[gme-users] RE: [udm-users] BON2 vs. UDM

Peter Volgyesi peter.volgyesi at vanderbilt.edu
Tue Apr 13 20:04:57 CDT 2004


> I think that the point is not to find fault in the different 
> tools but to consolidate them to form a better tool, not have 
> redundant effort. I may be missing the big picture and I will 
> appreciate it if some one could fill me in on the larger goal 
> of having both approaches.
Integration or consolidation is not always justified. As Akos pointed out,
BONx is our GME specific thin API. Many of the current components cannot be
implemented by UDM (at all, or not economically). I would not like Microsoft
to consolidate Notepad and MS Word just because both tools are for text
editing.

> C++ and Java are different languages developed and propagated by
> different vendors. 
If you think that the main reason of having C++ AND Java is because of
different vendors then we dissent on fundamental concepts.

> > It is simple: if you do not like/want to use BON2, then don't.
> 
> [Adi]
> This is true. However, a considerable effort is being spend 
> on both these tools/API's and if we merge them we can get the 
> best of both worlds as well as have a more people working on 
> complementing each other rather than competing. Another issue 
> is that new students and staff have to potentially lean both 
> these interfaces.

I don't compete with UDM. Never will.  :) 
I have my own problem with BON2 however. It is way too complicated. As I
said, I consider BONx as a thin layer (thin means: on-demand object lookup
AND straightforward object model). I'm afraid BON2 missed the second
criteria, and I agree with you that it must be _learnt_ now. This was not
true for BON1. Once a student understood the concepts in GME, BON1's object
model was self evident. 

> [Adi]
> 1. UDM can also be used to write Add Ons 2. UDM allows you to 
> access the "name" and "position" registry values.
> We are yet to come up with a project where we need more.
Probably I missed the point of UDM but can you implement domain independent
components in UDM ?

> 4. Being Generic and decoupled from GME and MetaGME has 
> always been an advantage.
Why always ? If you have a generic framework, than our users will have to
learn the concepts of GME AND the concepts of UDM. It is the same problem
you mentioned earlier about the two parallel frameworks.

> 5. Having event notification I believe is an attribute of 
> gme's interpreter interface and so theoretically even UDM 
> should be able to use it.
Note that event notification is not just a new feature, it heavily
influences the overall architecture of the component. 

--
peter



More information about the gme-users mailing list