|Re: Look to the 33s for answers (and "800+ registers"?)|
Message #13 Posted by Les Wright on 5 June 2007, 12:56 p.m.,
in response to message #12 by Paul Brogger
I've been able to get past the 33s' looks and limitations, and have enjoyed it immensely.
I do to! I have peaceably lived with the HP33s's more embarrassing bugs, like the errors involved when cos and tan take arguments close to 90 degrees. I am looking forward to the increased register number. But I am especially hoping that the restricted number of program labels is rectified, perhaps with an alphanumeric labelling system akin to how the registers are directly addressed.
If the 35s uses 33s processor electronics, or better, then I am particularly looking forward to its speed. I have a number of special function routines that basically iterate a series or continued fraction to convergence. They are agonizingly slow on the otherwise revered 15C, a bit faster on the 11C, a bit faster still on the 41CX, somewhat quicker on the 42S, but quite peppy on the 33S. I really hope for more labels and registers so I can easily port some of my favourite 41C programs and watch them run much faster.
Indeed, despite its many flaws, it is the speed of the otherwise maligned 33s that makes the revered 15C, for example, seem embarrassingly archaic and irrelevant in comparison. If the 35s extends the strengths of 33s and incorporates the improved quality observed in the 50G, then we could have a pretty darn good RPN programmable scientific.