Re: [svs-devel] some issue during testing

[R. R. <ren...@ba...>]

Hi Sebastién!

On Monday 03 May 2010 05:15:34 pm Sebastien Caty wrote:
> I wanted to test NO_SYSLOG=1 (even if it's not recommended in the
> readme). Well, it doesn't compile :
> 
> /source3/include -I../source3/../popt -I../lib/replace -I../lib/talloc
> -I../lib/tevent -I../lib/tdb/include -I../source3/librpc -I..
> -I../source3 -I. -o svs-main.o svs-main.cpp
> svs-main.cpp: In function ?void svs_log(int, QString)?:
> svs-main.cpp:179: error: expected `;' before ?arg?
> make: *** [svs-main.o] Error 1
Ok, that was a syntax error... fixed now in SVN rev. 35.
 
> svs-main.cpp line 179 :
> svsLogStream << QString("%1: %2: %3:
> %4\n")arg(timeString).arg(svsSessionID).arg(svs_msgtype_string(messageType)
> ).arg(message);
> 
> Should be this I think :
> svsLogStream << QString("%1: %2: %3:
> %4\n").arg(timeString).arg(svsSessionID).arg(svs_msgtype_string(messageType
> )).arg(message);
Yes, correct!

> I had an "invalid" share configured in samba (home path for my user
> doesn't actually exist). SVS doesn't seem to like this and throw this
> error :
> 
> svs_clamav[31616]: canonicalize_connect_path failed for service
> catse1, path /home/catse1
Well, it's more that smbd doesn't like it ;)! This message isn't produced by 
the SVS plugin... when SVS receives the "connect" it doesn't yet check if the 
service is "valid", it's simply assumed that it is ;). I don't think it's even 
possible within the VFS API...

> One more thing, when quarantineDirectory isn't valid, I also get some
> fault in the log but svs still seems to "work". Did not try a test
> virus but it might be good to test if the quarantineDirectory can be
> writen to during startup. If not, output a more clear warning.
I've added some checks in SVN now, hope this helps. Thanks for the suggestion!

Have fun, René
-- 
René Reucher
ren...@ba...
http://www.batcom-it.net/

No matter how subtle the wizard, a knife in the shoulder blades will
seriously cramp his style.