Re: ds:system-subsystems ?

Architecturally, I require my systems to explicitly use 'require' 
statements to ensure that their dependencies are loaded, but I can 
see the value of parsing system definitions. When you have a system 
packaged into a "fasl", then you may not have access to the system 
definition. I place my require statements and defpackage forms into a 
single file that I 'concatenate' into the final fasl module, but I 
don't include the system definition; I may want to try that.

At 10:35 PM +0100 4/6/05, Tiago Maduro-Dias wrote:
>Hi David,
>
>Well, you are right, I am (was) re-implementing part of that 
>functionality. The documentation I originally read (in HyperSpec) 
>didn't have quite the behaviour I wanted. But after reading your 
>e-mail I went investigating and just found out that Allegro 
>implements a sys:*require-search-list* which can easily be extended 
>to do part of what I want in a cleaner way, with minimal changes to 
>my code (thank god for abstraction).
>
>This does not solve all my problems, however. The main issue with 
>knowing a system's subsystems is being able to automagically load a 
>system's dependencies (by this I mean systems this system depends 
>of), as well as perform other operations such as test-runs. Even if 
>I'm using require to automatically look for and load the system 
>definitions, wherever they are located, I still need to know which 
>systems are those...
>
>Tiago Maduro-Dias.
>
>
>On 6 Apr 2005, at 21:00, David Mittman wrote:
>
>>It almost sounds as if you are re-implementing the functionality 
>>provided by the "require" function. Can you explain how what you 
>>are looking for differs from "require" and "provide"?
>
>---
>[This E-mail scanned for viruses by Declude Virus]


-- 
David Mittman ([email protected])