Re: galculator should be included in Gnome

[Andrew Sobala <aes gnome org>]

On Sat, 2003-04-26 at 08:39, Anand Kumria wrote:
> On Sat, Apr 26, 2003 at 02:26:29PM +1000, Jeff Waugh wrote:
> > Hey,
> > 
> > The release team received this suggestion - seems there are quite a few
> > people agreeing that galculator "feels" a lot nicer than gcalctool, though
> > there have also been suggestions that the code is less than desireable.
> 
> Kcalc, gcalctool and galculator all abstract away the display from the
> numerical calculations and the code is all pretty similiar. I've got a
> version of kcalc with a GTK front-end on it for instance, not much work
> involved to change the front end. Both gcalctool and galculator use
> glade for display creation.
> 

gcalctool doesn't.

> Feature wise:
> 
> 	galculator
> 		- uses libm; which will do for 90% of uses but doesn't
> 		  do arbitary precision.
> 		- provides better visual feedback on what it is doing
> 	
> 	gcalctool
> 		- uses MP (Fortran arbitary precision package), which
> 		  will do for about 95% of people
> 		<URL: http://web.comlab.ox.ac.uk/oucl/work/richard.brent/pub/pub043.html>

IIRC it's a port of MP to C, rather than MP itself.

> 		- has financial mode (but it doesn't compare to a HP 14B)
> 
> Both should probably use MPFR instead to do arbitary precision.
> <URL: http://www.loria.fr/projets/mpfr/>
> 
> gcalctool has odd key bindings (for example), hexadecimal can only be
> entered in lowercase on the keyboard. The shifted variants of A, B, C,
> D, E and F do things like turn off hexdecimal and change signs, etc.
> Very unintuitative.
> 

It's very hard to do anything apart from enter numbers via the keyboard
because it's all unintuitive. Still, this is really an easy-fix.

> galculator doesn't do scientific / fixed precision mode (because it uses
> libm), nor does it provide 'canned' configurations like gcalctool's
> Basic, Scientific or Financial modes.
> 

That is fairly irritating. I use scientific mode all the time.

I think we've been using "crack feature" to mean "something only wanted
by computer geeks" [geek is not an insult disclaimer here]. Scientific
mode isn't even a crack feature in this sense, since any
scientific/engineering user is going to need it.

> Neither have a statistical mode like Kcalc.
> 
> Both of them are strictly "operations sequence mode" rather than
> being able to do "textbook mode" / "formula entry". For example:
> 
> 	1 + log 10 = 11 on all calculators (kcalc, gcalctool and
> 	galculator).
> 

Could someone satisfy my curiousity and tell me why? Is it totally
ignoring the log? What are you expected to type?

> 	On a Casio FX-100V (1992 vintage!) it knows enough to output
> 	correct answer as 2. There doesn't appear to be a image of a
> 	100V, but the closest model is probably a fx4000p.
> 
> 	<URL: http://www.voidware.com/calcs/casio.htm>
> 	<URL: http://www.rskey.org/fx4000p.htm>
> 
> Casio FX's are probably good candidate calculators to test against since
> they are 'approved' calculator in both Australia and UK secondary schools 
> and universities.
> 

Oh, we're being religious. Sharps are better :-), and also approved for
just about anything in the UK.

I think the concept of "Typing something in as it's written on paper
should give the correct answer straight off" is probably common to all
modern calculators now. Or by "testing against" a calculator do you mean
something else?

-- 
Andrew Sobala <aes gnome org>

"A freudian slip is when you say one thing but you mean your mother." -- unknown