Mail Archives: geda-user/2015/02/13/17:31:20
Bob Paddock wrote:
>> "C" : 11800
>> "python": 890
>> "lua" : 150
>> "guile" : 3 however, "scheme": 68
>
> The optimistic view is C is good because it has lots of books, however
> it is also the oldest of the lot so had more time to have books
> written.
The number of books does not say anything about the quality of a language.
It does say even less about the appropriateness for a specific task.
Actually, C is a pretty bad choice for scripting ;-)
But the number of books does say something about popularity. After all,
there is no economic point in publishing 500 books about a language that
is used by 1000 coders worldwide. Even more important, it says something
about relative popularity.
An open source project like geda relies on volunteers to get ahead. Few
people feel comfortable to learn a new language from scratch just to start
contributing. Choose a rarely used language and you get less contributions
than you could. IMHO, this is what we see with guile.
> It is like job ads I see "We need C developer".
In a way, yes. Currently, it would read "We need contributors willing to
contribute guile code". Turns out, the response is saddening quiet.
Actually, we also need someone who manages to fix the cross compile issue
of guile. Ever since geda requires guile >2.0 we effectively lost the
ability to produce a windows binary. Note, the road block is the cross
compile of guile, not the cross compile of geda itself.
> There is no one 'best' language. Use the most suitable for the
> problem domain, of which there are many in EDA.
IMHO, in an open source context a popular language that is OK for the
problem beats a perfect but unknown language hands down.
---<)kaimartin(>---
--
Kai-Martin Knaak tel: +49-511-762-2895
Universität Hannover, Inst. für Quantenoptik fax: +49-511-762-2211
Welfengarten 1, 30167 Hannover http://www.iqo.uni-hannover.de
GPG key: http://pgp.mit.edu:11371/pks/lookup?search=Knaak+kmk&op=get
- Raw text -