Attendees: Sue Probert, Mike Adcock, Marion Royal, Peter Yim, Lisa Seaburg, Bill Meadows, Monica Martin, Nigel Woods
This is a special meeting of the LCSC scheduled because of the amount of work ahead of us. Because it is important to capture what is happening between the members of the group, I have typed as fast as I could and captured this meeting almost word for word. If you see errors, please let me know. There are gaps in some of the discussion that I just couldn't keep up with. Happy Reading.
1. Welcome from Chair and appointment of Secretary to take minutes - Lisa volunteered.
2. Release 0p70
Tim: I would like to go through the material, clarifying and having discussion about the materials.
Sue: I think we should take the extra time needed to go through this material. It will probably make a time difference to our schedule.
Tim: Yes, I agree, this is a significant moment in our work, and we need to take the time needed to really understand where we are.
Tim: Let me run through a perspective of what I have attempted to do:
[Change from Tim] The first thing we had on our schedule was do to modeling based on our structures. The process I went through was to take the spreadsheet 0p66 Burlington, that contained "MIKE'S" comments and updates, also the edits made during the plenary. From "NORMALIZING" that material I developed the class diagram (0p70), this describes all of the objects "AND THEIR ASSOCIATIONS" that "EMERGED" from the spreadsheet we had. In the process of doing that some things disappeared and some new things appeared. I do admit I may have got some of those things wrong. This is not a definitive model, but is a good starting point. It is in need of review.
This class diagram is a key artifact, it is the framework that we develop everything else from. This is still open to improvement also.
The one thing this class diagram does not show us is the actual properties for each element.
Mike: I would like to flesh out this model using Rational Rose. We can use this tool to help us build a more complex model, it gives us more info.
Tim: I used a different tool, didn't have RR. There are pluses and minus for this, and one minus is that the model I presented confused people a lot.
Tim:The spreadsheet called "normalised components8 0p70.xls", is that class diagram in spreadsheet form. If we look at this spreadsheet, every pink line is a box on the model (or should be). The plus side of using spreadsheet is when we transfer this into a UBL spreadsheet, we have the same columns and can transcribe it over.
Mike: My concern, Tim, is when we get to the green lines in the normalized spreadsheet.
Tim: I will try to explain. If you look at Party in the class diagram. You will find that Party has lots of arrows that connect it to other boxes (all arrows point out of Party), the associations. The green rows show an association.
Mike: Everything we have done so far shows it the other way around.
Tim: Previously we were modeling hierarchy, now we are trying to model relationships.
Mike: We question the relationships and are very confused by them.
Sue: We would appreciate it if the green row would represent something we are more familiar with.
Tim: I have no trouble showing a two way relationship, but I don't think it is important. The important artifact we are discussing right now is the class diagram and the normalized components.
Mike: your class diagram and mine are exactly the same only mine had attributes.
Tim: I disagree, there are subtle differences.
Discussion on objects and changes and similarities between the two class diagrams.
Marion: I think what you have discovered a way to make the model more flexible that it was before and I am interested in understand further.
Sue: We knew the model as of 0p66 was not perfect in its relationships, in fact the first attempt in Mike drawing a class diagram. You have helped with doing this piece of work. Now we are discussing how to take this work and review it and move it forward.
Tim: Lets take those things one at a time. If everybody is more comfortable with two way relationships, then lets do that. It really just adds to the size of the model. To present the material like this was my decision, its just one way of showing it.
Sue: So how do we move forward?
Tim: Mike and Sue have started to review this model, and it is important that we get the description of the things like PartyAddress, Names, ShipmentStage, etc. all correct. This is the point that we fix all of these things. When we get to the point where we develop documents, we will be using these. So it is important that we can all agree and sign off on these at this point.
Tim: The green represents associations, it is what we use to have when we had an aggregate within an aggregate. All it means is there is something else in here. Any box that has an arrow stuck in it has a green line. Parent objects. It does not show you everything underneath that aggregate, it will be somewhere else on the spreadsheet.
Sue: Its the principle of the relationship we are concerned with.
Tim: What I would be comfortable with is an understanding from everybody on how to get the normalized model reviewed and accepted by the group. Going from this model to building documents is not a complex process.
Sue: we are encouraged by the fact that all of the structures needed are already there and to use them will not be that far off.
Tim:
First: normalized model needs to be gotten right.
Second: assembly is not as important as we had first thought, it will come naturally as the other work comes together.
Tim: We will have this discussion on the assembly process. There is a slide presentation attached to the agenda for this meeting that goes through this process.
Lisa: Isn't document assembly really part of the context methodology group. This is being looked at by this group and we need to make sure we are all together on this.
Sue: Another thing to think about is we are really collating and documenting experience gained which will go into the context work.
Tim: We are breaking through into an environment that gives us a clear path to the future.
3. Action Plan
Tim: We need to discuss where we go from here, I am encouraged by Sue's comments. Because this is a significant process that we are going through, we need to make sure everyone is comfortable with it.
Tim: QA Team, do you have a feeling for this and the artifacts and also an idea of turn around time on this?
Sue: We have work to do with Peer Review before this goes to QA. We have identified that we need to do the peer review and separate that concept from the QA review.
Tim: Okay, I suggest we allow ourselves another 7 days to do this review and then submit to the QA Team.
Tim: Before we get to the Face to Face we can hopefully get this done.
Tim: By Nov 7th, we submit to QA. Peer Review Team: everybody in group.
Marion: Lets use next Tuesdays session as a work meeting.
Tim: Mike, how hard would it be for you to capture this work in Rational Rose?
Mike: Capture what?
Tim: Objects and their properties.
Mike: I was viewing the RR as part of the review process.
Tim: I am asking you to go the other way.
Mike: My suggestion is that comments, go to one person as master of the spreadsheet, and Mike will be master of RR model. These two should be in sync with each other by the end of this review.
Tim: How would it be if I maintain the spreadsheet and you maintain the RR model.
Mike: I thought Lisa was the editor.
Tim: Lisa will maintain the spreadsheet and Mike the model.
Lisa: I will make sure the most recent documents are on the WorkingDocs webpage.
Tim: We need the discussion to happen on the email list.
Sue: We need to be clear about what we are discussing. Tim, would maintain Issues list.
Marion: We can use Netmeeting on Tuesdays to make actual changes.
Lisa: Agh!!!! But okay.
Tim: We make Netmeeting mandatory if you want to see the changes happen.
Nigel: Regrets for next week anyway, I will have to figure out the Netmeeting after this.
Peter: I will work with each individual to help them get up on the Netmeeting.
Bill: I will work from home to get on.
4. QA report
Marion: I agree with that, we are reaching a final stage of the Scope document. The model needs to stay in Peer Review for another 7 days.
Marion: Fridays call will be finalizing the Scope document.
Mike: Marion, I have a few more changes for the Scope document to send to you.
5. Other Business - No other issues.
Next Meeting - 5 Nov 2002, original phone number and time.