[Axiom-developer] bootstrap Meta/Boot

[daly]

Mike was certainly very helpful in the early effort to get things
running and I intend no criticism of either his efforts or NAG.

What you said was:
> Other emails on this list from Mike Dewar of NAG have made it clear
> that NAG would have had no objections to releasing an open source
> version of Axiom that required a running Axiom to compile.

Well, clearly NAG would have no opinion about the structure of
the open source building process. Why could they possibly care?
And why is this relevant?




But it is certainly true that there were "restrictive licensing
conditions which apparently prevented any of the previously
COMMERCIAL binary version of Axiom to be distributed along with
the Axiom source code". Unfortunately that conversation is in
private emails.

As for "I provided copies of the Axiom product to several people on
the list so you would have no problem bootstrapping the first
open-source versions from the NAG code"... I did receive a NAG version
of Axiom. I don't remember anyone else who got copies directly from NAG
so I'm unable to comment.  I'm unaware of anyone else working on
open-sourcing Axiom. But Mike and NAG were very helpful here and I 
again publicly thank them for their support.



The bootstrap-from-Axiom issue is not related to the CCL change.
Building Axiom from Axiom is ok if you own the whole project.  But
requiring a running Axiom to build Axiom was not a viable open source
strategy. I didn't see a way to distribute sources that included
pre-compiled binaries.

As for the change from CCL to GCL I've already explained that I was
unable to build a CCL version of Axiom. I spent several months (Sept
to Jan) on the problem.  




In any case I've distributed the original sources to the world.  Since
you raised the issue it appears that you would rather have a different
building strategy than we currently use.  You have the original
sources so you're welcome to create your own branch and try.

Tim