[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Axiom-developer] RE: algebra Makefiles with explicit dependencies,

From: root
Subject: Re: [Axiom-developer] RE: algebra Makefiles with explicit dependencies, bootstrap, fixed-points etc.
Date: Fri, 7 Jan 2005 23:54:44 -0500

The lisp code in the BOOTSTRAP files (for the most part)
has been pretty-printed. Since the NRLIB .lsp files are
not normally read by humans they are not pretty-printed.

In some cases the BOOTSTRAP code has been hand-formatted.
The intention was that it would correctly tex and fit on
the page. 

Both of these reasons will keep you from diffing the file.
You'd need to use a lisp routine to (prettyprint (read them
before diffing.

I did receive your email that contained the differences
but since all of my email comes into an emacs RMAIL buffer
it generally looks like:


that is, base64. I have to decode it by hand to see what it is and I failed
to do that.

re: propagation

When you write out a domain in Axiom it contains a large vector which
describes all of the domain. Some of those vector elements are functions
and some are variables in the domain.

When you use a function from domain A in domain B the domain vector for
B will contain an index into the domain vector for A to get at the 
function. So if you inspect the domain vector for B you'll see hard-coded
integers which are indexes into the domain vector for A.

If A changes then the domain vector will be reshaped and the slot index
for functions will get renumbered. If B depends on A then it will get
recompiled and have a new slot number. 

So changes to A don't propagate quite the way I think you mean.
What you will find is that a bunch of magic integers have changed.


reply via email to

[Prev in Thread] Current Thread [Next in Thread]