Featured Post
These days, I mostly post my tech musings on Linkedin. https://www.linkedin.com/in/seanmcgrath/
Thursday, July 31, 2003
Towards Jython 2.2
Jython, lest you do not know of it, is the most compelling weapon the Java platform has for its survival into the 21st century:-) Jython 2.2 alpha release now available.
Wednesday, July 30, 2003
le voila
Python 2.3 released. List of some of the new stuff here. My favourites: csv module and zip file imports.
Monday, July 28, 2003
Service Oriented Integration
Principles of Service Oriented Integration is a white paper by myself and my colleague Fergal Murray of Propylon. [5 page PDF]. Think you know what BANDSTAND means? Think again :-)
Sunday, July 27, 2003
Name based polymorphism in programming languages
Reading the (very useful) article: Typing: Strong vs. Weak, Static vs. Dynamic has lead me to re-formulate my understanding of name-based polymorphism.
Simply put, Python has what is sometimes referred to as 'duck typing'. That is, if an object has the attributes one associated with a duck, then its a duck, *regardless* of its inheritance hierarchy.
In Aspect Oriented Programming-speak, it is as if Python "interfaces" (Java speak) are implemented as cross-cutting concerns rather than tucked into the syntax of the language.
In philosophy-of-language-speak, it is as if Python does not require a rigid designator for a collection of object attributes.
i.e. You don't have to name the interface. Any collection of attributes is readily usable as an interface - no formal naming mechanism required
Descriptivists probably approve:-)
Simply put, Python has what is sometimes referred to as 'duck typing'. That is, if an object has the attributes one associated with a duck, then its a duck, *regardless* of its inheritance hierarchy.
In Aspect Oriented Programming-speak, it is as if Python "interfaces" (Java speak) are implemented as cross-cutting concerns rather than tucked into the syntax of the language.
In philosophy-of-language-speak, it is as if Python does not require a rigid designator for a collection of object attributes.
i.e. You don't have to name the interface. Any collection of attributes is readily usable as an interface - no formal naming mechanism required
Descriptivists probably approve:-)
Subscribe to:
Posts (Atom)