[imp] Re: Segfaults with IMP

[Anders Persson]

I also have problems with segfaults. I'm running an Apache created with 
Apache Toolbox v1.5.64 (PHP/4.3.0 mod_gzip/1.3.19.1a mod_perl/1.27 
mod_ssl/2.8.12 OpenSSL/0.9.6b) on a Redhat 8.0 server, and I get a lot of 
segfaults when using horde (I use it mostly with the https protocol), but no 
segfaults otherwise. 

Besides the segfaults I also had problems with large attachments not working 
even though I turned every parameter I could find (both in the .php config 
files and php.ini) that might have something to do with it up really high. 
When I tried to use imapproxy v1.0, attachments didn't seem to work at all. 

Oh well, I've run out of time I can spend on this so I'm putting the whole 
project on ice for the moment. Horde and Imp look really nice though, so 
I'll look into it again when new versions of Apache Toolbox and horde/imp 
come out. I'll also be lurking on this list to see if someone finds 
solutions to these issues. 

Cheers, 

/Anders 

 

Hans de Groot writes: 

> Hi, 
> 
> I have asked this question before here. 
> 
> The answer was that it could not be horde or imp but something with apache or php. 
> 
> My problem with this answer is that I changed redhat and apache and php and
> still imp 3.1 gives me segfault ( and "page cannot be displayed" errors in
> internet explorer. Netscape 7 works great I never have the problem in netscape
> but I noticed netscape tries to fetch a page again if it fails ie does not) 
> 
> Since more people have this problem and I my knowledge is not good enough to fix
> it my self I hoped there was a general answer to this problem.  
> 
> Does any one have any suggestions? (besides that it's ot imps fault.) I got the
> impression turba or kronolith is the cause since the page cannot be displayed
> message happens most on the intro screen. (for me it displays a summary of imp ,
> kronolith nag. Also the lower frame which displays the menu is tha one that most
> often has this problem.  
> 
> I also heard that segfault 11 has something to do with hardware. could it just
> be out of memory? 
> 
> sorry to bother every one again with this problem. 
> 
> regards 
> 
> Hans de Groot 
> 
>