Hi Chris,
Can't wait to try this out, but.....
With the Alpha & Beta versions, of both Dictionary & Standard edition, I receive the message below on startup. Appears milliseconds after the flash dialog.
I have Stackz Dictionary Ed. 2006 (full), and it works fine.
(Also tried holding shift, no dice...)
[Microsoft Visual C++ Runtime Library]
Runtime Error!
Program: C:\Program Files\MindD...
This application has requested the Runtime to terminate it in an unusual way.
Please contact the application's support team for more information.
Let me know how I can help try to resolve (feel free to contact by e-mail.)
Running XP Pro SP2, .NET 1.1 & 2.0.
First time I've encountered the error.
Thanks!
Quote from: Abi on May 31, 2007, 02:45:29 AM
[Microsoft Visual C++ Runtime Library]
Runtime Error!
Program: C:\Program Files\MindD...
This application has requested the Runtime to terminate it in an unusual way.
Please contact the application's support team for more information.
This error is new to me, and it's very difficult to give an answer from here. Googling shows that this error used to happen with internet explorer which had the google toolbar installed, and many other applications that had some sort of conflicts with other apps...
http://www.google.ch/search?hl=en&q=Microsoft+Visual+C%2B%2B+Runtime+Library+This+application+has+requested+the+Runtime+to+terminate+it+in+an+unusual+way&btnG=Search&meta=
It would be very interesting to know what is the cause in the case of Stackz, but it might be difficult to find out. Maybe you could try stopping all other running tasks and see if it helps?
If I understand correctly, the application runs fine until you open the test dialog. Is that the only cause of the crash, or does it also happen with other actions? (e.g. open the Learn / Match dialog, open the ListImportWizard, open the Stackz Edit Dialog etc).
Thanks for the response! Yeah, I had done some searching myself. No IE/FX toolbars of any sort here.
Contacted Microsoft & received a hotfix (KB 884538) that mentions the error - I'm wondering now though if it's something that has to be done on the development machine, & not the client's? Makes some mention of compiler switches.. Didn't help at any rate.
Just in case, I also grabbed vcredist.exe (VC Runtime Library (KB 259403)), but it's probably too old to have done anything.
I stopped & closed almost all running processes & services, down to just 11 system processes (trying to execute along the way) with no luck unfortunately.
Actually, I can't get past the flash screen "MindDate Software" "STACKZ" on startup. Before it disappears (but after it fades in) the error appears. Pressing OK terminates the application. Haven't even seen the updated GUI yet :(. Also tried renaming the "Chrtschnbrr" key to no avail - other than it popped the up the registration dialog, then showed the flash screen & the error.
Regmon & Filemon didn't show anything too fishy either, fwiw.. Ah, also temporarily denied read on %systemroot%\WinSxS; no luck.
I'm running out of ideas here, lol.. tasukete~
Thanks for your investigations!
I have seen the hotfix thing, but I'm not using the compiler switches that are mentioned in this article so this can't be it...
I will contact you directly to continue this issue. I'll send you some special diagnostic apps to find out more.
That would be great!!
I await your email.
I posted more in detail about this on the "crash on start up" list, but this is exactly the same problem I am having.
Paul
Hi Paul - I'm glad to hear that I'm not the only one!
Chris - unfortunately still a problem as of Beta 2.
I tried it on another PC, and it worked okay. Comparing the two, the following occurs on the one that crashes.
Exception after/during the load of:
C:\WINDOWS\IME\IMJP8_1\Dicts\IMJPCD.DIC
C++ EH exception - code e06d7363 (first chance)
C++ EH exception - code e06d7363 (!!! second chance !!!)
Initial stack trace:
kernel32!RaiseException+0x52
MSVCR80!CxxThrowException+0x46
image00400000+0x6ce78
image00400000+0x6d151
image00400000+0x4adaa
ntdll!RtlAllocateHeap+0x117
MSVCR80!malloc+0x7a
image00400000+0x23920
image00400000+0x23ab5
image00400000+0xd054b
MFC80U+0x3d31a
kernel32!RegisterWaitForInputIdle+0x49
kernel32!ValidateLocale+0x2b0
kernel32!RegisterWaitForInputIdle+0x52
|
After tracing a bit into the exception:
kernel32!RaiseException+0x52
MSVCR80!CxxThrowException+0x46
MSVCP80!std::_String_base::_Xlen+0x3e
MSVCP80!std::basic_string<wchar_t,std::char_traits<wchar_t>,std::allocator<wchar_t> >::append+0x17
MSVCP80!std::basic_string<unsigned short,std::char_traits<unsigned short>,std::allocator<unsigned short> >::resize+0x21
MSVCP80!std::basic_string<wchar_t,std::char_traits<wchar_t>,std::allocator<wchar_t> >::resize+0xb
|
Maybe all red-herrings, maybe useless, but just in case it helps.
Also to add to the list of things-tried:
Uninstalled & reinstalled the IME. Removed custom IME dictionaries. Copied .DLLs (All VC++ related MFC & CRT .DLLs (\WinSxS)) and IME files from working PC to Stackz install dir (still reads some from Windows source however during startup.) Installed the VC++ 2005 redistributable package (in vain).
Quote from: Abi on June 02, 2007, 01:25:11 AM
That would be great!!
I await your email.
Have you seen the email I sent you?
I created a special version (not the beta 2!) that should allow us to find the origin of the problem. Please follow the link in the email I sent you.
Quote from: Abi on May 31, 2007, 02:45:29 AM
Running XP Pro SP2, .NET 1.1 & 2.0.
Maybe the ".NET 1.1 & 2.0" part is the reason?
In some article on a different forum someone had a problem where this C++ EH exception - code e06d7363 is involved as well.
He could solve it after repairing his .NET 2.0 setup... (see very bottom)
http://forum.sysinternals.com/forum_posts.asp?TID=8829&PN=2 (http://forum.sysinternals.com/forum_posts.asp?TID=8829&PN=2)
Edit:Another (very recent!) post from the sysinternals forum:
Quote
Posted: 14 May 2007 at 8:25am
YES!!! It works. That was in fact the problem. I've installed .NET v2 with "repair" option and everything went OK after that.
I didn't know that the code 0x80040154 was "Class not registered". Thanks for that one.
http://forum.sysinternals.com/forum_posts.asp?TID=9636 (http://forum.sysinternals.com/forum_posts.asp?TID=9636)
Who knows, maybe that helps? I don't have .NET 2.0...
I can't test it on my system, but chances are not bad that with the latest fix in Stackz 07 beta 3 this problem is solved as well. Please let us know what happens on your computer!
beta 3 did not fix the problem for me.
Paul King
Too bad - I hoped it would.
Currently only Abi and you report this "terminate in unusual way" problem. Since it only seems to appear on your machines, it's difficult for me to find the reason...
I am working with Abi to debug the stuff on his PC. It will take a while though...
Is anybode else besides Abi and kanemiya experiencing this "terminate in an unusual way"? ???
I just uninstalled Beta 3 and installed Beta 4 and the "terminate it in an unusual way ..." error still persists. I hope you can resolve this as I am looking forward to seeing version 7.
Thanks for all you do.
Paul
(http://arqui3d.com/JMdict_e.gif)
This happens when trying to generate an index from the latest version of the JMdict file (English glosses only version). I "un-gzipped" the dic before loading it into Stackz!.
The same thing happened in two machines: A Core 2 Duo w/ 2GB of RAM, and a Pentium 4 w/ 512 MB of RAM, both running Windows XP SP2.
I have removed beta 4 and tried Beta 5a and the problem still persists on my machine.
Does the latest beta (07 beta 6) solve the issue?
Abi just emailed me, telling that the problem was gone.
The initialization of the language sample strings in the language dialog (menu options -> stackz options, language tab) involved a conversion from their stored non-unicode encoding to the displayed unicode encoding. This conversion mysteriously fails on a few computers with the new version, although it worked fine on the same computers in the old version.
Stackz did not correctly handle the failure of this erroneous conversion, which caused the "close in an unusual way" problem.
Now the erroneous conversion is not leading to a crash anymore, but the reason for the error is still unknown - it might be due to some other problem that will show up in a different place...
You can check if your installation has this problem by inspecting the language sample texts in the language dialog. If one of these displays the text "conversion error", then the conversion for some reason did fail on your system.
I just removed beta 5 and installed beta 6 and it now opens fine. I will start experimenting with this version and see how it works on my machine. I already like the look and some of the new features that I have tried so far. I am getting some dictionary errors, but will make sure that I have the paths and settings are correct and see what happens. Thanks for all of your work.
Paul King