HP 48 Series -- interseries compatibility?

[johnb]

I recently snagged a nice HP-48s in nearly new condition.
(Photos to follow later!)

I have an extensive suite of programs and menus for my 48g's (I own two), but I really like the color scheme of the 48s better, and would like to swap it into my daily driver lineup. The programs are all 100% user RPL. No SysRPL was used.

Can I back up my 48g with the interface cable and just restore the RPL onto the 48s with no changes? (It's a lot of tedious and error prone typing in otherwise.)


Thanks!

[rprosperi]

Fundamentally, yes you can backup from one and restore to another, but there definitely are some compatibility issues to be aware of, even in UserRPL. The GX added some new commands and many SYSEVAL addresses changed as well.. Look at the HP-48 FAQ here:

https://www.hpcalc.org/hp48/docs/faq/48faq.html

Section 10 covers GX changes, as this FAQ goes back to when the GX was relatively new.

Also, be warned that if you are speed sensitive and you like to run huge programs or ones then run for many hours, etc. and attempt to boil the ocean as so many others around here seem to love, you will definitely feel the difference, as the 48SX is definitely slower and you will see it.

But most often, for haughty calculation challenges like household chores and checkbook balancing, I prefer the 48SX too, especially those with the LCD replaced by a Black one from a later machine. Perfect aesthetics...

[johnb]

Thanks, Bob!!

(10-02-2023 11:09 PM)rprosperi Wrote:  Also, be warned that if you are speed sensitive and you like to run huge programs or ones then run for many hours, etc. and attempt to boil the ocean as so many others around here seem to love, you will definitely feel the difference, as the 48SX is definitely slower and you will see it.

But most often, for haughty calculation challenges like household chores and checkbook balancing, I prefer the 48SX too, especially those with the LCD replaced by a Black one from a later machine. Perfect aesthetics...

I can so agree about the aesthetics! And, no, I don't need to run huge multivariate integrations with a bazillion iterations to squeeze out every digit down to the ULP. There's PCs and Mathematica (and other tools) for that.

LOL, what I use my 48g for (and would use the 48s for) is to emulate a 16c but with a big stack display... or run my interval arithmetic library (which is pretty simplistic and computationally non-intensive). It's kind of spiffy for me to just enter approximate values and either a +/- error term or an asymmetric range about each value, and then do my computations and not have to think about how the uncertainty is calculated: it's done brute-force for me and I can see the actual uncertainty right there.

Oh, just remembered, I do date math on it, too. And simple complex arithmetic since it's so easy to both enter the numbers and read the results.

[johnb]

(10-02-2023 11:09 PM)rprosperi Wrote:  Look at the HP-48 FAQ here:

https://www.hpcalc.org/hp48/docs/faq/48faq.html

Section 10 covers GX changes, as this FAQ goes back to when the GX was relatively new.

Ugh. Nailed by section 10.1:
> Almost all commands that did not accept list(s) as their arguments can do so now.

The examples are exactly the kind of code I have in my self-written interval library. An interval value is a list: {best-approx, lowval, highval} and the library heavily relies on those commands working the way that the 48g does with lists.

Not an impossible hurdle: I can probably port the library. But it's not gonna run unchanged. (Everything else outside of the interval arithmetic probably will, however.)

Snif. Whimper.

[rprosperi]

Other than porting the library to handle lists, it seems these are all reasonable chores that the SX is up to, so come on down...

[Giuseppe Donnini]

You should give the following programs by the late and great John H. Meyers a try:

https://www.hpcalc.org/details/2889

Although written in User RPL, they provide nearly equivalent (and identically named) S/SX versions of the list processing commands introduced with the G/GX.

[John Keith]

(10-03-2023 03:25 PM)Giuseppe Donnini Wrote:  You should give the following programs by the late and great John H. Meyers a try:

https://www.hpcalc.org/details/2889

Thanks, I hadn't noticed them before. Worth a read even for users of later models- the notes explain the inner workings (including bugs!) in the GX list processing commands.

[johnb]

(10-03-2023 03:25 PM)Giuseppe Donnini Wrote:  You should give the following programs by the late and great John H. Meyers a try:

https://www.hpcalc.org/details/2889

Although written in User RPL, they provide nearly equivalent (and identically named) S/SX versions of the list processing commands introduced with the G/GX.

That's beautiful; I'll try adding those tonight.

I guess I am closer than I thought.

Anyone else here know what CRS disease is? Because I definitely have it! On my lunch break today, I remembered: the 48 series has IR ports! (Where's the dunce cap emoji, now?)

So I just transferred everything directly from my daily driver 48g to my "new" 48s. Most of what I've transferred Just Works (hooray!), and those that don't are simply giving "Unknown XLIB" errors, which don't seem harmful, just "I don't know what to do here." I will bet that adding in Mr. Meyers' implementation of the missing operations will get me to the 90-95% point.

All that would be left from there is to write functions to add/subtract/otherwise-apply a scalar to a list, and add in the hooks to access those, and I'll probably be at the 99% point. (There's no doubt I'll miss something... no telling what until I trip over it and do a face plant.)

[rprosperi]

Even if you add in the missing commands using replacements from John Meyer's routines into your device, you will have to recompile the programs that utilized the missing commands as the saved programs you transferred have the original ROM/library addresses compiled into them, which is why they are throwing the missing XLIB error.

Look at the original source code for the programs having these errors and compare to how the code looks on the 48S where you'll see the XLIB reference where each of the missing commands should be; this will reveal which commands you need to replace. Once they are in the 48S, recall the code, edit it to change the XLIB commands to the new command name (as seen in the original code and also the new one you created) and then save the revised program to new name (to save your old one in case of errors). Once you've got them all working, you can change the program names back to the original ones.

hth

[johnb]

(10-05-2023 02:03 AM)rprosperi Wrote:  Even if you add in the missing commands using replacements from John Meyer's routines into your device, you will have to recompile the programs...

Thanks for that important information, Bob!

Since there's nothing special on the 48s that needs saving, could I wipe the memory, then load the Meyers routines, then re-transfer all my UserRPL components from the 48g? With the referenced routines already being in place, would the transfer serve to get the recompilation done, or would I still need to take additional steps?

[rprosperi]

(10-05-2023 03:07 AM)johnb Wrote:  

(10-05-2023 02:03 AM)rprosperi Wrote:  Even if you add in the missing commands using replacements from John Meyer's routines into your device, you will have to recompile the programs...

Thanks for that important information, Bob!

Since there's nothing special on the 48s that needs saving, could I wipe the memory, then load the Meyers routines, then re-transfer all my UserRPL components from the 48g? With the referenced routines already being in place, would the transfer serve to get the recompilation done, or would I still need to take additional steps?

More knowledgeable experts will have to opine here to be sure, but I think that you will still have to do what I suggested. I believe the programs are transferred as-is (binary) and no recompilation is done, but could be wrong. I'll follow too to find out...

[John Keith]

(10-05-2023 03:07 AM)johnb Wrote:  Since there's nothing special on the 48s that needs saving, could I wipe the memory, then load the Meyers routines, then re-transfer all my UserRPL components from the 48g? With the referenced routines already being in place, would the transfer serve to get the recompilation done, or would I still need to take additional steps?

You could also use the Connectivity Kit (provided that you have the cable) to transfer the programs from the 48G to a PC where editing would be easier, then transfer the programs to the 48S when done.

[DavidM]

This sounds like a fun project. Let us know how you're progressing!

[johnb]

(10-06-2023 10:36 PM)DavidM Wrote:  This sounds like a fun project. Let us know how you're progressing!

Well, fiddling with the 48s is on hold for a moment.

I've just found out that my proposed panel, "Anime Blinded Me With Science," has been accepted (again!) for this year's Anime Weekend Atlanta.

I need to "Get Crackin'!"

The last 3 times I did this, they gave me a 100 seat room, then 150 seats, then 200. Each time, we filled the room and had to turn people away! <3

This time, they gave me "prime time" = Saturday, 28 Oct, 3:15 PM, in the middle of the most-attended half-day of the convention.

And they gave me the Habersham ballroom. 500 seats!

Presentation to 500 people is not the same as 150-200. Not everyone can see you clearly, so you can't use a whiteboard unless you have a live camera feed to the overhead screen.

I'm digging out my green pointer laser (bright enough to see the beam in the night sky to point out planets and constellations. [Question: TOO BRIGHT?]

If any of you will be in the Atlanta area during that time, please consider attending at least on Saturday so you can join me and cheer me on! (And prevent my attendance from being 210, which would be REAL embarrassing!) The entire convention runs Thursday 26 through Sunday 29, if you're interested --- it's loads of fun and you might discover a new thing or two.

[johnb]

The best part (for HP fans) is when I present a scene from "Rocket Girls."

A young prospective astronaut is in a classroom, and they're covering Hohmann transfer orbits (and the equations on the whiteboard in the scene are correct!)

The prof asks her for the value of a couple of variables. "Use your calculator!"

And she says, "What the Heck? This calculator has no equals key!"

And they've clearly rendered an HP-41cv.

   

[Steve Simpkin]

(10-08-2023 05:20 PM)johnb Wrote:  The best part (for HP fans) is when I present a scene from "Rocket Girls."

A young prospective astronaut is in a classroom, and they're covering Hohmann transfer orbits (and the equations on the whiteboard in the scene are correct!)

The prof asks her for the value of a couple of variables. "Use your calculator!"

And she says, "What the Heck? This calculator has no equals key!"

And they've clearly rendered an HP-41cv.

That scene was very funny Thank you for finding that johnb.
Here is the above described sequence shown in still frames.

   

[Massimo Gnerucci]

(10-08-2023 05:20 PM)johnb Wrote:  The best part (for HP fans) is when I present a scene from "Rocket Girls."

A young prospective astronaut is in a classroom, and they're covering Hohmann transfer orbits (and the equations on the whiteboard in the scene are correct!)

The prof asks her for the value of a couple of variables. "Use your calculator!"

And she says, "What the Heck? This calculator has no equals key!"

And they've clearly rendered an HP-41cv.

Yeah John, you already told us about that anime here.

Thank you!

[johnb]

(10-08-2023 06:18 PM)Massimo Gnerucci Wrote:  Yeah John, you already told us about that anime here.

ROTFL! I am clearly getting old, and forgetful!!
Thanks for keeping me honest!

Steve Simpkin Wrote:That scene was very funny :-) Thank you for finding that johnb.

It's one of the places where I've gotten my best laughs previous years' talks. I show the scene. Then I show a side-by-side photo comparison of the anime artwork versus the real thing. Then I say that I am, of course, a nerd, and needed no help recognizing it [I hold high my own 41cv] "because I have my own right here."

[johnb]

LOL. And Steve posted the very same response in the earlier thread!

[Steve Simpkin]

(10-08-2023 06:33 PM)johnb Wrote:  LOL. And Steve posted the very same response in the earlier thread!

Well, in my defense, that was like 9 months ago.
I often have trouble remembering what I had for dinner last night...