httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Randy Terbush <>
Subject Re: minor features and changes
Date Sun, 08 Jun 1997 16:18:57 GMT
> Paul Sutton wrote:
> > As an example, the call_exec function in the module API currently does not
> > return (it does an exec*()). In the NT port it does a spawn*() and returns
> > the child PID. So all modules which use call_exec (including mod_cgi,
> > mod_include, mod_rewrite) have to be updated with more #ifdef WIN32
> > sections. These leads to a large number of changes throughout the code.
> > This will make the job on module writing more complex, and testing very
> > difficult (since a typical module writer on Unix may not have access to MS
> > operating systems and MSVC, for example). So in a merged Apache for Unix
> > and NT I would like to see a *consistent* module API interface (it may
> > well be that call_exec functionality has to be updated to work more like
> > the NT version, but at least it would be consistent across all platforms). 
> > There are other cases similar to this (socket functions, timeouts,
> > signals, etc). 
> Although you give call_exec as an example, if my memory serves me, it is the
> only such case. Not that many modules use it.
> I'd rather integrate NT into 1.3, even if it is a bit messy, simply because the
> time wasted trying to keep the two trees in sync will vastly exceed the time
> wasted by Unix persons having to navigate around Windows code. Of course, if
> a Unix module doesn't work on Windows because of differences, its up to the
> Windows coders to fix it, if they want it.

I agree. I see no other reason to release a 1.3 other than a major 
integration/platform support change. We can happily add performance 
enhancements, etc. under the 1.2.x versioning if that is all we 
intend to do.

Since it sounds like we are going to start somewhat from scratch on 
2.0, and we have safely branched the 1.2.x code, I say lets start 
looking at some patches for NT integration onto the HEAD. We can 
always back it out.

> Or, we could attempt to fix call_exec so it can be used on either...

call_exec() needs some work as I am not completely comfortable with 
what we arrived at in this first attempt to provide setuser 
execution. There are some other very interesting tools becoming 
available on this issue that could be very useful in the 2.0 API.
I think one of the issues/features to grapple with in the 2.0 API
will be a secure execution interface.

> Cheers,
> Ben.
> -- 
> Ben Laurie                Phone: +44 (181) 994 6435  Email:
> Freelance Consultant and  Fax:   +44 (181) 994 6472
> Technical Director        URL:
> A.L. Digital Ltd,         Apache Group member (
> London, England.          Apache-SSL author

View raw message