Comment by ssl-3
9 days ago
> I spent ~2 years actively editing Wikipedia for multiple hours every day. I remember taking my laptop out at airports for 20 minutes between transfers, just to tweak an article or improve a source. While I originally started because I found some articles lackluster, I quickly realized how vigorous the editing process could be on controversial topics.
That brings to mind an interesting parallel: I spent over ~2 years actively editing Waze for multiple hours every day. I don't fly much, but I remember taking notes about changes and taking my laptop out when I had a chance (wherever I was) to correct the map to better-match reality. While I originally started because the area I edited had way too much basemap[1], I quickly realized how vigorous the editing process could be when the end goal was to provide a map of driveable roads.
In some cases, the signal-to-noise ratio was so bad that I selectively nuked large parts of whole cities just to redraw them more-correctly.
I was producing good results that unambiguously had better validity than what I started with. The flurry of activity had me rise up quickly through both the editor ranks, and also the role-rankings.
It felt good.
But eventually, I got to see my careful well-researched edits be reverted by either stupid people or stupid bots. I didn't like this; I started editing Waze to make better maps of my area so I could have better maps with which to navigate with. That was awesome, but I finished those maps. So I branched out to improve adjacent areas and finished those maps, too. That also felt awesome.
I was motivated by improvement, not by competition.
When the competition showed up to re-arrange my work in ways that didn't make sense, I dropped out of editing maps on Waze as a serious pastime. I don't want to actively compete; I just want to passively fix.
I still fix things here or there, but months (instead of hours) will go by between edits.
And that's OK, I think: It still works better than it did before I put the effort in.
[1]: Oh, right. Basemap. We don't really have a single, official, freely-usable/government-supplied road map source in the US. We instead have counties doing their things with their formats, and 50 differnt states doing whatever they do, and sometimes cities with their own ways, with only the US Census Bureau's old TIGER database covering the whole gamut.
That conglomerate dataset is a damned mess, and that damnednness of that mess varies from place to place, but that damned mess is what Waze had to work with for the initial map import.
That initial import is known as "basemap."
And TIGER is cool and all (I remember an Internet where online census-provided TIGER maps were the only online maps), but it's really geared towards census-takers. It can include every private driveway, and every cowpath -- and it can include them as regular roads. I've cleaned up thousands of square miles of basemap in my area.
Please consider diverting your efforts to OpenStreetMap! The TIGER-derived basemap still needs fixup, but your work will benefit everyone, not just the Google shareholders who ultimately own Waze.
I would love do that. OSM is awesome, and basemap is a problem that I think I am good at solving.
But AFAIK OSM still has no mechanism that helps me get from A to B with dynamic and otherwise-unforeseeable roadway conditions. Waze still provides that, and improving the utility of a system that provides this kind of navigation aid has always been a primary motivation for working on Waze.
So my personal reward/payoff was/is high with Waze, because I could put those edits to use. It's not very good at all with OSM.
Besides. OSM seems to get bogged down in weird shit, like: Who cares how wide a public street is, in fractional meters? Maybe someone cares, but that someone isn't me: I drive a fairly large vehicle with fairly low ground clearance, but I've never had any difficulty driving it on urban roads in my neck of the woods. Those details don't matter to me.
So I'm not motivated to go out and measure these things like road width, and I'm also not motivated to provide assumed data as a presumed source of truth.
Falsification is bad. That's lies.
Superfluously-precise extrapolation is also bad. That's also lies.