Re: Cronolog jumbo patch fix

[Andrew Ford <A.Ford@xxxxxxxxxxxxxxxx>]

Thanks to both of you for your emails.  I am acutely aware that I 
haven't done much on cronolog for a while.  I do have plans for 
cronolog, including the option to parse the log entries to extract a 
token (e.g. virtual host tag) then split on that tag as well as the 
time.  I envisage a configuration file that would provide options and 
associate tokens with file templates, so that in Apache you would just 
specify a single custom log.  This would eliminate the stampeding horde 
of cronolog processes in large virtual hosting environments.

I will get round to incorporating the jumbo patch at some point, but I'm 
not quite sure when yet.

BTW I changed the mailing list to "moderated" due to the amount of spam 
on the list, but have a backlog of messages waiting to be approved.

Andrew

Matthew Grosso wrote:

> Hi,
>
> Unfortunately you werent the first to get bit by this bug. I've fixed it to
> respect the compile time defaults and also added runtime command line options
> to override those.  Fortunately I've finally spent the few minutes it took to
> put together a tarball with this fix and the changes represented by the jumbo
> patch.
>
> I've made it available so people can test it at:
>
> http://www.nullo.org/~mgrosso/cronolog-1.7.0-beta.tar.gz
>
> still todo:
>
>    There are several new command line options that are only documented in the
>    usage message. I'd love to update the man page but havent spent the time
>    fiddling with it yet.   
>
>    I'd be interested to get build and test reports from windows, mac osx, and
>    other platforms.
>
> Its still Andrew's baby, so he may well have different thoughts on the
> approaches taken or the version number or qa or whatever.
>
> Later,
> Matt
>
>
> On Aug 18 1737, Richard van den Berg wrote:
>  
>
> > Hi there,
> >
> > I just started using cronolog in combination with the jumbo patch 
> > provided by Matthew Grosso. I immediately spotted one problem: the jumbo 
> > patch ignores the FILE_MODE define when creating new log files. As a 
> > result, all files are created with mode 700. It's a trivial fix, but I 
> > could not find any reports of it, so here it is.
> >
> >    
>
>