That is right! I keep thinking of you as a Calgarian. It turns out that there is also going to be an FME user meeting in the UK too and I will be there. This one in Southampton on April 29th. I know that is out of the way for you but if you were there it would be great. I will also definitely send you a link for a GoToMeeting when I return into the office in April.

It would be great to see you in one of the Canadian cities if possible. We will be in Vancouver, Calgary, and Toronto. If none are possible I would really like to connect via GoToMeeting after the Odyssey is over.

Here are two links that explain the NSGIC Twitter activity demo that is being used.

Here is the page that shows to the page for NSGIC-Twitter activity . This is a fun one that uses the ability of FME to pull data from web services. In this case we are searching Twitter and then plotting the location of where the tweets are made if geo-tagging is enabled. If geo-tagging is not enabled we plot that location.

Here is also a link to the PowerPoint deck from Learon Dalby that briefly explains what is going on.
http://www.slideshare.net/learondalby/safe-twitter-demo

I am now going to read your blog.

What was Dale’s link (I think) to a show, which you’re posting geotagged tweets to until 31 March, using your FME engine to post on Gmap?

I would love to show you some examples. Here is a synopsis of one thing that we are doing:

a) Installed a CSW such as con terra’s CSW or DeeGree or GeoNetwork.

b) Created a workspace and published to server. This enables any dataset in a format that FME supports to be published to the server and populate the dataset’s metadata to the CSW.

c) Created another workspace that enables a user to find all of the data in the catalogue for a bounding box and also get the bounding box of the dataset from the CSW. The entire CSW record is extracted so this query could easily be restricted to any information that is in the CSW.

d) Wrote a workspace that makes it easy to publish a large number of datasets to this data server while also populating the CSW.

e) Created a KML Network Link workspace so that a user can navigate using Google Earth (as an example) and dynamically see the bounding boxes of the dataset.

f) User is then able to simply click on a dataset and can either download it in its original format or perform a more complex data extraction. Again this is also defined by a workspace.

If you or anyone else wants to see a demo of this example then I encourage you to go to one of the events in North America or Europe (Dale blogged about these events last week). If you can’t make it, then I would be happy to show it to you then provide you with all of the demos, etc.

Is this what you are looking for?

Great point! Metadata is indeed important and must be part of any data repository. I didn’t make the point in this post, but should have stated that the first thing that needs to be done is to verify that the data sources are in fact of sufficient recency and quality to meet the needs. FME can read the metadata and a user could in fact verify through the metadata that the data being used does meet the requirements of the task. Right now this is all in the early days and we are seeing metadata being created and extracted but it is being primarily viewed by users rather than software. Pulling different datasets together with no regard for the metadata is not a good idea. We do have some examples of FME using metadata to do things based on the metadata. As you suggested, one of the problems is getting users in the habit of creating the metadata. I am confident that this will change as more and more data is used across organizations.