why is this still happening?
kgp at nethere.com
Thu Dec 2 23:05:54 PST 2004
i am in hardcoded requirement hell...
it has happened again with an rpm.
this time with libpcap... like that isn't going to change... like that
doesn't have about a bazillion undiscovered security holes in it... (and
it won't believe a symlink)
and lo and behold, the version they want has a huge buffer overflow
problem... and it's for a false positive generator for Snort...
well, if you use it you could get a positive positive ... that counts as
i hate everyone but me... at least until i stop speaking to myself.
On Thu, 2004-12-02 at 22:22, Andrew P. Lentvorski, Jr. wrote:
> On Dec 2, 2004, at 3:41 PM, Mike Marion wrote:
> > Quoting "Andrew P. Lentvorski, Jr." <bsder at mail.allcaps.org>:
> >> But, the stupidity is using autoconf/automake in the first place.
> > What? Are you nuts? Would you rather go back to xmkmf and Imakefiles?
> <sarcasm> M4 is such an improvement! </sarcasm>
> Give me a real build system that uses a real programming language,
> SCons has been working pretty well for me lately. Even in those times
> when I find SCons too heavy, Python works very well.
> Even *Perl* would be better than autoconf/automake.
> > Unless the person does something stupid in their autoconf/automake
> > scripts, it
> > works probably 99.9% of the time.
> Maybe on Linux. My success rate when I have to regenerate a configure
> script on anything else approaches 0%.
> > At least in my experience it does, and I
> > build things from source on linux and solaris all the time at work.
> Build things from source != autoconf works
> Most of the time you never have to rebuild the configure script.
More information about the KPLUG-List