Re: [SIP-devel] colormaps

[Ricardo F. <rf...@if...>]

On Thu, 19 Feb 2004, DRUEL Jocelyn wrote:

> Hi,
>
> >
> > Bruno Pincon's colormap stuff seem to be really neat.
> >
> > My only restrictions to their inclusion into SIP is that the final
> > functions be as similar to M4t1Ab as possible, in case there exists an
> > equivalent in that software.
>
> Aren't you afraid that one day, M4t1Ab says SIP plagiarize their software ?

Not at all ;)
Simply because it doesn't plagiarize. It is just my opinion that SIP should
try to keep some compatibility with other well known software. (read
bellow)

> I remember a very nasty story about a software called Krayon in Suse that
> caused Suse to be prevented from selling its boxes in Germany:
>
>  http://news.com.com/2100-1001-803724.html?legacy=cnet

Interesting.

>
> Personaly, I think giving same names and syntaxes as M4t1Ab will cause
> problems (although it's easier for users to migrate) if SIP becomes more
> popular. This year, I had a sale representative of Mathworks on the phone for
> my school. He was not quite happy of the existence of Scilab.

That makes /me/ quite happy :)

> Common points are natural, but SIP is not a (cheap) clone of the M4t1Ab Image
> Processing toolbox. I consider it as a complementary tool, or as another tool
> which allows image manipulation.
>

You are right, SIP is really not a mtlb image toolbox clone. I'm just
suggesting that, whenever possible, we keep compatibility. Here is
what I suppose is reasonable about this, please tell me if you agree:

Whenever we invent an interface to a function that is clearly better
than the interface of the corresp. function in mtlb, I think we should use
our new interface and forget about the matlab interface.

However, I think we should always improve, never do inferior function
interfaces. If you have a different function interface that is clearly
worst than the corresp. in mtlb, we should use something at least similar
to the better one.

On the other hand, if you have a function interface that is better
than mtlb, then, sure, lets use it, since what we want here is not a
clone, but a good software package above all.  If a function interface
is different from mtlb, and we are not sure if it is better or not,
then we should discuss it or merely use the one that users are more
familiar with.

I'm not doing anything wrong, but I confess I'm taking a slight risk
here. You always risk yourself whenever you get in the way of big
companies, even if you are right under law.

I am pretty sure that most of the interface issues cannot be
copyrighted nor be intelectual property in any ways. Except for some
countries, of course, but I'm not afraid of that. I'm willing to to
the best for an image processing software that we're crafting, and this
does not means cloning some other tool that, we all know, is not
perfect.

cheers,
Ricardo.

--
  Ricardo Fabbri, Cybernetic Vision Research Group, USP, Brazil.
  WEB: cyvision.if.sc.usp.br/~rfabbri   ICQ: 208974212