Archive for February, 2009

New CA Mainframe 2.0 Webcast

Why should Gen customers be interested in “Mainframe 2.0″ ? Simply because Mainframe 2.0 is written in CA Gen !

Support this initiative and sign up to the webcast – a great example of CA committing to CA Gen as a strategic product:

“Mainframe 2.0 and Electronic Software Delivery “

10th March 2009 at  2:00 pm – 3:00 pm EST

At CA World 2008 Mainframe 2.0 was launched, an ambitious effort to redefine how software is installed and managed on z/OS. One of the first Mainframe 2.0 deliverables is a new Electronic Software Delivery (ESD) process that allows CA products and service to be downloaded from the internet and installed directly from DASD without the need to use intermediate tape volumes.  This is available now for most of the z/OS products.

Please join CA’s David Luft and David Danner as they give an overview of Mainframe 2.0 and explain the new ESD process. The presentation will be approximately 45 minutes with a Q&A at the end. It will show you how to simplify and improve your installing and maintaining of CA software on z/OS. The presentation will be technical in nature and will be of interest to systems programmers, managers, and others interested in learning about the new ESD process.

Register HERE

Blog: QAT’s CA Gen Plugin Library Series #3

The 3rd in the series of the QAT Blog’s “OLE Automation versus C API” is here . At this point, Alex has done all the framework that needs to be in place for writing the CA Gen plugins.

I’ve been following this series with interest, as I’ve always found the crafting of plugins a little bit of a “black art”  – putting it simply, it just seems too hard for words  – maybe I’m not looking in the right places as a starting point!

Does anyone have any pointers to creating plugins using the “traditional” OLE Automation route or the C API route ?

Discussion on MDD

In the Model Driven Software Network”s forum, there is a current discussion on regeneration of code from models.

The discussion centres around whether the only way to get maximum benefit from Model Based Development (like Gen) is to regenerate completely from the model. From reading the threads, it seems that many people have tools and/or processes which allow 90+% generation from model, and that they have to allow for a certain amount of hand-coding and tweaking the generated code.

Of course, in a Gen environment (EABs aside) we get 100% code generation ready for deployment.  The tooling is geared up for this, so I can’t understand why anyone would want to have tools that don’t support 100% code generation. Many parts of the discussion seem to hint that some code generated by the tools that the participants use is wrong and NEEDS manual overriding – something which is alien in the Gen world.

Check out the discussion and see what you think !

Migration to Java … but why ?

There have been a few posts recently on the duick.com mailing list about moving away from Gen and recreating the systems in quetion in raw Java – maybe using some tooling to achieve the initial migration.

For the life of me, I can’t see any rationale behind this notion. Why take a system that is proven to work and has forward-compatibility with ANY technology that arises in the future and lock it in to Java ?

Once the system is written in Java, it would be an onerous task to convert to another technology or paradigm, one which would cost a lot and introduce risk into the system lifecycle.

I’m not aware of any other tooling that has the capability of Gen when it comes to targetting technologies as diverse as Mainframe, Client/Server and the web – so why would anyone do this?

Answers on a postcard please ???

Urgent update – CA Webcast date changed

CA Webcast: Multi-User Diagram Trace Utility for C servers

This webcast will now take place on 18th Feb 2009 at 11:00 a.m. – 12:00 p.m. EST for approximately 1 hour .

The date has changed due to circumstances beyond CA’s control.

To read the original abstract, see here

Register now for the webcast to gain benefit from CA’s experience.