Difference between revisions of "D45 Scalability"

From Event-B
Jump to navigationJump to search
imported>Pascal
imported>Pascal
Line 32: Line 32:
 
== Edition ==
 
== Edition ==
 
* Rodin Editor
 
* Rodin Editor
As previously said, the legacy editor drawbacks mainly come from its greedy components and structure. However, a structured way to edit the models (based on a database storage) seams still particularly appropriate. This mainly explains why a new structured editor, the Rodin Editor, has been build to keep a structured edition of the model contents on the basis of the improvements made to the Proving UI. In fact, the textual component, the SWT StyledText, used in this latter UI element significantly improved the modelling experience. Hence, every element modelled in the Rodin editor is described as text. Not only this lightened the presentation, but also side stepped the use of a significant amount of greedy graphical components. Although using this technique makes models look like plain text, it is just a sort of pretty printing and they remain edited and stored by the database. The edition is then possible through the use of a unique additional graphical component that overlays the presentation of the element to edit. This drastically reduces the memory consumption as well as the number of needed graphical components because at most two of them are needed!<br>
+
As previously said, the legacy editor drawbacks mainly come from its greedy components and structure. However, a structured way to edit the models (based on a database storage) still seems particularly appropriate. This mainly explains why a new structured editor, the Rodin Editor, has been built to keep a structured edition of the model contents on the basis of the improvements made to the Proving UI. In fact, the textual component, the SWT StyledText, used in this latter UI element significantly improved the modelling experience. Hence, every element modelled in the Rodin editor is described as text. It lightened the presentation, but also side stepped the use of a significant amount of greedy graphical components. Although using this technique makes models look like plain text, it is just a sort of pretty printing and they remain edited and stored by the database. The edition is then possible through the use of a unique additional graphical component that overlays the presentation of the element to edit. This drastically reduces the memory consumption as well as the number of needed graphical components because at most two of them are needed!<br>
The first public version (0.5.0) of the Rodin Editor as been released on the 13/07/2011 as a plug-in of the Rodin platform. This decision has been made in order to let the plug-in in incubation for the time it is being tested and stated that no regression is introduced. The Rodin Editor since its version xx is part of the Rodin Core Platform.
+
The first public version (0.5.0) of the Rodin Editor as been released on the 13/07/2011 as a plug-in of the Rodin platform. This decision has been made in order to let the plug-in incubation for the time it is being tested and stated that no regression is introduced. The Rodin Editor since its version xx is part of the Rodin Core Platform.
 
* Camille  
 
* Camille  
 
{{TODO}} Ingo Weigelt
 
{{TODO}} Ingo Weigelt

Revision as of 16:09, 23 November 2011

Overview

  • Improved Performance According to the refocus mentioned in the General Platform Maintenance chapter, much of the reworking efforts performed on the core platform basically aimed to overcome Rodin scalability weaknesses, and the continuous need of seamless proving experience. The whole performance was enhanced by core implementation and user interface refactorings. Improvements made to the proving experience will be detailed in a separate chapter.
  • Design pattern management / Generic Instantiation TODO An overview of the contribution about the design pattern management / Generic Instantiation (Thai Son Hoang)
  • Edition It appears along DEPLOY - the models defined by pilots becoming industrial sized due to the level of complexity reached - that the edition became a central concern.
  • The legacy structured editor based on a form edition specific architecture reached its limits in responsiveness and hence in usability when being constrained to display a huge amount of graphical elements. Industrial partners found this issue significant regarding the adoption of the Rodin platform in industry and that is why providing a new structured editor that would solve them has been a main task during this last year of the DEPLOY project.
  • Camille textual editor had to tackle challenges related to resources mapping and management for projects of industrial size mentioned above or even the design of extension capabilities, that became a major concern since the grammar could be extended with theories.

Motivations

Improved Performance

It appeared that the various DEPLOY partners encountered several major issues while editing large models. Some were related to the core code of the Rodin platform, causing crashes, loss of data, corruption in models. Some other were related to the UI causing platform hanging, and sometimes leading to its freezing which required sometimes to kill the Rodin process, thus also leading to potential loss of data and corruption in models. Hence, it appeared necessary to solve such issues before the end of DEPLOY.

Design Pattern Management / Generic Instantiation

TODO To be completed by Thai Son Hoang

Edition

  • The motivations behind the creation of the new textual structured editor mainly concern performance but not only. From the growing size of complexity reached in models inducing a growing of their physical size lead the platform to its limits. Two major kinds of failures were faced when developing such models:
  • Failures occurring on windows platforms, the mostly used system, when reaching the operating system limit number of graphical elements in memory allowed to be displayed at once. This bug used to crash the Rodin platform, and it appeared necessary to control the number of graphical elements needed to display the model elements. Architectures sparing these graphical elements should be thus favoured.
  • Failures due to the high consumption of memory by the heavy graphical elements used by the forms of legacy editor.
Moreover, it appeared important to visualize some elements that are not part of the current level of modelling, but are linked to it, for example, the actions in an abstract event, or its guards. Such elements are called the "inherited" elements, or "implicit children". The legacy structured editor being directly interfaced with the underlying Event-B models in database, it was difficult and tricky to modify it in order to display inherited elements. This was one more argument to go for a new editor based on a intermediary representation of the models.
Another motivation to go for another editor was to get a more modest and ergonomic way to visualize and edit the models. The models being presented through styled text pretty printing and the edition tightly derived from a textual edition.
  • The motivations about Camille evolution TODO Ingo Weigelt.

Choices / Decisions

Improved Performance

SYSTEREL lead a two phase investigation to have a better idea of the work to be done. Each phase being followed by some refactoring of the code. Out of the early investigation, a cause and effect relationship has been found between performance loss and the various reported bugs, such as "platform hanging" bugs or even "no more handle" bugs related to the high consumption of graphical elements on Windows platforms. Indeed, it appeared that solving the performance issues sometimes solved induced bugs as well which made the scalability improvement tasks encompass the maintenance goals.
Later, a deeper investigation was performed, to identify and tackle the remaining performance issues. Profiling and code review were the two techniques used. The profiling strategy allowed to get a better localisation of the performance loss in both UI and core code while the code review helped to understand the intrinsic misuses or drawbacks of particular components and/or architectures. A good example, was the Event-B built-in editor based on form editors with a high use of greedy graphical components. Such architecture appears to be weak when it was needed to display industrial size models. This affected the modelling experience with some long, and really annoying to the user, reaction lags. To solve such issue, it has been chosen to refactor the editors using a textual representation which was a light-weight graphical alternative to lower the number of needed components.

Design Pattern Management / Generic Instantiation

TODO To be completed by Thai Son Hoang

Edition

  • Rodin Editor

As previously said, the legacy editor drawbacks mainly come from its greedy components and structure. However, a structured way to edit the models (based on a database storage) still seems particularly appropriate. This mainly explains why a new structured editor, the Rodin Editor, has been built to keep a structured edition of the model contents on the basis of the improvements made to the Proving UI. In fact, the textual component, the SWT StyledText, used in this latter UI element significantly improved the modelling experience. Hence, every element modelled in the Rodin editor is described as text. It lightened the presentation, but also side stepped the use of a significant amount of greedy graphical components. Although using this technique makes models look like plain text, it is just a sort of pretty printing and they remain edited and stored by the database. The edition is then possible through the use of a unique additional graphical component that overlays the presentation of the element to edit. This drastically reduces the memory consumption as well as the number of needed graphical components because at most two of them are needed!
The first public version (0.5.0) of the Rodin Editor as been released on the 13/07/2011 as a plug-in of the Rodin platform. This decision has been made in order to let the plug-in incubation for the time it is being tested and stated that no regression is introduced. The Rodin Editor since its version xx is part of the Rodin Core Platform.

  • Camille

TODO Ingo Weigelt

Available Documentation

  • TODO Links for Improved Performance
  • TODO Links for Design Pattern Management / Generic Instantiation
  • Links for Edition
Two documentation pages have been created for the Rodin Editor:
  • A general page describing the editor and how it works [1]
  • A user guide page [2]
Moreover, a special category has been created on both SourceForge feature request[3] and bug[4] trackers.
TODO Ingo Camille

Status

Improved Performance

The refactoring made on both core code and UI code allowed to gain up to 25 times speed-up on the UI, and almost a 2 times speed-up in the core code, making the platform usable in an industrial sized context.

Design Pattern Management / Generic Instantiation

TODO To be completed by Thai Son Hoang

Edition

TODO To be completed by Thomas Muller, Ingo Weigelt


References