- Forum
- General C++ Programming
- C++ Road Trip Planner
C++ Road Trip Planner
Hello all!
I am an applied mathematician that has always used C++ for purely computational purposes. I am trying to branch out and learn how to use C++ in more versatile ways. The idea I have in my mind is to write a road trip planner that fetches directions, forecasts, fueling points with pricing information, traffic projections, and constructions updates of interest.
Clearly, this will require lots of website interaction--including querying--but I am not sure if this is feasible using C++. I am not wanting anyone to do this learning work for me, but if anyone has any libraries or tutorials that might be of use I would really appreciate it.
Best,
Ben
I am an applied mathematician that has always used C++ for purely computational purposes. I am trying to branch out and learn how to use C++ in more versatile ways. The idea I have in my mind is to write a road trip planner that fetches directions, forecasts, fueling points with pricing information, traffic projections, and constructions updates of interest.
Clearly, this will require lots of website interaction--including querying--but I am not sure if this is feasible using C++. I am not wanting anyone to do this learning work for me, but if anyone has any libraries or tutorials that might be of use I would really appreciate it.
Best,
Ben
Sounds like a massive project.
Really C++ would only be used for the heavy lifting on a server that provides your mapping service. (you may have some graph algorithms in mind)
All the client stuff would be javascript, or something. A lot of time could be spent getting that stuff to work.
How about starting with a native application, using a C++ graphics library? What you could do is "decouple the abstraction from the implementation" (i.e. the bridge pattern) This means you define an interface between your core stuff and your presentation stuff. You then write one implementation that presents your stuff as a native application (so you have something to show). Later another implementation that is able to present it over the web (if you feel like it). This might allow you to get started without dealing with the horrors of the web.
For example, suppose you want to draw maps. You define an abstract object with methods like drawRoad(), drawTree() whatever.
You have one C++ implementation that just puts it straight on the screen, this allows you to get on with the project. When you want
to do the web version you write another implementation that can draw these things in a browser on a client machine. (you may need to write a layer to generate a bitmap from the commands and then send it over the web)
In designing the interface you need to take account of the capabilities of the main target devices. A web based interface is a lot slower and more cumbersome than a native one.
Just some thoughts
Really C++ would only be used for the heavy lifting on a server that provides your mapping service. (you may have some graph algorithms in mind)
All the client stuff would be javascript, or something. A lot of time could be spent getting that stuff to work.
How about starting with a native application, using a C++ graphics library? What you could do is "decouple the abstraction from the implementation" (i.e. the bridge pattern) This means you define an interface between your core stuff and your presentation stuff. You then write one implementation that presents your stuff as a native application (so you have something to show). Later another implementation that is able to present it over the web (if you feel like it). This might allow you to get started without dealing with the horrors of the web.
For example, suppose you want to draw maps. You define an abstract object with methods like drawRoad(), drawTree() whatever.
You have one C++ implementation that just puts it straight on the screen, this allows you to get on with the project. When you want
to do the web version you write another implementation that can draw these things in a browser on a client machine. (you may need to write a layer to generate a bitmap from the commands and then send it over the web)
In designing the interface you need to take account of the capabilities of the main target devices. A web based interface is a lot slower and more cumbersome than a native one.
Just some thoughts
Topic archived. No new replies allowed.