#52 [Bug] Failed to read real value in short form (0.5 -> .5)

[GHPS]

Using MathMod 3.1 with the included scripts I get the error that real values can't be read in the abbreviated form, e.g. 0.5 -> .5. This makes using the examples almost impossible. See enclosed picture.

Platform

• Kubuntu 14.04 LTS
• MathMod

[Abderrahman Taha]

Humm...I can't reproduce and I have Ubuntu 14.04 (64Bit)
Please add mode informations about your settings, ie: Qt version,64/32bit...
Also, can you try the provided MathMod-3.1(Linux64) build, i'll be curious to know if it's working.

[GHPS]

Also, can you try the provided MathMod-3.1(Linux64) build, i'll
be curious to know if it's working.

I tried that before compiling the source - since this is a 32 bit machine it expectedly failed.

This machine has Qt5 installed: qmake -version

QMake version 3.0
Using Qt version 5.2.1 in /usr/lib/i386-linux-gnu

The script engine has a problem with any kind of floating value. Take for example the Schwartz example: Out of the box it runs fine since the model doesn't contain any floats. If I try to alter the formular in the "Script Edit" tab and add a float (1.5, .5, 1.) to it (say "cos(y*2.5)"), running the script throws the error shown above. An integer works fine.

The problem is therefore more serious than initially described.

Perhaps it's my fault but I can't see a clue how to fix this...

[GHPS]

To be honest I wasn't convinced that Qt was to blame - but I was wrong: I just booted into Kubuntu 15.04 Beta-2 Live and here MathMod works a treat.

[Abderrahman Taha]

Cool! I was thinking about a bug in the Clib since the error message comes from the math parser code, which is writing in pure C/C++ langage. I'll consider this bug as resolved.
Greeting

[GHPS]

Sorry that I have to come back to this issue: After upgrading to Kubuntu 16.04 it has become apparent that the initial problem was not caused by Qt4 vs. Qt5.

It's caused by the language setting of the respective system. On a system with different number formatting (e.g. in Germany the decimal separator is a comma, not a dot) MathMod tries react appropriately - which fails.

Therefore floating point numbers can't be used.

Is there a way to set the number format in Qt5 to "English"?

[GHPS]

Because nothing has changed since initial report and because the problem is rather severe (rendering MathMod unusable on non-English systems) I searched around for a way to patch this bug.

The hotfix is quite simple - starting MathMod on the command line with "LC_ALL=C ./mathmod" forces the number format to English.

[Abderrahman Taha]

Hi,
Please upgrade to MathMod-6.1 beause this version had already fixed this issue : https://sourceforge.net/projects/mathmod/files/MathMod-6.1/
Also, you're missing a LOT of new features and new scripts not included in MathMod-3.1