[Pipet Devel] locus vs. node naming

Brad Chapman chapmanb at arches.uga.edu
Sat Apr 1 11:36:31 EST 2000


> Brad Chapman wrote:
>>     Can we either use "locus/loci" in the definition layer, or think
>> up a new term that won't clash with DOM so much?

Jeff wrote:
 > This is a 'simple' namespace problem, which can usually be solved by
> appending
> the application's name.  So, instead of 'node', you can use...
> 
>    vshnode

Well, I wasn't so worried about the namespace problem as the problem 
with clarity of code. For instance, let's take this current snippet o' 
code from the middle:

		middle_node = dom_tree.createElement('middle')
		dom_tree.appendChild(middle_node)
        
		success_node = dom_tree.createElement('success')
		middle_node.appendChild(success_node)
        
		locus_node = dom_tree.createElement('locus')
		locus_node.setAttribute('id', locus_id)
		locus_node.setAttribute('num_inputs', str(num_inputs))
		locus_node.setAttribute('num_outputs', str(num_outputs))
		locus_node.setAttribute('gui_name', gui_name)
		success_node.appendChild(locus_node)

I think you can see the problem if I started referring to a locus as a 
dlnode (dlnode_node = ...).
    The additional problem is that I am redoing the middle code and am 
thinking of organizing the loci structure in a DOM-like manner. Ie. 
loci can have parent loci (the container or composite locus holding 
them) which will make things additionally confusing. I just want to 
see the code be as clear as possible and the node/node stuff is 
confusing, IMO.

> BTW, about the BL nodes, I was thinking that since these always have 
an
> Internet location, they could be referred to as a node of type 
'locus'.

I'm confused by what you are talking about here. Do you mean the 'GMS' 
layer?

Brad








More information about the Pipet-Devel mailing list