[Spambayes] Win 7, Outlook 2007 issues

Mark Hammond mhammond at skippinet.com.au
Wed Mar 31 07:50:23 CEST 2010


On 31/03/2010 3:37 PM, Ocean wrote:
>> The binary is built from source, so it should work fine as long as you
>> have the latest pywin32 installed.

> 	In an earlier email to the newsgroup, you gave a link for the binary
> (since I can't find 1.1a6 on the official site, I'll assume that link is
> still the most recent version).  Where can I grab the source version of
> 1.1a6?

 From SVN - although there are 2 files not checked in which relate only 
to the reported version number.

> 	You explained to me that running the Trace Collector externally had
> known issues with Vista and Win7.

Yeah - but they should all be resolved with the most recent pywin32 - 
from which the most recent binary was built.

>  That rules that out, leaving the only
> option to be trying to modify the source to allow file logging.  I posted
> the exact changes I made.  At this point, you'd have a far better idea than
> I as to why it would still fail to create the log files.
>
> 	But for what it's worth, in prior emails, I gave you some of the
> error messages that were being logged (via the internally run Trace
> Collector).  The only thing I can think of is that those errors might be
> causing these issues - perhaps all of them (loading, executing, and
> logging).
>
> 	To my recollection, you never mentioned pursuing any of them.
> Perhaps they're worth looking into, at this point.

I would if I could.  I don't see any such problems with that binary and 
Win7x64 + Outlook 2010 beta and there is only so much I can debug via 
speculation.

>> Alternatively, try again with the binary to confirm things work as
>> expected, then move to the source version with no changes to ensure
>> things *still* work in an identical fashion, and then make your code
>> changes to change the functionality.
>
> 	I did that with 1.1a4, at your request.  There was no difference
> whatsoever.  I'll give 1.1a6 a shot once I get the source.

Please try the 1.1a6 binary first - that will give us more information 
upon which to try and diagnose the problems you see running from source.

Cheers,

Mark


More information about the SpamBayes mailing list