Re: Extending Gnumeric Documentation

[Donald Permezel <djper1 student monash edu>]

> I would like to implement all analysis tools with the `formula' version.
> (The "value" version is then a simple consequence and doesn't require
> any new code.) So if you would like to modify any of the existing tools
> in that way it would be greatly appreciated.

> Similarly we would love any additional tools.

> I realize that some of the tools we should add as well as some of the
> existing tools may require some new functions.

Great!  Once I've had a chance to look over the 'formula' versions from 1.3.x, I'll see if I can convert any 
more and notify you on the IRC channel or e-mail.



> The only note would be that for pure `value' versions the data
> collection functions in analysis-tools.c are discouraged. The similar
> implementations in collect.c should be used. (This is one of the places
> where we currently duplicate code in gnumeric and this duplication
> should hopefully be removed.)

I'll change the documentation to detail 'formula' implementation again when I've read over 1.3.x, and also 
modify it to note the use of collect.c functions when dealing only with 'values'.

Presently as the 'value' type tools in my 1.2.x Gnumeric use the analysis-tools.c functions, I don't yet 
understand how the collect.c functions should be used instead, but over the next few days I will read over 
all the functions, experiment, and then update the document accordingly, aswell as make the correction with 
respect to the Glade objects.


Many thanks again for your help,
Donald Permezel