Re: [papyrus-ic] Papyrus: performance improvment.

["Garcia-Dominguez, Antonio" <a.garcia-dominguez@xxxxxxxxxxx>]

Dear all,

Thank you for letting us know about this! If I may, I am one of the developers of Hawk, an Eclipse-based model indexing tool based on graph databases with support for queries in an OCL-like language (the Epsilon Object Language):

https://github.com/mondo-project/mondo-hawk

Hawk can already index EMF models and we've used it to index UML models in the past. We can also automatically recompute derived attributes and index model elements by the value of a certain attribute, and do so in an incremental manner while reacting automatically to any changes in the currently open Eclipse workspace.

I would be very interested in hearing more about these requirements for indexing that you may have, and seeing if Hawk can meet those requirements.

Kind regards,

Antonio

--

Antonio Garcia-Dominguez

Lecturer in Computer Science

School of Engineering & Applied Science, Aston University

Aston Triangle, Birmingham, B4 7ET

Room: MB211Q

---

From: papyrus-ic-bounces@xxxxxxxxxxx [papyrus-ic-bounces@xxxxxxxxxxx] on behalf of Philip Langer [planger@xxxxxxxxxxxxxxxxx]
Sent: 14 November 2016 15:15
To: papyrus-ic discussions
Cc: Thierry LESERGENT (Thierry.LESERGENT@xxxxxxxxxxxxxxxxxxxxxxxx); papyrus-ic-steering-committee; Papyrus Project list; Papyrus SIG; Alain Le Guennec
Subject: Re: [papyrus-ic] Papyrus: performance improvment.

Hi Séb,

thanks for the information!

I'd like to add an additional requirement on the Papyrus-specific indexing component; that is, capturing the cross-file model dependencies among Papyrus models.

Currently, we have to do that in EMF Compare for diff/merge by crawling the workspace building up an own cache. It would make much more sense to have a Papyrus-indexing-component, which provides to EMF Compare the information on which models are actually linked to each other (incoming and outgoing cross-file references), and which is kept up-to-date with Papyrus-specific mechanisms, such as updating the index when relevant changes that affect the cross-file dependencies are applied to the model.

Do you think this fits here and could be added to this work as well? I'm happy to add more detailed requirements in the wiki page if you agree.

Thanks a lot and best wishes,

Philip

On Mon, Nov 14, 2016 at 12:11 PM, GERARD Sebastien 166342 <Sebastien.GERARD@xxxxxx> wrote:

Deal all,

 

Performance is a working area we are working on regular basis to improve Papyrus. In this context, as announced earlier by Florian on the Papyrus dev email list, we are starting a collaboration with IncQuery. This work will consist of the three following tasks:

·         Task 1. Setting up a continuous scalability evaluation testbed for Papyrus. This testbed will be based on RcpTT tests and Hudson jobs

·         Task 2. Identifying and sharing existing models (UML, notation, css, profile) related to performance bottlenecks in Papyrus. The Papyrus community is expected to share some models, but many performance issues are only present in big proprietary models (Obfuscation will probably be required)

·         Task 3. Integrate a Papyrus specific indexing module based on the VIATRA base-indexer and query engine to improve model navigation and derived feature usage. Based on the 2 previous tasks, there will be an evaluation before and after the model integration to check performance improvements.

You can see all the details here: https://wiki.eclipse.org/Papyrus/Oxygen_Work_Description/Refactoring/PerformancesImprovements.

 

We would appreciate a lot to receive your feedback, comments, suggestions, use cases and also to get some examples of model illustrating performance issues.

 

Thanks,

Best,

Sébastien.

 

			Sébastien Gérard Head of the LISE labs CEA Research Director Papyrus project Leader (www.eclipse.org/papyrus) Commissariat à l’énergie atomique et aux énergies alternatives Institut List | CEA Saclay Nano-INNOV | Bât. 862- PC174 F-91191 Gif-sur-Yvette Cedex M. +33 6 88 20 00 47 T. +33 1 69 08 58 24 sebastien.gerard@xxxxxx  www-list.cea.fr

 

 

_______________________________________________
papyrus-ic mailing list
papyrus-ic@xxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://dev.eclipse.org/mailman/listinfo/papyrus-ic

--

Dr. Philip Langer

Senior Software Architect / General Manager
EclipseSource Services GmbH