Hi,

i think I found a nasty bug in the SMath desktop (windows) v0.85.3578.

let us consider an expression:
arccos(-0,3536)=
we get a correct result: 1,9322

however, if we do the same, but with an intermediate assignment to a variable:
a:=arccos(-0,3536)
a=
we get a result: 1,2094

after performing the reversed calculations we get:
cos(1,9322)=-0,3536 - correct
cos(1,2094)=0,3536 - invalid sign


Witold Andrzejewski

From SMath Studio Live
[LIVE width=287 height=292]http://smath.info/live/?file=767[/LIVE]

On the other hand, in 0.85_3578 there is no arccos but acos function
[MATH]acos(-0,3536)=1,9322[/MATH]
[MATH]a←acos(-0,3536)[/MATH]
[MATH]a=1,2094[/MATH]

Regards,
Radovan

Ok, so the bug is confirmed. I'm glad it isn't some configuration specific glitch

I've reproduced it too (and I'm not glad about that ). Thank you! Will check and fix.

P.S.: omorr, sorry I'm not answered o you in PM, but I've got a bug with undo/redo as well, so, many thanks to you - will fix it.

Updated: about acos vs arccos. It is an option of the SMath Studio. You can set functions name type via Options > Interface > Functions style (...and program knows about right value of this option for 22 countries).

Wrote

P.S.: omorr, sorry I'm not answered o you in PM, but I've got a bug with undo/redo as well, so, many thanks to you - will fix it.

Not at all

Wrote

Updated: about acos vs arccos. It is an option of the SMath Studio. You can set functions name type via Options > Interface > Functions style (...and program knows about right value of this option for 22 countries).

Oh yes, you are right. I forgot about it, sorry. I could (or someone else) extend the wiki page concerning multilanguage support.

Regards,
Radovan

Wrote

a:=arccos(-0,3536)
a=
we get a result: 1,2094

Error fixed:


Fix will be available into the next SMath Studio version. Thank you for the report.