Virtual Identity Dialogue

As many of you know - I'm really into improving presentations - aka moving beyond bullet points and slideuments (aka slides as documents).

The Oracle Middleware team just released a new video on YouTube describing the application grid.

What I like about this:

1. Shows how you can give effective presentation without needing to show the audience your cue-cards (after all that's really what the bullets are meant for)
2. I love how using the whiteboard gives the visual cues without feeling overly slick. If you don't have a whiteboard -just a couple of slides could be used
3. Leverages the fact that there is an entire website dedicated to give you much more detail - thus can give concise information.

It very easily could have been a 20 minute "drink from the fire-hose" presentation. Now instead it's more like drinking a fine glass of your favorite beverage.

Eve Maler (btw who pronounces her last name differently than Sophia Maler who is our PM for federation. I suspect but have not yet confirmed that any female involved with SAML must have the last name of "Maler" but I digress) recently wrote up her ideas on a new protocol "ProtectServe". In particular to enable a new type of experience around "relationship management" (which sounds like an episode of Oprah about marriage but again I digress).

Which from the best I can figure involves some-type of automation to help manage what data is sent automatically on your behalf to different entities.

And it also relates to the concept of "Vendor Relationship Management" aka VRM.

Now, I happen to be at Oracle aka home of Siebel and related CRM products. Now, I don't know if we have an official position on VRM and if we do, what it is. If I find out, I'll share.

But until then, it occurred to me - we actually could support this type of scenario today. And with some of the Social CRM projects they have cooking - could possibly do more in the future.

However, here is how the scenario could be worked out today.

1. The RM sends us some data via SAML (whether this is a traditional IDP or InfoCard)
2. Java application gets the SAML data and converts into a Java object
3. Java application then makes a call to Siebel's Web Service APIs (either directory or via an Application Integration Architecture pack)

Today there would be lots of protocol translations and related activities but our future plan is that ArisID Beans (part of the IGF implementation) would allow us to simplify the programming model. But that will have to wait for a future post.

I was listening to episode 47 of StackOverflow and Joel & Jeff celebrated the demise of Microsoft FrontPage. While I'm sure many others feel the same way - I do want to provide to the future Web historians - why FrontPage was important. And then where it went wrong.

I first encountered FrontPage in November 1996 when I became the first official "Webmaster" for the University of North Texas. My job was to not only manage the UNT central web servers and home page (until campus marketing woke up to the Web) but also help get other departments content up on the Web. And my team consisted of me plus 1 half-time person. We were housed in a room slightly smaller than a broom closet.

This was well before blogs and where Notepad was the de-facto standard for developing Web pages. And since we were running the central Web server on Sun boxes so to get content published - required FTP.

Or in other words - there were severe roadblocks to getting content up on the UNT web.

When I started a new upstart OS - Linux had just come into existence. And one Linux flavor - Caldera had support for Novell (the primary campus PC network) shares. So we cobbled together a PC, with an old network card (it only supported coax-cable "thinnet" not cat5 , so we had a converter) and Caldera. It worked (I knew Linux was going to be success when I went years without needing to touch the core OS, except the Novell shares kept dropping) and solved my FTP problem (learning FTP was beyond most people) but didn't address the hand-editing HTML problem.

When I first started - I was really hoping AOLPress - which was to my knowledge the first true WYSWIG HTML editor, would stick around. Unfortunately it was killed off in early 97. Thus I turned my attention to FrontPage.

FrontPage was not an original Microsoft product - it was an acquisition. Thus it actually initially worked better with UNIX systems. And it was exactly the tool I needed to get UNT Web off the ground.

It provided a simple, WYSWIG interface analogous to a word-processor.

It allowed you to create simple forms without needing to do any programming (a major sticking point for us since we didn't have a box we could dedicate to CGI scripting run amok).

And to publish you did "File->Save".

I cannot underestimate the success this was for us. I was able to grow the UNT web rapidly - I think at some point we were doing like 10 new sites a week or something similar. It also helped us pick our choice in our online course software (WebCT) because it was basically just Apache+Perl.  Thus I could hack it to make it work with FrontPage (which was my initial claim to fame in that community. As in at the first WebCT conference, my presentation on FrontPage integration was attended by as many people as the keynote).

Of course FrontPage got sucked into the "we'll pretend the entire Web can run entirely on Microsoft only" insanity. And of course FrontPage couldn't really decide (IMHO) what it wanted to be. It wasn't tailored for the hardcore designed geeks like Macromedia Dreamweaver and then Web 2.0 cut out the bottom. Web 2.0 (blogs, MySpace/Facebook,Flickr, etc) made it simpler to create content without alot of special tools.

Thus the need for FrontPage was diminished though not completely forgotten. I mean I'm using Microsoft Live Writer to write this blog. Which in the end is much nicer and faster than cranking out tags.

So while in the end FrontPage had lots of warts - it's importance to the early Web shouldn't be completely forgotten.

There is a new webcast on how to provide security for SharePoint using Oracle Identity Management products.

And we have additionally published a paper on how to secure SharePoint with Oracle Virtual Directory.

I was recently published in SC Magazine:

"Virtual identity services enable merger business process integration for M&A"

It's basically another intro-article on virtual directories.

Combing through email that I had put away to read until after RSA - I discovered Deb Volk from Identigral has written a couple of nice posts on virtual directories and OVD.

Thanks Deb!

I've discovered that my comment filter for my email - um actually works. And it has been sticking it into my internal folder that I only occasionally peek at. So - thanks for the comments -sorry took long to reply.

Eve responded to my earlier post on VRM and CRM. Until the Oracle/Sun deal is finalized - I can't really respond to her but did want to post a comment I got from Iain Henderson at Mydex:

Hi Mark,

just building on your VRM to CRM post; have a look at the link below - a post from Asa Hardcastle about sending VRM data into CRM environments via Information Cards and/ or ID WSF.

http://www.openliberty.org/blog/2009/03/13/mydex-cic-bootstrapping-id-wsf-20-from-an-information-card/

I've still to write up the business side view of this, Asa has done the tech view.

The Mydex proof of concept is an early form of 'relationship manager' as Eve articulates the term.

We also have IGF/ Aris issues in our sites in order to look after the VRM data once it has left the individuals personal data store.

Iain

I didn't figure my idea was that radical (I've long believed that for all of the hype of infocards - they are really evolution of Web forms) and glad to see validation.

Good work Iain and Asa. Looking forward to seeing more of what y'all turn up.

OTN and Amazon are teeming up together to host a free "Virtual Developer Day". If you are interested in learning how to build cool, Web 2.0 style applications (aka Ajax) using Oracle technologies - you should look into this.

But I'm bringing this up because I really like the delivery mechanism - it's going to be taught via webcast using lab machines hosted on Amazon EC2 infrastructure.

This is a great example of what cloud computing can do that is very hard to do otherwise - scale up developer training without breaking the bank.

By using EC2 it's possible for the training team to build images that are pre-configured to run but because it's trivial (and cheap) to split the app server and database machines (though I'm not sure this is how they will do it) you don't have to have the constraints you might have in a normal lab (e.g. how much stuff can you cram into a single vm on a desktop pc). And unlike a traditional in-person class - you can keep your labs on the EC2 (though you do have to pay for the EC2 time, but again, that's cheap).

If you would be interested in training like this around identity management - please leave me a comment.