[Ecls-list] ASDF revolt

[james anderson]

On 2010-04-11, at 23:29 , Pascal J. Bourguignon wrote:

> james anderson <james.anderson at setf.de> writes:
>
>> On 2010-04-11, at 20:19 , Pascal J. Bourguignon wrote:
>>
>>> james anderson <james.anderson at setf.de> writes:
>>>
>>>> if the dominant goal is to simplify, it is demonstrated, that there
>>>> is sufficient information in package declarations to build lisp
>>>> programs.
>>>
>>>
>>> This is wrong.
>>
>> that you can describe another method says nothing more than that the
>> truth value of the sentence above is qualified.
>> but then, it contained no universal and was itself qualified.
>
> Perhaps you meant "system declarations"?

no the statement was "package declarations".

>
> By "package declarations" I understand defpackage forms, and I meant
> that it was not enough, since to do what you want, that is, build lisp
> programs, you need to add information over the bare defpackage forms.

more correctly, evidently, to do what _you_ want.

> The proof is in the fact that the defpackage form doesn't list all the
> qualified symbols used by the code in that package.

you may need to do that for some purposes, but the requirement is not  
universal.
were one to set the "simplify" goal as dominant, then programs with  
undeclared package cross references would be excluded.
the statement was not that it is possible to build all lisp programs.
and one also needs a file system, and the source files, and a  
protocol to impute designators to the elements of the package  
definitions and another to resolve designators to files, but that  
goes without saying.