Subject:unrecoverable error Author:Gerald Nelson Date:22 Aug 2001 08:55 PM
Here is the log file for an unrecoverable error with version 3.1 (I think). Regards, Jerry Nelson//=====================================================
Date: Wed Aug 22 19:58:16 2001
Exception code: C0000005 ACCESS_VIOLATION
Fault address: 12076BEB 01:00075BEB C:\Program Files\eXcelon Corp\Stylus Studio\bin\xmlengine.dll
Subject:Re: unrecoverable error Author:Minollo I. Date:22 Aug 2001 09:37 PM
>Here is the log file for an unrecoverable error with version 3.1 (I think).
Looks like a crash inside Xerces; can you give us a few details about what
you were doing when the crash happened, and if that is something that you
are able to reproduce?
Subject:Re: unrecoverable error Author:Gerald Nelson Date:27 Aug 2001 09:14 PM
I get crashes pretty regularly with the xml/dtd/xsl files I'm using. I don't know that there is anything specific I can point. Just all of a sudden I get a crash. I can send more log files if you want.
I have a big (I think) xml file (14,854 lines).
I am a real neophyte with this stuff so maybe that makes me a good tester ;).
A couple of usability comments.
1. It would be nice if I could interrupt the validation process. When I do something like forget to uncomment the dtd line in the xml, being told that every line in my xml file is invalid isn't very helpful. I have to kill the process.
2. Wordwrap in the output file would make it easier to see what is going on, as would regular copy (cntl-c, instead of right-click, copy).
Subject:Re: unrecoverable error Author:Minollo I. Date:27 Aug 2001 09:57 PM
>I get crashes pretty regularly with the xml/dtd/xsl files I'm using. I
>don't know that there is anything specific I can point. Just all of a
>sudden I get a crash. I can send more log files if you want.
>
>I have a big (I think) xml file (14,854 lines).
Gerald,
the issue you are reporting seems to be related to DTD validation; if
you can, please send us (you can use my email address) the latest log file
you have in your \bin directory, and the XSLT/XML/DTD file
you are using.
Your XML file doesn't sound big; we regularly run tests with files in the
order of a few dozens of megabytes; but it also depends on how much
available memory you have on your system; that information will help too.
>A couple of usability comments.
>
>1. It would be nice if I could interrupt the validation process. When I do
>something like forget to uncomment the dtd line in the xml, being told
>that every line in my xml file is invalid isn't very helpful. I have to
>kill the process.
I agree with you that being able to interrupt the validation process would
be nice (it's already in our PCR list).
But why in your case you have an error for every line of your XML file?
Isn't you XML file valid according to the DTD it uses?
>2. Wordwrap in the output file would make it easier to see what is going
>on, as would regular copy (cntl-c, instead of right-click, copy).
I'm adding these two to our PCR list.
Thanks for your feedback. And we look forward to testing your documents to
understand the source of the crashes you are experiencing.
Subject:Re: unrecoverable error Author:Gerald Nelson Date:02 Sep 2001 04:26 PM
I sent the log, xml, and dtd to your email.
The reason every line of the xml is flagged is because I have to comment out the dtd statement before I send it off to a different site where it gets parsed by something someone else set up. If I forget to uncomment it out, and try to validate, ...
So the cntl-c is more to help with my forgetfulness than anything wrong with your software.
Subject:Re: unrecoverable error Author:Minollo I. Date:02 Sep 2001 08:58 PM
>I sent the log, xml, and dtd to your email.
Thanks, we'll work on them.
>The reason every line of the xml is flagged is because I have to comment
>out the dtd statement before I send it off to a different site where it
>gets parsed by something someone else set up. If I forget to uncomment it
>out, and try to validate, ...
>
>So the cntl-c is more to help with my forgetfulness than anything wrong
>with your software.
In 3.1 beta 2 (which will be available Tuesday or Wednesday) you will be
able to set the maximum number of errors after which the validation process
is aborted; that should help.