[Logcheck-devel] #207795: logcheck: fails to clean up when interrupted

[Eric Evans]

On Wed, Apr 28, 2004 at 10:25:16AM -0400, Todd Troxell muttered these words:
> I don't like the dependence on bash either, but I think it's more
> important to have the patch in at this point.  I don't know of a way to an
> onexit trap for other shells.  There is onintr for csh, but it's not the
> same.  Hmm.

I'm not sure if this particular patch warrants making logcheck shell
specific, but I am fairly confident that similar issues are lurking in
the not too distant future. Locking in a specific shell now will make
things easier in the long run.

I updated CVS to reflect the change in intrepreters from #!/bin/sh to 
#!/bin/bash.

-- Eric

> A perl/python rewrite would be nice too, but right now, I think it's more
> important to get these bugs squashed.
> 
> -Todd
> 
> On Tue, Apr 27, 2004 at 03:34:45PM -0500, Eric Evans wrote:
> > I applied the patch attached to bug #207795 to CVS yesterday, and then 
> > some hours later it dawned on me that the trap builtin might not be 
> > present in all shells, (and sure enough it isn't).
> > 
> > Does anyone know of a more portable way to handle signals in a shell
> > script?
> > 
> > -- 
> > Eric Evans
> > eevans at sym-link.com
> 
> 
> 
> -- 
> [   Todd J. Troxell 	                                        ,''`.
>       PSU Student, Debian GNU/Linux Developer, 100% Geek       : :' :
>       31B8 BF83 4898 E26D 1935  4C5D E2ED E843 7147 3F66       `. `' 
>       http://www.rapidpacket.com:81  || http://www.debian.org    `-     ] 
>       * Join the EFF & secure your digital freedoms today! http://eff.org *



-- 
Eric Evans
eevans at sym-link.com