From: Sam S. <sd...@gn...> - 2002-02-18 17:51:09
|
> * In message <154...@ho...> > * On the subject of "Re: eval-when & *.lib" > * Sent on Fri, 15 Feb 2002 14:35:09 +0100 (CET) > * Honorable Bruno Haible <ha...@il...> writes: > > Sam writes: > > > I would appreciate your attention. > > Maybe this whole *.lib thing should be removed? > > With such extremist propositions you indeed get my attention. :-( Cool - unless you answer my questions about the Todd's patches (why did you have separate buffered_full_flush() buffered_half_flush()?) I will remove buffered i/o altogether :-) Also, unless you answer my question about TRACE (why TRACE1 returns code instead of evaluating it), I will remove TRACE functionality. Hey - this is the "RMS way": remove all code one does not understand (I am referring to the ange-ftp episode, when RMS removed support of platforms he did not understand from the code he included in Emacs). > > We got a bug: > > https://sourceforge.net/tracker/index.php?func=detail&aid=515604&group_id=1355&atid=101355 > > I think it is not a bug > > (eval-when (:compile-toplevel :execute) > (defvar *crap* #.(lambda (arg) (* 10 arg)))) > > because a literal function object is not an externalizable > object. Please read CLHS chapter 3, in particular 3.2.4. this variable is defined only during compilation and evaluation, not loading, so there is no reason for it to be either in *.fas nor in *.lib. > If it is a bug in clisp and not in the 2-line snippet (I'm not sure), > then the solution would be to be more careful when writing objects of > unknown type to a .lib file. I don't think this is correct. we should be careful _when_ we write to .lib, not _what_ we write there. > The "poor man's system construction" is indeed to have a makefile that > looks like this (pardon my mix of make syntax and Lisp syntax): > > .lib files are to .lisp files like what .h files are to .c/.cc files > in the C/C++ world, except that they are automatically generated. I bet _I_ am the only one using this nice feature (and I am using it since this morning - see the 'wrap.lisp' patch). Okay, I will document this better. -- Sam Steingold (http://www.podval.org/~sds) running RedHat7.2 GNU/Linux Keep Jerusalem united! <http://www.onejerusalem.org/Petition.asp> Read, think and remember! <http://www.iris.org.il> <http://www.memri.org/> Hard work has a future payoff. Laziness pays off NOW. |