z-dude's Diary

Recent diary entries

added a bunch of villages to Mali, then noticed that someone else does an edit which exactly copies a node, resulting in duplicate nodes. if you look at the area around this node, the 3 villages to the south and 2 villages to the north overlap my old village nodes.

is this some editor bug where software is deleting other people’s nodes, and rewriting those nodes with the exact same info? Is this a caused by an edit bot?

Read the BBC article where cellphone data was used to see how bus routes could be made more effective… noticed a osm map in the article.

There should be a tag for lakes which have wet and dry seasons, such as the intermittent tag.

Location: Tin Aicha, Goundam Cercle, Timbuktu, Mali

China hates mappers.

Posted by z-dude on 14 March 2013 in English (English).

“The perils of doing business in China were highlighted again recently after it was revealed that Coca Cola is under investigation following allegations it illegally mapped parts of Yunnan province.”

“China is also planning to increase fines for providers which fail to include its “full territory” when compiling maps, as sometimes happens when cartographers decide disputed lands such as the Diaoyu Islands don’t belong in Chinese territory.”

Bing Maps updated.

Posted by z-dude on 1 March 2013 in English (English).

Just saw this on Engadget.

I took a quick look at an African area which had no Bing coverage, and now there is coverage.

I think this is good.

Thought I’d share the link to this Google Tech talk.

It’s more about using imagery to get data about a site, like the movement of glaciers, ice thickness, the surface of the moon.

Deja Vu

Posted by z-dude on 30 December 2012 in English (English).

sometimes I have the feeling that I edited a place before, but when I look back on it, the stuff I added disappears, so I have to re-add it again.

Location: Machias Seal Island, Washington County, Maine, United States

Marine Boundaries need some work.

Posted by z-dude on 18 December 2012 in English (English).

The current marine boundary in OSM uses bot generated 12nm circles which does not always correctly show the proper international boundary which is reached by an international treaty.

The international Boundary Commission lists the actual marine boundary between Canada and the US. The boundary by Maine needs to be updated.

Also, the Canada / St. Pierre Miquelon boundary needs to be updated from the bot generated lines.

The US State department has a good list of resources for marine boundaries all over the world.

Looking at the current map, some marine boundaries show the 200 mile Exclusive Economic zone, while most of the world shows the robot generated 12 nm line. Link to google book.

An early surveyor surveyed a 110 mile line to an accuracy of 16.8 feet.

edit, tried new link to book.

Location: Victoria, Australia

Someone expects Canadians to sail their boats into the rocks.

It appears that the Alaska / BC border has been modified. In this area, the US side of the border takes up the whole channel. I think the border may be inaccurate between US and Canada.

Screenshot of the border area.

Here’s the Canvec rail lines in Smithers BC.

I really appreciate that people are importing canvec data, and appreciate the streams and lakes, but there was already a nice set of rail lines in northern BC made by ipodolsk and that data has been overwritten by what appears to be less accurate canvec data as the canvec rail lines have exaggerated shapes.

In Northern BC, other keepright errors that I fixed seem to have Canvec and Geobase imported data overlapping each other. In most of the cases I have seen, Canvec and Geobase had identical nodes, so we didn’t gain much by re-importing roads.

I’m sure that importing data from Canvec and Geobase is a huge task, but I think that in a lot of cases, we don’t need to delete old road, or old rail lines to re-import more Geogratis data from the Government of Canada.

Best of luck.

Location: Smithers, Regional District of Bulkley-Nechako, British Columbia, Canada

I think I spotted a whale.

Posted by z-dude on 28 September 2012 in English (English).

the old bike route through the myra canyon doesn’t show up on opencyclemap any more.

This will be a bit of work. At the least, I’d like to re-add the Myra Canyon area back to this relation.

The onramps and offramps will be changing a lot over the next few months.

On Youtube, I saw a Hak5 interview with a person who uses balloons to map events using a high altitude balloon - interview starts around 2 minutes. He mentions how to use balloons to map, and what the restrictions are.

Referenced links in the article

Is this a redaction bug?

Posted by z-dude on 25 July 2012 in English (English).

This way was hit by the Redaction bot, but only myself and Mbiker’s imports user touched this way.

in the potlatch editor, Mbiker’s import account worked on rev 2, and I worked on rev3 and rev4. In the way history, revs 2,3,and 4 are redacted.

What gives?

I always find these edits fun to see. Typically a node gets moved a few feet.

With my Garmin Oregon, I typically get a very repeatable track with a clear sky. It’s when you get 200 foot tall giant trees that tracks differ by more than 10 feet.

I do have a question: with the Glasnos GPS systems, is the accuracy better? I hear that it’s supposed to be better in the areas where you ave tall trees and canyons. Has anyone tried mapping with the new Garmin Glonass devices yet?

Location: Suncrest, Burnaby, Metro Vancouver Regional District, British Columbia, V5J 1Z3, Canada

Will a Garmin route cars along tracks and dirt roads by default? Is it an issue with compiling the map to garmin format?
Should we tag tracks as 'motorcar=no' if the track is for jeeps and 4 wheel drive vehicles?

For summertime use, the whistler cross country area is used for walking, biking, and horseback riding. Does adding a 'highway = track' tag to the ways in this area mess up the piste routing and rendering?