Inicio Nosotros Búsquedas
Buscar en nuestra Base de Datos:     
Autor: =Henning, Michi
2 registros cumplieron la condición especificada en la base de información BIBCYT. ()
Registro 1 de 2, Base de información BIBCYT
Publicación seriada
Referencias AnalíticasReferencias Analíticas
Autor: Henning, Michi michi@dstc.edu.au
Oprima aquí para enviar un correo electrónico a esta dirección
Título: API Design Matters
Páginas/Colación: p. 46
Fecha: Mayo
Communications of the ACM Vol. 52, no.5 May 2009
Información de existenciaInformación de existencia

Resumen
There seems to be something elusive about application programming interface (API ) design that, despite years of progress, programmers have yet to master. For every way to design an API correctly, there are usually dozens of ways to design it incorrectly. Even minor and quite innocent design flaws have a tendency to get magnified out of all proportion because APIs are provided once, but are called many times. Poor APIs are difficult to program with and often require additional code to be written. Moreover, poor design can lead to inherently inefficient code by forcing unnecessary data copies. There are a few guidelines to use when designing an API. Here are a few of the author's favorite things to think about when creating an API: 1. An API must provide sufficient functionality for the caller to achieve its task. 2. An API should be minimal, without imposing undue inconvenience on the caller. 3. APIs cannot be designed without an understanding of their context.

Registro 2 de 2, Base de información BIBCYT
Publicación seriada
Referencias AnalíticasReferencias Analíticas
Autor: Henning, Michi michi@dstc.edu.au
Oprima aquí para enviar un correo electrónico a esta dirección
Título: Binding, Migration, and Scalability in CORBA
Páginas/Colación: pp.62-71.; 28cm.; il.
Communications of the ACM Vol. 41, no. 10 October 1998
Información de existenciaInformación de existencia

Resumen
Common Object Request Broker Architecture (CORBA) object model relies to a large degree on semantics of object references. An object reference uniquely identifies a local or remote object instance, clients can only invoke an operation on an object if they hold a reference to it. Apart from the ability to denote remote objects and to convert to and from strings, CORBA references have semantics that are similar to those of C++ class pointers, references can dangle and they can be nil. Object references are opaque. Application code is not allowed to look inside a reference or to make any assumptions about its internals. The details of how an object reference identifies an object and how requests are dispatched to their correct destination are carefully hidden away in ORBS run-time support. Semantics of object references have profound implications on the design of an ORB, in particular its ability to support object migration and to scale to very large number of objects. Some mechanisms and design trade-offs involved are explained in this article, including some speculation on how CORBA should evolve to address needs of ever-larger distributed systems.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

UCLA - Biblioteca de Ciencias y Tecnologia Felix Morales Bueno

Generados por el servidor 'bibcyt.ucla.edu.ve' (216.73.216.138)
Adaptive Server Anywhere (07.00.0000)
ODBC
Sesión="" Sesión anterior=""
ejecutando Back-end Alejandría BE 7.0.7b0 ** * *
216.73.216.138 (NTM) bajo el ambiente Apache/2.2.4 (Win32) PHP/5.2.2.
usando una conexión ODBC (RowCount) al manejador de bases de datos..
Versión de la base de información BIBCYT: 7.0.0 (con listas invertidas [2.0])

Cliente: 216.73.216.138
Salida con Javascript


** Back-end Alejandría BE 7.0.7b0 *