Java Restart with WebFX
Download as PPTX, PDF1 like1,329 views
1. WebFX and Java ReStart aim to combine the advantages of web and native clients by allowing Java applications to run directly in a browser with no installation, built-in updates, and lazy downloading of required parts only. 2. A live demo showed a native-like user interface running in a browser using WebFX's FXML markup and Java code downloaded on demand via Java ReStart. 3. The system addresses limitations of both web apps and native clients by providing a simple, fast, and responsive experience that works across browsers and platforms and has transparent auto-updating.
Java Restart with WebFX
- 1. Java Restart with WebFX A new web and a new web browser Nikita Lipsky (twitter: @pjBooms) Excelsior
- 2. Agenda • Client side: Web vs. Native clients • WebFX live demo • WebFX and Java ReStart technical highlights • FAQ – WebFX vs. Java applets – Java ReStart vs. Java Web Start – WebFX and Java ReStart vs. RDP – Security • Future Development • Q&A
- 7. Client side today • Web • Mobile • Desktop Three different worlds
- 8. What’s Good with Web
- 9. Web Advantages 1. No client side installation
- 10. Web Advantages 1. No client side installation 2. Built-in free “auto updates“
- 11. Web Advantages 1. No client side installation 2. Built-in free “auto updates“ 3. Lazy partial downloading
- 12. Web Advantages 1. No client side installation 2. Built-in free “auto updates“ 3. Lazy partial downloading 4. Simple expressive means: HTML + JavaScript
- 13. Web Advantages 1. No client side installation 2. Built-in free “auto updates“ 3. Lazy partial downloading 4. Simple expressive means: HTML + JavaScript 5. Runs anywhere (where browser exists)
- 14. What’s Wrong with Web
- 15. Browser Hell
- 16. Irresponsive UI
- 18. Mobile and Desktop • Two worlds, same property: installable native clients
- 19. What’s Good with Native Clients
- 22. Look (almost) the same on any PC • Do not depend on multiple external renderers
- 23. What’s Wrong with Native Clients
- 24. Segmentation
- 25. Large systems weigh too much
- 27. Summary We need a system with the advantages of both worlds and without their disadvantages.
- 28. Summary • No client side installation
- 29. Summary • Runs anywhere (where “browser” exists)
- 30. Summary • The same in all browsers
- 31. Summary • With built-in transparent auto updates
- 32. Summary • With lazy downloading of required parts
- 34. Summary • With fast responsive expressive UI
- 35. Summary • Written in effective programming languages
- 36. Next Big Client Wish list 1. No installation (one click access) 2. Transparent “auto updates“ 3. Instant startup (lazy partial downloading) 4. Fast, expressive UI that is the same in all browsers (written in effective languages) 5. Runs anywhere (where “browser” exists)
- 37. Science fiction?
- 40. WebFX Key Features • FXML as a new HTML • Any JVM scripting language can be integrated with FXML • Arbitrary Java byte-code downloaded on demand via Java ReStart
- 41. Java ReStart Technical Highlights • Client downloads classes/resources on demand • App executes in parallel with downloading (streaming a la YouTube) • Server collects app downloading profile • Subsequent clients can prefetch required classes/resources with a single HTTP request Result: startup of an application from the web can be at least as fast as local startup!
- 42. Check list No client side installation
- 43. Check list Runs anywhere (where “browser” exists)
- 44. Check list The same in all browsers
- 45. Check list With built-in transparent auto updates
- 46. Check list With lazy downloading of only required parts
- 48. Check list With fast responsive UI (native UI is allowed)
- 49. Check list In effective programming languages
- 50. What’s the Trick?! • Web with native client look-n-feel OR • Native client with dynamics & flexibility of the Web
- 51. Java Applets • Have not evolved since 1995 • Applets are foreigners in the traditional web • Applets are not supported by browser vendors • Applets are monolithic • No “download-on-demand” out-of-the-box • Poor startup time as a result
- 52. Java Web Start • Cool idea, but implementation not so cool • There is lazy jar downloading but no “lazy” downloading of jar contents • As a result, applications are monolithic with poor startup time • Very unusual for end-user: apps are not in a browser, but are not native either • Too many security warnings!
- 53. RDP-like Approaches • Tight and constant communication with the server provokes lags • Session context on the server is the whole program – Result: poor scalability on the server
- 54. Security
- 55. Security • Security is often a social, not technical problem • Do you feel secure when you run a just- downloaded native client? • Are you sure that you can’t get malware when surfing the Internet? • I feel secure only when I trust the vendor of an internet resource
- 56. Security • WebFX reuses Java security model – At least more secure than C/C++ • Can run on top of HTTPS/SSL, to ensure the source of the remote content • Can employ mobile like security: – an app describes what it needs, an end-user has to provide access for the app to what it needs • Application stores as secure place for apps
- 57. Future Developments • Create Java application store – Start with Java apps hosting for JavaFX community • Run Java IDEs via Java ReStart (to employ the hype around cloud-based IDEs) • WebFX browser as rebrandable standalone “small” native client.
- 58. Conclusion • Web and native clients have their advantages in comparison with each other • It is possible to erase the borders between web/mobile/desktop • We can run Java applications from the web as fast as locally • So the network is the computer! Finally!