Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [] Offer of GMF-T to help with their resources: what is the answer of the MDT OCL project?


Below is update from the GMF-T:

For last 3 weeks we were working on the integration of the ImpactAnalyzer into the GMFT-generated diagrams, 
as part of the Bugzilla #368398 -- modeling support for diagram properties auto-updated based on semantic properties of the element. 

As part of this work we have implemented 50+ tests for IA (right now they depends on the GMF-T), and manually tested the generated diagrams (that uses IA) with fairly complex OCL expressions. We were focusing on the use cases affecting less than a 100 semantic elements, as it is the primary use case for diagramming. 

So far we have't found any problems and we would like to promote the work done for #368398 and # into the GMF-T master. 

Even more, in case if IA would be promoted from an examples to be a part of the MDT OCL, we will consider rewriting the base diagram updater mechanism for using with IA, because it is now clear that it would lead to a simpler and more stable diagram code than the one we are actually generating for diagram update. 

Thus, I am definitely +1 for promoting the IA to the main OCL distribution. 


Michael "Borlander" Golubev
Eclipse Committer (GMF, UML2Tools)
at Montages Think Tank, 
Prague, Czech Republic

Montages AG
Stampfenbachstr. 48, CH-8006 Zürich

tel:    +41 44 260 75 57
mob: +420 602 483 463

On Thu, Feb 2, 2012 at 11:55 AM, Axel Uhl <eclipse@xxxxxxxxxxxxxxxxxx> wrote:

I can only say that I fully agree with your assessment. We know what to do: get the IA review done and work on promoting it from examples/ to plugins/ for the "mature" ecore implementation.

-- Axel

On 1/4/2012 11:51 AM, Philipp W. Kutter | Montages AG wrote:
GMF-T offered to help with the review of the IA and to promote it to the
MDT OCL main component.

AFAIK the IA only works for the Stable/Old implementation, which is as
well used in GMF-T currently. Thus while it will of course help in the
future to extend the IA to the Pivot/New, it would currently improve the
Stable/Old implementation.

Has the offer been accepted by the MDT OCL project? I have not seen a
clear answer in this mailing list. I have seen prototypes of simply
using the example component (thus code duplication!), and an own GMF-T
IA implemenation, both of them working.

I assume, we all agree that using the IA from the MDT OCL project in
GMF-T would be the ideal solution.

A clear answer to the offer of the GMF-T would be very good!

Regards, Philipp

_______________________________________________ mailing list

_______________________________________________ mailing list

Back to the top