I will try with a newer version of ICU, if there is one, and see what that does. I think we should turn off ICU use on Windows by default until we locate the root cause of the issue.
GC On Wed, Nov 28, 2012 at 5:39 PM, Stefan Bidi <[email protected]> wrote: > I think that might be the case. The only thing I can think of is that a > call to malloc is failing inside ICU. > On Nov 28, 2012 11:34 AM, "Gregory Casamento" <[email protected]> > wrote: > >> Sorry, forgot to forward to the list... >> >> GC >> >> ---------- Forwarded message ---------- >> From: Gregory Casamento <[email protected]> >> Date: Wed, Nov 28, 2012 at 12:31 PM >> Subject: Re: Issue with NSDateFormatter on Windows with ICU... >> To: Stefan Bidi <[email protected]> >> >> >> Yes, that's correct I am using mingw. Here is the information: >> >> >> heron@glados-vm-win7 ~ >> $ gcc test_udat.c -lpthread -lm -L/mingw/bin -licui18n -licuuc >> -licudata -l >> pthread -lm >> >> heron@glados-vm-win7 ~ >> $ a.exe >> (1) Error Code: U_MEMORY_ALLOCATION_ERROR >> (2) Error Code: U_MEMORY_ALLOCATION_ERROR >> >> heron@glados-vm-win7 ~ >> $ icu-config --version >> 4.6 >> >> heron@glados-vm-win7 ~ >> $ >> >> >> On Wed, Nov 28, 2012 at 12:20 PM, Stefan Bidi <[email protected]>wrote: >> >>> Assuming you are running the gnustep mingw environment, you should be >>> able to use icu-config --ldflags. I can't remember where icu-config is on >>> windows and I no longer have a windows install to check. >>> On Nov 28, 2012 11:14 AM, "Gregory Casamento" <[email protected]> >>> wrote: >>> >>>> Sorry I was delayed on getting back to this. Work and sleep >>>> interfered. I ran the tests, same issue. The problem is in the >>>> initialization code of NSDateFormatter itself, not in how Gorm is >>>> initializing it. >>>> >>>> I got your test program... the only issue is that I can't figure out >>>> WHERE ICU is installed on windows so that I can give the correct directive >>>> to gcc via -L. Any idea? >>>> >>>> GC >>>> >>>> >>>> On Tue, Nov 27, 2012 at 9:10 PM, Stefan Bidi <[email protected]>wrote: >>>> >>>>> Greg, >>>>> Can you compile and run the attached test. It is really simple and >>>>> will say if the ICU lib on your system is working or not. >>>>> >>>>> On Tue, Nov 27, 2012 at 6:53 PM, Stefan Bidi <[email protected]> >>>>> wrote: >>>>> > Greg, what happens if you run the NSDateFormatter tests? Does >>>>> > everything still crash? >>>>> > >>>>> > I'll prepare a set of tests to exercise the ICU functionality. That >>>>> > might help figure out what is going on here. Which version of the >>>>> ICU >>>>> > and base are you using? >>>>> > >>>>> > On Tue, Nov 27, 2012 at 3:43 PM, Gregory Casamento >>>>> > <[email protected]> wrote: >>>>> >> Same result. >>>>> >> >>>>> >> >>>>> >> On Tue, Nov 27, 2012 at 2:02 PM, Stefan Bidi <[email protected]> >>>>> wrote: >>>>> >>> >>>>> >>> No, this is correct. ICU requires Unicode for time zones. >>>>> >>> >>>>> >>> The thing is that ICU does not recognize the Windows timezone >>>>> names, they >>>>> >>> must be in the Olson format. >>>>> >>> >>>>> >>> A work around might be to set that to null for now and see if that >>>>> works. >>>>> >>> >>>>> >>> On Nov 27, 2012 12:25 PM, "Richard Frith-Macdonald" >>>>> >>> <[email protected]> wrote: >>>>> >>>> >>>>> >>>> >>>>> >>>> On 27 Nov 2012, at 18:18, Gregory Casamento wrote: >>>>> >>>> >>>>> >>>> > >>>>> >>>> > Here is what is being sent.... >>>>> >>>> > >>>>> >>>> > (gdb) p *(NSDateFormatterInternal *)_internal >>>>> >>>> > $3 = {{isa = 0x6702f740}, _behavior = 0, _locale = 0x41f9a28, >>>>> >>>> > _tz = 0x278ce08, _timeStyle = 0, _dateStyle = 0, _formatter = >>>>> 0x0} >>>>> >>>> > (gdb) p NSToUDateFormatStyle(0) >>>>> >>>> > $4 = -1 >>>>> >>>> > (gdb) po ((NSDateFormatterInternal *)_internal)->_locale >>>>> >>>> > en_US >>>>> >>>> > (gdb) p tzID >>>>> >>>> > $6 = ( >>>>> >>>> > UChar *) 0x4202b80 "E\000a\000s\000t\000e\000r\000n\000 >>>>> >>>> > \000S\000t\000a\000n >>>>> >>>> > \000d\000a\000r" >>>>> >>>> >>>>> >>>> The thing that strikes me instantly is that we are looking at >>>>> unicode >>>>> >>>> (16bit characters). >>>>> >>>> Could it be that the ICU library routine is expecting ascii or >>>>> utf-8, but >>>>> >>>> because windows uses 16bit characters natively it's getting >>>>> confused? >>>>> >>>> If so, that would be easy to fix. >>>>> >>>> >>>>> >> >>>>> >> >>>>> >> >>>>> >> -- >>>>> >> Gregory Casamento >>>>> >> Open Logic Corporation, Principal Consultant >>>>> >> yahoo/skype: greg_casamento, aol: gjcasa >>>>> >> (240)274-9630 (Cell) >>>>> >> http://www.gnustep.org >>>>> >> http://heronsperch.blogspot.com >>>>> >>>> >>>> >>>> >>>> -- >>>> Gregory Casamento >>>> Open Logic Corporation, Principal Consultant >>>> yahoo/skype: greg_casamento, aol: gjcasa >>>> (240)274-9630 (Cell) >>>> http://www.gnustep.org >>>> http://heronsperch.blogspot.com >>>> >>> >> >> >> -- >> Gregory Casamento >> Open Logic Corporation, Principal Consultant >> yahoo/skype: greg_casamento, aol: gjcasa >> (240)274-9630 (Cell) >> http://www.gnustep.org >> http://heronsperch.blogspot.com >> >> >> >> -- >> Gregory Casamento >> Open Logic Corporation, Principal Consultant >> yahoo/skype: greg_casamento, aol: gjcasa >> (240)274-9630 (Cell) >> http://www.gnustep.org >> http://heronsperch.blogspot.com >> >> _______________________________________________ >> Gnustep-dev mailing list >> [email protected] >> https://lists.gnu.org/mailman/listinfo/gnustep-dev >> >> -- Gregory Casamento Open Logic Corporation, Principal Consultant yahoo/skype: greg_casamento, aol: gjcasa (240)274-9630 (Cell) http://www.gnustep.org http://heronsperch.blogspot.com
_______________________________________________ Gnustep-dev mailing list [email protected] https://lists.gnu.org/mailman/listinfo/gnustep-dev
