Wednesday, 14 December 2011

Ducking the wrecking ball

The long saga that is the OSM licence change process is picking up pace again. The plan now is to move to the new licence on April 1st next year. That will mean that anything in the database needs to have been added by someone who actively accepted the new licence and contributor terms. So anything else will be archived and removed from the live database.

Declaring the date has been like a light to moths for the few but vociferous deniers. Spreading FUD and generating bonkers conclusions from scraps of truth continues to be their main weapon against the will of the majority. Tricks like calling the change "the deletion" would be worthy of tabloid journalists. One tactic seems to be neither accepting nor declining the new licence to continue to spread uncertainty. I respect people's choices, I deplore the underhand tactics. The LWG continues to hold the high ground of not responding to the provocation.

I have been reviewing the area I'm most interested in, so that by next April there will be no problem areas here. The first weapon is to simply get as many mappers to accept the new licence and CT. So far everyone I have contacted have happily agreed. I have been as neutral as possible in contacting people, not wanting to agree is very much their choice and I respect that. The problem remains that a few people are not responding to messages sent to them. The main reason I suspect is that the email address they registered with is no longer valid. They may have used a work email address and no longer work there for example. I have tried to find people with Google+ accounts, Facebook and Twitter. I've searched for people by name in an area with Google but so far a few stubbornly remain as unresponsive.

There are a few objects created by people who have declined the new licence. These are easy to deal with: simply remap them. I risk losing details added later by acceptors, but having reviewed the bulk of the stuff, the loss is minimal. Indeed, gathering what I normally gather sometimes adds more than is there now, such as speed limits, weight restrictions, bridge heights, lit and so on. I'm leaving some stuff added by people I haven't been able to contact yet, in case they accept. I expect that there will be some places to replace, but I don't mind that. Some of the footpaths will take longer to replace and some will be easier with the better weather in the spring and summer.

I will be remapping stuff created by the FUD mongers too. If they accept at the last minute (which I doubt) there will be nothing of their work left to see, which will be a shame. I was out today gathering some traces and photos (in the gathering gloom) to start that process. I will be happy and a bit smug when I see an email on the lists that is clearly trying to be a wrecking ball and I know that my area is free from such problems.


Craig Loftus said...

> I have been reviewing the area I'm most
> interested in, so that by next April
> there will be no problem areas here.

What is the best way to make this comparison? Is there a particularly good/up-to-date visualisation?

Chris Hill said...

I use the tool provided by geofabrik you can see here.

gom1 said...

The trolls are certainly frustrating, and the impression they leave for newbies is a worry. But the patient, sensible and constructive response of the bulk of the OSM community is encouragement enough.

vdp said...

It's great if your area is relatively clean (mine is too; we're even hoping to have the country 100% clean for april), but not all areas are that lucky.

Some countries risk losing a lot of data, often due to disagreeing top-commiters. That just means we've still got a lot of contacting and remapping work to do, I guess :/

Tom Chance said...

Chris, thanks for this.

On re-mapping, is there any guidance? I recently saw a reference to an email with a complicated flow diagram that completely confused me. so two questions:

1. Is it just a matter of identifying objects that were originally created by somebody who hasn't acepted the CTs, or any objects that have been modified by them?

2. Do you then delete and re-create that object from a ground survey / other source to "decontaminate"?

Chris Hill said...

Still some discussion going one about that. The area I map has a low number of mappers but there are still problems. My suggestions below are my own ideas and really need confirming with the LWG.

If a decliner created an object with no further edits then delete it and recreate it, preferably from a survey. Don't take anything from the original object. This is rare. Usually there have been edits by other people. If the creator is a decliner I don't see how you have any choice but to delete and recreate the object as above, but you lose the history and tags from others in the process.

If you have an object created by an acceptor and changed by a decliner then reverting to the version that is accepted is the thing to do I think. Again you risk losing the changes added by other acceptors after the decliner's change. I think the LWG-driven process will do this in April, but I suggest that a more individual, managed approach by locals might recover more legitimate data.

My primary suggestion is get people to accept. Find them and message them. The more acceptors the easier it is. In parallel you can examine an area and deal with the decliners in it, starting with any easy stuff. Maybe if an acceptor's work will be lost, message them, explain what is going on and see if they want to add their changes again (I know that is a painful message, but we will possibly create a tighter community from it).

I think we should share ideas and results too. Maybe email talk-gb with good / bad experiences.

Sorry if I've just stated the obvious.

Robert said...

Well said. Voices of sanity are all too often drowned out.

Wankmann said...


I used the Geofabrik inspector tool, too.

First, it scared the hell out of me when I first saw my area in northern Bavaria, Germany. The area is very well mapped (all buildings with addresses, all marked hiking routes, etc). I notice that most of the data is OK, but the trainted notes and ways are the most critical once on the map. About 40% of all counterpoints on the highways were marked. 100% of the relations were at risk.
I have spent several hours remapping the critical notes and ways of my area. Thanks to good BING maps nowadays and my local knowledge it was not a difficult task but a very time consuming task. I do not mind if some data gets lost due to the licences change, but from my experience I can tell that the data we may be losing is the most important we have.

If you like to remap some of your local data, you may pick any relations and fix the notes and ways in it. The will ensure you save the most critical data for our map.

Happy mapping


Gregory Marler said...

I've found this site.
You can change the layer to only see the 'bad'(declined) areas too.

What you'll often notice as you zoom in, the only stuff that is lost is landuse areas. Well they are only useful to be visually pleasing and can easily be recreated.