Welcome to Inqwell
- What Is Inq?
- What About Clients?
- What Else Does Inq Do?
- Inq in a Bottle
- Inq Roadmap
What Is Inq?
Inq is a dynamic scripting language aimed at the rapid development and streamlined maintenance of enterprise systems.
As a language, Inq is procedural though it has specific code blocks for application instance life-cycle stages and a general purpose closure concept.
As a server-side container, Inq manages transactions, caching, cross-vendor database persistence and events. As a minimum, these concepts require no user-written code. Inq also supports communicating processes and Hoare style monitors.
At the application level Inq has a type system, where things such as persistence bindings, labels, preferred rendering widths and formats are defined. These additional attributes are captured within the application type and available for clients and Inq's reporting engine to use.
At the language level Inq is typeless. Everything is represented as a node of type any, which may contain other nodes. Because of this distinction between the application and language levels, Inq is able to offer high levels of functionality already built in. Examples are aggregate operations (for example weighted average) and sort.
Inq is parsed once then executed. Applications can be modified while they are executing with no requirement for a restart.
Inq is is written in Java (requires Java 1.6) and is therefore fully cross-platform. It treads very lightly on the various Java APIs it uses (for example its transaction mechanism is its own; JMS integration adheres strictly to the specification) so combining Inq with vendor implementations of (say) JDBC and JMS holds no surprises.
What About Clients?
The Desktop Client
At present Inq includes a desktop client that works closely with the server. This client uses the server to maintain its state; events are pushed both ways. This client also employs the Inq language and presents the Swing components.
The client has access to all the application type meta data which it uses to label and provide default dimensions for components. Together with its layout models and syntax, these aspects make GUI construction and maintenance easy.
As well as the standard Swing component set, Inq includes Docking Frames and a calendar component. Try out the examples from the distribution, in ./examples/gui.
With its typeless node structure and events pushed from the server, Inq automatically implements MVC between the node space and the GUI components. This is another example of Inq's built-in high level functionality that the users do not have to write themselves.
A servlet runs in the same VM as the Inq environment. Requests from the browser are executed as Inq service calls and the node structure produced is returned as JSON. By extending the qooxdoo components to reference appropriate subnodes and meta data, rich internet applications can be rapidly developed. Application types and code can be reparsed in the server at any time.
What Else Does Inq Do?
Because the Inq language is typeless yet an application type's meta data is available, many complex functions can be supported completely independently of specific applications. For example, the Item Chooser is a desktop client dialog that accepts some function arguments (Inq's closure concept) and cooperates with the server to query and select amongst a number of items. We expect this type of component to translate to the web client environment as that is developed further.
There is a plugin architecture that allows existing Java code to access the transaction environment. Such code can create application type instances, mutate or destroy them within the Inq server's transaction model.
Inq can call Java code via its callmethod function.
It is very quick and straightforward to write reports in Inq. Once more, the meta data kept in the application types is made available to templates when a node structure is transformed into XML. Such an XML document can be transformed using external tools, for example Apache FOP.
Examples can be found in the reports produced in the Inq version of petstore.
If you like the look of Inq and think it is missing something, or if you would like to help with the web client integration then we'd love to hear from you.
Get in touch at email@example.com
Inq in a Bottle
Whether writing and running client or server code, much of what the programmer wants to happen is handled automatically by the Inq runtime. Inq has been written to allow the developers of complex applications to concentrate entirely on exactly that - only the application. Inq removes two major facets of application development that are often traumatic:
- The engraining of the chosen view of the "real world" in the code. Inq is not an OO language. It has user-defined types but its simple procedural approach and typeless language means that application areas are not tightly coupled to each other. Maintenance for anything unanticipated is unlikely to cause major rework or result in duplicated or spaghetti code.
- There are no multiple layers of software to integrate and write against the APIs of. So no JDBC, no endless classes implementing this or that interface, no transaction calls to make. No table models to continually implement and to adapters to write. Inq is RAD in the extreme.
Inq Protects Your Investment in Legacy Systems
A "big bang" approach to upgrading your systems to the latest technologies is rarely an option. Because Inq cleanly separates database schemas from the application code, clients are able to migrate their existing systems in discrete sections of whatever size they choose.
Inq will tolerate database schema changes much more readily than code whose dependency on the SQL is engrained. As larger sections are migrated, so the schema can be updated as required and the legacy code retired. The pace and size of the migration is entirely at the discretion of the user.
The preferred method of implementing server-side application logic is by expressing it as Inq code. However, legacy processing held as database stored-procedures can be maintained as long as required. Inq can be notified that a class or defined set of objects have been modified externally, upon which it will resynchronise with the external data source, raising events about what has happened as required.
Inq Instantly Distributes Your Application
Inq is a scripted language - it needs no compilation. The source files are maintained centrally and downloaded to connected clients by the Inq server.
Inq clients are indistinguishable from their traditionally coded Java counterparts. Once parsed, the source is not needed again. With the runtime available locally, Inq combines the instant distribution of browser-based solutions with the rich functionality of desktop applications.
Inq Efficiently Distributes Your Data
Inq propagates events between server and client according to whether the client is observing the data raising the event. By only notifying clients of the data they are interested in, Inq makes efficient use of network bandwidth.
When retrieving large datasets, Inq can use compression to reduce its network usage by up to seven times. Inq solutions solve the problems of centrally managed applications often suffering from network bottlenecks.
The Inq server caches objects it has read from a database. Together with the efficiency of its own client-server protocol, Inq operates effectively when client, server and database are coupled over a wide-area network.
Inq Brings Users and Developers Together
The speed and simplicity of Inq development means that users and developers can work much more closely together. Developers can rapidly produce and refine prototypes that users can see and try for themselves.
This highly iterative process becomes the de facto methodology of development. The users and developers have confidence in each other, leading to much better project management and a more assured delivery schedule.