Survey: Catalysis Templates

[Didier PH Martin]

Hi Eliot,

Eliot said:
I ask because I think that Frameworks as defined by Catalysis are
similar to, if not identical to the concept of document architectures as
defined in the HyTime standard. I'm wondering whether or not talking
about "frameworks" or "model templates" would mean more to the typical
XML developer than does talking about "architectures".

Didier reply:
I am currently studying Catalyst so, before saying anything about it I'll
have to master it. However, concerning the architecture, we can say that we
have more an inheritance relationship between our particular document
structure and the "Architecture". Because of this kind of relationship, we
can say that we have here something that resemble a lot a particular pattern
(ref: Gamma & al.): the interface pattern. For some, a Framework is a
particular assembly of patterns targeted to a particular domain. So, I guess
that it all depends on the architecture scope to call it a "pattern", a
"base structure" (that we inherit from so this makes the architecture
resemble the interface notion) and finally a "framework". I guess that
because of the scope of Hytime we can probably call that a framework applied
to the linkage domain.

This said, thanks for bringing the subject because, for me, "architectures"
are closer to "interface" like found in C++ base classes or Java interfaces.
And what is an interface after all? In its simplest form, we can say that it
is a contract that assure the client that, even if an implementation does
something totally differently than an other one, we still have the same way
to interact with all the implementations. So, for example, if I have a
particular document structure (all the elements have names totally different
than Hytime) an Hytime apps can still interact with my document structure.
Thus, the concept is similar to the notion of interface.

Thinking loud, I would say that a particular Hytime link is like an
interface and if I want to interact with a varlink, I interact with the
"interface" instead of the implementation (which could be for instance a
topic map "assoc" element). However, the collection of interfaces that
compose the linkage domain could probably be called a framework. But because
we deal here more with data than methods, a "data template framework" seems
a good way to call this. We have to be cautious, framework for Object
oriented guys also implies behavior because a framework, most of the time is
an aggregate of interfaces and these interfaces are objects. Therefore the
access to the object is a mix of properties and methods. So, finally, "data
template framework" seems a good candidate here (no implicit notion of
behavior).

Simple opinion of a guys with a bad cold and who hate to be sick :-)

Cheers
Didier PH Martin
mailto:martind at netfolder.com
http://www.netfolder.com


xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To unsubscribe, mailto:majordomo at ic.ac.uk the following message;
unsubscribe xml-dev
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)