Q1: Andre, acquiring SofTouch with its formidable CrossPlex 2.1 offering was certainly a great coup for SEAGULL and let me at this juncture congratulate you before we get distracted.  So congratulations, well done and I wish you great success.  Andre, I know that Transidiom already had well proven CICS, IMS, 3270 and database (including IDMS) connectors, and being Java-based was also, in theory, mainframe deployable.  So what are your short- to mid-term plans, say over the next 18 months, to position and integrate CrossPlex 2.1 with Transidiom, vis-à-vis LegaSuite?

A1: The addition of the CrossPlex mainframe resident integration engine to Legasuite will offer our customers the
      unique benefit of a choice between a mainframe resident [q.v. large picture on front page], a distributed
       or a hybrid integration solution [see pictures below].

      Having the option of a mainframe resident integration engine will give a huge boost to performance, scalability and throughput by placing the integration point as close to the source application as possible, thus minimizing expensive roundtrips and chattyness. At the same time a Hybrid solution will proove to be effective in keeping MIPS consumption on the mainframe under control by allowing XML parsing to be offloaded to a distributed platform.

      Our plans for the next 18 months, are to closely integrate the tooling in an Eclipse based Legasuite Workbench, and to  further enhance bi-directional - mainframe app as webservice client - capabilities, as well as to further strengthen support for IMS based systems.

LegaSuite Distributed Solution

 

LegaSuite Hybrid Solution

Q2: Andre, what does the SofTouch acquisition to do your relationship with HostBridge (and I find it vaguely ironic that they are, by chance, both out of Oklahoma) given that CrossPlex now gives you an impeccable CICS story?

A2: The relationship between Hostbridge and Seagull has ended with the acquisition of Softouch. The fact that both businesses reside in Oklahoma is merely a coincidence, and may be attributed to a relative high concentration of real mainfram skills in that area.

Q3: Andre, what is SEAGULL’s current stance on non-host adapters … or lets refer to them as contemporary applications adapters … for example PeopleSoft … relative to the other big players in the market … and also given that you don’t have an explicit relationship with any well known adapter vendors?

A3: Unlike host based systems, Application systems from SAP and Oracle/Peoplesoft and the likes have traditionally been equipped with what I would characterize as a set of API's - that typically allow you to write integration logic to feed information into these systems, or retrieve information from them, thus reducing the 'challenge' of integration and therefore the value of an adapter. This trend has recently accelerated as the industry is widely adopting Web services as the universal means for integration. As a result Seagull has chosen not to enter the market of application adapters. This however does not mean that we do not have relationships with SAP, Oracle/Peoplesoft and others. On the contrary, Seagull's technology is certified under these vendors respective programs as the integration technology that allows customers to interface between the Application systems and host based systems.

Q4: Andre, again sticking to the short- to mid-term, what is going to be SEAGULL’s response to customers, who having it heard it bandied about, with abandon, by some others, want to know whether they should be looking at an “Enterprise Service Bus” (ESB) oriented approach to implementing their SOA-based solutions?

 A4: ESB is a relatively new term, and at this stage it is not uncommon to find quite large variations in the definition of what it actually is. At the end of the day one will need certain network and infrastructure  / middleware support for handling, optimizing and managing the increased volume of webservices  traffic, and I think this is where ESB's will  eventually find their place to provide value add, while at the same time some of the support will just ubiquitously  find its way into the network. So my recommendation for now would be: If you can't wait till this area matures and you need the infrastructure now, stick to a proven piece of Message Oriented Middleware  - that maybe has some specific webservices capabilities added to it -, while at the same time keeping your eyes open on maturing products and capabilities to further assist in managing webservices in a Service Oriented Architecture.

Q5: Andre, given SEAGULL’s impressive reach and recognition in Europe, what are your plans to exploit CrossPlex 2.1’s VSE support in that market – noted for its large VSE installed base?

 A5: Seagull's initial focus will be on CICS customers, and this includes CICS customers running on VSE.

Q6: Andre, as my last question (at least for the time being), can you please share with us what guidelines and criteria that SEAGULL will recommend to multi-host customers if they want to know whether they should opt for a Transidiom on a Unix platform … with the necessary mainframe adapters … or an hybrid configuration of CrossPlex on the mainframe for the mainframe applications and Transidiom for the non-mainframe applications?

A6: The optimal architecture for a heterogeneous enterprise with multiple kinds of host systems is what we at Seagull refer to as "the hybrid deployment option". This means that for those platforms where this is supported, the integration points runs as close to the applications as possible - example: Crossplex in the CICS region -. For all other platforms AND acting as the enterprise front-door for all webservices and XML parsing ,  the distributed integration engine would be deployed on the platform of choice ( Linux, Unix, Windows, Mainframe).