Re: XMLSERVICE with .Net

Nathan

It would be useful if you tested XMLSERVICE itself, right? If you have time!!

I'm still trying to remember the details, but I think XMLSERVICE was written to replace another remote interface that had limitations - I suspect it was jt400, but I am probably mistaken.

The use of XML is meant to make this interface more flexible, IIRC.

And it did begin, as I recall, from an impetus from Zend - yet it does not require PHP - but Zend uses it for its interface to the i.

If we find out that performance is a wash, there still may be other reasons to consider it.

Again, someone else has to clarify my vague recollections. Or get Brian's session handout from his COMMON presentation.

Vern

On 8/6/2012 12:32 AM, Nathan Andelin wrote:

From: "Dean, Robert"
I'm actually reserving judgment. I'd like a little more detail on the tests.
Based on the descriptions Nathan's provided, I have a concern that it
might not be an apples-to-apples comparison.

I tried to make it clear that I wasn't testing calls to XMLSERVICE procedures or comparing its CGI interface vs. its QZDASOINIT interface. I was just bench marking my own CGI interface vs. a QZDASOINIT interface for downloading SQL result sets to my browser, Visual Foxpro, and IBM i Navigator.

The CGI interface performs slightly better than the QZDASOINIT interface, in my benchmark.

So I question Alan's and Brian's performance "warnings" with respect to the XMLSERVICE CGI Interface.

-Nathan