[cad-linux] Re: data format vs data system

> The following was supposedly scribed by
> Janek Kozicki
> on Thursday 18 September 2003 07:26 pm:

>> > format <-> read/write library <-> networkable access system <-> client
>> > library
>>
>> or:
>>
>> format <-> read/write library <-> client tools that can do
>>                                   interesting network stuff if they
>>                                   like
>
>I bet that Bruno is right - 'the simpler at first - the better' can be
>of course enchanced in the future.

Yes, and possibly the client library contains the functions for the network 
access system.

With this multi-tiered abstraction model, I am just trying to think on a 
long-term and large scale in terms of adoption and true universality.

There is no shortage of geometric data formats.  I think the way to really 
universalize is to create a data system (network of libraries and programs.)

For large complex programs, the more abstraction the better.

I think it is great to start simple, but we should also try to envision the 
future uses and requirements, which will allow the design to evolve without 
huge setbacks.

One of the issues that comes up immediately is how to expand this 
format/system to parametric data.

Maybe a more important question, which would likely provide the answers to a 
lot of these details, is what would be the central defining metaphor for the 
format?  (example:  with Unix, "everything is a file" has been central to the 
design of all things Unix)

Maybe we should come up with a label other than "format"?   

Database Specification?

daSpec?

--Eric
-- 
"Chess is a foolish expedient for making idle people believe 
they are doing something very clever when they are only wasting 
their time."
                                        --George Bernard Shaw

Other related posts: