OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

vel message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Meeting minutes 2018-12-10


Dear VEL TC members,

attached you find the meeting minutes of the TC meeting held last Monday.

Best regards,
Uwe Ryssel

--
Dr.-Ing. Uwe Ryssel
Phone: +49-391-54456939 Fax: +49-391-54456990
--
pure-systems GmbH
GeschÃftsfÃhrung: Danilo Beuche, Holger Papajewski
Sitz der Gesellschaft: Magdeburg
Registergericht: Amtsgericht Stendal, HRB 113044


Meeting Minutes - Regular Call
Date: 10.12.2018
Time: 16:00 CET - 17:00 CET

Participants (TC Members):

Attendant:

Ryssel, Dr. Uwe, pure-systems
Nesic, Mr. Damir, Swedish Royal Institute of Technology
Malfreyt, Mr. Gilles, Thales
Forlingieri, Mr. Marco, Accenture
Chawla, Mr. Himanshu, Dassault Systemes

Non-attendant:

Schulze, Dr. Michael, pure-systems (absent with valid excuse)
Côrte-Real, Dr. Joana, Siemens AG
Mattern, Mr. Stefan, Intel Corporation
Ensign, Mr. Chet, OASIS
Schur, Mrs. Dee, OASIS
Ollerton, Mr. Patrick, PTC
Apperly, Mr. Hedley, PTC
Chown, Bill, Siemens AG
Gauthier, Mr. Eric, Thales
Orhant, Mr. Jean-Christophe, Thales

Host:
Ryssel, Dr. Uwe, pure-systems

Minute taker:
Nesic, Mr. Damir, Swedish Royal Institute of Technology

Agenda items:

- Approval of minutes for the meeting on 26.11.2018
  - There are no objections from TC members, so the minutes of the last meetings are approved.

- Suggestion about the time for the next meeting
  - Uwe suggests 7th of January. Humanshu suggest to postpone one week because he won?t be able to attend. Damir says that he cannot attend on the 7th either. 
  - Uwe will discuss with Michael and send a decision/new proposal via email.

- Discussion about the draft. Are there any suggestions for conceptual changes?
  
1) Gilles: asks about the UML model on page 14 of the draft. Why is the attribute ?condition? for classes VariationDependency and Variation optional? 
    Uwe: explained that this is the case because there could be an additional use-case of VEL when variation points are defined but their conditions are not defined. In other words, a VEL model acts as variation points description without mapping to any variability model.
	Proposal by Gilles: Add OCL constraints to detail the UML model.

2) Damir: suggest that we need a section with clear definitions/glossary and clear description of the considered use cases in order to be able to discuss the details of the standard. Another suggestion is to structure the introductory text (until Section 4) according to problem/solution space dichotomy. 

3) Himanshu: continues from Damir?s arguments and asks do we want to clearly define the use cases of VEL or try to make it as general as possible. According to Humanshu, VEL currently considers software and systems but not products. By product he means, for example, the deployment infrastructure and other aspects that are not the system itself. For example, let?s say I?m building a software that can be sold in several countries. I can update the software through different servers in different countries. So depending on the country, I choose the correct version of the software and the correct server.
Uwe: Can you present such a use case for use (in a ppt)? In general, we don?t define the semantics of variability points so it should be quite general. But you should present and then we can discuss.
Himanshu: yes, I can present the case. Maybe not at the next meeting but in one of the forthcoming meetings.

4) Uwe: From pure systems side, we have some ideas on the needed changes. Firstly, when we use Enums for ExpressionTypeEnum and BindingTimeEnum, because of the XML schema principles, these Enums are not extensible. In other words, in that way VEL locks the possible expression types and binding times. Uwe suggests to switch to Strings instead of Enums so that each user can have a custom value, but VEL should define several standard values which are the same as in the current draft.
Damir: why do we have exactly these types of expressions or binding times? Why was this decision made? 
Uwe: these are probably the most common ones. But of course we can discuss these. 
Damir: also, why do we need Binding time?
Uwe: we can also discuss this. This is legacy from older projects but we should discuss if we need it. Pure systems currently does not have a use for this attribute.

5) Uwe: second suggestion is to define support within VEL for partial configurations. Current use of variation point description and variation point configuration do not support that. Suggestion is to either to add a tristate VariationPointPartialConfiguration attribute or add an additional Boolean attribute.

6) Damir: suggest to start using an issue tracking system so we can put up our comments somewhere and systematically address them? 
Uwe: Yes, Oasis provides with Jira and Github. [Addition: OASIS website offers "Actions Items" tab to add tasks to do]


[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]