Bah, c'est quand meme rapide le python. Et puis avec un peu de profiling, on peut détecter les parties les plus lentes et les réimplémenter en C[1].
Et comme ils le disent dans l'intro/FAQ :
In its current implementation, Python's execution is not nearly as fast as most languages traditionally used in OS development. However, this is a problem of the implementation, not the language itself. The decision to use Python was actually made after considering the creation of a new language to address frustrations with existing languages. Thus, developing a faster Python implementation remains a future goal of Unununium.
Unununium is not alone in the quest for a faster Python. Future implementations of the standard interpreter are planned to include more agressive optimization throughout. Psyco is a promising program which compiles Python bytecode to native code. Pyrex compiles a subset of Python to C.
Et puis il y a plein de projets visant à améliorer les performances de ce langage (ouvert et très réactif, tout comme les dévs, d'ailleurs).
[^] # Re: Les OS objets bien sûr
Posté par Sebastien . En réponse au journal Guerre des Gourous :: Episode 3 :: le Finnois tire à vue. Évalué à 2.
Et comme ils le disent dans l'intro/FAQ :
In its current implementation, Python's execution is not nearly as fast as most languages traditionally used in OS development. However, this is a problem of the implementation, not the language itself. The decision to use Python was actually made after considering the creation of a new language to address frustrations with existing languages. Thus, developing a faster Python implementation remains a future goal of Unununium.
Unununium is not alone in the quest for a faster Python. Future implementations of the standard interpreter are planned to include more agressive optimization throughout. Psyco is a promising program which compiles Python bytecode to native code. Pyrex compiles a subset of Python to C.
Et puis il y a plein de projets visant à améliorer les performances de ce langage (ouvert et très réactif, tout comme les dévs, d'ailleurs).
[1] http://www.wag.caltech.edu/home/rpm/python_course/Lecture_5.htm(...)