Re: bit-count broken

["Dale P. Smith" <dpsm at en dot com>]

Dirk Herrmann wrote:
> 
> On Fri, 19 May 2000, Dale P. Smith wrote:
> 
> > I also got very disgusted with the differences in the documentation for
> > procedures and the arguments.  A (help bit-set*!) is very confusing!
> 
> So, what's _your_ suggestion for an improved description?  Since you seem
> to be using it, you could probably tell best what it does.  I am willing
> to put your documentation into guile.
> 
> > Better names should be given to both.
> 
> Again:  What's _your_ suggestion?  Maybe you can even provide patches?
> This should be easy to do.  BTW:  Does the FSF have copyright
> assignment papers from you?

Here is what I mean:

guile> (help bit-set*!)
(bit-set*! v kv obj)
If uve is a bit-vector @var{bv} and uve must be of the same length.  If
@var{bool} is @code{#t}, uve is OR'ed into @var{bv}; If @var{bool} is @code{#f}, the
inversion of uve is AND'ed into @var{bv}.

If uve is a unsigned integer vector all the elements of uve must be
between 0 and the @code{LENGTH} of @var{bv}.  The bits of @var{bv}
corresponding to the indexes in uve are set to @var{bool}.

The return value is unspecified.
[unif.c:1836]
guile> 

The arguments to bit-set*! are shown as v, kv and obj, generated from the actual C arguments.  The documentation references bool, uve and bv.  What goes with what?  The patch I sent in changes the
docs to match the arguments.  I didn't want
to change all the variable names.  But is should probably be done.  With names that
mean something.  It happens that v==bv kv=uve and obj==bool.

I don't have papers in yet, sorry.  It's hard getting me and my boss together lately.  It's not a problem, just a matter of getting together and signing the thing.

> Never get disgusted - provide patches and enjoy seeing how they get
> applied :-)

Well, I *did* send in a patch. ;-)

-Dale

-- 
Dale P. Smith
Altus Technologies Corp.
dsmith@altustech.com
400-746-9000 x309