(34S) Integration (using the double-exponential method)

[emece67]

(03-25-2017 12:33 AM)Paul Dale Wrote:  What do people think about including the last one in the 34S firmware?

Perhaps a little bit more time for people other than me to test the program is advisable.

(03-25-2017 12:33 AM)Paul Dale Wrote:  Does it do the right thing with the final stack?

It does the very same thing, in this respect, that the built in integrator: corrupt all the stack. I was able to write a version (not to be used in XROM) preserving the stack, but I found some trouble when porting it to XROM. Perhaps I need some advice here. My approach was:

• Use 2 local register to backup Z & T
• When returning, build a stack with the backed up T & Z, the result in Y and a dummy value in X
• Then do a DROP and return
  

It worked if the user was working in SSIZE4, but not in SSIZE8.

I do not want to use xIN/xOUT because it forces DBLON mode and because I'm not sure if XEQUSR/POPUSR can be used inside xIN/xOUT.

In any case, I'm still trying to fix this (looking into ->A..D->).

(03-25-2017 10:21 PM)Luigi Vampa Wrote:  BTW, the license and copyright header reads: "M. César". That "M" must stand for marvellous... or may be mathematician... or both in one single letter, by some sort of arcane&syncopated algebraic expression ;O)

Jaja, Luigi, la M solo es de Manuel.