I've been thinking a lot about multi-app web projects, and how none of
the Ruby web application frameworks I know offers a clean solution to
build a site that conveniently integrates multiple apps (e.g. a core
site and a forum).
Integrating a rails app with beast (a rails forum engine) for example is
just as cumbersome as integrating a rails app with a php forum engine.
This shouldn't be like that, when Ruby can offer so much more elegance.
I would like to get some insights from the community and stimulate a
productive discussion of multi-app capabilities for ruby web frameworks.
I wrote a longer blog post detailing what I think sucks about current
multi-app integration (exemplified by my works using rails), and how to
fix some of it.
There are currently a lot of promising, growing ruby web frameworks, and
right now seems like a good time to put this out there for
consideration.
Camping has some support for this built-in. Like most things Camping it's simple stuff, but still kind of neat.
James Edward Gray II
···
On Jan 3, 2008, at 6:44 PM, Ben (ducktyped) wrote:
I've been thinking a lot about multi-app web projects, and how none of
the Ruby web application frameworks I know offers a clean solution to
build a site that conveniently integrates multiple apps (e.g. a core
site and a forum).
(*shrug*) I do it in IOWA all the time. Core site. Forum. Tools for managing site information. Reports. Dynamically generated tables. Multiple different apps for the customer's customers. Whatever.
Kirk Haines
···
On Fri, 4 Jan 2008, Ben (ducktyped) wrote:
I've been thinking a lot about multi-app web projects, and how none of
the Ruby web application frameworks I know offers a clean solution to
build a site that conveniently integrates multiple apps (e.g. a core
site and a forum).
Integrating a rails app with beast (a rails forum engine) for example is
just as cumbersome as integrating a rails app with a php forum engine.
This shouldn't be like that, when Ruby can offer so much more elegance.
I would like to get some insights from the community and stimulate a
productive discussion of multi-app capabilities for ruby web frameworks.
I wrote a longer blog post detailing what I think sucks about current
multi-app integration (exemplified by my works using rails), and how to
fix some of it.
I tend to do it with Rack. Either a bare metal app, Mongrel hosted, A
Camping app, or the like. If there's an adapter then rack can pipe to it. No
restrictions on the toolchain.
···
On Jan 3, 2008 6:31 PM, <khaines@enigo.com> wrote:
On Fri, 4 Jan 2008, Ben (ducktyped) wrote:
> I would like to get some insights from the community and stimulate a
> productive discussion of multi-app capabilities for ruby web frameworks.
> I wrote a longer blog post detailing what I think sucks about current
> multi-app integration (exemplified by my works using rails), and how to
> fix some of it.
(*shrug*) I do it in IOWA all the time. Core site. Forum. Tools for
managing site information. Reports. Dynamically generated tables.
Multiple different apps for the customer's customers. Whatever.
(*shrug*) I do it in IOWA all the time. Core site. Forum. Tools for
managing site information. Reports. Dynamically generated tables.
Multiple different apps for the customer's customers. Whatever.
Kirk Haines
Kirk,
I also saw your twitter message ("Multiple apps in 1 Rails instance is a
problem? Maybe I should write an IOWA demo that does multiple apps +
multiple domains in 1 instance.")
I think that's a great idea (it could even be the start of IOWA's home
at swiftcore). Throwing in some ajax might help its marketing as well.
I think part of the problem is the MVC approach taken by Rails.
While it does allow for a relatively clean abstraction, it is
"controller-driven".
This works well enough when all the logic is handled by *a single controller*.
Wicket (Java) and Lift (Scala), on the other hand, follow a
view-driven approach.
I'm just diving into Lift, but the approach seems promising: It may
work better for me. Or it may not.
At the very least, it's another approach to the same problem.