Am Sonntag, 14. Februar 2010 21:19:57 schrieb Torsten Bergmann: > Hi Danny, > > > - What is the reason to have HelpBuilder and to create a hierarchy of > > HelpTopic instances mirroring the layout of the actual content classes? > > Why > > not use the class objects directly? > > HelpBuilder builds the topics/books from code. I just wanted to decouple > the storage mechanism (here ST source code). Topics/books could also > be stored in an external database/files or even created on the fly from > dynamic information. > I've thought a bit about this. I see two problems with this. Let's say I want to create a book with high level documentation about a package. But one chapter should be the complete API documentation. The easiest way would be if he could write documentation in the current way and for the sub books return in addition one object that generates the needed nodes automatically. It seems to me that it would be easier if the content node knows how to retrieve its data. Secondly, let's say I want to browse the whole API of the system. Because building the nodes is a separate step and the HelpTopic/HelpBook instances do not know how to do this, I would need to create the whole help content before browsing, and this takes quite some time I guess.
I would prefer to have a simple protocol to query a node for its contents and its sub books. Then specific subclasses could implement the required fetching behaviour. I will try to find some time to write actual code to show what I mean. Danny _______________________________________________ Pharo-project mailing list [email protected] http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
