WeakFinalizationList class (or WeakFinalizationItem))

Most of the failures are in DecompilerTests. Hm, it'd be nice if we
wrote out the failures/errors to the Transcript so I could just select &
copy the list of failures. Hasn't someone done this, actually?

Failures:
DecompilerTests>>#testDecompileLoopWithMovingLimit
DecompilerTests>>#testDecompilerInClassesBAtoBM
DecompilerTests>>#testDecompilerInClassesBNtoBZ
DecompilerTests>>#testDecompilerInClassesCAtoCM
DecompilerTests>>#testDecompilerInClassesCNtoCZ
DecompilerTests>>#testDecompilerInClassesDAtoDM
DecompilerTests>>#testDecompilerInClassesENtoEZ
DecompilerTests>>#testDecompilerInClassesGNtoGZ
DecompilerTests>>#testDecompilerInClassesIAtoIM
DecompilerTests>>#testDecompilerInClassesINtoIZ
DecompilerTests>>#testDecompilerInClassesLAtoLM
DecompilerTests>>#testDecompilerInClassesMAtoMM
DecompilerTests>>#testDecompilerInClassesMNtoMZ
DecompilerTests>>#testDecompilerInClassesOAtoOM
DecompilerTests>>#testDecompilerInClassesPAtoPM
DecompilerTests>>#testDecompilerInClassesPNtoPZ
DecompilerTests>>#testDecompilerInClassesSAtoSM
DecompilerTests>>#testDecompilerInClassesSNtoSZ
DecompilerTests>>#testDecompilerInClassesTAtoTM
DecompilerTests>>#testDecompilerInClassesUNtoUZ
DecompilerTests>>#testDecompilerInClassesWAtoWM
FileStreamText>>#TestPositionPastEndIsAtEnd
MCPackageTest>>#testUnload
PNGReadWriterTest>>#testPngDecodingColors32
PNGReadWriterTest>>#testPngEncodingColors32
PureBehaviorTest>>#testClassesWithTraits
RectangleTest>>#testRoundingAfterHalfPixelTranslation
ReleaseTest>>#testUndeclared
SMDependencyTest>>#test2
StandardSystemFontsTest>>#testRestoreDefaultFonts
TestURI>>#testDirWithHash
TraitCompositionTest>>#testAliasCompositions
TraitFileOutTest>>#testCondenseChanges
TraitFileOutTest>>#testFileOutCategory
TraitFileOutTest>>#testAllSentMessages

Transcript shows:
Removal of class named TUTU ignored because TUTU does not exist.
Removal of class named TUTU ignored because TUTU does not exist.
Removal of class named TUTU ignored because TUTU does not exist.
Removal of class named TUTU ignored because TUTU does not exist.
Removal of class named TUTU ignored because TUTU does not exist.
Removal of class named TUTU ignored because TUTU does not exist.Fetch:
3684 ms
Size: 712383 bytes
Save checkpoint to disk: 22 ms

CA
CB
Object>>foo (GeniePlugin is Undeclared)
ClosureCompilerTest>>DoIt (AbstractInstructionTests is Undeclared)
ClosureCompilerTest>>DoIt (AbstractInstructionTests is Undeclared)
ClosureCompilerTest>>DoIt (AbstractInstructionTests is Undeclared)
ClosureCompilerTest>>DoIt (AbstractInstructionTests is Undeclared)
TraitsResource>>createTraitNamed:uses:(t3 is shadowed)
TraitsResource>>tearDown(t1 is shadowed)
TraitsResource>>tearDown(t2 is shadowed)
TraitsResource>>tearDown(t3 is shadowed)
TraitsResource>>createClassNamed:superclass:uses:(t4 is shadowed)
WeakFinalizersTest>>testNewFinalizationSupported (WeakFinalizationList
is Undeclared)
WeakFinalizersTest>>testNewFinalizationSupported (WeakFinalizerItem is
Undeclared)
Time for Heap: 97 msecs
Time for SortedCollection: 154 msecs
Time for heap-sort: 118 msecs
Time for quick-sort: 81 msecs
Time for merge-sort: 99 msecs
WeakFinalizersTest>>testNewFinalizationSupported (WeakFinalizationList
is Undeclared)
WeakFinalizersTest>>testNewFinalizationSupported (WeakFinalizerItem is
Undeclared)
WeakFinalizersTest>>testNewFinalizationSupported (WeakFinalizationList
is Undeclared)
WeakFinalizersTest>>testNewFinalizationSupported (WeakFinalizerItem is
Undeclared)
WeakFinalizersTest>>testNewFinalizationSupported (WeakFinalizationList
is Undeclared)
WeakFinalizersTest>>testNewFinalizationSupported (WeakFinalizerItem is
Undeclared)

> Grab it here: http://ftp.squeak.org/trunk/Squeak4.2-10548-alpha.zip
> Brave and noble souls will download it on various platforms, run the
> tests, and generally make sure the image is good.

Stumbled right away on a bug: fix attached
(variables theOne, left at nil, seems a leftover from a previous
implementation)

best,

Stef

On 2010/09/22 09:31, Frank Shearar wrote: <snip>
> DecompilerTests>>#testDecompilerInClassesBAtoBM

The (first, at least) failure in this test is BasicButton label:font:,
where the parameter names and local variables have differing names:

label: t1 font: t2
        | t4 t5 |
        (self findA: StringMorph)
                ifNotNil: [:t3 | t3 delete].
        "rest of method body snipped"

versus

label: t1 font: t2
        | t3 t4 |
        (self findA: StringMorph)
                ifNotNil: [:t5 | t5 delete].
        "rest of method body snipped"

Otherwise, the two versions are isomorphic.

Maybe the test's being too restrictive and, instead of comparing the
methods as text, we need to compare the bytecodes? Oh, heh. Someone put
two handy comments in there showing the diffs between the sources and
the diffs between the bytecodes.

So the (non-)issue here is just that the temps are reordered.

frank

On 2010/09/22 10:53, Frank Shearar wrote:
> On 2010/09/22 09:31, Frank Shearar wrote:
>> On 2010/09/22 05:31, Casey Ransberger wrote:
>>> Grab it here: http://ftp.squeak.org/trunk/Squeak4.2-10548-alpha.zip

OK, new experiment:
* Start up the image,
* Run all tests,
* Run all tests _again_.

The failing Decompiler tests now all pass.

frank

On Wed, 22 Sep 2010, Frank Shearar wrote:

No they don't. They just don't run, because those are LongTestCases and
there's a test, which turns off the preference that enables the running
of LongTestCases. So after the first run, the preference is off, therefore
LongTestCases won't be run the second time.


Levente

>
> frank
>
>

Wait, what? The second time you run, you get less tests?

That's completely counter intuitive. Tests should never affect future runs... seems like someone left off a tearDown method somewhere.

On Sep 22, 2010, at 6:53 AM, Levente Uzonyi <[hidden email]> wrote:

I'm surprised too.

I can see the point - long-running tests, well, take too long. Could we
have some kind've warning though? "35 long-running tests not run."

frank

On 2010/09/22 16:25, Casey Ransberger wrote:

Hello Casey

7 weeks have passed and 100 updates have come in since yo did the
trunk image mentioned in the subject line.

The new upddate include the ones which make it possible to have the
OmniBrowser with refactoring and Ocompletion.

So my question -- would it be possible that you could do another trunk
image this or next week?

As for the Hudson build server effort. Bert tried to install the Java
runtime on the Squeak server but ran into problems doing so on the
Debian box we currently have. If Java is there setting up Hudson may
be done pretty fast -- Alexander Lazarević set it up quickly on his
laptop based on the documentation by Lukas [*]

So as a summary -- wee need another build done manually.

And please do
MCWorkingCopy flushObsoletePackageInfos   [**]
and rebuild the menus, so that the App menu is already there

Regards

Hannes

[*]
> Yannis Hudson Server at http://hudson.jooshr.org
> Lukas Hudson Server at http://hudson.lukas-renggli.ch
> Sources and Doku at http://github.com/renggli/builder

[**]
See thread Obsolete PackageInfos started by Bernhard Pieber , Oct 31,
2010  -- Bert's answer.



On 9/22/10, Casey Ransberger <[hidden email]> wrote:

I'll start looking for some time to do a test run and cut an image. Not sure exactly when I'll be able to get it done, but I've heard you.

Have you run the tests lately? How are they looking?

On Nov 2, 2010, at 2:37 AM, Hannes Hirzel <[hidden email]> wrote: