Re: How to reduce binding load time

[Andreas Rottmann]

Marius Vollmer <address@hidden> writes:

> Andreas Rottmann <address@hidden> writes:
>
>> Hi!
>>
>> Andy, i suggest you try to use 'scm_call_1 (scm_sym_make,
>> scm_generic)' instead of using 'scm_make(scm_generic)'; this should
>> speed up generic function creation a fair bit (and this is an
>> understatement). I did so with some sucess in my g-wrap code with
>> great success (reducing binding load time by 1/3 -- 1/2).
>
> Why is that?  Could we speed up scm_make to give the same kind of
> performance as scm_call_1 (scm_sym_make, ...)?
>
> What is scm_sym_make, incidentally?  You can't call symbols...
>
scm_sym_make is what you get from resolving "make"
(i.e. make-instance). scm_make is dogslow for some reason... This also
relates to the last message I sent to guile-user, since I thought C
scm_make(...) and Scheme (make ...) were equivalent.

Andy
-- 
Andreas Rottmann         | address@hidden      | address@hidden | address@hidden
http://www.8ung.at/rotty | GnuPG Key: http://www.8ung.at/rotty/gpg.asc
Fingerprint              | DFB4 4EB4 78A4 5EEE 6219  F228 F92F CFC5 01FD 5B62

Python is executable pseudocode, Perl is executable line-noise.