Regression with calc-eval

Regression with calc-eval

[Carsten Dominik]

Hi,

I just noticed a regression with calc-eval.

Calc-eval can no longer be called out of the blue.  Calc-eval calls calc-do-calc-eval which calls calc-check-defines.  calc-check-defines does a set-buffer to "*Calculator*", so it needs this buffer present and initialized, as if M-x calc had already been called interactively.  This did not be the case earlier, and I have not been able to see when this change occurred.

This should be fixed.  I would like to know how to work around it.  Org-mode calls calc-eval to do its table calculations.

Any suggestions on how to go about fixing this issue?

Thanks!

- Carsten

Re: Regression with calc-eval

[Jay Belanger]

> I just noticed a regression with calc-eval.
>
> Calc-eval can no longer be called out of the blue.  Calc-eval calls
> calc-do-calc-eval which calls calc-check-defines.  calc-check-defines
> does a set-buffer to "*Calculator*", so it needs this buffer present
> and initialized, as if M-x calc had already been called interactively.

calc-check-defines only does anything when calc-define has a property
list, which isn't always the case.  But calc-do-calc-eval also calls
calc-create-buffer, and always has.

But the *Calculator* buffer is not shown by default, so what problems
does this cause?

> This did not be the case earlier,

Are you sure?
When I get the chance I'll try it out on older versions of emacs, but do
you have a specific version where the *Calculator* buffer wasn't
created?

> This should be fixed.  I would like to know how to work around it.

I'll look at it when I get the chance.

Re: Regression with calc-eval

[Carsten Dominik]

On 25 mrt. 2013, at 15:33, Jay Belanger <jay.p.belanger@gmail.com> wrote:

> 
>> I just noticed a regression with calc-eval.
>> 
>> Calc-eval can no longer be called out of the blue.  Calc-eval calls
>> calc-do-calc-eval which calls calc-check-defines.  calc-check-defines
>> does a set-buffer to "*Calculator*", so it needs this buffer present
>> and initialized, as if M-x calc had already been called interactively.
> 
> calc-check-defines only does anything when calc-define has a property
> list, which isn't always the case.  But calc-do-calc-eval also calls
> calc-create-buffer, and always has.

Ahh, you brought me on the right track, thank you.  I caused
the problem myself by restructuring my init file and by
loading the .calc file explicitly, by mistake.  This
adds properties to calc-define and causes the problems in the wake.

No loading .calc explicitly fixes the issue - sorry for the noise.

- Carsten