Running a Plone product on Substance D
1 like1,990 views
This document summarizes Éric Bréhault's talk at PloneConf 2014 about running a Plone product on Substance D. It discusses why one may want to experiment with running Plone on Substance D, describes the structure and components of Rapido which was developed to work on both Plone and Substance D, and outlines some of the challenges and solutions explored in making Rapido work across both platforms. It concludes that while Plone should not fully migrate away from its current architecture, there is value in experiments that explore flexibility and preservation of its core components.
Running a Plone product on Substance D
- 1. Running a Plone product on Substance D Éric Bréhault - PloneConf 2014
- 2. Let's be specific about it "running" is not "migrating".
- 3. Why would we do that? No obvious reason.
- 4. Why would I do that? I have free time. It sounds fun. It might be a good experiment for the future of Plone.
- 5. Good things in Substance D All the good things from Pyramid PLUS stores data in a ZODB, provides a management interface (the SDI), "Substance D owes much of its spirit to the Zope application server" (sic)
- 6. Good things in Substance D The D initial! "So it still lives... The will of D."
- 7. First attempt to integrate Plone $ pcreate -s substanced ./plone Make sure to load everything in __init__.py: config.load_zcml("plone-d.zcml") DEMO!
- 8. The real experiment case: Rapido Rapido is the next Plomino version. It is a complete rewrite.
- 9. Plomino It is an old Plone product (started in 2006). Still based on Archetypes. Stores data into CMF objects. Uses extensively ZCatalog and PythonScript.
- 10. Rapido (Originally) it is a Plone 5 product. Based on Dexterity.
- 11. Rapido structure rapido.core totally independent on Plone provides adapters able to produce the expected behaviors requires a storage service
- 12. Rapido structure rapido.souper provides a storage service based on Souper (Souper does work on Plone AND Pyramid)
- 13. Rapido structure rapido.plone standard Dexterity contents adapts them using rapido.core (ideally) uses nothing but plone.api
- 14. Rapido structure rapido.substanced standard Dexterity substanced.content classes adapts them using rapido.core uses nothing but plone.api Substance D API
- 15. DEMO (the real one)
- 16. How does that work In core, only use components working in both Plone and Substance D.
- 17. TTW scripting TTW scripting is what Rapido is about. (more globally, a rich TTW experience is vital to any CMS) PythonScript ➜ zope.untrustedpython
- 18. Catalog Substance D has its own catalog, and Plone too... ➜ repoze.catalog is just fine.
- 20. Settings persistence Sharing schemas between Plone and Substance D sounds scary. ➜ Annotations!! Both Plone and Substance D contents can be IAttributeAnnotatable.
- 21. Forms & widgets Deform is not rich enough. I did not considered running z3c.form on Substance D. ➜ client-side rendering!! with Angular Schema Form
- 22. Access control Both systems have a granular ACL service. Probably possible to support them transparently from core, but for now: custom security implementation
- 23. My experience with Substance D PROS: Fun!! Happy to find all the good ingredients.
- 24. My experience with Substance D CONS: Not 100% ZCA-ready. Need to call config.hook_zca(), it works fine, no problem. Just not confortable with the "hook" term here. We would probably need a local registry.
- 25. Conclusion about Plone future ZCA + buildout + ZODB makes our identity, we must preserve it. We can find clever approaches to avoid a full rewrite.
- 26. Conclusion about Plone future Can we easily migrate to Substance D? No Should we migrate to something else? No
- 27. Thank you!