Heavy critic regarding "transactions page"

Dear Linden,

A)
Why do the server cache my requests (altought I send no-cache headers everytime...)?

I've paid 20L$ to my alt at 10:45.
I did following request a few minutes later, logged in as my alt:

(1)https://secondlife.com/downloads/transactions.php?date_start=2006-08-13&date_end=2006-08-14&type=xml&include_zero=no
(2)https://secondlife.com/downloads/transactions.php?date_start=2006-08-14&date_end=2006-08-14&type=xml&include_zero=no

One request contains this last transaction, the other not.
My program isn't caching the request.
If I post both request in IE or FireFox I will get the same result.


I would even pay for posting non-cached requests...

B)
Please use standardized XML, i.e. not <transaction_39439> but <transaction>
Where's the point of such thing? No XML parser can read this.

Yes I can fix it with 4 lines of code, but why can't you use standardized XML?

C#

Regex regex = new Regex("<transaction_[0-9]*>");
responseData = regex.Replace(responseData, "<transaction>");

regex = new Regex("</transaction_[0-9]*>");
responseData = regex.Replace(responseData, "</transaction>");

Asking about this...

A) The queries are cached because of database limitations. The queries to get all of your transactions for a given range are too expensive to allow a real-time view into the database that way.

B) This is a side effect of the interaction between PHP and the XML library. We knew about it when we put it in place, but figured that people who wanted to use the XML could do the simple parsing adjustment just like you've done!