Maude licensed under GPL, 2.0 approaching

[Brian T Rice]

The latest alpha test e-mail reveals the following. I'll put together
binary and source packages for it and host it on the tunes website until
it's officially released.

On Sat, 31 May 2003, Steven Eker wrote:

> Dear Maude Abusers,
>
> A new alpha release can be accessed at SRI-CSL in:
[omitted details]
>
> This version claims to be Maude 2.0 at startup - this is because
> unless any show stoppers are reported before I leave for Spain it is
> what we will release at RTA 2003.
>
> Bug fixes:
> ===========
> (1) A bug where the bug reporting address was not always printed when
>     Maude crashed. This was caused by the line wrapper being in a bad
>     state - it is now bypassed with a system call. Bug reported by
>     Ambarish.
>
> (2) The special hook mechanism no longer crashes if there is a missing
>     sort; but bad hooks is still a good way to corrupt memory. Bug
>     reported by Christiano Braga.
>
> Changes:
> =========
> The most important change is that this version is released under the
> GNU GPL - see the source tree for details - so you may redistribute it.
>
> The other major change is that the persistent AC/ACU rewriting now
> supports conditional rewriting for certain patterns - though this is
> more of a proof of concept than comprehensive implementation.
>
> There are several minor changes:
>
>   op extending_. : ModuleExpression -> Import [ctor] .
>
> is added to fmod META-MODULE.
>
> For associative and iterated operators (assoc & iter attributes) the
> operational semantics of mbs/cmbs is tidied up. It is now an error to
> have both non-kind declaractions and mbs/cmbs for the same associative
> or iterated operator but non-kind declaractions can be converted to
> mbs as long as there is a kind declaractions. So instead of
>
> fmod ASSOC-MB1 is
>   sort Foo .
>   op f : Foo Foo -> Foo [assoc comm] .
>   op e : -> [Foo] .
>   ops a b c d : -> Foo .
>
>   mb f(a, e) : Foo .
> endfm
>
> red f(a, b, e) .
> red f(a, b, e, a) .
> red f(e, b, e, a) .
> red f(a, b, e, e, a) .
> red f(a, a, b, e, e, a) .
>
> write
>
> fmod ASSOC-MB2 is
>   sort Foo .
>   op f : [Foo] [Foo] -> [Foo] [assoc comm] .
>   op e : -> [Foo] .
>   ops a b c d : -> Foo .
>
>   mb f(X:Foo, Y:Foo) : Foo .
>   mb f(a, e) : Foo .
> endfm
>
> red f(a, b, e) .
> red f(a, b, e, a) .
> red f(e, b, e, a) .
> red f(a, b, e, e, a) .
> red f(a, a, b, e, e, a) .
>
> and instead of
>
> fmod ITER-MB1 is
>   sort Foo .
>   op f : Foo -> Foo [iter] .
>   op e : -> [Foo] .
>
>   mb f(e) : Foo .
> endfm
>
> red f(e) .
> red f(f(e)) .
> red f(f(f(e))) .
>
> write
>
> fmod ITER-MB2 is
>   sort Foo .
>   op f : [Foo] -> [Foo] [iter] .
>   op e : -> [Foo] .
>
>   mb f(X:Foo) : Foo .
>   mb f(e) : Foo .
> endfm
>
> red f(e) .
> red f(f(e)) .
> red f(f(f(e))) .
>
> For this build I have removed the integrated compiler and MOS oracles
> - though the code is still in the source tree. These features are not
> officially part of the release and will not be documented in the
> manual. I have also removed the Full Maude hack from prelude.maude -
> the 2.0 release of Full Maude should not need it.
>
> Steven

-- 
Brian T. Rice
LOGOS Research and Development
http://tunes.org/~water/