This is the mail archive of the guile@sourceware.cygnus.com mailing list for the Guile project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: Guile, CORBA and GNOME


Keith Wright writes:
 > But now I'm confused.  The man who wrote it says it is
 > a binding to standard Scheme, while Jim says it goes through
 > YASOS.  (I did see that word in the docs too.)
 > What is YASOS?  Guess: Yet Another S___ Object System.
 > But who wrote it,

Ken Dickey

 > who uses it,

those who want to explain the connection between objects and closures,
as well as of the role of syntax transforms here in.

 > why?
In order to show the connection between closures and objects, and to
teach usage of syntax macros.

 > Is it an OO library written in Standard Scheme?

core R4RS + syntax extensions 

 > (That might reconcile what Bill and Jim are saying.)
 > 
 > What is the advantage of going through GOOPS instead of
 > straight to Scheme?
 > 
one should make benchmark test among others. Yasos loads fairly slowly,
and it can't be hobbitted because hobbit can't munch syntax transformers
(at least not yet , Bernard might know better where are the problems).
Goops is loads fairly fast because it is largely written in the guile
core C API. So though it is larger and more flexible due to the MOP,
it might be faster than yasos for many purposes.
The new Slib switched to Wade Humeniuk's mos (in object.scm) and built
a yasos compatibility syntax layer upon it (yasyn.scm) for backward
compatibility reasons. mos is deprived from syntax transforms and loads
pretty fast , while retaining the portability , even increasing it due
to dropping need of syntax transformers. Alas it lacks some syntactic
sugar thus making maintenance of scheme scripts using it a bit harder.

Klaus Schilling

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]