Re: Autoloading of toolboxes?

[David Bateman]

Søren Hauberg wrote:
> John W. Eaton skrev:
>> I don't think it is a good idea to load all packages by default.  That
>> Octave Forge was typically packaged and distributed this way is
>> something that I've never really liked, and was the reason I wanted to
>> have a package system that would make it easy to install add-ons in
>> smaller chunks.
> I didn't like that octave-forge loaded ALL functions after an
> installation either. I do however not mind that all packages are
> loaded by default because I only install the packages I use. This
> logic does however fail in multiuser environments where global
> packages will be loaded for all users. So, perhaps we should autoload
> local packages and not global packages?
>
> Søren
>
It doesn't bother me personally as I'm aware of the issue, but if the
signal processing toolbox is installed globally and it isn't autoloaded
we'll have every matlab user transitioning to octave complaining that
octave doesn't support X, when they'd have this function is the package
was autoloaded. In terms of number of headaches for us I think
autoloading everything is better than not autoloading global packages...
However, ideally there should be some means to avoid autoloading
particularly of globally installed packages...

D.


-- 
David Bateman                                address@hidden
Motorola Labs - Paris                        +33 1 69 35 48 04 (Ph) 
Parc Les Algorithmes, Commune de St Aubin    +33 6 72 01 06 33 (Mob) 
91193 Gif-Sur-Yvette FRANCE                  +33 1 69 35 77 01 (Fax) 

The information contained in this communication has been classified as: 

[x] General Business Information 
[ ] Motorola Internal Use Only 
[ ] Motorola Confidential Proprietary