Now, looking at the different menus here are some additional observations:

In the CLEAR menu we have CLStk, in the PRINT menu we have ⎙Stck, it would be more consistent to have ⎙Stk instead

In the ARROW (conversion) menu, it may be more explicit to have an arrow in front of the the names to make clear that for ex. DEG is a conversion to degree and is not the DEG command changing the angle mode. There is already an arrow in ->HMS and HMS->.

In the Setup 2 menu we have LdPRG which loads a program file in flash, from where the program can be copied to RAM with PRCL. It may be useful to have PRCL in the Setting 2 menu, above LdPRG to avoid having to search it in the P.FCN catalogue, then it would make sense to have also PSTO in the Setting 2 menu above SvCPRG.

I agree with all of these changes. I've implemented them and uploaded a new version.

[*]In the STAT menu there used to be a special key sequence to light all the dots in the display: press and release shift, and then press and hold CPX. Now that we have Power-Fit on f-CPX in the STAT menu, this doesn't work anymore. Have you moved or do you plan to move this feature to another key sequence?[/list]

The new location of this super-secret special sequence is f press-and-hold CPX on the version of the default menu with only two entries. Sorry for not mentioning the move!

I've created a QRG for complex lock mode using your QRG as a model. A link to a pdf version is below.

I agree with all of these changes. I've implemented them and uploaded a new version.
[...]
The new location of this super-secret special sequence is f press-and-hold CPX on the version of the default menu with only two entries. Sorry for not mentioning the move!

When doing an integration, it seems as if the WP 34S on the DM 42 stays in run mode well longer than when the integral is achieving the accuracy that number of fixed digits would imply. Is there a way to have the integration stop when the accuracy reaches the point where the result will be correct for the fixed number of digits chosen to display?

When doing an integration, it seems as if the WP 34S on the DM 42 stays in run mode well longer than when the integral is achieving the accuracy that number of fixed digits would imply. Is there a way to have the integration stop when the accuracy reaches the point where the result will be correct for the fixed number of digits chosen to display?

So far as I can see the DM42 behaves in exactly the same way as the physical WP34S in this regard - it keeps on refining its result until two successive approximations agree to within the precision of the current display format. This means that if the step 3 result (for example) is correct, the calculator won't know this until it has calculated the step 4 result. This means that the correct result can appear to "hang" in the display for a while before the calculator finishes.

Do you have a specific example where the integration routine isn't working as it should be?

I find that the WP 34S continues to display “RUN” on pretty much all of the simple integrations that I have done. I am in fix 4. I am in double precision. When I hit enter, the answer will often be shown accurate to well beyond the first 16 digits, even though I am only in fix 4. An example is int of (x-1)/sqrt x from 1 to 9. Quickly converges on 13 1/3, but continues to display RUN.

I think that the integration was finishing as it is supposed to, but RUN was not turning off. I don't (yet?) understand why this was happening; I've added a line after the "turn run annunciator off" command to make sure that this change is displayed, but I don't like not understanding why this extra line is necessary.

Anyway, I've uploaded a fixed version. Let's hope that this change hasn't broken anything else!

Thank you for noticing this. Let me know of any remaining problems.

Nigel (UK)

(Edit: I now understand what was going on. If it does fix the problem, this fix won't have broken anything else.)

Our IT person at work used his administrative privilege, and flashed my DM 42/WP 34S with your latest version. Integration works now as it should! Much appreciated!

I was trying to use SLV f’(x)’F0’ with my F(0) program of the function x/(2x+1)^3 to find an extremum. An extremum is at .25, but my unit is coming up with .2520. My physical WP 34S (repurposed HP 30b) comes up with the correct .25 Is it possible that I need to specify a small epsilon (and how do I do so?)? Same program in both units!

In the original WP34S, said 'epsilon' is defined in a routine called δx (default is 0.1, see p. 176 of 410 in the manual for more). No idea how this is solved in the migrated edition.

I bought your earlier 211 page manual when I got the HP 30b flashed to the WP 34S years ago! Do you know if the epsilon was discussed in that manual (and where)? Otherwise, can you tell me how to get to the epsilon so that I can change it to a much smaller value?

Your work is quite remarkable, and that first manual is still quite useful to me!