Re: [Axiom-legal] RE: Aldor and Axiom

[Ralf Hemmecke]

On 02/14/2006 05:16 AM, C Y wrote:
> --- Bill Page <address@hidden> wrote:

> > > No, GPL won't quite work - Aldor.org requires complete rights
> > > to derivative works, IIRC.
> > No where does it say that Aldor.org reserves *exclusive* rights.

Well as I understand the license it says. If a person A distributes 
modified forms of whatever you take from aldor.org, he must follow the 
modified BSD that is stated in the first 3 items of the Aldor Public 
License 1.0.

The additional fourth item says that A has to submit the modified work 
to aldor.org and let them do with it whatever they like. They could 
release it under any license close it and/or sell it.

If A chooses to release the modifications (and the orignal source code)
again under APL1.0 (I think the fourth item carries over) in source 
form, then whatever aldor.org does with the modified work, it stays open 
source. That basically sounds like a fork or like distributing identical 
software under different licenses.

I am not really agains giving back modifications to aldor.org. The idea 
probably is that there should not be diverging versions of the Aldor 
language. Note that the aldor language could still be slightly modified 
to meet peoples needs. Stephen Watt already had some ideas when I last 
spoke to him two years ago. (Unfortunately there are no people who 
implement that.)

> Hmm.  I guess that's true, so long as Aldor.org didn't want to ever
> release a commercial binary only version using someone else's
> derivative work.  My recollection was that they do have that right
> under the Aldor license, but IANAL (as they say on slashdot).

Aldor has the concept of libraries and there is libaldor and libalgebra, 
basically which are under APL1.0. If one changes them, then the code 
currently must be given back to aldor.org. If you write another library 
which just uses those libraries, then this new library can be released 
without the need to hand it to aldor.org. So that basically says, if we 
turn libaldor and libalgebra into pamphlets (correct bugs, add a 
testsuite) that should be given back to aldor.org. If we build libaxiom 
on top of libalgebra, then aldor.org cannot claim any rights on that code.

Who interprets the Aldor Public License differently?

> > The issue might eventually be who can incorporate Aldor into
> > non-open source software.

Anyone can do this (close the source) under the APL1.0 as long as they 
give the sources to aldor.org. However, aldor.org is allowed to release 
these sources.

> I think Aldor.org has complete rights to any derivative work based off
> of aldor, not just the right to open up the code.

No. the APL does not state derivative work, only "modifications". A new 
library (libaxiom) based on either libaldor or libalgebra is not a 
modification of of them.

> > I am also in favour of pushing the availability of Aldor as open
> > source by "jumping the gun" (to use an Olympics-style analogy :)
> > and releasing Aldor source now with the Axiom distribution.

Given that Stephen Watt is in favour of opening the Aldor compiler, we 
should discuss of how the Aldor library sources could be integrated as a 
basis of Axiom. I would like that.

But even if the APL allows us redistribution, I would inform aldor.org, 
in particular Stephen Watt that we are going to do this. The Aldor and 
Axiom projects should stay on good terms with each other and eventually, 
maybe, become just one project (consisting of several subprojects).

> > But I think we have to consider this strategy carefully - 
> > particularly the licensing issues.

Leave the code from aldor.org under the APL, that doesn't hurt. But we 
modify it by putting it into pamphlet form. Oh, I'd love to see ALLPROSE 
be of some help here. :-)

> For the moment, I would say we make it easy to install against an
> existing Aldor, provide good instructions how how to get it installed,
> and leave it at that, I guess.  Anybody messing with Aldor at this
> point will have the skillset to get it working from the instructions to
> begin with.

I think that one could assume that people who want to program in Aldor 
know how to install it. But it would also be possible to take the .tgz 
files from aldor.org and install the right binaries (the Aldor compiler) 
during the installation of Axiom. The library sources would be in a 
separate aldor/src subtree of Axiom as .pamphlets. The APL would allow 
that as I see it. But it would always be friendly to tell Stephen Watt 
about what we are going to do.

> P.S.  I don't suppose we could try to decompile
> (http://boomerang.sourceforge.net/,
> http://www.backerstreet.com/rec/rec.htm) Aldor?  Does the license or
> legal situation forbid that?  It wouldn't be pretty but at least it
> would be source code.

I am against decompilation. Not that I see that APL would forbid it, but 
the aldor compiler surely has some bugs and they would remain there for 
centuries probably because the source code is then even more unreadable 
as it is now, I guess.

Ralf