Alarm Functions in DM42 request

Discussion around the Swiss Micros DM42 calculator.
Post Reply
User avatar
TomC
Posts: 56
Joined: Mon Apr 24, 2017 7:00 pm
Location: Detroit/Palo Alto, USA/Third Stone

Alarm Functions in DM42 request

Post by TomC » Thu Jul 25, 2019 9:18 pm

I would like to formally request that the alarm functionality(ies) be added to the DM42.

This functionality was in the 41, and this would make the DM42 a more worthy successor to the 41!

Thank you,
TomC
ps: Thanks again for the DM42 - a great machine!
TomC
DM42:00068/03961

User avatar
ab_normall
Posts: 16
Joined: Sat Sep 07, 2019 11:50 pm
Location: geneva

Re: Alarm Functions in DM42 request

Post by ab_normall » Wed Sep 11, 2019 3:13 am

it seems that it is Hewlett-Packard's themself have left out the scheduling of hardware interrupt at the use of calling program/function or simply messages with snooze (TONE] when they conceived the new HP-42S, on other basis and mostly oriented for calculations (my hypothesys) IDK(!)

But against XYZALM and ALMCAT /etc functions as well as the timing function (SW) -mode (original Time Module or the CX version of the HP-41) all this seems to work on the DM41L. For the 'SW? it is set like a clock it works perfectly, there is also the time factor adjustment function (SETAF, RCLAF), but all this on the DM41L, I did not see any problem, but I do not butvery short tests of those functions.

But to return to the DM42, these functions were already "zapped" from the original HP-42S, between the last HP-41CX (including clock interrupt (~ RTC or some) from HardWare to to HP-41CX-program level/stockage or the HP41CV + TimeModule) these functions have completely disappeared since the original HP-42S which seems to be focused on mathematics skills, like it seems (???)

So the DM42 which is already better than the original could potentially implement in addition these functions, but I confirm that it will be a lot of hours of work to redo exactly what was able to do with clock/interrupt for the HP-41CX. (or HP-41CV + TimeModule)
So with DM41L it seems all right in working with clock at "highlevel", but it is also normal that the DM42 does not necessarily do what the HP-42S's choice themselves NOT to implement the clock/alarm/SW, etc. In fact this could lead the machine to a very tricky state especially including the fact the XYZALM have a "Repeat Option" and could just call[back] in any time a "highlevel" function (like for a HL-slice). Remember the XYZALM function is able to call function even when the Machine is "OFF" (at hard level, including the RTC, its never completly "OFF" so it can be awake..... each day? ....ok , but each seconde.... could be like You can just HARD-RESET and havig the epic message "MEMORY LOST" of the HP-41C/CV/CX)
Normally I think that these kind of options can be programmed either by the basic functions simili-HP-42S / DM42 but you have to ask HOW TO SCHEDULE A CALL of one of your OWN existent program/function. (This will be just the first step, before asking an clock interrupt on a TimeStamp and after repeating each X minutes or hours or Day, etc, but the trap here, is if you ask a repeat less than the time you can stop it manually, it will make your machine falling into a "High Level Time Slicing" un controllable if you called (as callback) function is not accurate and only RESET will stop this bad boat.

But really it seem like Hewlett-Packard themself had abandon those function for the HP-42S which I've neverk new the existence since some few months, compare with I knew XYZALM and ALMCAT since 80's on HP-41CV+TimeModule and was sometime in big trouble if the repeat factor (trying to make kindof TimeSlice for as very short time function, but in any case even HP-41CV was to slow to do such thing in High Level of langage, .... I havent trayed in Synthetic Langage (notice just I don't).

Resume: Ask for a scheduling DM42's function level (LBL "myInt"...) as a callback on clock interruption if the ARM lets do that (normally yes) but there is other element of complexity here. I confirm just that realTime interruption from the Hardware till the Highlevel langage of the machine is a complex process to program and adjust/synchronize/etc
#RealTimeProgramming catch an interruption at hardlevel (electronic signal and stock timestamp also managing with process-context swaping, in deal with a "HighLevel" stocked interruption (timestamp of any), etc. To do taht, its System Programmation in any case. (and a huge work! but I don't know if its on their list)
about XYZALM usage its on page 32 of the HP-41CX Quick Ref Guide ...
And, at my side, this little problem interest me too (!) so I continue to investigate with the two DM42 which I have actually,
But I'm also very curious about the completness of those functionnalities in my both DM41L , ... with XYZALM normally easy to check :shock:
sorry for my very brainwashed english spoken :oops:
  • -----------
Geneva, 6h22pm edition and try to sleep #zzzZzzzzzz....RzzzzzZzzzzzzz <----- this is not a Bee!
EDIT/1 OK, but not OK, I've just tryed the DM41L's XYZALM with the machine OFF .... and NO INTERRUPTION OCCURS since I've turn ON my machine DM41L. I confirm absolutly it was NOT the case with the original HP-41CV with Time Module ... even "OFF" all ALARM was acting as ALARM as to do with an OFF machine especially which machine having a timeout to get OFF themself after the.... timeout, lol. #Complexity of this let me suspicious about making all same functionnalities in the DM42. (in fact this is a long work and adjustment, try-test retry-test with time and at HardWare level till HighLevel functionnalities. So... (?)

ALARM system of TimeModul or HP-41CX was able to callback a specified HLL LBL or just send a MESSAGE with a TONE, or without any indication it BEEP a TONE with DATE + TIME as message. Im absolutly sur about this.

Is it important? I can say, to me I use other machines to be awake in time for medical purpose, but i'm just curiouse if the HP-50g is scheduled with OFF too ... normally yes, like the HP-48SX I haved 2 pieces, but dead by bad stocking by ma fault. So I've then taked (buy) 3 times the HP-50g and actually no more on HP's market, not even the HP-Prime (?) but the HP-Prime does'nt interest me for electronical reasons.
I'm too brainwashed to not ask something about this. #calmdown (says all my doctors)

User avatar
ab_normall
Posts: 16
Joined: Sat Sep 07, 2019 11:50 pm
Location: geneva

Re: Alarm Functions in DM42 request

Post by ab_normall » Wed Sep 11, 2019 7:23 pm

some other "blind" investigations...


About the XYZALM function of the HP41C[V]+TimeModule or the HP-41CX, I confirm that the original behavior implemented by Hewlett-Packard is one to cause a "strong" interruption and able to deliver a message in date+time+seconde (with repeat delay possibilities) as much as calling a "high-level" routine (e.g. LBL "MYINT") even when it is OFF, including if it is running another "High-Level" program you can check it. Notably I was working with a browser that had some digital control option plugged in that could perform navigation fixes after computations produced eg when sleeping, etc. These features were clearly mentioned by the manufacturer.

singlest like this: (with DM41L)

Code: Select all

0 ENTER 
0 ENTER
[ALPHA ^^MYINT]
[b]XYZALM[/b]
...
check with: [b]ALMCAT[/b],
...
.../etc
...
Check with original documentation of HP-41CX documentation or just the Time Module documentation delivered as <Libraries> with RTC (in 80's context) alway running with very low current intensity (ask your favorite electronician about RTC and microcontrollers able to invoke hard interruption to the processor after having get "ON" (or insurring the machine is already "ON" in time delay. Including context switching, interruption handlers, synchronizations, etc. (RealTime Programming for further handling the interruption in respect of ALL internal data structures, registers, etc #VeryTechnical Point here I let you maker your choice of interpretation because i'm too old school to say more details about it #SystemProgramming like for UNIX or Embed programming, using POSIX norms whatever(...), etc) LOOOOOL AND after all investigations this option is embed inside the hard of the Cortex (ARM STM32L476xx so.... Eagle? VHDL? Whatelse? Nescaffe? )

So I see that these scheduled interruptions are incomplete for the DM41L but the SW work well :shock:

By stating that in my case it is not a vital problem since I like these SwissMicros GmbH machines unconditionally and that the concept of alarms does not really matter to me as my tiny part of fun activities also to old aged to restore a kind of new world.

This was to respond indirectly to a request for this kind of interruption for the DM42, pointing out that even Hewlett-Packard themself had not re-implement these clock functionnalitises and I understand their choice: That very tricky to drive from hard (electronic level passthrough a (lets say) "a microKernel" (idk?), and an Interpretor (may Free42) to finally reach a "highlevel" routine without synchronization (POSIX?) just unthinkable by construction (or? ... not so impossible but I suppose other engagments are yet involved for further)

[refs_to_request about DM42]
viewtopic.php?p=11299#p11299

But anyway, the machine DM41L works wonderfully for example for its original function SW (StopWatch?) But without (any?) doubt it has certainly been written directly in "synthetic language" or just in hexadecimal (C) probably including a / Microcontroller (s) assigned to the tasks of management of the electronic timer to be "raised" to the "high level" of this very specific use to the HP-41C/++[TimeModule] ...

Thus, the "SW" function works very well on the DM41L ...

Just try with

Code: Select all

XEQ "SW"
...I don't remember if HP-41C/++TimeModule had also shortcut keys for that, I mean just to call this function. In running "SW" their is many keys able to register intermediate time lap when StopWatch is started, or to erase all, or just stop and restart, etc. :roll:

It was just to say (after some request about this kind of functionnality but to me is ok, I dont need more then that or I will do in ADA/GNAT 2005 (because the higher ADA(s) are just out of my "pocket money" even if I have a perfect running Sun Blade 100 Station, but actually without ADA so (or it will be very costly by ADACORE for SPARC architectures, etc.).... yes it work but this machine actually is just in stock. #MyLife My "Decoration" are Machines (sorry, but poster of Michael Jackson could hurt my feeling so I do prefere surrounding me by MACHINES many MACHINES and over my bed is other MACHINES I love MACHINES,) [and here my doctor (psychioatrist is talking to me: please could you calmdown? and I say Yes!]

Bad english is for me: so "SW" could mean "StopWatch" ??? (not important) but "Chronomètre" (in French) is "Chronometer Watch" in english .... I'm lost in translation. By the way XEQ "SW" is working in DM41L. <---- that's bad, this is the third time I say something like this #brainwashed

Anyway this subject interest me for some old school reason...
One life,
One Old School,
Many Deceases,
Tomorrow I've other doctors to diagnose things...

And I love Bees
Imagehow much does an anthropologist make a year

STM32L476 seems like Knuth's could be the "father" of RISC-architecture (versus CISC) I could be wrong.

[datasheets]
https://www.st.com/resource/en/datashee ... l476je.pdf
https://www.st.com/resource/en/datashee ... l476rg.pdf
https://www.st.com/resource/en/programm ... 046982.pdf ~ about interruption with 16 level of priority
https://infocenter.arm.com/help/topic/c ... 01_UAL.pdf
https://simplemachines.it/doc/arm_inst.pdf

I've checked the ARM-Cortex abilities: 2 ALARM(s) could be "tamper" inside the ARM itself, using an external 32768Hz RTC, so ... are those options already used for another thing (?), ALM1+ALM2 seems programmable at HardwareLevel but of course it could be a very long work to synchronize the hard interuption till all stage in between to reach the "highlevel" stage operating without RT-synchronizing all level of the Interpretor. (e.g. POSIX) ( talking about DM42 ) may other project(s) further on bench ? (idk, only Thomas Okken, Michael Steinmann and David TheManBehindTheCurtain could say more about global team's intention)
All RTC events (Alarm, WakeUp Timer, Timestamp or Tamper) can generate an interrupt and wakeup the device from the low-power modes.
page 41/270 in [STM32L476xx/pdf] document, but at this level, the last time i've programmed an assembly from an old Pascal from Niklaus Wirth itself was in 80's in Microproicessor Laboratory of the school, and with a Motorola 6809 driving an articulated arm in an industrial production chain frame, it was just before we approach the epic Motorola 68000 (RISC) times are a changin... (6809 was used in one model of the TRS-80 should be actually found in some Museum or Privates)

I can't help any, for this ALARM in this context, but I've just notice there is 2 programmables in this ARM STM32L476xx (as they sayd) I don't even understand how they match the usage of those specific alarm of this Cortex, and the official Instruction set for ARM ... and wich one? ARM v7? idk... :oops:

The STM32L476xx (ARM) is Ultra-low-power Cortex addr 32bit, so those choice make not so "obvious" compromises to deal with, before using one of both ALRM with this internal RTC/calendar managment. oups calendar managment is another specificity using very old BCD binary mode to represent decimals (IBM use them for their favorite client: The Bankers) -- I was working for some, so -- Do we Speak EBCDIC ? ... hum I prefer ASCII thanks.

P.S.: Excuse me in advance, but I have some pretty severe medical obligations just tomorrow morning (till now) associated with some heavy administration, and I could not follow that thread "too late" until probably this Weekend. But any way I love to read ALL your FORUM slowly. (I'm actually in hospitalization/diagnoses by appointment concerning several diseases that fall under medical surveillance).

XYZALM will stay a mysteryyyyyyyyyyyyyyyyyyyyyyyyyyy (to me) :o
I'm too brainwashed to not ask something about this. #calmdown (says all my doctors)

Thomas Okken
Posts: 600
Joined: Tue May 02, 2017 3:48 pm
Location: New Jersey, USA
Contact:

Re: Alarm Functions in DM42 request

Post by Thomas Okken » Thu Sep 12, 2019 1:22 am

This topic has been discussed before on this forum. Not in as much depth as it could be, but still, look here: viewtopic.php?f=19&t=2217

Good luck with your health issues!

User avatar
ab_normall
Posts: 16
Joined: Sat Sep 07, 2019 11:50 pm
Location: geneva

Re: Alarm Functions in DM42 request

Post by ab_normall » Fri Sep 13, 2019 9:30 pm

Thomas Okken wrote:
Thu Sep 12, 2019 1:22 am
This topic has been discussed before on this forum. Not in as much depth as it could be, but still, look here: viewtopic.php?f=19&t=2217

Good luck with your health issues!
Thank you very much, and I took note of it.

It seems that this is what is at stake in the complexity of solving this problem (including the multiplateform restrictions this is strong quasi unsolvable point since Free42 seems to be a restricted system access (APP) for several plateform, beyond solving synchonization's points, this will stay a "Dead End" state of solving this case.

To me that sound correct. HP-41(C/CV/CX are just ONE plateform and they (HP) was owning One and only One model's category/electronic scheme [...] And also it seems the fact they choose to not continue to reimplement XYZALM especially in their next specific HP-42S generation, etc.

Exception later with very high refactoring (or some turn over of team involved) returned from scratch for those series of "RPL" (CAS, XCAS, YCAS or any) I think those generations of HP (from 48 to 50 and I suppose HP-Prime was too (idk) if they work on another basis with an embedded RTOS make all that things back again, but always the common fact is: One plateform, They OWN it, and it was certainly some new and fresh conceptors involved for (turn over or idk).
I was at the hospital this morning for a preamble of analysis towards a choice that gives me a maximum theoretical life expectancy of 2 years, or less. So i'm not anymore doing to huge projects involving more than few months.
I always having fun with those DM-machines and supporting your works and idea, including of course SwissMicro GmbH (by stay an inconditionnal client of all machines involved in DM) including yourself, do a work of a nature almost against the current modes of a certain ignorant global consumerism. By the way I will stay a kind of weird consumerist till my own and very last "game over".

Only passion involved here is important to my eyes.

....and sorry for my "GoogleTranslated" way of writing english... (LOL)
I'm too brainwashed to not ask something about this. #calmdown (says all my doctors)

Post Reply