Re: [O] Table rows and ranges as LHS of formulas

[Carsten Dominik]

On 2.3.2011, at 19:54, Nick Dokos wrote:

> Carsten Dominik <address@hidden> wrote:
> 
>> 
>> On 2.3.2011, at 18:21, Bastien wrote:
>> 
>>> Hi Carsten,
>>> 
>>> Carsten Dominik <address@hidden> writes:
>>> 
>>>> Bastien, please let me know if you want to have this integrated
>>>> before the release, then I will do so.
>>> 
>>> Please go ahead!  This looks like a nice addition, thanks for that.
>> 
>> I just pushed it.  The patch I pushed contains another small thing:
>> you can use @L to mean the last row.  So
>> 
>>    @L=vsum(@I..II)
>> 
>> is now a great way to sum columns in the last row. 
>> To skip the First column (in case it contains unsummable
>> labels or so, use
>> 
>>   @address@hidden@8=vsum(@I..II)
>> 
> 
> Both of these and Christian's multiplication table are great
> examples. Thanks!
> 
> One nit: symmetry dictates that $L should be the last column, but it
> isn't.  Trying @address@hidden in Christian's example works fine, but
> @address@hidden changes the first column to all zeroes (not sure why).

Yes, of cause, this symmetry would be nice, I have thought about it.
However, $L is ambiguous syntax, there can be a name that is $L, so
this might even break existing tables.  I could do is so that if
there is no name $L defined, then use it.  But this is also
confusing, and of cause not stable if someone does define new names.

Maybe we should use something else to achieve symmetry, like @>  and $>.

Comments?

- Carsten