DailyNotes/2005/06/23 - Raymond Yee's Wiki

Notelets
Sakai, XOSIDS, and cross-language issues
(Not so early) reflections
Biotagging our plant pictures
Today's pictures
Yesterday's pictures
Open Threads

Notelets

I complained that I could not write <tt> in Manila but couldn't.

Sakai, XOSIDS, and cross-language issues

I have learned that the web services support in Sakai 2.0 is only an embryonic framework, not the working product for which I was hoping. The announcement of web services in Sakai 2.0 made me hope for something like Flickr Services, which has enabled interfacing to Flickr from Java, PHP, Python, Perl, Actionscript, Ruby, .NET, among other languages. Maybe that will be forthcoming soon.

On a related front, I've been wondering how the release of the XOSIDs will enable the development of "OKI-compliant" Python tools. Wilbert Kraan answers a few of my questions with his piece CETIS-OKI and IMS, wires and sockets revisited (emphasis mine):

Note that implementing an OSID in such a language doesn't mean that it will then be interoperable with OSID adapters that have been written in other languages. Language bridges such as the Mac world's widely used Objective C to Java bridge needs to be used for such a purpose.

Still, the XML based nature of the XOSIDs also opens the possibility of the creation of new XOSIDs as part of the new IMS specification workflow. In the latter, a new specification's datamodel and behaviour are specified in abstract UML first, and then transformed to the XML Schema definition and WSDL files that Web Service enabled systems use to interoperate. Deriving OKI style API descriptions (i.e. XOSIDs) at the same time should be perfectly possible, provided that the IMS spec fits with the existing OSIDs and how they are factored.

Which of course raises the older question of how OSIDs relate to Web Services in practice. Part of the answer will be demonstrated at alt-i-lab in Sheffield, were a fair number of search tools will be coupled to an even larger number of repositories via the repository OSID. The not-so-secret sauce that makes this possible is the fact that each of the relatively simple and innovative search tools only has to implement the one OSID, and that each repository needs to supply only one OSID adapter to the search tool.

I will write Wilbert for some followup to his last intriguing paragraph.

(Thanks to Scott Leslie for pointing to Wilbert's article.)

(Not so early) reflections

Yesterday, I wrote about how I want to begin my days with early-morning reflections. Because of the need to get out of the house early after a lot of sleep and an early morning meeting, I'm only now getting to my coveted reflection time. I have a strong preference to writing "in the open" whenver poosible, though clearly there are times for private meditation and writing. I do happen to have somethings to share with the world.

Today, I am pausing at the same spot I did yesterday with the question of what to write when there is so much to write. I will take the same strategy of writing a bit about my picture-taking since it's become an integral part of my daily life. I will also cut myself off without exhausting the topic.

One question I've been pondering is why I'm so into taking pictures and presenting them in Flickr. At the beginning, I was just playing around, experimenting with the features of Flickr. (There is, of course, a whole pre-history of my getting into digital photography in July 2000, a story for another time.) To experiment with image collections, I figured that I should have some of my own pictures with which to work. My early experiments with Flickr coincided with my getting a cameraphone. I started taking pictures with my Treo 600 on the premise that the cameraphone would be an easy way to create a collection of meaningful images, even if the images were not about grand themes but only about my own daily life. I did set out to create images along certain thematic lines (e.g., Milosz, garden plants around Berkeley, Chinese art.) But so many other things caught my attention that I essentially started "life logging/life blogging" (to echo the name of commercial products for managing a lot of mobile media.)

Along the way, my relationship with my images and my camera evolved. At first, I was largely content with the poor VGA level Treo 600 pictures from my camera phone, but once I started carrying around a 5 MP Pentax Optio S5i, I would use my camera phone as a last resort. With my 5 MP camera, I started to generate too much data to sustain a daily process (overloading the processor time, bandwidth, and disk storage I had for daily picture taking). Consequently, I have set my 5 MP camera to take VGA pictures by default! (I'd like to come back to the issue of how many pixels are enough for my purposes....) Still, at the current rate, I am generating more data than I can process satisfactorily (i.e., at the very least tag). I continue taking pictures, thinking of the process as gathering raw data for further processing recontextualization at some later time.

In midst of taking daily pictures, I have become a "daily media producer" (a concept I learned from Marc Davis' Garage Cinema Research Group.) For someone who never thought of working with images in any serious way, I'm astounded to learn how pictures have become a major part of my vocabulary for telling stories. I've had to take on technical challenges to produce images that begin to be along the lines that I want. (e.g., creating a rigorous time sequence for my images, calculating some amount of geo-referencing, writing my own uploading tools).

Now, is any of my picture taking relevant to research and teaching? I think so, and I will want to articulate reasons why some other time. Many researchers are working with collections of their own digital images and need to use and reuse them in various contexts. So I believe a lot of what I'm trying to accomplish with my own picture taking, sharing, and archiving will help us to provide useful services to researchers and teachers on this campus.