> Um, distribution under the GPL has to include offers of source.
>
> In fact the terms of the GPL are all designed to promote a very
> specific philosophy that is counter to traditional commercial
> practices!
True, but it hasn't always happened.
> >If perl is to be called free software, there can be no limitation on
> >redistribution of compiled binaries. This incorrigible business practice
> >has
> >become an epitome of how some open source licenses do not work.
>
> If Perl is to meet either the Debian or OSI definitions of
> free or open source (respectively) software it cannot restrict
> the sale of compiled binaries.
I'm not talking about the sale of binaries made with Perl, but the sale of Perl
itself. No, not even the sale of it, but the redistributability of it. To my
understanding, that IS the GNU license. However, you make one tiny change and
it becomes something totally new and the source can be privatized under the AL
to the point where you can require that it not be redistributed under your
license.
Example, company X takes perl, adds a nominal gizmo to it, and distributes the
binaries along with an installer and a helpfile index. Little more. This is
marketed as Perl, but the user is forbidden to redistribute it. Actually this
isn't an example and it isn't theoretical. It's true and it breaks the
principles of free software. Under the AL, they appear to have the right to do
this, whereas under the GPL, I don't believe they could. My legal
interpretations of the licenses may be off, but the actions are true, and need
to be prohibited.
Basically, "if you add a gizmo and want to limit it, it isn't perl. If you want
to call it perl, then it's redistributable and you need to provide source".
However, I've seen another message last night that said this issue is being
approached from another angle also, and since I started in here late, I'll read
back a little.
> Please read the draft that I put out of an AL. I suspect
> that it does indeed provide the restrictions you are looking
> for. In fact I think it is the only idea under discussion
> which could be palatable to Perl developers that comes close
> to doing so.
Yes, I agree that I have to catch up a bit. I've simply voiced an ongoing
concern of mine. We have an opportunity to correct a long-standing error that
has allowed bad things to happen, and I spoke out of turn.