RFE: dietlibc review

[Bill Nottingham]

Enrico Scholz (enrico.scholz at informatik.tu-chemnitz.de) said: 
> > Wasn't looking at the specs, actually; just the source; it certainly
> > doesn't require dietlibc in that sense.
> 
> As said, rpm does not know something like 'recommends'. For ip-sentinel
> and dhcp-forwarder it should not make much difference (except different
> resource usage at runtime) whether they are compiled/linked against
> dietlibc or glibc. But dietlibc is recommended for them.
> 
> But for util-vserver, I hope that I made the warnings big and fat enough
> when the tools will be built against glibc. ;)

I suppose; it just seems to me that building against an alternate
libc exposes you to many of the same problems that building with
an alternate compiler does; it's more likely to experience odd bugs
that other things don't, it limits you from various features that
are in the standard libc (including some of the memory protection
ones), etc.

Bill