I have the same issue, and the file has mysteriously appeared in my
image directory.  Usually you can Shift+Command+e to see references to
substrings in the code, but doing that for FontJapaneseEnvironment.sar
does not produce any results.  Strange indeed..!

On Sat, Jan 8, 2011 at 8:56 AM, Bernhard Pieber <bernhard at pieber.com> wrote:
> It is the LocaleTest. When I run it all the tests are green but the debugger appears. I have a FontJapaneseEnvironment.sar in the folder where my image resides. I have no idea where I got that from. When I remove it and run the LocaleTest the debugger does not appear.
>
> Cheers,
> - Bernhard
>
> Am 08.01.2011 um 15:27 schrieb Bernhard Pieber:
>> Hi Andreas,
>>
>> Thanks for the tips. Inserting self halt shows that it is called from the preamble of FontJapaneseEnvironment.sar by the SARInstaller. Where that is called from I have yet to figure out. Strangely my image hung after proceeding from the halt. Alt+. does not work.
>>
>> Am 08.01.2011 um 12:02 schrieb Andreas Raab:
>>> This might be the result of a fileIn (possibly one of the LanguageEnvironment tests). You can try two things:
>>> 1) Check your recent changes if anything got logged. A fileIn from loading a project file or so should be logged.
>>> 2) Insert a "self halt" into the method (before the fork) to see where it comes from.
>>> On 1/8/2011 11:47 AM, Bernhard Pieber wrote:
>>>> I did the following:
>>>> 1. I opened the latest trunk image Squeak4.2-10779-alpha.image with the Mac OS X VM Squeak 5.8b12.
>>>> 2. Update Squeak.
>>>> 3. Run all tests. After all the tests I get a debugger with "Error: Instances of StrikeFont are not indexable"
>>>>
>> snip
>>>> Is there a way find out where this process is created? Tracing senders of StrikeFontSet class>>installExternalFontOn:encoding:encodingName:textStyleName: leads ultimately to no senders. In a Process Browser I do not see anything unusual.
>
>
>