Also, AFAIK, it’s also good practice to not use “inner text” on nodes that have child nodes. You’ll notice the XML file will have kind of strage formatting if you do this, even though it can be done.
great, thanks rich – this thread has been really useful. I was thinking that depending on the data type would depend on whether things were nested or not – ie a stored morpher has channel,keytime,in/out interp type etc so could be nested and a stored TM could just be placed into the node on it’s own, perhaps as an attribute of the xml node.
Yup, ZB, LR, PEN, that’s too strongly put … I’m being a purist :).
nothing wrong with that! thanks for your input, as always Dr.
If anything, I prefer the innertext method purely for readability, even if it does make the file larger. I’ve not done a huge amount of XML parsing myself, but for eye-scanning I find the attribution method frustrating.