Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [gef-dev] Graph Structure Proposal

Hi

The Suppress-EMF-Types genmodel option ensures you use List rather than EList.

You might manage to use some variant of unsettable to make

getCollection() internal so that you just have a getSafeCollection() that wrap an immutable list around it.

Be very very wary of introducing improved generator templates; you will find bit rot sets in very very rapidly; chasing EMF changes is very tedious.

If you want a simple API, it is really easy to write a custom Ecore to MyJava M2T with Acceleo or Xtend, so once you're happy with your EMF-based development do your own M2T so that you are independent of EMF evolution.

    Regards

        Ed Willink




On 23/07/2014 14:43, Ujhelyi Zoltán wrote:
Hi,

I have looked at the EMF model, and played a bit with the code generation settings for a start. My results are in my fork of the Github repository: https://github.com/ujhelyiz/kgraph

Results/problems:
  1. The interfaces do not refer to EMF stuff, e.g. there is no base interface, and simple ELists are used instead of Lists, etc.
  2. The Map implementation is handled differently, so in the implementation class there is some related compile error related the EMF class is used incorrectly. I am not sure how large effort is needed to update the EMF generator to handle this case correctly, but it is definitely doable.
  3. If we want to have an EMF-independent implementation, there is a single API functionality that is hard to implement using EMF-style: multi-valued references are handled by a single getter that returns an _editable_ collection. These editable collections are problematic, as it is expected that the internal model stays consistent with these editing steps, so we will most likely need custom collections to handle this case.

As both steps 2 and 3 require significant amount of work, at first I would like to discuss whether/how would we like to use EMF, and in which way. Some arguments and counterarguments for using EMF:

  + EMF maintains several internal consistency, e.g. supports bi-directional traversal; additionally extra validation is possible if needed.
  + EMF provides a notification mechanism that will be required by our tools as well (e.g. updating the visualization on model changes).
  + Serialization/export support - either built-in (via XMI), or custom (e.g. using code generators).
  - EMF dependency. Depending on who we ask, this might be tough (larger application size, dependency management) or minor (e.g. Eclipse 4 Platform already depends on it, and works outside Eclipse as well).

Cheers
Zoltán
-- Zoltán Ujhelyi
https://www.inf.mit.bme.hu/en/members/ujhelyiz

Fault Tolerant Systems Research Group
Budapest University of Technology and Economics

On 2014.07.22., at 9:58, Alexander Nyßen <alexander.nyssen@xxxxxxxxx> wrote:

Hi Miro,

I could take a first short glance and some (conceptual) questions already came to my mind, while I unfortunately did not have the time to look into all details yet:

- What exactly was the motivation of factoring our EdgeLayout into own interface? Does it make sense to haven an Edge without the properties contained therein?
- What was the motivation of doing the same for ShapeLayout? It seems all GraphElements extend also this interface. Why is there then no "Shape" abstraction but only a ShapeLayout?
- Why exactly do we need the PropertyMapping? It seems transient properties are already handled by a map. Why not use the same mechanism here? What is the notion of "persisting" them?

@Zoltán: Could you take a look into how we could realize the separation between EMF-unrelated interfaces and EMF-based implementation?

Cheers
Alexander

Am 17.07.2014 um 14:42 schrieb Alexander Nyßen <Alexander.Nyssen@xxxxxxxxx>:

Thanks Miro. I will probably have to for a detailed look tomorrow to give a first feedback.

I would like to do the discussion here or in a related bug (but not outside of the Eclipse infrastructure), so our community can participate.

Cheers
Alexander

Am 16.07.2014 um 13:21 schrieb Miro Spönemann <msp@xxxxxxxxxxxxxxxxxxxxxx>:

Hi all,

I created an Ecore-based graph structure proposal in a GitHub repository:

http://github.com/franchi82/kgraph

Feel free to experiment with it. I suggest that discussion on this proposal should be continued using GitHub issues.

Regards
  Miro


On 09.07.2014, at 23:10, Alexander Nyßen <alexander.nyssen@xxxxxxxxx> wrote:
Regarding the question whether to use EMF or not, we decided that we
(read: the KIELER project) need EMF in any case. We have two options:
* specify the data structure directly in Ecore, or
* create interfaces and an EMF-independent implementation in one plugin
and an Ecore-based implementation in another plugin.
Even if EMF is used, it is possible to use the data structure in
Eclipse-independent applications by shipping the plugins
org.eclipse.emf.common and org.eclipse.emf.ecore with it. This gives an
overhead of about 1.6 MB.
I do not object to using EMF and I think that the overhead is something that we could well live with. Actually the approach you describe would facilitate the split of GEF4 Graph into interfaces and implementation, as I proposed earlier. I think from our viewpoint, the following would be desirable:

1) the generated interfaces should not make use of the EMF collection classes (i.e. use a plain List instead of an EList) and (if possible) should not extend EObject, so we could put them into a bundle without any EMF dependencies (I think this can be achieved by certain EMF genmodel options).
2) the implementation classes could be put into a separate bundle that depends on EMF. I do not think we would have to provide a non-EMF-based implementation in addition, at least I do not see a direct use case for this yet.

This way, the use case of writing adapters between GEF4 Graph interfaces and a custom model would remain pretty much lightweight (no EMF contracts to follow), while a transformation scenario would also be best supported (here, EMF seems to be advantageous).
It is possible to remove all EMF classes from the interface. The only thing required is having some javadoc tags defined for the interface, after that EMF can generate everything required. I have first-hand experience with this - if we can follow the (naming) conventions of EMF, it works really well. The most problematic point I see is providing a modifiable list for multivalued attributes - this requires specific list implementations that can send change notifications (at least if we want to make the graph model dynamic).
Would it be feasible that I send a proposal consisting of an Ecore model
file, and you generate EMF-free interfaces + EMF-bound implementation
out of it?
That would probably be useful (1) to concretize our discussions about the actual changes that would have to be applied to GEF4 Graph to match your requirements and (2) to confirm whether the intended approach of separating GEF4 Graph into a set of EMF-indenpendent interfaces and related EMF-based implementation is feasible. As such, I would welcome this procedure.
_______________________________________________
gef-dev mailing list
gef-dev@xxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://dev.eclipse.org/mailman/listinfo/gef-dev
--
Dr. Alexander Nyßen
Dipl.-Inform.
Software-Engineer

Telefon: +49 (0) 231 / 98 60-210
Telefax: +49 (0) 231 / 98 60-211
Mobil: +49 (0) 151 /  17396743

http://www.itemis.de
alexander.nyssen@xxxxxxxxx

itemis AG
Am Brambusch 15-24
44536 Lünen

Rechtlicher Hinweis:

Amtsgericht Dortmund, HRB 20621

Vorstand: Jens Wagener (Vors.), Wolfgang Neuhaus, Dr. Georg Pietrek, Jens Trompeter, Sebastian Neus

Aufsichtsrat: Dr. Burkhard Igel (Vors.), Stephan Grollmann, Michael Neuhaus


_______________________________________________
gef-dev mailing list
gef-dev@xxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://dev.eclipse.org/mailman/listinfo/gef-dev
--
Dr. Alexander Nyßen
Dipl.-Inform.
Software-Engineer

Telefon: +49 (0) 231 / 98 60-210
Telefax: +49 (0) 231 / 98 60-211
Mobil: +49 (0) 151 /  17396743

http://www.itemis.de
alexander.nyssen@xxxxxxxxx

itemis AG
Am Brambusch 15-24
44536 Lünen

Rechtlicher Hinweis:

Amtsgericht Dortmund, HRB 20621

Vorstand: Jens Wagener (Vors.), Wolfgang Neuhaus, Dr. Georg Pietrek, Jens Trompeter, Sebastian Neus

Aufsichtsrat: Dr. Burkhard Igel (Vors.), Stephan Grollmann, Michael Neuhaus


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


-----
No virus found in this message.
Checked by AVG - www.avg.com
Version: 2014.0.4716 / Virus Database: 3986/7903 - Release Date: 07/22/14






Back to the top