BTW, for us this work started out mining yet another of those many gems 
in the Smalltalk way of doing things - the way the model for Smalltalk 
code in the development environment is connected to the runtime 
representation (so natural to everyone here, its hardly worth 
mentioning, but rather rare in other worlds).


One thing that happened as Andrew and I were trying to add some smarts 
to the tools, is we began to ask ourselves how the code analysis 
algorithms should be split between the tools and the code model. On the 
one hand, we can't put everything in the code model. On the other, 
placing algorithms inside tools makes it harder to reuse them (ask 
anyone working on alternate browsers). Turns out this gets more 
complicated when the algorithms are pretty demanding in terms of run 
time, and we think these are problems that we share with other 
communities, and that probably none of us have a single complete answer 
to, so we might learn interesting ideas from one another. So this 
workshop might be interesting for anyone interested in code analysis, 
development tools, and what's between them.


Daniel Vainsencher


Andrew P. Black wrote:

>
> Those involved in building tools and program analyses may be 
> interested in the following workshop.  It evolved out of some work 
> that Daniel and I started on figuring out how to make the "required 
> methods" analysis modular in Squeak.  We are hoping to involve the 
> tool-building communities from a number of languages, but a couple 
> more Squeakers would always be welcome.
>
> If you have connections with other tool-building or program analysis 
> communities, please pass along this invitation to participate --- or 
> give me the contact information.
> * *
> *Call for Participation: Workshop on Extensible Program 
> Representations for Interactive Development Environments (EPRIDE)*
> *
> *
> The EPRIDE 2007 Workshop will be held on Monday July 30th, immediately 
> preceding ECOOP (European conference on Object-Oriented Programming) 
> in Berlin.
>
> The thesis of the workshop is that no single, pre-defined 
> representation of programs in an IDE can possibly be adequate for all 
> of the uses that may emerge; instead, we must focus on extensible 
> representations. How can we engineer representations that are 
> extensible and efficient, and yet provide clients with a simple way of 
> getting the information that they need?
>
> The workshop aims to bring together those involved in building plugins 
> and restructuring tools for IDEs — the consumers of information about 
> the program — and those involved in developing parsers and program 
> analyses that /create/ the information needed for the tools — the 
> providers of information.
>
> For more information, and to find out how to participate, please go to 
> http://c2.com/w4/epride . The deadline for applications is 13th May.
> ------------------------------------------------------------------------
>
>
>