On Thursday, October 17, 2002, at 11:12 AM, goran.hultgren@bluefish.se wrote:
And almost all of the above has nothing to do with classic CM problems which more or less focus on "how do we build a BIG system using 20 developers". That is where Envy comes in. And Team/V - just listen to the name of the product.
For me, Squeak IS the BIG system where much more than 20 developers cooperate... For example to implement something like Modules...
Disclaimer: this does not mean that I am especially in favor for Ginsu, or for other things. But this statement is for me the exact motiviation for having modules in Squeak. My problem with the current module system was that I never was able to understand it, even not when looking at the code and the tests or spend a lot of time discussing with Henrik. This is not necessarily to say that the code is bad or anything, just that I am not able to get it. I am pretty sure that this is the reason that nobody really was able to take control of the code and move it forward. I agree with Daniel's mail in that respect where he challenges people to make this change to the modules.
So, to be constructive, what do I think that there is to do?
This is what I think: 1- make a paper/webpage/Swiki page explaining the model and the reasons for the decisions. I really like Goran for all the effort he is putting into this: it is needed to have someone to do this. 2- let people comment on the proposal. 3- Reimplement the model. Having looked at the code myself a couple of times, I think the current implementation is not the simplest that can be. It was a prototype and served very well as that. But a prototype is not meant to go in production/extensively used by people. So throw it away and built a new one taking into account everything that was learned building the first one. I cannot stress this enough: I like what Henrik did (since he built something that served as thinking and experimentation food), but it is now time to move this forward.
PS: Yes, I know, it is easy for me to say this because I am not going to do this. There is only so much hours one can spend working, and we spend a lot of time trying to come up with a module system for Squeak but we are not sure about the end result. We'll continue working on that in some form or another. But by having a paper we were able to talk to a lot of people about it at Camp Smalltalk in Douai, and get valuable feedback. It would be great if we could do the same at OOPSLA.
PS: I am going to OOPSLA and I am willing to talk about this any time. I'll probably attend the Squeak BOF, where I guess this will be a discussion topic. That is, if emotions can be kept down and Henrich still wants to talk to me :-) I thought he knew Stephane's crude approach of expressing things better. But let's not dwell on that, I guess.
-- Roel Wuyts Software Composition Group roel.wuyts@iam.unibe.ch University of Bern, Switzerland http://www.iam.unibe.ch/~wuyts/ Board Member of the European Smalltalk User Group: www.esug.org