I recently received notice that a changeset I had made had been reverted due to not following the protocols in place for importing data. First the facts of the matter. My family was going to vacation at Folly Beach in South Carolina and I thought it an excellent opportunity to do some mapping while in the area to verify local businesses and add information about them to the map. When viewing the area in OSM I saw that only a few buildings had been mapped and that the majority of POIs were nodes. I decided that I would see if I could find a building footprint file already created from the county so that I could add the buildings to the map and then replace the node information with tags connected to the structures where a single occupant resided within the buildings. I find this to be cleaner when it is able to be done rather than have closed ways and nodes trying to describe the very same thing. Now I will admit that I had some confusion and very much a lack of knowledge of what constituted an import and the process that was in place to regulate them. I found a file with the county for building footprints, exported the buildings for Folly Beach using Qgis, cleaned up the attribute data so that when bringing them into JOSM they would have the tag building = yes and proceeded to copy them into the main layer. After this was done I scoured the area for duplicates and resolved any there were, removed nodes that had POI information and instead attributed that information the the closed ways, and used the error checker to insure that I had not introduced any errors. I believe it was necessary to move some roads that ran through buildings due to geographical error but other than that it was nearly a blank canvas. I posted the changeset and readied myself for my vacation.
On arrival and with the mind to do some mapping in the wild I loaded up Vespucci and began downloading the data so that I could get to it. I was amazed to find that the buildings I had placed and properly attributed had been removed. When I arrived back at home I checked my email and found that a user had redacted my changes and rolled everything back to as it was before I began editing, citing a lack of a wiki page or discussion within the import mailing list. I searched out the import page to see if I had been in violation and it would seem that I had been. Mea culpa, mea culpa.

Now based on the directions at the import site I can see that if a group was to be doing very large and numerous imports the instructions make complete sense, especially in regards to large geographic areas. The need to insure that data is duplicated or error is introduced has been addressed and I see that there were lessons learned the hard way. Based on the instructions included it seems that the idea is about very large imports along a county, province, state, country, or large city where the potential for error is far greater and much data may already exist. Folly Beach is, according to Wikipedia, 18.9 square miles, the data that existed were a few nodes, the beaches, the wetlands, the roads, and a couple of buildings. The chances of error were minuscule and the import, which I acknowledge failed the guidelines, was checked over for error. My problem is that the wholesale redaction of the changeset did not actually look at whether there was an improvement to the map (I believe throughout the OSM community it is generally found desirable to have buildings represented) nor if any error had been introduced. It would seem rather that a prominent individual within the community has created an account that is solely dedicated to a vigilante style role of cleaning up what he describes as vandalism and what he deems to be bad imports and mechanical edits. Now while this role may be a valid and needed one I do find it intriguing that while there is so much oversight in the import process, one individual can decide for himself without discussion or debate that a changeset is either vandalism or an illegal import or is damaging to the map. Be that as it may I submit myself to that editing as I had seemed to violate the protocols in place.

As to the protocols themselves I find that while it does seem to make sense to prevent error or duplication, it has a chilling effect on mappers and on entities that may wish to upload their data to map. In the United States there are 50 states, 3,007 counties, and 19,354 incorporated places that are, for the most part, independent of one another. This also does not take into account that each of these entities would have multiple commissions or departments that could have interest in uploading data to the map. What the import guidelines would mean for these groups, should they wish to share data with the map, is that they would each create a wiki page for each import they wish to do, they each create an account specific for the import, and that they would submit their import to a mailing list for approval. Having worked in the public sector for small government I can verify that the majority would not want to go through this hassle every single time they wished to share data with the community at large, nor would they have the time at their disposal to do such. This leads to less adoption of the map and a loss of great data for the map. The question would be what to do about this.

We could hope that the many different entities would work in cooperation to limit the amount of import wikis and requests by combining their efforts or even forming a committee to combine their efforts thus making the requests less numerous (presupposing mass adoption of the map). This could happen and it would be great, but I do not see it in the near future as many have no idea about the map or about contributing to it. The second proposal that I would humbly submit is this, allow a trusted partner status to communities and groups much in the same way that Google has done with their map. Communities have a great desire to see their communities represented accurately in online maps. In two of the local governments I worked for we had this partnership with Google, which did not exclude us from scrutiny but rather gave us the benefit of the doubt that we knew what we were talking about. With this process an entity would need only submit once for the right to import, committees that accept the applications can make a stress of what imports are allowed and the requirements the group must follow, verify that the group is legitimate, and then approve them to maintain their AOI on the map. I think this process could lead to greater adoption and more accuracy for the map in the long run.

Whatever may be decided on the future of imports I still acknowledge my error and humbly offer my apologies for breaking protocol in my efforts to strive for the bettering of the map.