New Eclipse project to marry the desktop IDE and the cloud takes flight
Proposal submitted for project Flight, which aims to provide infrastructure for integrating dev tools across desktop, browser, and servers.
Software development is gradually pivoting
towards the web – and in many ways, the cloud has now come to
resemble the desktop IDE. In response to this trend, two members of
the Eclipse community, Martin Lippert (check out his talk Browser & Cloud – The Future of
IDEs? in the video at the bottom of the page) and
John Arthorne, have proposed a new project: Flight.
Currently, there seems to be some cognitive dissonance
between approaches for cloud and desktop-bound IDEs. Existing
cloud-approaches require devs to haul their work into the cloud for
development, leaving useful existing tools behind that don’t exist
in the cloud.
Ultimately, the overarching goal for Project Flight is
to end this practice by providing an extremely flexible platform
and infrastructure capable of allowing new cloud-based tooling
components to be built highly decoupled from each other, which
simultaneously bridge the gap to existing desktop IDEs.
It’s not about reinventing the wheel and
re-implementing a desktop IDE with all the bells and whistles into
the browser, nor does it plan to re-implement the majority of the
existing language tooling already there in the Eclipse universe.
What it will do is re-apply the capabilities of what’s
already there to cloud based services.
Not only would this new emerging platform and
infrastructure allow for independent components to be produced – it
could also provide connections to the most popular development
environments. The focus will be on the integration of legacy tools,
not on the re-implementation of a full-fledged desktop IDE.
There will be cloud-based tooling for a “limited” set
of languages on top of this infrastructure, which will serve to
demonstrate how language tooling could be implemented for the
cloudy tooling architecture.
The foundations for Flight will be laid with Java and
infrastructure. It’ll be put into effect by re-cycling parts of JDT
and Orion for implementing cloud services. In the future,
additional services for other languages may follow.
In contrast to Project Orion, which focuses on the
client side, Flight will have its sights fixed on the server side.
The team view the architectures of Flight and Orion “strongly
complementary”, and envision that one day, the best idea may be to
combine these technologies under a common top-level project. To
this end, Flight will combine Orion’s browser-side integration
technology as much as possible, and reuse Orion project developed
The potential of this ambitious project is obvious:
With Flight, the omnipresent gap between the traditional and the
browser IDE would effectively be sealed. The mission is now on for
Lipper and Arthorne to comprehensively integrate the tools for the
desktop, browser and server.