[squeak-dev] Re: Mantis reports? [was: Completing the 3.10 release process]

Klaus D. Witzel klaus.witzel at cobss.com
Fri Apr 11 21:00:28 UTC 2008


On Fri, 11 Apr 2008 21:59:31 +0200, Edgar J. De Cleene wrote:

>
> El 4/11/08 2:52 PM, "Klaus D. Witzel" escribió:
>
>> The wiki page mentions http://bugs.squeak.org/view.php?id=6336 which is
>> still open on Mantis. I confirm I can see the patch at work in Damien's
>> 3.10 based Squeak-dev .images.
>>
>
> If you read the History.morph (no need to go Mantis) found this
>>> 7125Unicode-1-kwl.cs
>>> Reporter: kwl
>>> Summary:    0006336: Unicode membership test reduced to String #=  
>>> (String
>>> #compare:caseSensitive:) ?
>>> Description     The fix below has been included in the OLPC image.
>>>
>>> I suggest to include the fix into mainstream Squeak too, so if in the  
>>> future
>>> there are other fixes and improvements to I18N they can be applied to  
>>> both
>>> branches.
>>>
>>> The files must be loaded in order:
>>>
>>> CodeInCs ReleaseBuilderFor3dot10 new loadTogether: (Array with:
>>> 'Multilingual-edc.27.mcz') merge: false.
>>> ReleaseBuilderFor3dot10 new updatePackages:  
>>> 'Collections-edc.87(86).mcd'
>
> So your contribution, marked as 3.9 in Mantis , should be in all 3.10
> derivate images from 7125 and newer.

Right, but this is *not* the concern. I just wanted to ask/say something  
about the Mantis report(s) which are not *closed* as part of the  
"completing the 3.10 release process" subject matter.

Will you close http://bugs.squeak.org/view.php?id=6336 ?

> I miss add a note in Mantis at the time but send feedback ....
>
>> Do you want me to check http://wiki.squeak.org/squeak/5975 against  
>> Mantis?
>>
>> /Klaus
>
> I wish (not want !) more needed FIX and ENH from you ,

I try to do what I can :)

> this time with
> appropriate test showing your fix works

I'm sorry that there are none which test Unicode character membership. I  
mentioned that in my mail to the original author and, since he already  
included my patches in the OLPC Squeak release I believe that he had a way  
to validate them.

> and not break any of 2200 plus test we have now :=)
>
> Edgar
>




More information about the Squeak-dev mailing list