Serious bug in beta 4/16

[jonmoore]

Reporting this here as it's a real showstopper.

All command instantiations on the CAS home are prepended by the word CAS.



Obviously, this should only happen on the Numeric home screen. If the user doesn't first delete the CAS header, all functions fail on the CAS home screen.

This bug is present on both the hardware and emulator.

[trojdor]

Yeah, I'm seeing the same thing...

[TheLastMillennial]

Everything looks and works normally on my G2. I even tried multiple CAS settings but still could not reproduce. Perhaps try resetting the calculator?

[toml_12953]

(04-18-2021 06:58 PM)jonmoore Wrote:  Reporting this here as it's a real showstopper.

All command instantiations on the CAS home are prepended by the word CAS.



Obviously, this should only happen on the Numeric home screen. If the user doesn't first delete the CAS header, all functions fail on the CAS home screen.

This bug is present on both the hardware and emulator.

I don't have that problem on my G2 or my emulator. I'm using these settings:

Angle Measure: Radians
Number Format: Standard
Integers: Decimal
Simplify: Minimum
Exact: on
Complex: on
Use √: on
use i: on
Principal: on
Increasing: off
Recursive Evaluation: 25
Recursive Replacement: 1
Recursive Function: 100
Epsilon: 0.000000000001
Probability: 1e-15
Newton: 60

[akmon]

I do have the same problem on my G2 unit. Same example CAS.solve in CAS mode. Tried to insert same settings above described but still no changes. How strange.

[jonmoore]

Tried resetting and cold rebooting still the same problem. BTW should have mentioned this before, my unit is a G2 too.

I’ve rolled back to an earlier firmware for now and will start testing again with the next build.

[cyrille de brébisson]

Hello,

found and fixed for next version.
Thanks.
Cyrille

[toml_12953]

(04-19-2021 06:36 AM)cyrille de brébisson Wrote:  Hello,

found and fixed for next version.
Thanks.
Cyrille

If I may ask, why doesn't the bug appear on some calculators? The reason I ask is because I wonder if the fix will break it on those machines on which it works now.

[jonmoore]

(04-19-2021 06:36 AM)cyrille de brébisson Wrote:  Hello,

found and fixed for next version.
Thanks.
Cyrille

Thanks Cyrille.

[cyrille de brébisson]

Hello,

It was an uninitialized variable issue..
a flag that was meant to indicate if the CAS. prefix was needed was not always initialized...
Depending on what you had in RAM prior to using the function, you could get this or that happening...

Cyrille

[David Hayden]

(04-19-2021 02:17 PM)cyrille de brébisson Wrote:  Depending on what you had in RAM prior to using the function, you could get this or that happening...

If you (or anyone else) can come up with a workaround, it might encourage people to continue testing this beta.