Ported over to WordPress
New URL is http://fantomplanet.wordpress.com. The feed is the same as it still should be http://feeds.feedburner.com/FantomPlanet.
FANTOM PLANET is "some random geographer's" opinionated output about all things geospatial, geotechnical, and whatever else on location based information and services.
New URL is http://fantomplanet.wordpress.com. The feed is the same as it still should be http://feeds.feedburner.com/FantomPlanet.
James' post tonight won't be the last comment about how ESRI users and developers see their server-based products. It just seems that over the past few weeks more and more people, especially developers friendly to ESRI, have seriously complained about ArcIMS, ArcSDE, and ArcGIS Server and how they're all "too little too late." In the case of ArcIMS, it's a matter of the services and the lack of OGC compliance of it. It's not advanced enough for the people who I talk to. Granted, they haven't looked at ArcGIS Server, but still, there are some OGC specs that would be nice to have now until later. Even though WMS and KML are available, there's something missing that goes with those outputs. For ArcSDE, a number of people have told me. If you can afford Oracle Spatial, buy it, skip SDE. If you can't afford Oracle Spatial, PostgreSQL with PostGIS. Just stay as far away from SDE if you transactional times that won't cause you to loose 80% of your day. Finally, there's ArcGIS Server. This product gets no love because it's the red-headed stepchild of the bunch. Not enough people are in need of it, thus a lack of understanding about it. If no one is implementing it, then there's not enough visibility, and again it could be something worthwhile, but no community equals no success or use. ArcGIS Explorer could save it, then again, you have to be building tasks, globes, and services with it. If no one is, then it's DOA. Sorry, Jeremey. I mean, I feel bad writing this. I like ESRI. Desktop is great—except for my export map problem on my laptop. Yet, this state of affairs for ESRI server apps are really sad. I just hope they can either 1) improve, or, 2) get out without betting the farm on server apps.
Tomorrow the city of Bangalore (ಬೆಂಗಳೂರು), India, will be changing its name to Bengaluru, the city's name in Kannada, the local language. Now I have go buy a new map. Shoot!
Chris notes his thoughts today about the WMS-C tiling spec that's currently being worked on through OSGeo. He proposes a fairly cheap solution by using Amazon S3 to host the tiles with all it's capabilities. So, Chris—with Schyuler's notes—broke down the cost of caching the Landsat 7 dataset with S3.
"Sure its not free, but it sure as hell is cheap … Based on Schuylers calculation from the wiki article,Taking 15m pan-sharpened Landsat-7 composites as an example, at a tile size of 512 x 512 pixels, each tile would be about 7,680 meters on a side, or about .0625 degrees across. Plugging in the other values, we get a maximum of 22,118,400 tiles in the layer.
"Assuming the optimum size of 64kb is reached per tile, we’re looking at 1415.577gb of physical storage. Lets take a wild guess of 50gb of transfer per month, with the actual tiles only be updated annually and we have the following,
$0.20 * 1415 = $283 to initially upload the cache
$0.15 * 1415 * 12 = $2547 for a years worth of storage
$0.20 * 50 * 12 = $120 of user transfers (eg. downloads)
Total: $2950"
Chris goes on to make a hint that Refractions perhaps get in on sponsoring this. I commented on Chris' blog that "why doesn't USGS or NASA just pay for it?" It's a drop in the bucket for them.
A friend of mine at NYU's Interactive Telecommunications Program sent me the link last night to Outside.in. Another pretty cool collaborative atlas that is intended to link and build communities through the web. So, what is Outside.in? Author and site founder, Steven Johnson, explains in his blog:
"So what is outside.in? In a phrase, it's an attempt to collectively build the geographic Web, neighborhood by neighborhood. I wrote up a mini-essay describing the original inspiration for the site, and explaining some of our core principles, which I've included below. But you can also just go visit the site and explore..."Some of those principles are (which sound like they came from me):
I just uploaded Firefox 2.o like everyone said I should. So, I go to Google Maps. Looks good. Yahoo Local. Looks good. Local.Live.com. . . WTF? Actually, Local.Live works fine. The toolbars and drop downs seem a little off though. Note to Redmond, check it out. Getting at what I wanted to note about the new features of 2.0 that I though would be of use to 'slippy maps' and neogeographers. Shoot me dead, now. I used buzzwords! They were:
Client-side session and persistent storage: New support for storing structured data on the client side, to enable better handling of online transactions and improved performance when dealing with large amounts of data, such as documents and mailboxes. This is based on the WHATWG specification for client-side session and persistent storage.
SVG text: Support for the svg:textpath specification enables SVG text to follow a curve or shape.
It's nice to know people who do cool stuff. I was talking to Chris Schmidt via Meebo today when he sent me the URL to a Metacarta Labs project he has been working on. It's called PageMapper, and it's something that's not up on the Lab's page yet. It's pretty cool. It's a bookmarklet that when you click on it—once you've dragged it to your Bookmark toolbar—it will locate the places in the webpage onto an OpenLayers map. Being privvy to the experiment, I decided to try it out on three unsuspecting websites that I thought would be great tests for PageMapper. 1) USCHO.com's team page for the University of North Dakota hockey team. There just has to be stuff in here that can be mapped?
2) The football team page for the University of California, Berkeley. Another team schedule that should be great for place names. 3) Finally, WikiTravel.org's main page. WikiTravel and geographic names, duh. As you can see. . . it's still a lab project. Chris tells me there's something about websites with cookies that causes PageMapper not to read everything. Still, an f'in cool lab project. Check it out. It does have some seriously potential for those interested in geolocating news or geographically searching for events and how they relate to one another.I replied, "We need to complain about 9.2 before we give you input for 9.3."
Just another ignorant question fueled by the existence of Datum Shift:
"How do I migrate from ArcIMS to MapServer or to Ka-Map? Is there a utility to "easily" convert AXL to MapScript?"Seriously, though, I had a bad day with ArcGIS. It may have been my three year haitus from GIS? It may have been gremlins? But whatever it was yesterday, it sucked. I tried to export a project to PDF, GeoPDF, and even AI for printing and all I got was a self-closing version of ArcGIS. (Steve had a similar issue yesterday too.) This all happened late yesterday and my ESRI rep was able to recommend something that required a ScanDisk and a defragment. So, I put the wheels in motion last night before I left the office. Hopefully, this morning what I did worked. Oh! The other thing that confused the hell out of me and some colleagues was: measuring the area of GRID rasters in ArcMap. I know that I need to use the Zonal Statistics tool, but I needed to convert the Grids to something else. Anyway, long story short, a friend hacked it out for me and got my area measurement. Note to ESRI for 9.3 & 10.0 devleopment: The above—Not as easy as it should be.