3 ms·

What novel software features have Uber and Lyft added in the last two years?

zaroth 6y ago
I would imagine the vast majority of the work is in mapping metadata/annotations and functions to customize the features of a route based on that data. Airports would be the first example that comes to mind. Guiding both driver and passenger to the designated pickup location, which might be separate from the normal passenger loading/unloading areas. But throughout big cities you will have designated taxi/ride share POIs that need to be coded. I’m sure the pricing algorithm is also quite complex. It’s an easy mistake to make to look at something from the outside and assume the details are all pretty simple.
- abdullahkhalids 6y ago
  Both of those are important improvements, and the first is more of a life-long process than a task. A coop ride-sharing app would benefit from a "wiki-map" like OpenStreetMaps which can be directly added edited by the drivers, whose pockets will see more money if maps were better, and riders who will spend less money if maps were better. The pricing algorithm, I can see, will require a lot of advanced mathematics, and will involve a lot of politics, which might make it hard for a democratic organization to function. The politics comes from the fact that different pricing algorithm will have different losers and winners.
stu2b50 6y ago
I think good UberPool functionality is one. The routing for that is a lot harder technologically and prepandemic I did a lot of Uber pool rides, since in many cases I wasn't in a hurry and it was quite a bit cheaper.
- abdullahkhalids 6y ago
  Pooling goes well with the ethos of coop ride-sharing. The success of a coop ride-sharing app seems to depend a lot on how much open-source work will programmers be willing to volunteer to the project.
  - jtsiskin 6y ago
    I can’t imagine it would work as mostly volunteer work, the amount of code that needs to be written is just too high.

I would imagine the vast majority of the work is in mapping metadata/annotations and functions to customize the features of a route based on that data. Airports would be the first example that comes to mind. Guiding both driver and passenger to the designated pickup location, which might be separate from the normal passenger loading/unloading areas. But throughout big cities you will have designated taxi/ride share POIs that need to be coded. I’m sure the pricing algorithm is also quite complex. It’s an easy mistake to make to look at something from the outside and assume the details are all pretty simple.

abdullahkhalids 6y ago
Both of those are important improvements, and the first is more of a life-long process than a task. A coop ride-sharing app would benefit from a "wiki-map" like OpenStreetMaps which can be directly added edited by the drivers, whose pockets will see more money if maps were better, and riders who will spend less money if maps were better. The pricing algorithm, I can see, will require a lot of advanced mathematics, and will involve a lot of politics, which might make it hard for a democratic organization to function. The politics comes from the fact that different pricing algorithm will have different losers and winners.

stu2b50 6y ago

I think good UberPool functionality is one. The routing for that is a lot harder technologically and prepandemic I did a lot of Uber pool rides, since in many cases I wasn't in a hurry and it was quite a bit cheaper.

abdullahkhalids 6y ago
Pooling goes well with the ethos of coop ride-sharing. The success of a coop ride-sharing app seems to depend a lot on how much open-source work will programmers be willing to volunteer to the project.
- jtsiskin 6y ago
  I can’t imagine it would work as mostly volunteer work, the amount of code that needs to be written is just too high.