Next Tuesday, June 15, at 1pm EDT (10am PDT, 17:00 GMT) our next edition of CTO Tuesdays, the BPM podcast, will present “BPMSs and Disaster Recovery”

Running a BPMS in a cluster of machines can help a system continue to operate even when a single machine goes down, but what can you do if an entire data center goes offline? In this talk, Michael Rowley, Active Endpoints’ CTO, will describe a deployment option that permits the BPMS runs in multiple geographically distributed clusters so that the loss of a data center does not prevent access to the BPMS. Michael will also describe how such a configuration can also be used to handle increased loads when all sites are online.

Register for CTO Tuesdays here: http://www.activevos.com/ctot. Please note our special time this week. Instead of noon EDT (9am PDT), we are holding the live recording of the podcast at 1pm EDT (10am PDT).

Episode 26 of CTO Tuesdays, the BPM podcast, is a must see.

Michael Rowley discusses how WS-HumanTask, coupled with capabilities provided by JSON and AJAX make it possible to eliminate presentation tier services on application servers for worklist management and process initiation in BPMSs.

It’s a look into how the browser, courtesy of AJAX techniques, has become so powerful for presentation logic that it has obviated the need for extra server-side work. While that may not be the latest news, what is fascinating is the application of these capabilities to BPM — and the inside look at the protocol enabled by the move of presentation logic to the client.

A note: I had a problem with the recoding of this podcast. The bad news is for the first couple of minutes, you’ll hear me give my welcome and introduction to the webinar, but you won’t see my PowerPoint slides. The good news is, once Michael begins his talk, you will see everything perfectly.

Update: During the Q&A in this episode, a user asks about parsing JSON on the server side. One of our engineers, PJ, heard that question and offers this answer:

Here’s a link to the JSON specs and libs and links to for parsing in Java, C, C++, ASP, PHP, etc.: http://json.org/

In ActiveVOS, you do not need to parse JSON because it is automatically converted to XML by the ActiveVOS engine and passed into the Receive (or onEvent) activity in your process. So, in ActiveVOS, one works with XML using XQuery & XPath.

For those using REST based services and want to parse JSON or respond with JSON, they can use the following extension functions:abx:jsonToXml(jsonStr) -> returns XML element
abx:xmlToJson(xmlElement) -> return JSON string

For more information, in ActiveVOS’s Expression dialog picklist, see Functions->BPEL->ActiveVOS->JSON

PJ also notes that we have documentation on using JSON in ActiveVOS at http://www.activevos.com/dev/sdks/XML-JSON-Binding/docs/Part1-ActiveVOS-XML-JSON-API.html and sample code in our SDKs at http://www.activevos.com/developers-sdks.php.

Thanks, PJ.

On the next episode of CTO Tuesdays, the BPM podcast, Active Endpoints CTO Michael Rowley will discuss how to eliminate the presentation tier when access worklists.

Typical web applications have a service tier and a presentation tier on the application server. However, AJAX technologies have made it possible to move more presentation logic to the web browser. Michael will describe how ActiveVOS has used these techniques to completely eliminate the presentation tier for our worklist application.

Sign up (free!) for CTO Tuesdays at http://www.activevos.com/ctot.

On CTO Tuesdays #25, Active Endpoints CTO Michael Rowley discusses BPMS alert monitors and services which can be used when the BPMS detects issues in running processes.

On episode #24 of CTO Tuesdays, the BPM podcast, Michael Rowley delivers a very interesting talk on how to manage services that might be unavailable when your process is running. A modern BPMS consists of web services running “all over the place” — which might make you think given the vicissitudes of networks and applications, building a robust BPM process application would be difficult, if not impossible.

But in this BPM podcast, you’ll see how well-thought-out BPMSs tackle this very fundamental problem — and how easy it can be to manage “problem” services. In fact, there’s so much choice in both design and deployment of BPM applications, next week’s CTO Tuesdays will expand on the ideas presented in this episode.

Be sure to watch this episode if you are working with a BPMS — and join us for part 2 next week. Register at http://www.activevos.com/ctot.

Tomorrow on CTO Tuesdays #24, Michael Rowley will cover the crucial topic of what to do when a web service that is required by a process application isn’t available. Register for the webinar here.

We’ve long argued that processes defined using a BPM (business process management) discipline which are automated should be developed using a services-based environment. But what happens when those services aren’t available? Join us on Tuesday, May 18, 2010 at noon EDT, 9am PDT, 16:00 GMT to explore your BPMS options for this eventuality.

This episode of CTO Tuesdays details with how escalation works in a BPMS. In particular, the discussion and demonstration tackle how to use process-leveland task-level escalation, including deadlines, in the design and deployment of processes.

As always, you can join us live for CTO Tuesdays by signing up at http://www.activevos.com/ctot.

We hope you will join us.

Tomorrow on CTO Tuesdays #23, we will explore how a business process management system (BPMS) handles escalation. Register for this episode at http://www.activevos.com/ctot.

We hope you can join us at noon EDT, 9am PDT, 16:00 GMT for our discussion of this important topic. When you consider the kinds of core processes that BPM systems are used for — processes which almost always include human tasks — being able to escalate work when something doesn’t happen is an important capability.

As always, Michael Rowley, Active Endpoints’ CTO, will present an educational overview of escalation followed by a panel discussion to answer your questions.

In the loosely-coupled world of today’s process applications, how does the BPMS know which running process instance to route incoming messages to? It’s an important question…and one CTO Tuesdays has explored before (see CTOT #5 on engine-managed correlation).

In this episode, Active Endpoints CTO Michael Rowley tackles a more complex case for correlation: when you can’t use engine-managed correlation because the developer doesn’t have control of the client in a business process. This episode reviews the terms used in correlation and then walks through a simplified procurement process to illustrate the concept.

As always, we welcome your feedback and suggestions. And we hope you will be able to join us for the live recording of CTO Tuesdays ever Tuesday at noon ET.

Correlation sets in BPEL allow designers to manage asynchronous conversations in their business processes. Correlations sets are both a simple idea and very useful. By design, BPEL engines can manage a wide range of situations automatically, as we explored in CTOT #5: Engine-managed correlations. But what can designers and developers do when things get more complicated?

That’s what we’ll explore  on CTO Tuesdays for Tuesday, May 4, 2010. We’ll be discussing more complicated correlation examples and offer tips and techniques for using BPEL correlation to support a very wide range of processes.

Please join us at noon ET, 9am PT and 16:00 GMT. You can register to watch the podcast live and participate in the Q&A discussion by registering at http://www.activevos.com/ctot.

We look forward to seeing you on the podcast.

Well, I’ve been waiting a very long time to write a blog post that mentions that in the ABC TV series “Lost” the famous hatch where the button had to be pushed every 108 minutes had some very retro, old-school tech in it.

After searching and searching, I finally found a still on a fan website from the show’s second season that clearly shows a pair of IBM 3420 mainframe tape drives. Say what?

The Dharma Initiative was using IBM mainframe technology on the island in 1974? What did they need data storage for? Why are the tape drives always empty? What do those cool lights shining up from the bottom of the unit mean? (There were no lights there on the original 3420s.) How come I am the only person to have noticed these antiques in the hatch? (OK, not the only one. But a lot more people are wondering about the meaning of the smoke monster, or whether the Island is Hell — and all I want to know is what OS they were running on these things..MVS, VSE or the progenitor of all things virtual, VM/370.)

Lemme tell ya, these babies could store up to 150MB per 10″ tape reel! And to speed access, the drive “sucked” the tape down into the vacuum columns you can see on either side of the lighted area on the bottom of the tape drive. And, believe it or not, these things “burped” when you unloaded them — which depending on the version of the operating system the mainframe was running, sometimes had to be done manually.

So, what’s all this have to do with the content of CTO Tuesdays #21? Easy: for all the jokes about outdated mainframe(rs) and their never-very-cool reputation compared to the then-nascent PC, mainframes matter. They mattered then…and, to the surprise of no one who’s ever worked on them, they matter today.

On this episode of CTO Tuesdays, Michael Rowley turns the con over to Rob Morris and Dusty Rivers of GT Software, who describe an elegant and practical way of linking everything on a mainframe — CICS transactions, IMS queues, VSAM keyed datasets, raw 3270 data streams…you name it — to a modern BPMS like ActiveVOS. This is done via standard web services, using WSDLs to define available mainframe services.

Simply, if you have mainframe technology in your organization (and, frankly if you are in financial services, telecommunications and/or entertainment & media, we’d bet you do) and you want to write business processes that live up to the promise of flexibly integrating both people and systems, you gotta watch this podcast.

Tomorrow on CTO Tuesdays #21 for April 27, 2010, we are pleased to present a topic that is on everyone’s lips: how to integrate mainframe systems like CICS, IMS and 3270 apps into a the new world of process applications.

Our guests tomorrow are Rob Morris and Dusty Rivers of GT Software who will show how web services can quickly and easily integrate these two worlds.

Register for the live webinar here.

Attached to this post are the recordings and a PDF of the slides presented on CTO Tuesdays #20 for 13 April 2010.

Completing a “trilogy” of sorts that started with CTOT #18 and continued in CTOT #19, Michael Rowley covers the concepts — and politics — behind REST, the representational state transfer protocol.

This episode is a must-see for anyone interested in protocols and comparing and contrasting REST with SOAP.

We promised to include this link to the WS-* standards that was shown in the discussion: http://www.innoq.com/soa/ws-standards/poster/innoQ%20WS-Standards%20Poster%202007-02.pdf

Our topic tomorrow is “Using REST with business processes.”

REST (for “representational state transfer“) has become a popular style of creating applications in a services-oriented environment. But many development teams are less familiar with REST than they could be. On CTO Tuesdays for April 13, 2010, Active Endpoints CTO Michael Rowley will provide an easy-to-understand overview of REST and show examples of its use. Register at http://www.activevos.com/ctot to join for this exciting look at REST. CTO Tuesdays is presented live at noon ET, 9am PT, 16:00 GMT.

CTO Tuesdays is a weekly podcast about the technology of BPM. Each week CTO Tuesdays tackles a single technical topic and delivers the education development teams want to learn how to create a new generation of process applications. Register for the next episode of CTO Tuesdays here.”

Attached to this post are the recordings of CTO Tuesdays #19 for 6 April 2010.

In this very interesting episode, Michael Rowley covers the (somewhat sad) history of URNs (complete with a screen shot of the entire web, circa 1991). Rowley then details how and why URNs are useful in process applications. A panel discussion follows the formal presentation.

Next week on CTOT: Rowley completes the story arc he started on CTOT #18 (using email-based approvals for processes) and which continued in CTOT #19 with a discussion of REST — the representational state transfer protocol. Don’t miss it: register at http://www.activevos.com/ctot.