I first learned of Street-Complete here and I really like it.
It’s satisfying to walk around, complete little tasks, and get prizes, scratching a similar itch to Pokemon Go.
Stuck waiting for someone? Add opening hours for a few local businesses.
Have a long walk ahead of you? See if you can add/check house addresses as fast as you can walk.
Want to walk off a few beers before heading home? Complete some tasks in the bar street.
Its a very constructive way to “be right” on the internet.
Why does F-Droid tell me that the app has features that I may not like?
If you expand the description it will be written out for you.
It is due to being bound to a specific mapping Ecosystem and using specific non Foss middleware
It is because it has what F-Droid considers anti-features. In this case, even tho the code is open source, it seems to require a non-libre dependency to measure distances. See https://gitlab.com/fdroid/fdroiddata/-/issues/2627 and https://github.com/streetcomplete/StreetComplete/pull/3709#issuecomment-1039710672.
F-Droid is very strict with what it considers an anti-feature, and Android is very restrictive to properly work without at least one closed source library (thanks, Google), so I say you can ignore this, but it depends on you.
@Danitos @AngryCommieKender @openstreetmap
IIRC that was pulled out of the app into a separate companion app (StreetMeasure) ages ago.
F-Droid is complaining about “non-changeable or non-free network service”.
It doesn’t seem to say _which_ network service this is, which I think is quite scummy behaviour TBH.
to me it looks like it’s pointing out the network “services”:
and to me it also looks like they also say what the sercives are needed for. however its in german for this device.
update:
it sounds a bit weird that according to this info not the app is uploading new informations, but westnordost.de would, which suggests that they also upload/steal your openstreetmap password to their servers instead of using the app with locally stored access credentials to do that. but this could also be just a bad wording/misunderstanding/translation thing as obviously the party where you download the quests also need to know which quests are already solved which would be the data that is uploaded to them.
@smb @openstreetmap
In my browser that can’t be expanded and clicking on it just takes you to their definition of “Non-Free Network Services” https://f-droid.org/en/docs/Anti-Features/#NonFreeNet
So they’re complaining that an OpenStreetMap app is tied to OpenStreetMap related services and hiding this in a way that makes it look worse than it is.
This dramatically lowers my opinion of F-Droid.
maybe the browser version isn’t fully complete, or there might be a thing with css, bad browser specialities, just a bug in the webpage (did you check if it does show up in page source plaintext maybe?)
i would see the warning "bound to a specific non-changeable service as a good thing, as it would promote completely-free apps like (i think) fluffy chat, which does neither depend on a specific matrix service nor on a specific push service (i.e. it uses ntfy for push when installed, and if i choosed to use my own instance it would - i think - use that instead of the default ntfy.sh) in theory everyone could run something like osm and in practice this could come in very handy in the future one day, but currently IMHO only one osm exists, so maybe its not that bad for an app to be bound to it, but its always good to know if an app is bound or free to choice, same i would like to know if its maybe multi-account (which fluffy-chat is), but single-account-apps are so very common and thus this is not seen as an anti-feature which is completely ok for me.
but also in this case here the app is not only bound to osm, but to two other non-changeable services too, thus the warning is due even if label-app-bound-to-label-service would NOT be considered an anti-feature.
@smb @openstreetmap
If I view source I get “phrase not found” on a search for “Jawg” . It also doesn’t show up in Firefox’s Inspector, which I _think_ includes dynamically loaded stuff.
Just checked Chrome and Microsoft Edge and they don’t show it either. What are you using to actually get the detail?
I wouldn’t mind F-Droid’s warning if they gave me the details, but for some reason they don’t seems to be accessible to me.
1/2
@smb @openstreetmap
I don’t know how you’d get round the dev using his own site to pre-process (already published) data or host a list of “banned versions” with bugs that could corrupt data. You can’t really choose to have a different developer on the fly.
Maybe you could let people chose a different host for a published photo, but I think if there were privacy respecting ones that easily allowed that the dev probably would have use those instead?
https://streetcomplete.app/privacy
2/2
i am using the F-Droid “app” (currently 1.20.0), not a browser now. But i asumed css or such could be a cause as sometimes nonvisible content is behind some bad (ad?) layer. i thought i had seen such infos also on their website, but that is then too long ago and i might just be wrong with that.
maybe you can only get that info from the app which would feel a bit like an anti feature *haha
@smb @openstreetmap
Apparently they’re using a more detailed index for the app that isn’t supported in the web.
(And yeah, I did call that an F-Droid Anti-Feature when I found out that you have to have to install something for this to be displayed in an intuitive way.)
https://github.com/streetcomplete/StreetComplete/discussions/5756#discussioncomment-10108447
@smb
RE the update. That seems odd that they’d do that. I thought the quests were now being generated on the fly after downloading the raw data direct from OSM (previously it was an overpass query per quest IIRC).
I don’t know why they’d be bouncing uploads through a remote server when OSM already has the capacity to reject a conflict.
I don’t currently have access to something that monitors raw traffic to check.
if they changed it at some day, maybe the change wasn’t updated in f-droid afterwards. anyway, my osm account was “empty” before, because i just didn’t manage to finish my first map update try (which was in 2012 o_O ) now i’ll try again and thus they get a trust advantage based on doing good and it looks easy. trust advantage until i read or experience something bad of course, but i’ll try it these days.
@smb @InsertUser
@AngryCommieKender @Danitos There is a more detailed writeup about #StreetComplete and #Fdroid antifeatures on https://github.com/streetcomplete/StreetComplete/discussions/5756#discussioncomment-10107158 but in short, it is benign: You only ever login to OSM with your username/password, which then passes the token to StreetComplete (which you can revoke any time on https://www.openstreetmap.org/oauth2/authorized_applications). There are other (less popular) OSM forks/APIs that could be used if they were not hardcoded, just like other vector tiles (instead of Jawg.io)
Also, openstreetmap itself is a centralized eco system, even though you could theoreticallly host one yourself.
Because you can’t change it from using OpenStreetMap.org to an alternative self hosted site.
Because it tracks real time location and uses the internet. Unless it’s an app like this where you explicitly want that functionality, that’s usually a sign of some sort of tracking mechanism for advertising or nefarious purposes.
That just is not what F-Droid is saying. Why would you even claim that?