User:Usul/NavitRefactoring

From Navit's Wiki
Jump to: navigation, search

Idea[edit]

This is not about forking!

  • enabling the NAVIT team to restart/refactor codebase to become the #1 OSM offline navigation application
  • fundraise resources to get this done

background[edit]

Error creating thumbnail: /bin/bash: convert: command not found
Aspects of NAVIT
Error creating thumbnail: /bin/bash: convert: command not found
Current software components
Error creating thumbnail: /bin/bash: convert: command not found
Possibly supported devices and platforms
Error creating thumbnail: /bin/bash: convert: command not found
Possibly usergroups
  • NAVIT had been started with different goals/scenarios long time ago (only few mobile platforms, no open geodata provider, ...)
  • internal structures reflect this and bring limits
  • makes only suboptimal use of OSM and get problems by the amount of the data
  • OSM would need more "consumers" (passive users) even to attract more editors worldwide - and to test new features / versions if they wish
  • therefore an offline app that uses state of art the technologies is nessesary but currently not on the way
  • commercial efforts have to fail (as they need to earn money in the long run)
  • NAVIT crew has superb knowledge on tuned file format, limited ressources, crossplatform development, customizations
  • but in last months there were no steps towards better usability/deployment

goals[edit]

  • NAVIT should be redesigned with OSM as major free data provider in mind (e.g. making downloads easier, make use of details, care about data volume/complexity, ...)
  • NAVIT should become user friendly for consumers, while stay easy to customize/expand/adapt for freaks
  • NAVIT should allow user feedback towards OSM as OSMBugs or similar
  • NAVIT should modularise/package OSD/map style/map colouring/routing profiles/GPS sources/speech output ...
  • NAVIT should support pedestrians...truckers, warn traffic lights / traffic signs, display speed limits, POI search, address search ... (making use of more OSM details)
  • NAVIT should support very different usecase e.g. just viewing map in train to find a restaurant when you hop off, finding the next rest area / gas station on a motorway ride etc.
  • NAVIT should try to offer equal or better functionality than commercial car navigation systems

pros[edit]

for OSM:

  • a working open source navi app for all platform might bring OSM to a lot of more endusers
  • can show how detailed OSM is and embed features, that only OSM can currently offer
  • let users give feedback on the drive in a way, that OSM mappers can really improve the map with usefull comments


for NAVIT:

  • gets the ressources for a sprint (money, fresh people including usability experts, designers, ...)
  • is getting able to correct legacy code and doing 'a big jump' forwards
  • gets the attention/users it (IMHO) deserves
  • starting from scratch offers max freedom of choice with keeping experiences in mind
  • with raising popularity, some tasks can be outsourced to the crowd (docs, platforms, designing new UI/styles/...).

cons[edit]

  • might be hard to accept, that some devs get paid and some not :(


limitations[edit]

What can't be done, even if people spend a lot of money?

  • no iPhone support? (GPL vs. iOS and appStore [1])
  • no Windows Phone support? (GPS vs. Windows Store [2] and no cross compiling possible?)

opinions[edit]

  • user:tryagain doesn't think that a general deep refactoring is nessesary
  • user:cp15 agrees that NAVIT is still highly modular

roadmap[edit]

Just ideas for a roadmap to improve NAVIT

  • Wiki cleanup
  • Wiki adding update/details
  • unify communication channels and information sources
  • poll about dev bottlenecks and design pitfalls
  • asking externals about user storys

money[edit]

  • for build server
  • for central map package convert server
  • for getting testing devices (smartphones, tablets, GPSreceivers, ...)

See also[edit]