Plus42 Equations, Preview Release

[Thomas Okken]

Hmmm... there is another possibility, which just occurred to me.
In the VARMENU shown by the solver, entering a value already takes place using a non-standard STO operation, one that keeps track of the previous value as well (so you can do, say, 2 X 3 X X, to provide starting guesses 2 and 3 for X, and then solve for X). This special STO operation could be modified to apply the previous unit, and you could force a unit-less value by pressing STO X instead of X, because using the STO key always gets you the standard STO operation.

[Marco Polo]

(03-04-2022 09:28 AM)Thomas Okken Wrote:  Hmmm... there is another possibility, which just occurred to me.
In the VARMENU shown by the solver, entering a value already takes place using a non-standard STO operation, one that keeps track of the previous value as well (so you can do, say, 2 X 3 X X, to provide starting guesses 2 and 3 for X, and then solve for X). This special STO operation could be modified to apply the previous unit, and you could force a unit-less value by pressing STO X instead of X, because using the STO key always gets you the standard STO operation.

I like it!

I think the hard part of the work will be to recap all the new features and write a brief manual

[Thomas Okken]

New update:

1. User-defined units.
2. Sticky units in VARMENU.

(03-04-2022 11:32 AM)Marco Polo Wrote:  I think the hard part of the work will be to recap all the new features and write a brief manual

There is certainly plenty of work to be done in that department, but I feel confident it will be easier than writing the code.

[rprosperi]

(03-04-2022 12:25 PM)Thomas Okken Wrote:  ... I feel confident it will be easier than writing the code.

Easier? Sure.

But as well done or enjoyed? No. But I suppose this is possibly a historical opportunity, where no software manual is ever as good as the software, with the notable exceptions of the extraordinary PPC manual and Angel Marten's superb manuals.

[Thomas Okken]

New update:

1. User-defined base units.
2. Unit support in the CUSTOM menu.

Both of these features make use of a new base unit named 'one', which is exactly what its name suggests. It is used as follows:

1. To create a base unit, store the value 0_one under the desired name.
2. To add a unit to the CUSTOM menu, store a variable with the desired name first. If you're assigning a user-defined unit, you had to do that anyway to, to define the unit itself, but if you want to assign a standard unit to the CUSTOM menu, you'll have to create a variable for it as well. For example, to assign the kilometer, do 1_km STO "km", followed by ASSIGN "km" TO <key>.
The 'one' unit comes in handy if you want to assign a dimensionless constant, for example: PI 2_one * STO "tau", followed by ASSIGN "tau" TO <key>. You need the 'one' unit here because when you assign a regular variable to CUSTOM, the menu key performs a RCL; in order to get CUSTOM to multiply or divide by a unit, the variable must be a unit object.

[Jlouis]

Hi Thomas,

MalwareBytes is flaging some file in the simulator anomalous, with their machine learning alghorithm. See their report below, in portuguese.

Just to let you know, it's a false positive for sure. Maybe you can contact them?

By the way, what a wonderful calculator, 42s + 27s + 50g (units)...

Cheers

Malwarebytes
http://www.malwarebytes.com

-Detalhes do Relatório-
Data da análise: 05/03/2022
Hora da análise: 04:05
Arquivo de relatório: b563b392-9c52-11ec-b7f9-000000000000.json

-Informações do Software-
Versão: 4.5.4.168
Versão de componentes: 1.0.1599
Versão do pacote de definições: 1.0.51887
Licença: Somente
Premium

-Informações do Sistema-
Sistema operacional: Windows 10 (Build 18363.1556)
Processador: x64
Sistema de arquivos: NTFS
Usuário: DESKTOP-0DI30N3\jloui

-Resumo da Análise-
Tipo de análise: Análise Customizada
Análise Iniciada Por: Manual
Resultado: Concluído
Objetos verificados: 1
Ameaças detectadas: 2
Ameaças em quarentena: 0
Tempo decorrido: 1 min, 12 seg

-Opções da Análise-
Memória: Desabilitado
Inicialização: Desabilitado
Sistema de arquivos: Habilitado
Arquivos compactados: Habilitado
Rootkits: Desabilitado
Heurística: Habilitado
Programa Potencialmente Indesejado: Detetar
PUM (modificação potencialmente indesejada): Detetar

-Detalhes da Análise-
Processo: 0
(Nenhum item malicioso detectado)

Módulo: 0
(Nenhum item malicioso detectado)

Chave de registro: 0
(Nenhum item malicioso detectado)

Valor de registro: 0
(Nenhum item malicioso detectado)

Dados de registro: 0
(Nenhum item malicioso detectado)

Fluxo de dados: 0
(Nenhum item malicioso detectado)

Pasta: 0
(Nenhum item malicioso detectado)

Arquivo: 2
MachineLearning/Anomalous.100%, C:\USERS\JLOUI\APPDATA\ROAMING\Microsoft\Windows\Recent\Plus42Windows1.zip.lnk, Nenhuma ação pelo usuário, 0, 392687, , , , , CD31047C8DB0C7CABE4BBB28AF9D2AA4, 73050CD38FD58CA2CB29F2BFA9AB6108D134E0BA20C6A14722543B216C017066
MachineLearning/Anomalous.100%, C:\USERS\JLOUI\DESKTOP\PROGRAMAS CALCULADORAS\PLUS42WINDOWS1.ZIP, Nenhuma ação pelo usuário, 0, 392687, 1.0.51887, , shuriken, , F3325DF0080A8D5634D5E5AFEE9116A8, A77D3CE09E3DFB9B18C4260971CA731B22B733A3052626E20E8EFBDCF7FF4ED6

Setor físico: 0
(Nenhum item malicioso detectado)

Instrumentação do Windows (WMI): 0
(Nenhum item malicioso detectado)


(end)

[Thomas Okken]

Well, they're in good company. Windows Defender has complained about Free42 in the past, and I have reported the false positives, more than once, but every once in a blue moon, it again decides it doesn't like Free42 Decimal. And then comes to its senses again with the next Defender update.

It doesn't have a problem with the binary version, interestingly enough, and that suggests there might be a bit pattern in the tables used by the Intel library that sets it off, since that's really the only difference between the two. But the trouble reports never specify what exactly it doesn't like, so I have no clue what I should be doing about it in the first place. Reporting the false positives over and over gets old.

(Or maybe Defender likes to complain about the binary version but leaves the decimal alone? Now I'm not sure, after enough of this the details get hazy. I also had problems with Norton anti-virus, it would sometimes even delete the executables right after Visual Studio finished building them. (And always just one of the binary and decimal, never both.) I had to stop using Norton in the end because I couldn't get it to stop doing that.)

[Thomas Okken]

New update:

1. Unit support in PLOT.

That concludes the to-do list for features for 1.0. Only testing, bug fixing, and documentation left to do...

[Thomas Okken]

New update:

1. More unit support in PLOT (specifically, units weren't being passed to SOLVE and INTEG yet).

[Thomas Okken]

New update:

1. Stack preservation during SOLVE and INTEG from the plot viewer.
2. Smarter unit handling in SCAN.

[Jlouis]

Hj Thomas,

The latest download I made only X and Y appeared on the Stack. I look in preferences and could not find a way tô get the four Stack in the display.

How can I do It ?

TIA

Edit I mean the Android version

[Thomas Okken]

ROW+, in the DISP menu

[Jlouis]

(03-06-2022 11:22 PM)Thomas Okken Wrote:  ROW+, in the DISP menu

Thanxs

[Thomas Okken]

New update: some more plot-related improvements / bug fixes...

1. SCAN now handles the numerical solver a bit better: you can specify the starting guesses it should use with YMIN and YMAX, and it will update YMIN and YMAX based on the actual range it finds only when the scan is complete.
2. Results of function evaluation (when marking a point with ENTER), SOLVE (Shift-7), and INTEG (Shift-8), now displayed with the appropriate units.
3. SOLVE (Shift-7) failure message now reported in plot viewer.

[Thomas Okken]

New update: Minor PLOT bug fixes & cosmetic improvements...

[Eddie W. Shore]

A lot to update. Thank you, Thomas.

[Thomas Okken]

New update, Android only: Fixed crash caused by exiting the app while the equation editor was active. The root cause was that when re-launching the app, the blinking cursor would be started before the parts responsible for updating it were fully initialized.

The other versions don't seem to have this problem, but the fix is in the core, so the other versions will get it as well, with the next build.

[Thomas Okken]

New update, Windows only: Added support for non-integer display scales.

This means that changing the number of columns now works... However, I haven't yet figured out how to get it to reduce the display bitmap with interpolation, so on a standard-resolution display, the result looks quite ugly compared to the MacOS version, running on the same hardware.

It does work, though, and hopefully it does look OK when run on high-DPI screens.

[Thomas Okken]

New update: Minor stuff... PLOT number formatting, renamed GRAPH to PLOT on the keyboard, and a couple of minor bug fixes.

[Thomas Okken]

New update, Linux only: Added support for non-integer display scales and COL+/COL-.