I scrapped almost everything Ora sent you in favor of writing my own
high level overview. Ora skipped four of the six sessions and was way
too detailed in the 2 sessions he did cover. Feel free to edit this as
needed.
You know, it occurs to me as I write this that it doesn't sound nearly
so convoluted from an overview perspective as it actually was from a
participatory perspective. Is that always the case?
:) Adam
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
Internet technologies (especially the World Wide Web) and object technologies (especially distributed objects) comprise the core of the next generation of applications planned by the software industry. There is a great deal of common ground between these two areas. DOMC 1996 provided a forum for the "cross-fertilization" of ideas between people building web services and people using CORBA. Authors of the best of 64 position papers participated in sessions, in an effort to capture "what can Web people learn from ORB technologies." Please see the DOMC 1996 Accepted Position Paper Page for a list of participants.
The workshop was organized by the World Wide Web Consortium (W3C) and the Object Management Group (OMG). The workshop's format consisted of six relevant topic presentations (with each respective speaker introducing key questions in the topic), and break-out group meetings focusing on some aspect of the presented topic and its relevant questions.
The six topics were:
The World Wide Web represents a global, small standard for networked information retrieval. The goal of DOMC 1996 was to take steps toward developing a global, small standard for networked distributed objects and mobile code.
Connolly based his presentation on slides by Philippe Merle on integrating CORBA objects into the Web. Key questions in this area include:
Three break-out sessions focused on the following issues:
Rossum first talked about agent implementation languages (NB: Rossum is the principal architect of the Python language), including Python, CAML, Java and Tcl. He stressed the importance of defining the word "agent", and suggested the word "mobile agent" for this context.
Three break-out sessions focused on the following issues:
Paepcke speculated as to whether databases, object file systems, and CORBA could be used to implement the retrievability and access delivery necessary for digital libraries. There are fundamental tradeoffs between object persistence (rich structure, concurrent updates, simple querying, high speed) and document stores (fast retrieval, frequent reads, elaborate queries, high volume), but both want to provide distributed access to multiple client platforms using increasingly rich data types.
Three break-out sessions focused on the following issues:
Janssen talked about the components of a wire format (specifying messages sent on a channel) and a binding mechanism (an object description that allows a client to connect).
Three break-out sessions focused on the following issues:
Fulton proposed that CGI is not object-oriented (since there are problems with interfaces), that the CGI computing model is too low level, and that CGI performance is poor. Solutions to these problems range from server APIs that allow custom objects to be embedded directly into an information server, fast CGI that converts CGI scripts into long running processes with state, server-side CGI agents that can be downloaded dynamically, and distributed object gateways that use CORBA or ILU to publish Web resources by translating HTTP requests to method invocations on distributed objects. The future CGI needs to preserve the current CGI semantics and API, while adding an object-oriented programming model for distributed computing.
Three break-out sessions focused on the following issues:
Herbert overviewed the concepts of trading (making explicit contracts between importers and exporters) and naming (to bind to a service), and observed that URLs might not be sufficient for some uses of trading, binding, rebinding, and relocation. Key questions in this area include:
By Adam Rifkin, September 11, 1996.