PIP - coming soon

FIS PIP on GT.M on Linux on x86 hardware is a complete FOSS stack with superb transaction processing throughput and unique functionality for extreme levels of business continuity. PIP will soon be released as free / open source software (GPL v3).

PIP is the infrastructure on which FIS Profile's financial applications are built. Till now, the infrastructure has been joined at the hip with Profile's financial applications and has not been separately available. It is now being separated from the financial applications, and will soon be available as FOSS. PIP includes a SQL engine, PSL (Profile Scripting Language, a lightweight object oriented scripting language), a JDBC driver, and two IDEs - one built on Eclipse and another browser based and driven by Tomcat on the server. Applications built with PIP can use either FIS GT.M, or Oracle as the database engine, and in the future will be able to use other engines as well. Under the covers, PIP compiles PSL and SQL into M in two flavors - M code accessing an M database and M code calling the API of a relational database.

A core processing system is the legal system of record for a bank, and its single most mission critical application. Built on PIP, Profile runs the largest real time core processing system that is live at any bank anywhere in the world that we know of, and has recently been successfully benchmarked on an x86_64 Linux platform on a database three times the size of that largest live system. GT.M is the database engine on which Profile is most widely deployed around the world as the legal system of record for many tens of millions of accounts.

PIP will be released on February 10, 2008, at the Southern California Linux Expo (http://www.socallinuxexpo.org/scale6x/conference-info/speakers/Bhaskar/) and will be downloadable from Source Forge (http://sourceforge.net/projects/pip).

-- Bhaskar

Going beyond REST

On Wednesday 28th November, Skills Matter are presenting a session called Going beyond REST in Clerkenwell, London.

This session is FREE and is slated to describe Resource Oriented Computing (ROC)which is a model that goes beyond REST and unifies a number of diverse but similar principles for treating data as resources. If you enjoyed my talk at Slipstream then you will probably find this interesting.

The presentation will probably also include discussion of NetKernel which is an Open Source implementation of ROC that's based on original research from HP's Dexter project.

Conveniently, it is happening the evening before this event, and is just around the corner. So if you are planning to travel down to London to attend the CAMTA meeting (for my SVG training session, of course) then why not arrive a little earlier, and take in the Skills Matter session.

Register here - it's free.

Google's new OS

I was lucky enough to be one of a small group of people invited along to Google's London offices last Friday for a breakfast briefing. This was right after Thursday's announcement of OpenSocial. We didn't know for sure what was going to be covered, but a combination of leaks and educated guesses meant that most people had figured out that we were going to get more than just a free breakfast.

As I'm sure you know by now, OpenSocial is a social-networking application development API created by Google that is being implemented by pretty much all social networking sites (with the notable exception of Facebook). You can now write applications that will work without modification on MySpace, Orkut, Friendster, Bebo, LinkedIn, Ning, hi5, salesforce.com, and many more. The only social networking site that is not likely to appear in this list any time soon is Facebook, and that's an obvious consequence of Microsoft's recent $240m investment. But even Facebook might have to kowtow to OpenSocial in the long run.

The OpenSocial API gives you the ability to write social-networking applications that can be hosted on any site that supports the API. This "learn once, run anywhere" capability makes it easy to develop functionality that can be used by a combined audience of over 150 million users. This is a one-click reach that puts your application right on the user's home page with no clumsy client-side installation or fiddly account registration to go through.

Suddenly this is the new application development platform kid on the block. At the moment it is billed as a social-networking platform, but from what I've seen there's no impediment to writing any kind of application using the OpenSocial API. Google have chosen to leverage HTML, JavaScript and Ajax, all of which are open and ubiquitous, to create a very rich and flexible platform. By contrast Facebook's platform (FBML), launched a few months ago, is idiomatic, closed and proprietary.

You can now build full function applications that can run inside each of your user's chosen social networking site. No longer do you need to ask users to navigate to your web-site, lead them through tortuous registration and login procedures to a site with its own idiosyncratic behaviour and UI. Just get them to add your application to their own MySpace, Ning or Orkut page. Whether your goal is to provide financial information, retailing, patient appointment booking or any other kind of customer facing service, you can now bring it to them rather than requiring them to come to you. You can have your application right in front of them, logged-on and authenticated whenever and wherever they're on-line. Why would you want to make it any harder than that for your users to run your application?

It's early days and the dust has yet to settle, but it could be that what Google have unleashed here is not just a social-networking platform but something that could grow to become the platform for all application development.

Today Google OS means Google OpenSocial, but sometime in the future it might just mean Google Operating System.

Amazon S3 Client

Well I announced it in my Lightning talk at the conference. The Amazon S3 Client
for Caché is now finished and available from http://www.mgateway.com.

Here's the notes from the web site:

This application provides a Caché-based web interface to the Amazon Simple Storage Service (S3) which
offers very low-cost, secure online data storage on Amazon's infrastructure. You can use Amazon S3 for
private storage of data (eg for low-cost backup) or for making data and/or files publicly available at an
incredibly low cost.

You can use the pre-built, Ajax-based client interface to manage and maintain your Amazon S3 storage.

Alternatively you can also use the same low-level Caché API methods that were used to create
the pre-built client interface. These APIs allow you to quickly and simply build your own application
interfaces and to integrate Amazon S3 storage with your Caché applications. The M/Gateway S3 APIs make it
possible to use Amazon S3 storage as an extension to Caché's database storage.

Amazon S3 is VERY cool, and incredibly cheap. Check it out and try out our interface!

Syndicate content