Recently, I have been experimenting with new ways to create collaborative "rooms" for use with the Podcast Hotel event in Portland next week.  The general goal is to create ad-hoc web-pages that provide an environment where people can share media captured at the event and make it easily (and instantly) available for anyone to mash-up or re-mix and upload again for others to play with and comment on.  Of course, being a videoblogger, I want these "rooms" to have rich media feeds that I can use in tools like FireAnt.

It is easy to imagine how an enterprise system for this might be designed.  However, part of the challenge was to build on existing tools and infrastructure in order to design and build something that could be used with any kind of web page, particularly blogs like Blogger, WordPress and TypePad.

Another challenge was the the solution could not assume an underlying storage system; blogs like Blogger do not provide storage for media except pictures and those that do (i.e., Typepad) are too limited in bandwidth and storage to support significant uploading and downloading of media files.

I also wanted these collaborative rooms to be simple to add to a web page.  For the moment, this means that it should be no more difficult to use for the author/publisher than inserting a Flickr or Feedburner badge.

I used SpinXpress as the basis of a solution because it is a p2p (person-to-person/peer-to-peer) file sharing tool that allows you to form "groups" (like rooms, or spaces, or networks) where you can store files, bookmarks and comments.  You can also access content in SpinXpress via web services (i.e., HTTP) and, most importantly to a geek like me, you can extend SpinXpress functionality via Java servlets.

So I did that.  I’ll leave the technical details to future posts, but basically I created servlets that add XML (XHTML, mRSS, SMIL, XSPF, etc) feeds, forms and handlers that allow one to add a "room" or "group" to a post where people can share files and collaborate.

I will demonstrate this in the very next post.  In the meantime, here is my first SnapZ ProX screencast about SpinXpress.  It just covers the most basic concepts, but should help with the conversation.  Click the pic to view the vid.

Disclaimer: I receive a small sponsorship from Outhink and have equity in the company.

When I was a young mathematician, I worked with a person who always wrote his own textbook - his way - I must admit that they were far superior to others in numerous ways (no pun intended).  This was because he was a leader in that area of mathematical research.

Later, as we both got into personal computers and programming, his teaching and textbooks began to cover ground that many had already trodden.  Again, it was necessary to do it "his way".  Here, the benefits were less obvious.  Often, these books added a certain small mathematical twist that was interesting to a few; but I often wondered about why there was a need to write yet another book on Pascal.

One day, my mentor announced to me that he was going to write a "C" book.  I was abhorred.  I loved my K&R (still do - Kernigahan and Ritchie for those who do not know).  There is no other introductory book on the C programming language that stacks up to it.  Those who did not care for it were invariably light-weight’s when it came to programming.  It’s a book that I can still enjoy and benefit from re-reading today, even though I no longer program in C (or C++ for that matter). 

I told my friend and teacher this and asked him to carefully study and master K&R before adding yet another C programming book to the countless others that do not even come close to igniting the passion for this archaic language, and always relevant style and words of wisdom, that many of us had. 

This "little book" is to C programming, what Strunk and White is to writing style.

Today, I am tasked with building new communities.  But I see countless ones already and many that need help in various ways.

So why build yet another? 

Why not take all of the resources and content planned for new community development and offer it to OurMedia, as an example.  How much could an extra paid programmer or two help?  Or a graphic designer? 

We can all help guest moderate more - they are always asking.  We can participate more. 

Why develop the infrastructure for proprietary web services, when they could be integrated into others automatically?

Why build so many new communities?  Why not join others and help make them richer?

I once set up an account on Ta-Da lists.  Created a simple grocery list and then stopped using the product.  It never took hold (yet, you never know). 

I think SpinXpress faces a similar dilemma.  Once you have it setup, you need something to make you use it and see it’s value or you will just forget about it.  We need to address this on SpinFlow, but that’s another website and another story.

Back to Ta-Da…

During that short trial of Ta-Da, I did subscribe to my list via RSS, however, and the feed is still being pulled today.  I was experimenting with "digital lifestyle aggregation", which is Marc Canter’s meme (personally I don’t think the name will ever take, same with "vlogging" - but the ideas are right on!)

Except that I notice that there are new items (not grocery) showing up from other lists (and their usernames).  I cannot view the associated data, as it is password protected, but I can see the entire item name and who the people are.

So be careful on your item titles if you use Ta-Da (or anything else, like secure blogs); they will often show up in strange places and some people can add 1+1, especially if they know you (e.g., a co-worker).

I wish I wasn’t so busy that I could fill out a report for Ta-Da, but I can’t debug everyone’s software and services and so I hope this post will make it onto their radar.

Sounds like a GUID problem to me.  It might be because the account lapsed due to inactivity and some ID number was reused?  Not so GU in the ID in that case.

No matter, I should not be seeing this personal data.

Delivering quality web services to people can be very challanging and 37 Signals has been breaking new ground with their web applications.  However, the pace of web development these days is such that testing often takes a back seat to development.

I think testing should drive development.

I posted an article on "Collaborative Video" at spinflow.org, please check it out; I’d really appreciate your feedback (or pushback!  I like that word.)

It covers a wide rage of group efforts centered around video in general and videoblogging in particular.

Here is clip from yesterday’s videoblogging video conference in which Nathan Peters is discussing music and video collaboration.

Just one of the many topics dicussed.