Re: [talk-ph] better ways to coordinate coastline mapping?

I agree, but can we keep the sawtooth webmap running for a few more weeks/months? Another thing, I am thinking is a script to report whether an island node is enclosed by a coastline way. A point-in-polygon in GIS-speak. This should help us identify and map the smaller islands. Coders and other

Re: [talk-ph] better ways to coordinate coastline mapping?

Hmmm... I guess we can declare this coastline correction task complete? There are still red spots in the webmap but I think these are all residual errors and I'm guessing there are no more SRTM-based coastlines remaining. I think all are now based on Landsat or PGS (which has small sawtooth coasts

Re: [talk-ph] better ways to coordinate coastline mapping?

In some cases, we tend to forget the coastline way direction rule "Left-side of the way direction should be land (or counterclockwise around the island) ". The garmin map compiler detects this and correct them accordingly in the local file. You can then correct them in osm by looking at my debugg

Re: [talk-ph] better ways to coordinate coastline mapping?

here's how to use the sawtooth webpage and josm http://vimeo.com/13546210 On Wed, Jul 21, 2010 at 9:48 PM, Eugene Alvin Villar wrote: > Nice! The Philippines looks a whole lot less "bloody" since I first put up > the webmap. Good work! > > http://forge.codedgraphic.com/osm/sawtooth_coastlines/ >

Re: [talk-ph] better ways to coordinate coastline mapping?

Nice! The Philippines looks a whole lot less "bloody" since I first put up the webmap. Good work! http://forge.codedgraphic.com/osm/sawtooth_coastlines/ On Sun, Jul 18, 2010 at 10:15 PM, maning sambale wrote: > FYI, since we started monitoring the sawtooths, we nuked around 25% of > all jagged

Re: [talk-ph] better ways to coordinate coastline mapping?

FYI, since we started monitoring the sawtooths, we nuked around 25% of all jagged coasties Less a thousand sawtooth segments to go. Keep it up guys. On Sun, Jul 18, 2010 at 11:26 PM, Eugene Alvin Villar wrote: > Thanks to maning's diary entry >

Re: [talk-ph] better ways to coordinate coastline mapping?

Thanks to maning's diary entry < http://www.openstreetmap.org/user/maning/diary/11240> we're making good progress on fixing those sawtooth coastlines. We are actually getting help from foreigners (many are Germans) in "cleaning" up our coasts. :-) On Wed, Jul 14, 2010 at 4:02 PM, maning sambale w

Re: [talk-ph] better ways to coordinate coastline mapping?

Eugene Alvin Villar wrote, On Thursday, 15 July, 2010 10:06 PM: > to generate > closed shapefiles that Mapnik can use. This tool is only run > infrequently and that is why coastlines do not update often in Mapnik. I see what you mean ...

Re: [talk-ph] better ways to coordinate coastline mapping?

Here's a little bit more elaboration on the problem of coastlines in Mapnik. This information can be found at the OSM Wiki: http://wiki.openstreetmap.org/wiki/Coastline Unlike every other feature in the OSM database, coastlines are rendered differently in Mapnik. The Mapnik rendering software need

Re: [talk-ph] better ways to coordinate coastline mapping?

On Thu, Jul 15, 2010 at 1:41 PM, Jim Morgan wrote: > maning sambale wrote, On Thursday, 15 July, 2010 12:01 PM: >> What I mean here is the mainland of Palawan.  But yes the smaller >> islands needs more work. > > Hmmm, now I'm beginning to think I didn't understand. I thought you were > saying th

Re: [talk-ph] better ways to coordinate coastline mapping?

maning sambale wrote, On Thursday, 15 July, 2010 12:01 PM: > What I mean here is the mainland of Palawan. But yes the smaller > islands needs more work. Hmmm, now I'm beginning to think I didn't understand. I thought you were saying that the coastline of Palawan (yes, the main island) was 99% fr

Re: [talk-ph] better ways to coordinate coastline mapping?

jim, What I mean here is the mainland of Palawan. But yes the smaller islands needs more work. On Thu, Jul 15, 2010 at 11:27 AM, Jim Morgan wrote: > maning sambale wrote, On Wednesday, 14 July, 2010 04:02 PM: >> Note that some have a 99% status already, but, it is good for other >> eyeballs to

Re: [talk-ph] better ways to coordinate coastline mapping?

maning sambale wrote, On Wednesday, 14 July, 2010 04:02 PM: > Note that some have a 99% status already, but, it is good for other > eyeballs to have a look and comment on the actual %age and status. I was surprised to see Palawan as 99% complete, so I whizzed over there for a quick look. I'd have

Re: [talk-ph] better ways to coordinate coastline mapping?

The top ten islands is almost complete (not in the mandelbrotian sense). I added a new list of coastline bounty in the wiki (11-30 largest islands) http://wiki.openstreetmap.org/wiki/WikiProject_Philippines/Coastline_Corrections#Priorities Please edit the status as you start working on each islan

Re: [talk-ph] better ways to coordinate coastline mapping?

On Tue, Jul 6, 2010 at 2:42 PM, Jim Morgan wrote: > Eugene Alvin Villar wrote, On Tuesday, 06 July, 2010 11:57 AM: > > To clarify things, the sawtooth detection script is quite naive. It > > simply detects if there are at three or more series of nodes where each > > pair of adjacent nodes have th

Re: [talk-ph] better ways to coordinate coastline mapping?

Hehe ... nope I am not trying to defeat the script, but am trying to stop it from misdetecting an already cleaned up coastal area. I have cleaned up the whole of Marinduque a couple of weeks ago and curiously saw the script reporting a sawtoothed section on the northern coastline. I checked it ou

Re: [talk-ph] better ways to coordinate coastline mapping?

Eugene Alvin Villar wrote, On Tuesday, 06 July, 2010 11:57 AM: > To clarify things, the sawtooth detection script is quite naive. It > simply detects if there are at three or more series of nodes where each > pair of adjacent nodes have the same latitude or longitude. This will > also detect any th

Re: [talk-ph] better ways to coordinate coastline mapping?

Just to add, not everything "sawtoothed" is wrong, I've seen piers and reclamation areas with a similar geometry. The sawtooth webmap simply shows you areas of improvement. On Tue, Jul 6, 2010 at 11:57 AM, Eugene Alvin Villar wrote: > I'm working with maning to devise a way to have the map update

Re: [talk-ph] better ways to coordinate coastline mapping?

Anyway, it seems that Negros is actually the "cleanest" large island in terms of coastlines. :-) On Tue, Jul 6, 2010 at 11:57 AM, Eugene Alvin Villar wrote: > To clarify things, the sawtooth detection script is quite naive. It simply > detects if there are at three or more series of nodes where

Re: [talk-ph] better ways to coordinate coastline mapping?

To clarify things, the sawtooth detection script is quite naive. It simply detects if there are at three or more series of nodes where each pair of adjacent nodes have the same latitude or longitude. This will also detect any three linear nodes that all have the same latitude or longitude like this

Re: [talk-ph] better ways to coordinate coastline mapping?

Ah I get it ... So it detects node strings that form a "step" (a horizontal and a vertical combination). It may be good to make mappers aware that such a combination will trigger sawtooth detection by Eugene's script. For there are some coastline contours that are indeed on a step shape. Good wa

Re: [talk-ph] better ways to coordinate coastline mapping?

@ ed, The technique is outlined here: http://wiki.openstreetmap.org/wiki/Talk:WikiProject_Philippines/Coastline_Corrections Couple of suggestions: 1. short link to the coastline correction wikipage 2. enable link to editors only at higher zoom level (perhaps 15-17?) 3. should be an osm-ph image of

Re: [talk-ph] better ways to coordinate coastline mapping?

Wow! amazing ... how do you detect "sawtoothness"? can you teach this at the skillshare? OSM genius ka talaga! Anyway, I take note of the sawtooth patterns between Lucena and Pagbilao Quezon ... I have corrected this already but the sawtooth appears only on higher zoom-out levels now. Does this

Re: [talk-ph] better ways to coordinate coastline mapping?

HI guys, Here's an *initial* webmap showing where the jagged coastlines are as of July 3: http://forge.codedgraphic.com/osm/sawtooth_coastlines/ The larger the circle, the more jagged steps there are to clean up. Sorry, there's no link going to Potlatch or JOSM/Merkaartor. I'll add that up when

Re: [talk-ph] better ways to coordinate coastline mapping?

We missed the June target to to finish the 10 largest islands. http://wiki.openstreetmap.org/wiki/WikiProject_Philippines/Coastline_Corrections#Priorities That's OK. All we need is to finish mainland Mindanao and we can set our radar to the next 20 on the list. We are also working on some webmap

Re: [talk-ph] better ways to coordinate coastline mapping?

I can load it in my public dropbox. On Sun, Jul 4, 2010 at 2:23 PM, Eugene Alvin Villar wrote: > Hi guys, > > Once we are done with the large islands, the next step is to do the smaller > islands. Ian Haylock has helpfully provided the PGS coastlines[1] for the > Philippines last year[2] which he

Re: [talk-ph] better ways to coordinate coastline mapping?

Oh BTW, I don't think we should do this automatically. Human input is still needed just to quickly verify that the PGS coastline does indeed match the Landsat imagery. On Sun, Jul 4, 2010 at 2:23 PM, Eugene Alvin Villar wrote: > Hi guys, > > Once we are done with the large islands, the next step

Re: [talk-ph] better ways to coordinate coastline mapping?

Hi guys, Once we are done with the large islands, the next step is to do the smaller islands. Ian Haylock has helpfully provided the PGS coastlines[1] for the Philippines last year[2] which he edited to combine ways. PGS itself was automatically generated by a software by analyzing the Landsat ima

Re: [talk-ph] better ways to coordinate coastline mapping?

Hi, I'll start with the south east of Negros Totor --- On Thu, 6/17/10, maning sambale wrote: > From: maning sambale > Subject: Re: [talk-ph] better ways to coordinate coastline mapping? > To: "Ed Garcia" > Cc: "osm-ph" > Date: Thursday, June 17, 2

Re: [talk-ph] better ways to coordinate coastline mapping?

UPDATE. Only two islands from the top ten need some coasty coastline love. http://wiki.openstreetmap.org/wiki/WikiProject_Philippines/Coastline_Corrections Anyone working on Negros and Mainland Mindanao? On Fri, Jun 4, 2010 at 9:49 PM, maning sambale wrote: > yes, please do add those roads > > m

Re: [talk-ph] better ways to coordinate coastline mapping?

yes, please do add those roads make sure to add source:Landsat On Fri, Jun 4, 2010 at 7:44 PM, Ed Garcia wrote: > hehehe thanks!  just too careful not to mess things up. > >  btw, while I was doing the edits, I can see what looks like main roads on > the landsat image ... can I assume they are i

Re: [talk-ph] better ways to coordinate coastline mapping?

hehehe thanks! just too careful not to mess things up. btw, while I was doing the edits, I can see what looks like main roads on the landsat image ... can I assume they are indeed roads showing on landsat? On Fri, Jun 4, 2010 at 7:26 PM, maning sambale wrote: > Relax. > > Those are coastline

Re: [talk-ph] better ways to coordinate coastline mapping?

Relax. Those are coastline nodes with tags. The tags are: created_by:srtm_coastline natural:coastline note: Original way #418062 source:SRTM natural:coastline should be given tags in ways and not nodes. You can delete them. On Fri, Jun 4, 2010 at 7:19 PM, Ed Garcia wrote: > > Help!   as

Re: [talk-ph] better ways to coordinate coastline mapping?

Help! as I was going around Mindoro doing coastline edits, I encountered unfamiliar "nodes" just about south of Mamburao. The nodes appear as big white squares in JOSM, different from the usual small yellow square nodes. What are they? Did not want to touch them as they are not the usual nodes

Re: [talk-ph] better ways to coordinate coastline mapping?

On Fri, Jun 4, 2010 at 10:47 AM, Jim Morgan wrote: > maning sambale wrote, On Thursday, 03 June, 2010 06:22 PM: >> FYI, I also finished Bohol and Panay.  I am starting Samar Island >> (offline for the moment in order not to break anything). > > Hah. I went to finish off Bohol myself a few days ago

Re: [talk-ph] better ways to coordinate coastline mapping?

Excellent! I updated the wiki page to indicate your progress. http://wiki.openstreetmap.org/wiki/WikiProject_Philippines/Coastline_Corrections#Priorities FYI, I also finished Bohol and Panay. I am starting Samar Island (offline for the moment in order not to break anything). The hard ones are M

Re: [talk-ph] better ways to coordinate coastline mapping?

I have just finished the entire eastern coastline of Mindoro. Wondering when they will appear at mapnik? Wasn't there a way to request for re-rendering? Or was that for Osmarender? anyway, will work on the western part naman this weekend cheers On Tue, May 25, 2010 at 5:25 PM, Ed Garcia wrot

Re: [talk-ph] better ways to coordinate coastline mapping?

This is via JOSM/WMS Landsat, right? Just the other week, I have done some Agusan coastlines near Cabadbaran City. Ok, I will take Mindoro now ... cheers On Tue, May 25, 2010 at 5:10 PM, maning sambale wrote: > > The first proposed priority is to focus coastline mapping to the 10 > > largest

Re: [talk-ph] better ways to coordinate coastline mapping?

> The first proposed priority is to focus coastline mapping to the 10 > largest islands and then move on to the next. > http://wiki.openstreetmap.org/wiki/WikiProject_Philippines/Coastline_Corrections#Priorities I expected lower figures, but it seems we have great coverage already for the top 10 l