OpenStreetMap
Colonizing Hanoi
Posted by CactiStaccingCrane on 29 June 2023 in English. Last updated on 30 June 2023.
Ok, there is a ton to explain since my last “real” progress post here.
See, as shown in hdyc, my OSM contributions distribution is a bit… funny.
The first cluster of contributions was about Hanoi’s urban areas, the place I’m living at. The second cluster of contributions was about Mường Tè district, Lai Chau, and the third cluster of contributions was again about Hanoi, but in rural areas.
Why do I quit mapping Mường Tè then?
Well, it’s due to burnout. I was feeling that all of these heavy work spent at detailing the landscape is just not worth it. Mapping landuses is a fricking pain in the ass as you have to squint at low-res imagery and trying to understand what the heck is going on at the ground, staring at the screen for hours at a time.
I became demoralized. Every changeset just leave me more and more disappointed, feeling that I haven’t done enough when I saw American mappers zipping through counties. I wasn’t fast enough. But more importantly, I wasn’t helping anyone who truly needed it. No one will gonna make use of the landuses that I’ve mapped. To a lesser extent this is also the reason why I stopped mapping Hanoi urban areas, this time with blobby building imagery.
So I quit. For “real” this time.
What’s funny about me is that I never truly quit thinking about OSM. My head still flash “openstreetmap” whenever I saw a map. I’m not a professional cartographer, but I know for the fact that all of these maps are so so bad! Especially for Google Maps. They are apathetic enough to generate mangled messy roads on the countryside and residential roads that lead to nowhere, pretending that no one would notice it. It doesn’t help that I know for a fact can truly do better than these guys at Google, if I just spend my time on OSM. So, I returned.
But I don’t want to get burned out again. If this happens to me once again, I’m not sure if I will go back to OSM anymore. And more importantly, if I repeatingly pushing myself too far, my mental health would also in jeopardy. And this time, I want to do something that’s actually useful and impactful to me. Something that I can truly feel proud of. Something that I can show to others that I have beaten the hell out of Google Maps dev in their own game. And government cartographers too.
I want to one up them, just because I like it.
I picked a spot around hanoi that’s easy to map and easy to deal with. Initially, I picked the area around Suoi Hai lake (see https://www.openstreetmap.org/relation/15004726) and actually make a lot of progress in 2 days (pic), but then I hit a snag when I tried to expand the landuse.
That “easy” landuses that I was mapping on the right? That’s a military area. That bottom bit? Meadows. Beyond that is a big giant chaos: residential areas on farmlands, farmlands that are partially submerged in water, everything going on top of each other, all captured in a beautiful 2 meter per pixel resolution. I don’t want to get burned out yet again. So I moved north, way north, to the farmlands: (see https://www.openstreetmap.org/#map=13/21.2561/105.4062)
That day, I just sat down and map like a crazy madlad. Just kept on mapping farmlands and tracks. I deliberately ignore the hard bits and only map the easy ones first. No need to be a perfectionist for now. I literally stared at iD for hours. So… 2 days in, I got this:
Since then I make progress on smaller things, so the difference between then and now may not be that drastic. But I learned a lot during then. I learned what a recreation ground and cemetery looks like, what differentiate a basin with a lake with a transient water area on paddy fields, how to find websites online to geolocate features, how to properly classify highways. I also intentionally take breaks whenever I needed to and being more proactive in asking help from other people. I became a more balanced mapper, by not trying to cover as much landuse as possible, but rather trying to find joy on what I do.
So… what next for me? I will try to map the areas above to completion. It’s gonna be hard, but there will be a lot of satisfaction when it is done. I don’t intend to set a hard deadline on it. Also, I’m learning how to apply machine learning to try to speed up work on these areas (a Kaggle novice atm). If you want to help me map Hanoi, I certainly appreciate that by a ton. Here is the link again: https://www.openstreetmap.org/#map=13/21.2433/105.4018.
Discussion
Comment from kucai on 30 June 2023 at 02:30
I’d mostly map what I think OSM is initially made for - vehicle navigation and the corresponding poi/landmarks that’s useful for it. Don’t plan to make it pretty, but just functional and light enough so that it can be used on older phones. That’s how I try to avoid burnout.
Comment from CactiStaccingCrane on 30 June 2023 at 02:48
alina kleyafa is a sex website advert spammer. they probably used chatgpt for the response.
To kucai: yes, I try to make the map more functional than pretty by properly classifying highways and such, but there’s also an element of beauty to it that I couldn’t resist.
Comment from Koreller on 30 June 2023 at 10:46
It’s very interesting this notion of discouragement,
I understand your desire to do things perfectly, I have it too, but we need to find the right balance, because when I saw what “mapping perfectly” meant… I understood that it was a professional skill in my opinion very very complicated to obtain. I’m talking about this level of perfection: https://www.openstreetmap.org/#map=16/38.7709/125.5247 which is unreachable for me, so my level is more like https://www.openstreetmap.org/#map=16/42.9543/130.0124 and I’m in peace with that.
What we can do is map well, and that’s hard enough to achieve, but it’s a good job!
On the mental side, I think it’s important to do cartography for fun (as a non-professional cartographer), because the moment it becomes a self-imposed “obligation”, there’s a good chance it will become counter-productive. Knowing how to stop and do something else is an important skill if you don’t want to lose interest in cartography. Sometimes it’s important not to be too demanding on yourself.
Mapping work in poorly mapped areas is too big to be done alone, and we have to accept that.
In my opinion, it’s important to be satisfied with one’s work and also to know how to congratulate oneself on what one has done!
Thanks for this diary!
Comment from CactiStaccingCrane on 30 June 2023 at 12:08
Personally I think that we shouldn’t strive for perfection and rather just reach to a “good enough” point personal to all of us. It’s ok to skip certain areas that you don’t like to map. If you enjoy mapping individual paddy fields then yes, it might be something to strive for, but don’t force yourself over it if you don’t enjoy doing it.
Comment from CasGroenigen on 30 June 2023 at 12:17
You do amazing work! Glad I could help a little too with the orchards. If you need more help with something, let me know!
Comment from CactiStaccingCrane on 30 June 2023 at 12:37
Well, the district is dying to have its forest mapped. There is so many of them!
Comment from dressed-pleasekin on 3 July 2023 at 16:37
Beautiful work! Landuse and natural features indeed take a lot of time, clicks, and squinting - but seeing maps with no greenery somehow makes me sad, so I get your urge to map all that stuff. Still, take care not to burnout or strain yourself! I would also advise trying alternative setups for mapping - i.e. not using a mouse or touchpad but a touchscreen, trackball, or stylus. It can be a nice change of pace and helps guard against RSI when regularly mapping something large-scale.