Re: GDL2: prototypes and EOAttribute.m

[Mark Clements]

Note that WebObjects​ 5 uses EOCustomObject: 

http://wocommunity.org/documents/javadoc/WebObjects/5.4.2/com/webobjects/eocontrol/EOCustomObject.html

Changing from NSObject categories in EOF 4.5 to using EOCustomObject in
WebObjects 5 may reflect the Java implementation. Perhaps the "hack"
relates to the NSObject categories - is this considered to be a bad idea?

Irrespectively, if the principle is to re-implement EOF 4.5, then
reverting the patch is sensible - but this will involve substantial work.

-- Mark

On 06/09/2017 05:42 PM, Matt Rice wrote:
> yes, EOModel/EOEntity/EOAttribute do not inherit from EOCustomObject,
>
> previously, there was no need to inherit from EOCustomObject, as in
> you could take any type of NSObject, add some categories to it, and
> have a working thing you could use with EOF...
>
> If you look at the documentation for WO 4.5
>
> http://mirror.informatimago.com/next/developer.apple.com/documentation/LegacyTechnologies/WebObjects/WebObjects_4.5/System/Library/Frameworks/EOControl.framework/ObjC_classic/Protocols/EOEnterpriseObject.html#//apple_ref/occ/intfm/EOEnterpriseObject/awakeFromInsertionInEditingContext:
>
> "NSObject's implementation merely sends ..."
> for some reason I never fully understood, why this was changed,
> the ChangeLog says:
>
>         * EOControl/EOEditingContext.m: remove hacks, adapt to EOCustomObject
>         * EOControl/EOCustomObject.h/m: added code here
>
> I consider these real failures, since the removal of the hacks, gdl2
> no longer even conforms to the class hierarchy e.g..
>
> http://mirror.informatimago.com/next/developer.apple.com/documentation/LegacyTechnologies/WebObjects/WebObjects_4.5/System/Library/Frameworks/EOControl.framework/Java/Classes/EOClassDescription.html
>
> https://github.com/gnustep/gdl2/commit/51b22fda000b1046a5443d956b0be951da000381
>
> my opinion would be to revert the patch.
>
> On Fri, Jun 9, 2017 at 6:14 AM, Mark Clements <mark.clements@ki.se> wrote:
>> Next up for the gdl2 testsuite, the EOModel/ModelEditing.m file fails with:
>>
>> "EOModel(instance) does not recognize
>> awakeFromInsertionInEditingContext:" from line 80 calling the
>> -insertObject method. Line 83 would also fail with "EOEntity(instance)
>> does not recognize awakeFromInsertionInEditingContext:", and line 89
>> would fail with "EOAttribute(instance) does not recognize
>> awakeFromInsertionInEditingContext:".
>>
>> The -insertObject method should call [EOCustomObject
>> -awakeFromInsertionInEditingContext:], but I can't see where EOModel,
>> EOEntity or EOAttribute inherits from EOCustomObject. Sub-classes of
>> EOCustomObject include EOGenericObject and EODelayedObserver.
>>
>> Any suggestions? Can these tests be removed?
>>
>> Kindly, Mark.
>>
>> On 06/07/2017 11:29 PM, Mark Clements wrote:
>>> Next up: the gdl2 testsuite was failing for some files with nested
>>> TEST_SET's, with errors for "message sent to deallocated instance". Is
>>> it obvious why the nesting does this?
>>>
>>> If we remove the nested sets, then we resolve the "deallocated instance"
>>> errors. Proposed patch attached.
>>>
>>> -- Mark
>>>
>>> On 06/06/2017 05:34 PM, Mark Clements wrote:
>>>> See attached for an improved patch on quoting. This gives similar tests
>>>> successes for both quoted and unquoted external names.
>>>>
>>>> -- Mark
>>>>
>>>> On 06/05/2017 10:00 PM, Mark Clements wrote:
>>>>> Please see attached for a patch to fix quoting of column names and
>>>>> Postgresql sequences. I have also included a few formatting changes.
>>>>>
>>>>> Kindly, Mark.
>>>>>
>>>>>