Release candidate Squeak4.4-12319 ready

* http://ftp.squeak.org/4.4/Squeak4.4-12319.tgz
* http://ftp.squeak.org/4.4/Squeak4.4-12319.zip

This just
* remerges Eliot's findA... fix
* corrects (finally! thanks to Klaus for the reminder!) the URL to
Andreas Raab's Community Supported Packages link

frank

3272 run, 3228 passes, 18 expected failures, 16 failures, 10 errors, 0 unexpected passes
LocaleTest>>#testLocaleChanged
MCChangeNotificationTest>>#testCoreMethodModified
MCFileInTest>>#testStWriter
MCPackageTest>>#testUnload
MCPatchTest>>#testPatchContents
MCSnapshotBrowserTest>>#testNoSelection
MCSnapshotTest>>#testCreation
MCStWriterTest>>#testOrganizationDefinition
MCWorkingCopyTest>>#testBackport
MCWorkingCopyTest>>#testOptimizedLoad
MCWorkingCopyTest>>#testRepeatedMerge
MCWorkingCopyTest>>#testSelectiveBackport
MCWorkingCopyTest>>#testSimpleMerge
MCWorkingCopyTest>>#testSnapshotAndLoad
SocketTest>>#testSocketReuse
SocketTest>>#testUDP


MCChangeNotificationTest>>#testExtMethodModified
MCSnapshotBrowserTest>>#testCategorySelected
MCSnapshotBrowserTest>>#testClassSelected
MCSnapshotBrowserTest>>#testClassSideClassSelected
MCSnapshotBrowserTest>>#testComment
MCSnapshotBrowserTest>>#testMethodIsCleared
MCSnapshotBrowserTest>>#testMethodSelected
MCSnapshotBrowserTest>>#testProtocolIsCleared
MCSnapshotBrowserTest>>#testProtocolSelected
MCStWriterTest>>#testInitializerDefinition

Image
-----
/Users/eglenpaling/Documents/Smalltalk/Squeak 4.4 Testing/Squeak4.4-12317 Cog/Squeak4.4-12317.image
Squeak4.4
latest update: #12317
Current Change Set: Unnamed
Image format 6505 (32 bit)

Virtual Machine
---------------
/Users/eglenpaling/Documents/Smalltalk/Build/VMs/Cog 4.0.2637.app/Contents/MacOS/Croquet
Croquet Closure Cog VM [CoInterpreter VMMaker.oscog-eem.238] Croquet Cog 4.0.2637
Mac OS X built on Dec 17 2012 19:54:57 Compiler: 4.2.1 (Apple Inc. build 5666) (dot 3)
platform sources revision VM: r2637 http://www.squeakvm.org/svn/squeak/branches/Cog Plugins: r2545 http://squeakvm.org/svn/squeak/trunk/platforms/Cross/plugins
CoInterpreter VMMaker.oscog-eem.238 uuid: a3d10eab-6079-4c91-99f6-3dcf58d1446f Dec 17 2012
StackToRegisterMappingCogit VMMaker.oscog-eem.234 uuid: 66acafd1-cad0-4f20-b786-ab8f48201d82 Dec 17 2012

Operating System/Hardware
-------------------------
Mac OS 1068 intel

On 19 December 2012 19:42, glenpaling <[hidden email]> wrote:
How exactly do you run the tests? Download image, fire it up, open
TestRunner, run tests... and nothing else?

When you run the tests, I'll bet
MCMethodDefinitionTest>>testLoadAndUnload brings up a FITBM asking for
initials, and that breaks all these tests.
http://bugs.squeak.org/view.php?id=7700 will fix this, when it's done.

frank

3272 run, 3222 passes, 23 expected failures, 16 failures, 11 errors, 0 unexpected passes

LocaleTest>>#testLocaleChanged
MCChangeNotificationTest>>#testCoreMethodModified
MCFileInTest>>#testStWriter
MCPackageTest>>#testUnload
MCPatchTest>>#testPatchContents
MCSnapshotBrowserTest>>#testNoSelection
MCSnapshotTest>>#testCreation
MCStWriterTest>>#testOrganizationDefinition
MCWorkingCopyTest>>#testBackport
MCWorkingCopyTest>>#testOptimizedLoad
MCWorkingCopyTest>>#testRepeatedMerge
MCWorkingCopyTest>>#testSelectiveBackport
MCWorkingCopyTest>>#testSimpleMerge
MCWorkingCopyTest>>#testSnapshotAndLoad
ProcessTest>>#testAtomicSuspend
Win32VMTest>>#testWinVM3ButtonMousePreference

FileDirectoryTest>>#testRelativeNameIfAbsoluteFor
MCChangeNotificationTest>>#testExtMethodModified
MCSnapshotBrowserTest>>#testCategorySelected
MCSnapshotBrowserTest>>#testClassSelected
MCSnapshotBrowserTest>>#testClassSideClassSelected
MCSnapshotBrowserTest>>#testComment
MCSnapshotBrowserTest>>#testMethodIsCleared
MCSnapshotBrowserTest>>#testMethodSelected
MCSnapshotBrowserTest>>#testProtocolIsCleared
MCSnapshotBrowserTest>>#testProtocolSelected
MCStWriterTest>>#testInitializerDefinition

Image
-----
D:\Documents and Settings\palingg\My Documents\Programming\Smalltalk\Squeak 4.4\Squeak4.4-12319 interp\Squeak4.4-12319.image
Squeak4.4
latest update: #12319
Current Change Set: Unnamed

Virtual Machine
---------------
D:\Documents and Settings\palingg\My Documents\Programming\Smalltalk\Squeak 4.4\SqueakVM-Win32-4.1.1-bin\Squeak4.1.1.exe
Squeak3.10.2 of 11 February 2010 [latest update: #9314]
Win32 built on Jul 27 2010 20:35:19 Compiler: 2.95.2 19991024 (release)

Operating System Details
------------------------
Operating System: Microsoft Windows XP (Build 2600 Service Pack 3)
Registered Owner:
Registered Company:
SP major version: 3
SP minor version: 0
Suite mask: 100
Product type: 1

Frank Shearar-3 wrote

On 19 December 2012 19:42, glenpaling <[hidden email]> wrote:
> 3272 run, 3228 passes, 18 expected failures, 16 failures, 10 errors, 0
 snip

How exactly do you run the tests? Download image, fire it up, open
TestRunner, run tests... and nothing else?

When you run the tests, I'll bet
MCMethodDefinitionTest>>testLoadAndUnload brings up a FITBM asking for
initials, and that breaks all these tests.
http://bugs.squeak.org/view.php?id=7700 will fix this, when it's done.

frank

Since you mentioned it a few days ago, I've been setting the author initials before running the tests.

Other than CI, has anyone had a successful run of these tests?

I stuck the release candidate into the Squeak 4.3 All-In-One release to run on it's interpreter VM

Test Runner

3272 run, 3222 passes, 23 expected failures, 17 failures, 10 errors, 0 unexpected passes
LocaleTest>>#testLocaleChanged
MCChangeNotificationTest>>#testCoreMethodModified
MCFileInTest>>#testStWriter
MCPackageTest>>#testUnload
MCPatchTest>>#testPatchContents
MCSnapshotBrowserTest>>#testNoSelection
MCSnapshotTest>>#testCreation
MCStWriterTest>>#testOrganizationDefinition
MCWorkingCopyTest>>#testBackport
MCWorkingCopyTest>>#testOptimizedLoad
MCWorkingCopyTest>>#testRepeatedMerge
MCWorkingCopyTest>>#testSelectiveBackport
MCWorkingCopyTest>>#testSimpleMerge
MCWorkingCopyTest>>#testSnapshotAndLoad
ProcessTest>>#testAtomicSuspend
SocketTest>>#testSocketReuse
SocketTest>>#testUDP

MCChangeNotificationTest>>#testExtMethodModified
MCSnapshotBrowserTest>>#testCategorySelected
MCSnapshotBrowserTest>>#testClassSelected
MCSnapshotBrowserTest>>#testClassSideClassSelected
MCSnapshotBrowserTest>>#testComment
MCSnapshotBrowserTest>>#testMethodIsCleared
MCSnapshotBrowserTest>>#testMethodSelected
MCSnapshotBrowserTest>>#testProtocolIsCleared
MCSnapshotBrowserTest>>#testProtocolSelected
MCStWriterTest>>#testInitializerDefinition


Image
-----
/Users/eglenpaling/Documents/Smalltalk/Squeak 4.4 Testing/Squeak-4.3-All-in-One with 4.4RC inserted.app/Contents/Resources/Squeak4.4-12317.image
Squeak4.4
latest update: #12317
Current Change Set: Unnamed

Virtual Machine
---------------
/Users/eglenpaling/Documents/Smalltalk/Squeak 4.4 Testing/Squeak-4.3-All-in-One with 4.4RC inserted.app/Contents/MacOS/Squeak VM Opt
Squeak3.8.1 of '28 Aug 2006' [latest update: #6747] 4.3
Mac Carbon 4.2.4b1 28-Mar-10 >45CAAEAC-5A1E-4327-9702-7973E3473FDE<

Operating System/Hardware
-------------------------
Mac OS 1068 intel

> * http://ftp.squeak.org/4.4/Squeak4.4-12319.zip

Fixes the NetNameResolver problem on localhost; thanks !

Stef

> How exactly do you run the tests? Download image, fire it up, open
> TestRunner, run tests... and nothing else?

Go to Help | About this System and then select the "SUnit" category.

> When you run the tests, I'll bet
> MCMethodDefinitionTest>>testLoadAndUnload brings up a FITBM asking for
> initials, and that breaks all these tests.

Why do you say that's breaking the tests?  If you "Do It" on one of
the failed MC tests it opens the debugger showing the failed assertion
even though initials have already been set.

These tests weren't failing in 4.3 anyone know what happened?

> http://bugs.squeak.org/view.php?id=7700 will fix this, when it's done.

The way to handle this is for the "automation" (i.e., CI or whatever)
to handle the appropriate UIManager notification (sorry can't remember
what its called at the moment).

On 19 December 2012 21:05, Stéphane Rollandin <[hidden email]> wrote:
>> * http://ftp.squeak.org/4.4/Squeak4.4-12319.zip
>
>
> Fixes the NetNameResolver problem on localhost; thanks !

Excellent!

One thing I noticed was the ChangeSet looking wonky. Now
ReleaseBuilderFor4dot4 class >> #prepareNewBuild calls #cleanUp: on
anything that understands it, which means that, in particular, it
calls ChangeSet cleanUp: true. Nevertheless, the Simple Change Sorter
tells a different story. If you then run ChangeSet cleanUp: true
yourself, the SCS shows what I'd expect: an empty Unnamed1.

frank

On 19 December 2012 21:15, Chris Muller <[hidden email]> wrote:
If the author initials aren't set, you get a FITBM asking for it. The
proper cleanup and restoration of MC's test state will thus not
happen. In particular, MCMockClassA does not get its #one method back,
which later tests expect.

> These tests weren't failing in 4.3 anyone know what happened?

People usually don't run tests without author initials? Certainly
these tests have ALWAYS passed on CI, but CI sets the author initials.

I have yet to see evidence to contradict my hypothesis.

>> http://bugs.squeak.org/view.php?id=7700 will fix this, when it's done.
>
> The way to handle this is for the "automation" (i.e., CI or whatever)
> to handle the appropriate UIManager notification (sorry can't remember
> what its called at the moment).

You can supply answers to the UIManager queries. One hacky way to
solve 7700 is to set the initials if not set, and restore them after
the test. Or one could use a dynamic variable, or wrap the #perform:
that runs the test with these answering things, or something else.

At any rate, unless someone can come up with an alternative
hypothesis, or a way to falsify mine, I don't think this is reason to
delay 4.4's release.

frank

Eh, well I just opened the 4.3 release image and ran tests.  It asked
for my initials at the beginning and then ran all tests.  Those MC
tests didn't fail.  I don't think we can release without understanding
what's going on with them in 4.4.

Would someone volunteer?  I'm currently looking at Dave's SqueakMap
patch is needed now but not before and making sure we can, in fact,
deploy packages for 4.4 that will show up in SqueakMaps 4.4 list.

On 19 December 2012 22:15, Chris Muller <[hidden email]> wrote:
Yes, but did you _fill in_ the initials?

At any rate, theories aside, it _is_ the case that the
MCMethodDefinitionTest >> #tearDown doesn't restore state correctly.
Note that this is #testLoadAndUnload test is the test that pops up the
prompt. Subsequent tests fail because, for at least some of them,
MCMockClassA >> #one no longer exists. (MCMethodDefinitionTest >>
#testLoadAndUnload removes it.)

> Would someone volunteer?  I'm currently looking at Dave's SqueakMap
> patch is needed now but not before and making sure we can, in fact,
> deploy packages for 4.4 that will show up in SqueakMaps 4.4 list.

Yes, please. I would like some countervailing evidence and/or
alternate hypotheses.

Nevertheless, these tests definitely run on CI, and definitely pass.
Have always passed, in fact. It was only glenpaling's recent findings
that brought these tests to my/our attention.

frank

On 19 December 2012 22:44, Frank Shearar <[hidden email]> wrote:
OK right. It's something more than just author initials. Taking the
RC, it doesn't matter if you set the author initials or not. Right,
sorry Glen, you're quite right. My hypothesis is wrong.

frank

In looking at where one test fails I see that

    MCSnapshotResource current snapshot definitions

is empty in the 4.4 image while it has a bunch of entries in the 4.3 image.

Cheers,
Bob

On 19 December 2012 22:15, Chris Muller [hidden email] wrote:

Why do you say that's breaking the tests?  If you "Do It" on one of
the failed MC tests it opens the debugger showing the failed assertion
even though initials have already been set.

If the author initials aren't set, you get a FITBM asking for it. The
proper cleanup and restoration of MC's test state will thus not
happen. In particular, MCMockClassA does not get its #one method back,
which later tests expect.

These tests weren't failing in 4.3 anyone know what happened?

People usually don't run tests without author initials? Certainly
these tests have ALWAYS passed on CI, but CI sets the author initials.

I have yet to see evidence to contradict my hypothesis.

Eh, well I just opened the 4.3 release image and ran tests.  It asked
for my initials at the beginning and then ran all tests.  Those MC
tests didn't fail.  I don't think we can release without understanding
what's going on with them in 4.4.

Yes, but did you _fill in_ the initials?

At any rate, theories aside, it _is_ the case that the
MCMethodDefinitionTest >> #tearDown doesn't restore state correctly.
Note that this is #testLoadAndUnload test is the test that pops up the
prompt. Subsequent tests fail because, for at least some of them,
MCMockClassA >> #one no longer exists. (MCMethodDefinitionTest >>
#testLoadAndUnload removes it.)

Would someone volunteer?  I'm currently looking at Dave's SqueakMap
patch is needed now but not before and making sure we can, in fact,
deploy packages for 4.4 that will show up in SqueakMaps 4.4 list.

Yes, please. I would like some countervailing evidence and/or
alternate hypotheses.

Nevertheless, these tests definitely run on CI, and definitely pass.
Have always passed, in fact. It was only glenpaling's recent findings
that brought these tests to my/our attention.

frank

4.4 MCSnapshotResource mockPackageName ==> 'MonticelloMocks'

4.3 MCSnapshotResource mockPackageName ==> #'Tests-Monticello-Mocks'

which seems to be due to

'From Squeak4.4 of 18 December 2012 [latest update: #12305] on 19
December 2012 at 9:43:45 pm'!

!MCMockPackageInfo methodsFor: 'as yet unclassified' stamp: 'cwp
8/1/2003 20:31'!
packageName
     ^ 'MonticelloMocks'! !


No such method exists in 4.3, so some code pieces the package name together.

Cheers,
BOb

Well, that's cool. I also get 3 failures, only they are:

FileStreamTest>>#testPositionPastEndIsAtEnd
LargeNegativeIntegerTest>>#testMinimumNegativeIntegerArithmetic
LocaleTest>>#testLocaleChanged

Cheers,
Bob

On Wed, Dec 19, 2012 at 9:44 PM, Bob Arning <[hidden email]> wrote:

4.4 MCSnapshotResource mockPackageName ==> 'MonticelloMocks'

4.3 MCSnapshotResource mockPackageName ==> #'Tests-Monticello-Mocks'

which seems to be due to

'From Squeak4.4 of 18 December 2012 [latest update: #12305] on 19 December 2012 at 9:43:45 pm'!

!MCMockPackageInfo methodsFor: 'as yet unclassified' stamp: 'cwp 8/1/2003 20:31'!
packageName
    ^ 'MonticelloMocks'! !


No such method exists in 4.3, so some code pieces the package name together.

Yeah. This goes back to some changes made to PackageInfo last year. PackageInfo subclasses are no longer detected in by PackageInfo class>>named:, they have to be explicitly registered. If I execute "MCMockPackageInfo initialize" all the errors go away, and I get only 3 failures:

LocaleTest>>#testLocaleChanged

SocketTest>>#testSocketReuse

SocketTest>>#testUDP

Colin

On 20 December 2012 02:55, Colin Putney <[hidden email]> wrote:
So I tried an experiment last night where I manually re-added
MCMockClassA >> #one, which failed dismally. Would adding
"MCMockPackageInfo initialize" be the right thing to do in MCTestCase
>> #tearDown? (That seems kind've strange; I don't see why that would
re-add a deleted method, for instance. Unless initialising wiped out
local changes... I'll try this out on the train to work today.)

> LocaleTest>>#testLocaleChanged
> SocketTest>>#testSocketReuse
> SocketTest>>#testUDP

And these are expected: the top test's broken, and the bottom two do
break on some platforms.

frank