tech-userlevel archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: proposal: inetd improvements.

> You state that Autoconf-generated configure scripts "don't do what
> they try to, and pretend to, do: automatically configure the software
> correctly for the system it's being built on."

Well, in my experience they certainly don't.

> Your only attempt to prove this assertion is citing four examples of
> buggy configure scripts.

True, I left off something: those are my most recent four brushes with
autoconf-generated configure scripts.  At four for four, I can't help
feeling that there actually is something wrong.

It's also true I've had less disastrous encounters with
autoconf-generated configure scripts in the past.  However, I can't
recall ever having one do what it is supposed to do; even at the best
of times, when the script doesn't crash, I've had stuff to fix
manually.  I'm sure there's a point of view from which this is my
fault; if I would just drink the Linux (or whatever, but usually Linux)
kool-aid and do it their way, I'd have no problems.  But I thought a
configure script's job was to adapt the software to the system rather
than convince the admin to adopt the software author's paradigm.

> They don't prove what you suggest to prove; i.e., that Autoconf
> cannot be used (or at least is not a good tool) for creating sane
> configure scripts.  Do you think the latter is the case?  If so: why?

Well, first, I'm not convinced that such a thing as a "sane configure
script" exists.  But even assuming, arguendo, that the phrase is not an
oxymoron, yes, I believe that autoconf is, at the very least, not a
good tool for generating them; whether it is possible to bludgeon it
into doing so is perhaps an interesting question for some people, but
not particularly interesting to me.

Why do I think this?

Because every time I've tried to use an autoconf-generated configure
script, it has failed to some extent.  The four incidents listed in my
blah post are just the four most recent (and perhaps most spectacular)
failures.  Historically, most of the failures have been relatively
minor and been patchable up - configure completes and just gets
something wrong - but based on those last four the situation seems to
be getting worse rather than better.

/~\ The ASCII                             Mouse
\ / Ribbon Campaign
 X  Against HTML      
/ \ Email!           7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B

Home | Main Index | Thread Index | Old Index