Re: very small packages - merge into general/miscelleneous or move into core

[c.]

On 22 Jan 2014, at 13:55, Thomas Weber <address@hidden> wrote:

> One was Avogadro,

I looked up the Avogadro number on the database where the physical constants 
come from:
http://physics.nist.gov/cgi-bin/cuu/Value?na|search_for=avogadro

The value listed there is 6.022 141 29 x 10^23 
with relative uncertainty  4.4 x 10^-8

so any relative change less than 10^-8 would be insignifificant
as any existing code should not have been relying on that level 
of accuracy in the first place.


Anyway f I compare to physicalconstants, I get
(physical_constant ("AVOGADRO CONSTANT") - 6.02214129e23) / 6.02214129e23
ans =    0.00000000000000e+00

so really I don't see any change at all here ...


> the other I don't remember. The change in value I do
> not remember - it doesn't matter however. I am not an expert in whatever
> domain the constants are used, so I assume that every change might be
> important to someone working in a specific domain. 

That's why I whish Debian packagers would let pkg.m work 
so Octave users could just decide for themselves.

> My thinking was more like: does it make a difference if we continue with
> the old package - and yes, it did. So out it went. 

Let's pretend the Avogadro constant had indeed been changed.

Now the same function wth the same error is in the package "general".
Are you packaging "general" in Debian?
Do you think it should be kicked out because physical_constant is buggy?

>       Thomas

c.