Challenges of Street Maps

So I was talking to my brother Al today, who is one of the smarter people I have met, and he clearly thought I was too harsh on openstreetmap.org.

I really had no intention to sound too negative. I think that they are one the forefront of inovation. They are applying wiki ideas to geospacial data... that is huge. My point was only that the data they collect is very sensitive, and its very important that it all be right, not only geographically, but also in every minuet detail. If a contributer forgets to mention that a street is one way, or that it doesn't have a sidewalk, etc, a program would not be able to properly use the data, and it would get a bum rep.

I think that the consumer of trail data will initially just be humans. And humans are really flexible in their interpretation of the data they are given. Also, trail data is not as complex, nor is it as important to have 100% accurate. If it's a little off, it won't make a big deal.

Thanks for your input Al, the more I have to think about the idea, and the more we talk about it, the more likely I will be to follow through with it.

alwold: so i started looking at open street map
alwold: and i think it's really cool
doctor wold: yeah... it's real cool
doctor wold: did you read the nerd blog I posted?
alwold: yeah
doctor wold: ahh
alwold: you were talking shit about it
doctor wold: cool
alwold: so thats why i looked at it
doctor wold: nah, not shit
doctor wold: it's just that routable data is so hard
doctor wold: like having people contribute it correctly would be incredibly difficult
doctor wold: and trail data isn't so complex
alwold: right thats why wikipedia is so inaccurate
alwold: cause everyone is too dumb
doctor wold: no
doctor wold: words are easy
doctor wold: people know the right information
doctor wold: but its easy to put it in words
doctor wold: it's harder to turn knowledge of roads into data that is accurate enough to actually use in a program
doctor wold: I think that for prited maps, open street map is great
doctor wold: and could be used very soon
doctor wold: but for routable data, it's very sensitive
alwold: so like you're saying that this would be easy to write
alwold: http://en.wikipedia.org/wiki/Antibody
doctor wold: if you know the information then yes
doctor wold: and the information is being used by humans, which are very flexible
doctor wold: with map data it is being consumed by programs
doctor wold: which are not so smart, and flexible with input
doctor wold: anyway... you're opinion is good
doctor wold: and I think that it will be useful at somepoint soon
doctor wold: if they would have started with the tiger data, it would have maybe been useful from the start too
doctor wold: idk

  in HTML

The world is a funny place, full of funny people, who do funny things.

I remember at one point in my programming career, I used to worry more about how to get a task done, given what I knew at that point. Now I worry more about knowing the correct way to complete a task.

Here is a funny example of someone trying to get a task done given what they currently know, instead of figuring out the right way to do it:

samsonasu:
<td valign="top">&nbsp;&nbsp;&nbsp;&nbsp;
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
&nbsp;&nbsp;</td>
samsonasu: thats such shitty code

Though one "&nbsp;" is often good in HTML, using them to achieve your desired spacing is HORRIBLY GHETTO :-p. If you do this, and want to know a better alternative, try either using a spacer image (a 1x1 transparent gif), with a width and height, or better yet using the padding-left or margin-left css attributes.

Thanks for this one Samson :-p And Happy HTML'n to the rest of you.

OpenTrailMap.org

In one of my more recent impulses, I purchased a domain, opentrailmap.org/com, with the intent to start a project where people can contribute GIS data, and try to produce a map of the trails around the world.

I'm hoping that I will be able to leverage off of the technologies used by openstreetmap.org, and that it will mostly be a matter of figuring out what to leverage off of, and not so much how to implement it myself.

My theory is that openstreetmap.org, though very cool, is years away from being useful, and even farther away from being better than the data that people pay for. You may ask, "If you think it's so far out, why do the same thing for trails"? There are many compelling reasons:

• No Good Alternative. There is no good alternative for trail maps. Topo maps are from the 80's and trails are always shifting because of fires/floods/development/etc.
  
• Passion. Outdoor enthusiasts have passion. They care about the trails, and they are very into it.
  
• Hikers have GPS. Hikers are already equipped with the GPS necessary to collect and store the trail data need, and they are already accustom to using it, in case they get lost.
  
• Looser Requirements. Unlike street maps, there isn't going to be a turn by turn direction application that is trying to find you the best route based off this data. This allows the data to be more sparse, and less detailed (no one way steets, no HOV lane, etc)
  

These factors make this a project that can start off useful. From day 1 it would be better than what exists today right from. There is no need to collect all the data for it to work either. The idea would be to start local, and spread out. Work with local groups that need the data, and see if it takes off.

In addition to the collection of data, the site would have to have a way to view and print the maps, and I would probably want to partner with a Topo map printer, and allow users to print the trails they want over a high quality topo map too.

As far as the data is concerned, I would not want to hoard this data. I'd be interested in providing it to anyone who wanted it, at least for their own use. I'm not sure yet how I would fund this concept, but one initial idea is to find funding through people like the city of phoenix parks. I know that they make maps, and I wonder how much work it is. Having online maps and having users keeping those maps up to date may be very appealing to them, or may not be :).

As I come up with more on this idea, I will continue to post under the label : opentrailmap.org. Comment if you have any ideas/input.

m500 Crashes Loading Large Images

the Samsung m500 throws a java.lang.IllegalArgumentException when you try to load a large image using the Image.createImage method.

In the program I'm using, it tries to download an Image that is twice the width, and twice the height of screen. It downloads it through a socket, and tries to load it via a byte array. If you reduce the size to like 1.5 x the width and height, it loads... This issue is killing me because this exception just doesn't make any sense :)

Has anyone else seen this issue?