Bug in #! processing - "pear broken on current"

Garance A Drosehn gad at FreeBSD.org
Fri Jun 10 12:39:42 GMT 2005

At 12:48 PM +0200 6/10/05, Roman Neuhauser wrote:
># flz at xbsd.org / 2005-06-10 10:06:46 +0200:
>  > On Jun 10, 2005, at 8:24 AM, Roman Neuhauser wrote:
>  > >
>>  >    The pear people have hacked around the other OS's limitations.
>>  >
>>  >    This change makes FreeBSD lose one small but fine competitive
>>  >    advantage over other unix-like systems. Pity.
>>      FreeBSD needed special handling, no it doesn't anymore.
>>      I'm not sure that's losing a *competitive* advantage.
>     The previous behavior in FreeBSD allowed me to use things on
>     the shebang line that weren't possible in e. g. Linux, and I
>     enjoyed it, because it saved me from various hacks. Aiming for
>     the lowest common denominator means losing useful features. One
>     reason to prefer FreeBSD less.

Well, there's more than one way to get the job done, and it is much
too early to be wailing over the end of FreeBSD due to this change.
The recent change to the kernel-level parsing pretty much has to stay,
or FreeBSD users will continue to run into some other problems which
happen *only* on FreeBSD.  The parsing that we used to do was meant
to be helpful, but in some situations it was completely wrong.

However, I do agree that the previous behavior could be very useful
in other situations.  I will soon have a version of `env' which will
provide all the benefits that used to happen due to the previous
parsing behavior.  In fact, it will be even more flexible and support
even more options than the previous parsing behavior did.  And I think
I can even define it in such a way that other operating systems could
pick up these changes to `env' (if they wanted to), and enjoy the same
flexibility.  And more important to me, I could recompile this new
`env' on other operating systems, and at least *I* would have those
benefits on all platforms I work on!

I think these changes could even be MFC'ed to 5.x (and 4.x, if needed),
and then a single #!-line could be written which would work on all
those systems.  I'm not sure that MFC-ing would be worth it, though.

I actually have my changes written and mostly working, and right now I
am  reviewing the ideas to see if the design could be done any better.
Now I don't know if I can get everyone else to agree that my ideas
are wonderful, of course, but it sounds like I might get a 'yes' vote
from you.  :-)    More details soon, and then we shall see.

Garance Alistair Drosehn     =      gad at gilead.netel.rpi.edu
Senior Systems Programmer               or   gad at FreeBSD.org
Rensselaer Polytechnic Institute;             Troy, NY;  USA

More information about the freebsd-ports mailing list