this post was submitted on 21 Nov 2025
57 points (100.0% liked)

CoMaps

839 readers
93 users here now

Cooperative community open-source fork of Organic Maps, an offline privacy-respecting not-for-profit OpenStreetMap GPS app for your phone.

CoMaps website

Codeberg

Matrix

Mastodon

OpenCollective


founded 6 months ago
MODERATORS
 

As the title states. I really want to like and use open source maps, but it seems that whatever database they are pulling from is pretty bad and incomplete overall. You also have to put the exact name, as it is in the system, otherwise it will not come up.

Just to use a really simple and well known example, after downloading the appropriate maps for this search, if I search for "the liberty bell", the first four results are either roads or businesses near me, then the liberty bell museum in Allentown (which btw is permanently closed), and then the liberty bell center, then the liberty bell center again for some reason, and then finally the actual liberty bell itself. There are also suggestions down the list that no longer exist, such as the liberty bell pavilion. So this is clearly outdated data that is being used.

The suggested searches are also nonsense, only taking the last word into account, i.e suggesting "belles ave", "bellgrove rd", etc, rather than "liberty bell center", which would be the most logical suggestion for that search.

Why is this so bad, and what needs to happen to make it better?

you are viewing a single comment's thread
view the rest of the comments
[–] tyler@programming.dev 16 points 1 day ago (2 children)

None of that is true anymore. Fuzzy searching is incredibly easy, you don’t need billion dollar servers, and there are open source libraries that do it right out of the box. It has nothing to do with community submissions and everything to do with how they implemented search.

What’s most likely is that they don’t have a ranking system for each poi, so liberty bell just defaults to whatever is closest rather than whatever is most likely.

[–] Hupf@feddit.org 3 points 9 hours ago

https://en.wikipedia.org/wiki/DWIM is hard even for other humans sometimes.

[–] bobo1900@startrek.website 3 points 1 day ago (1 children)

I don't know much about the topic, but I guess even if that's the case, it's likely most open source projects just don't have contributors knowledgeable or willing to invest the time

[–] tyler@programming.dev 2 points 13 hours ago

Yes of course that is the underlying reason, for example i contribute to a lot of open source projects and even if i have the expertise on any given thing i can only put my knowledge in so many places at once. And not only that but interests wax and wane and oss contributors get shit on quite often.