Sim IJskes - QCG wrote:
On 12/22/2010 10:02 PM, Dan Creswell wrote:
What I would say is that if I were to do such a thing I'd be tempted
more by
JSON which is often a little less verbose than XML for many things and
has
plenty of libraries for parsing etc in almost any language. And the other
thing I'd say is that if you're doing the work and prefer XML, you should
get ultimate choice.
Could we use xml? XML based processing is already in the java rt and it
saves us another dependency. If things need to be tooled in a later
stage i would prefer xml. To be honest, coding in DOM will cause me to
tear at least a few hairs out, and JDOM is my favorite (which will cause
another dependency).
I also prefer JDOM for my Java XML processing. That is why I was not
taking into account the dependency issue.
Do you think performance tests should be part of the distribution? I was
thinking of them as an internal tool, but there is some value in getting
them into the hands of users.
To be honest, my preference for XML is a purely emotional one.
There is one idea I've found very effective that I don't think would
work so well with JSON. I used XML for both input and output language
for my simulations. That way, I could do things like including the input
parameters in the output, and trivially regenerate the input file if I
wanted to re-run a test.
I don't think JSON would be a good performance test input file format,
because of the lack of comments.
I really don't know. I'm still thinking about all this.
Patricia
