I recently got back from a short vacation with friends and family in Hawaii. Although I’ve been part of the LizardTech Engineering team for many years, I’m relatively new to the geospatial industry, so you’ll forgive me if wax a little misty here.

One of the things that made this trip particularly memorable for me was geocaching all over the Big Island.

My wife and I found one cache that brought us through an ancient petroglyph site. Others of us found a cache placed by a 6th grade class near a 100-foot seawall. There were others… all of them at sites conveniently close to us but uncrowded and beautiful. I’m grateful to the locals who posted these, becoming private tour guides and sharing their intimate knowledge of beautiful places close to their homes and their hearts.

I’m continually struck by how small a place the world has become. From our living room at home, we get online and look at the road from the airport to the rental home we’ll be staying in. The aerial image shows the roadway, the round-abouts, the beach, the pool and the rooftop over our suite. When we arrive in Hawaii – 2000 miles from home – we walk, for the first time, through a brush trail looking for a 10-inch package and find it as easily as (well, with no more frustration than) if it were a box of Christmas lights in the garage.

This is just the fun side of it, but “better living through geospatial technology” is a worthy goal and we at LizardTech are proud to pursue it.

GeoExpress 7 was released this past January, and despite the fact that it has met with enthusiastic approval by users and reviewers, it is a product made by humans, and humans on a tight schedule at that. One of the things we were dissatisfied with even as ‘Geo 7′ went out the door was that in one of the new features – publishing images directly to Express Server from GeoExpress – the application appears to hang.

It doesn’t really hang, but the progress meter reaches 100% and then nothing happens for a while.

What’s going on here is that once the images have been uploaded, Express Server then ports them to their specified target. That can take some time. In publishing really large images or lots of small ones, it can be a really long time.

The fix for the misalignment in the meter is not a fully known entity, but it looks like this is a server-side issue, something we’d need to address with a future release of Express Server. In the meantime we decided to add a message in Geo 7 to the effect that ‘Please wait…Express Server is still working’, but in the attempt to do this we found that it involved some incompatibilities with the way we present other messages. We could find no way to do the right thing in the time we had left.

We momentarily considered some inelegant hacks, but because this didn’t appear to be a dealbreaker, better sense prevailed. We’re tracking the issue and we wrote a knowledge base article about it, which basically says what I’ve told you here. That KB article will be updated when there’s more to tell.

For now, be aware that the extra time after the progress meter shows completion may be as much as 30% of the original upload time.

LizardTech isn’t in a position to be able to do the 20% Google thing, but over the past couple years here I have been fortunate to be able to spend a little of my time “giving something back” to the GIS community.

Back in February of 2006, I had the chance to spent a cold and windy day at an airport hotel in Chicago in a meeting with a couple dozen luminaries from the open source and activist wings of the geospatial world. The open source ecosystem for our industry was getting more attention and traction, and people were starting to form some sort of organization which could serve as a clearinghouse for these development efforts, promote the use of open source in industry, champion the use of free and open products in education, and so on. By the end of that day, the domain name “osgeo.com” had been registered and the Open Source Geospatial Foundation was born.

As a for-profit company, LizardTech and others have lots of reasons to support the open source movement. Esteemed coauthor Matt Fleagle and I wrote an article about this very topic last year.

This past year, OSGeo has seen the birth of a number of chapters, based around communities of regional interest or common language. The Cascadia Users of Geospatial Open Source (CUGOS) formed here in Seattle last spring, meeting every month here at our offices in Seattle – it’s nice to have a chance to meet other open source geo folks face-to-face for a change, instead of just trading emails and IRC messages.

We Lizards were also very fortunate this year to meet the up with the OSGeo tribes at the FOSS4G conference in Victoria. I and four other LizardTech engineers went up north for the week, and we all came home with a much better appreciation of the broad set of libraries and applications we can build on, as well as the people who work on them.

Speaking just as an individual, it’s been a great privilege to get to know and work with the people from across the globe that make up OSGeo – smart and principled people all, but with a refreshingly diverse set of backgrounds, skills, and passions.

For many years, LizardTech has relied on open source software as one of our strategies for more effective development of robust software. I’m proud that we’re able to associate ourselves with OSGeo, and we look forward to another year of collaboration and growth with the Foundation.

Optimizing is a trick we wish all users of GeoExpress knew about. If you have files in MrSID Generation 3 (MG3) format, and they were originally encoded as “optimizable” (it’s the default, so chances are good that they were), you can perform additional compression and a number of other operations on them without having to decode and reencode them.

Why is that useful? Not having to decode (i.e. bring the image out of wavelet space) has two advantages:

• Optimizing is much faster than encoding
• Optimize operations are free (if you’re using a data cartridge, they don’t charge against it)

Loading Images and Accessing Options
Among the six tabs on the Job list is one called Optimize (shown at right). You may have noticed this tab before and never felt like venturing into it. We urge you to gird up thy loins and use it. Images you want to optimize are staged on this tab.

After you load an image into the Optimize tab (and make sure it’s selected), you can access options in one of three ways:

• Right-click on the filename in the job list and choose Optimize Options… from the context menu.
• Choose Encode Options… from the Options menu.
• Click the More Optimize Options… button on the Properties tab.

In each case, a dialog box appears that has tabs for Input, Output, Optimize Settings, and Advanced Settings.

Basic optimization settings
On the Optimize Settings tab (shown below) you can specify additional compression by ratio or by target file size, and you can resample (reduce the dimensions) by factors of 2.

The options for compression based on encoding ratio and target file size are pretty self-explanatory. If you need to fit particular images on a CD, compress by target file size to make sure you stay under the limit. If you want your images to be uniformly compressed to a fifth of their current size, set a ratio of 20:1.

Resampling in GeoExpress means changing the resolution of an image. For example, if your image measures 1000 x 1000 and you resample by a factor of 2, the result will be a 500 x 500 image. Resampling is achieved by discarding zoom levels; thus, the number of zoom levels in the image (remember, MrSID is inherently multiresolutional) determines the factors available in the drop-down.

Advanced optimization settings
The Advanced Settings tab (shown below) gives you the following additional options, which are described further on:

• Frequency balance
• Sharpness
• One-pass or two-pass optimizer
• Optimizable
• Use temp file

Frequency Balance – This determines the emphasis given to edges and flat areas in color and grayscale images. Set frequency balance lower for images requiring precise edge definition. If precise edges are not as important as consistency in flat color areas, use a higher setting. Range of acceptable values is 0.0 – 10.0. Default is 2.

Sharpness – This determines the sharpness of boundaries between different areas of an image. Use a lower setting for images with large amounts of textured area (where color or intensity changes are occurring throughout a region, rather than just at a boundary). Use a higher setting for images with little textured area. Range of acceptable values is 0.0 – 1.0. Default is 0.0.

1-Pass or 2-Pass Optimizer – This option is only available for MrSID Generation 3 (MG3) images. The 1-pass and 2-pass optimizer parameters affect encode performance and memory usage. Although not as fast as the 1-pass optimizer, the 2-pass optimizer (used by default) requires much less memory because it splits the encode operation into two passes. This is required on very large files. Encoding with the 1-Pass Optimizer is faster but requires that the entire image be loaded into memory. If sufficient RAM is not available for the image selected, the encode job will fail.

The 2-Pass Optimizer is automatically used for Area of Interest encoding even if this checkbox has not been selected.

Note that the “Estimated Memory Usage” is only displayed when the 2-pass optimizer is used (look on the Properties tab or on the Advanced Settings tab of the Encode Options dialog box).

Optimizable – Select the Optimizable checkbox if you want the output MG3 file to be optimizable later on. By default this checkbox is selected.

Use Temp File – A temp file can be used to store necessary statistical data about the image data during processing. If your image is very large this can help alleviate the memory burden on a given job.

Let us know whether or not you find the optimization functionality useful and how you think it could be better.

Last November, together with two co-workers, laptops, projectors, tons of inflatable globes, CD holders, pencils and of course plastic lizards, I went to the Brighton School in Lynwood, WA, to share my enthusiasm for GIS Day with a class of 5th graders. To prepare for the hour-long presentation the teacher had been in contact with us about where the class was at in their current history lesson, the 13 colonies and Revolutionary War, and could we somehow incorporate that into our GIS lesson?

We began with a birdseye view of their school and how far away it was from Seattle’s Safeco Field. They loved it! I was amazed at how many of them quickly made the connection between ESRI technology and Google earth.

To incorporate their current Revolutionary War curriculum I gathered maps of the 13 colonies, Revolutionary War battle sites and other key geographic locations such as the Liberty Bell and Independence Hall and current aerial photos of Philadelphia. I digitized, or overlayed, the current aerial imagery with some of the old maps like the one shown above.

In class I used this to point out the extreme difficulties that individuals such as General George Washington experienced while creating maps during that era.

“The want of accurate maps of the Country which has hitherto been the Scene of War, has been a great disadvantage to me. I have in vain endeavored to procure them and have been obliged to make shift with such sketches as I could trace from my own Observations.”
- General George Washington

After our presentation the kids were really excited to show us some of the tools they had been using to learn about the Revolutionary War and the 13 colonies. One way was by playing Twister with a giant map of the 13 colonies that their teacher had made. They insisted that we play a few rounds. I never realized just how far New York is from Georgia.

I won…twice.

This was a great day with the kids. Many of them asked us to return next year and I have to say I’m looking forward to it. Check out the article about our trip to Brighton (PDF) in the American Congress on Surveying and Mapping’s online publication, ACSM Bulletin.