Is a conforming Common Lisp implementation allowed to remove COMPILE-
FILE from a Lisp image? More specifically, assuming an implementation
provides the ability to create "stand-alone" executables, is that
implementation allowed to arbitrarily remove COMPILE-FILE from the
executable?
-- david
From: Thomas A. Russ
Subject: Re: COMPILE-FILE and conforming implementations
Date:
Message-ID: <ymilkiz835s.fsf@sevak.isi.edu>
"David E. Young" <··········@gmail.com> writes:
> Is a conforming Common Lisp implementation allowed to remove COMPILE-
> FILE from a Lisp image? More specifically, assuming an implementation
> provides the ability to create "stand-alone" executables, is that
> implementation allowed to arbitrarily remove COMPILE-FILE from the
> executable?
Well, I would have to say this is allowed.
Mainly because there is no mandate in the standard about what it means
to create an "image" or a stand-alone executable. Since the creation of
these things are not mandated by the ANSI standard, there can't be any
conformance issues relating to what an implementation does or does not
do when it creates an implementation-dependent extension.
I do agree with Pascal, that the resulting image or executable should
not claim to be a full, conforming implementation Common Lisp.
In general, it seems that the distribution licenses and proceduresfor
most commercial lisp systems have some such effect, as part of the
commerical company's desire not to have you distributing full Common
Lisp systems, thus robbing them of potential sales....
--
Thomas A. Russ, USC/Information Sciences Institute
On Thu, 15 Feb 2007 23:30:23 +0100, Thomas A. Russ <···@sevak.isi.edu>
wrote:
> "David E. Young" <··········@gmail.com> writes:
>
>> Is a conforming Common Lisp implementation allowed to remove COMPILE-
>> FILE from a Lisp image? More specifically, assuming an implementation
>> provides the ability to create "stand-alone" executables, is that
>> implementation allowed to arbitrarily remove COMPILE-FILE from the
>> executable?
>
> Well, I would have to say this is allowed.
>
> Mainly because there is no mandate in the standard about what it means
> to create an "image" or a stand-alone executable. Since the creation of
> these things are not mandated by the ANSI standard, there can't be any
> conformance issues relating to what an implementation does or does not
> do when it creates an implementation-dependent extension.
>
> I do agree with Pascal, that the resulting image or executable should
> not claim to be a full, conforming implementation Common Lisp.
>
> In general, it seems that the distribution licenses and proceduresfor
> most commercial lisp systems have some such effect, as part of the
> commerical company's desire not to have you distributing full Common
> Lisp systems, thus robbing them of potential sales....
>
>
Oh ACL will let you sell programs with compiler and/or evaluate all right..
At a substantial fee.. (royalties remember)
But yeah, it can generate executables without evaluator and/or compiler
which also means compile-file won't work of course..
There is nothing that says that a executable made with lisp has to
support Common lisp after all.
--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
Yes, I understand why a vendor might want to disallow file compilation
in a deployed app. I'm ok with that. I guess I was just caught off
guard because the feature involved removing part of the "public"
Common Lisp interface.
In any event, we ended up making allowances in our app for this
situation without any real harm to the solution we were after.
Good information from everyone. Thanks again.
david
On Feb 15, 5:30 pm, ····@sevak.isi.edu (Thomas A. Russ) wrote:
> Well, I would have to say this is allowed.
"David E. Young" <··········@gmail.com> writes:
> Is a conforming Common Lisp implementation allowed to remove COMPILE-
> FILE from a Lisp image? More specifically, assuming an implementation
> provides the ability to create "stand-alone" executables, is that
> implementation allowed to arbitrarily remove COMPILE-FILE from the
> executable?
No. But an implementation is authorized to implement only a subset of
Common Lisp, while clearly documenting the fact. Then it won't be a
"Common Lisp implementation", but an "implementation of a subset of
Common Lisp".
See: http://www.lispworks.com/documentation/HyperSpec/Body/01_g.htm
--
__Pascal Bourguignon__ http://www.informatimago.com/
HEALTH WARNING: Care should be taken when lifting this product,
since its mass, and thus its weight, is dependent on its velocity
relative to the user.
"David E. Young" <··········@gmail.com> writes:
> Is a conforming Common Lisp implementation allowed to remove COMPILE-
> FILE from a Lisp image? More specifically, assuming an implementation
> provides the ability to create "stand-alone" executables, is that
> implementation allowed to arbitrarily remove COMPILE-FILE from the
> executable?
While a "Common Lisp implementation" cannot remove anything from
Common-Lisp, note that there is in general a lot of liberty,
implementation specific stuff. For example, AFAIUI, COMPILE-FILE
could only do the minimal compilation specified in 3.2.2.2. and output
to the "binary file" the sexps thus pre-processed. Then load could
read this "binary" file. Some care should be taken to handle
eval-when and load-time-value, etc.
--
__Pascal Bourguignon__ http://www.informatimago.com/
HANDLE WITH EXTREME CARE: This product contains minute electrically
charged particles moving at velocities in excess of five hundred
million miles per hour.
Yes, that's what I thought also. Ok, thanks for the information.
-- david
>
> While a "Common Lisp implementation" cannot remove anything from
> Common-Lisp, note that there is in general a lot of liberty,
> implementation specific stuff...
> --
> __Pascal Bourguignon__ http://www.informatimago.com/
>
> HANDLE WITH EXTREME CARE: This product contains minute electrically
> charged particles moving at velocities in excess of five hundred
> million miles per hour.