Newsgroups: comp.lang.smalltalk
Path: cantaloupe.srv.cs.cmu.edu!rochester!cornellcs!newsstand.cit.cornell.edu!portc01.blue.aol.com!newsxfer2.itd.umich.edu!howland.erols.net!newsfeed.internetmci.com!in1.uu.net!news2.new-york.net!not-for-mail
From: vlad@world2u.com (vlad)
Subject: Re: ST to C++ Intrerpreters?
X-Newsreader: Forte Free Agent 1.0.82
X-Nntp-Posting-User: (Unauthenticated)
Message-ID: <DxHuHI.8rw@news2.new-york.net>
References: <4ufpnf$8a5@blackice.winternet.com> <4v4olr$q1u@freeside.cls.de> <4v9hr6$mu8@m1.cs.man.ac.uk> <322C82F5.128Esmalltalk@merck.com> <3233A82E.1B54@dialip.sdn.dk>
X-Trace: 842320276/9824
X-Nntp-Posting-Host: i123.177.world2u.com
Date: Tue, 10 Sep 1996 02:01:02 GMT
Lines: 68

Soren Skogstad Nielsen <ddf8732@dialip.sdn.dk> wrote:

>Vlastik wrote:
>> 
>> Ivan Moore wrote:
>> >
>> >   >In the archive at www-st.uiuc.edu you will find Producer, an experimental
>> >   >Smalltalk to Objective-C translator and also I know of a Smalltalk to CLOS
>> >   >(the CommonLisp Object System) translator, but momentary don't remember an
>> >   >URL. (Mail me for further details)
>> >
>> > http://www.cs.man.ac.uk/~ivan/babel.html
>> >
>> > there's also some references to other work at this site.
>> >
>> >   [stuff deleted]
>> >
>> >   >bye
>> >   >--
>> >   >Stefan Matthias Aust // ...come on, kiss the frog
>> >   >        http://www.kiel.netsurf.de/users/s/sma/
>> >
>> 
>> I think what we need, is a language which would describe how the program will
>> behave when it will become a program.
>> 
>> Then, when a description is made, we would need only to click on a button and
>> generate a source code (C, C++, Jave etc...), which actually will implement the
>> behavior description.
>> 
>>   Vlastimil Adamovsky
>>    **** C++, Smalltalk and Java freak*****
>> 

>In response to this remark of Vlastimil Adamovsky I will say, YES, but
>when going from the first dull specification down to the tested and
>running code it is not easy to decide when to push that button.
>Nevertheless you can gradually sharpen your concepts and test their
>interplay in a controlled way. We have described that in the OoASE
>method supported by the OoASEtool.
>You can send me a mail if you want further information.

Please send info. Thanks.

The software crisis (most of people denies it exists) is caused by
plenty of langugages, that are mastered only  somehow. 
People jump from one language to other and don't understand, that it
is not language but it is a methodology, some kind of "meta-language",
which is the most important.

According to my point of view, the specs should be translated to the
"metalanguage" by "meta-experts" and then the "metalanguage" should
generate what ever runable language. All gradual improvements should
be done in that "meta language".

Since I don't have any time, I will leave this idea over to other meta
people.

I am looking forward to hear some metanews.

  Vlastimil Adamovsky

>Soren Skogstad Nielsen
>A-Informatik
>Bredgade 34 E Copenhagen Denmark
>email: soren.s.nielsen@ddf.dk


