[MVG] Remove key bindings for shooting rocket and stun rocket

[EvoLandEco]

To make a minimal viable game (MVG), we have to implement and also remove some functionalities.

We will remove rocket and stun rocket from the game by commenting out the key bindings. We think that this approach would be easier and has less impact on the whole program.

[EvoLandEco]

As @richelbilderbeek suggested, a good approach would be to write a function to create a custom/test key action map for the minimal viable game purpose.

[EvoLandEco]

Or, we could map the keys to shoot/stun to the action type "none"

[richelbilderbeek]

Hi @EvoLandEco: using a none type is nearly always a bad idea, as every switch statement has to deal with it. This means this architecture scales badly. Returning an optional without a value in the few functions this is needed is superior.

[EvoLandEco]

Hi @richelbilderbeek , could you please elaborate a bit more about optional? Do you mean std::optional?

[TheoPannetier]

I think @richelbilderbeek is referring to some kind of specific type that exists in cpp? @EvoLandEco was referring to action_type::none (one of the possible values of enum class action_type). Linking the keys to this value would result in these keys not triggering any action. @richelbilderbeek was there a misunderstanding here, or do you reckon this course of action would be a bad idea?

[TheoPannetier]

cf our conversation in #648

[richelbilderbeek]

@EvoLandEco well spotted, we model optional after std::optional. We wrote our own optional, as some time ago we used C++11, at which std::optional was not there yet (?maybe as std::experimental::optional?). Also, it was a good exercise!

If you can replace the code in optional.h to using optional = std::optional; would be great if that would work :-)

[richelbilderbeek]

@TheoPannetier yes, optional/std::optional is standard C++. A std::optional is basically a std::vector that has either zero or one element. It allows you to 'return nothing', instead of adding an extra enum class value (i.e. none).

[EvoLandEco]

@EvoLandEco well spotted, we model optional after std::optional. We wrote our own optional, as some time ago we used C++11, at which std::optional was not there yet (?maybe as std::experimental::optional?). Also, it was a good exercise!

If you can replace the code in optional.h to using optional = std::optional; would be great if that would work :-)

Could we just use std:optional without using? Some consider using namespace std a bad practice, I suppose using optional = std::optional would be somewhat similar?

[EvoLandEco]

std::optional: How, when, and why, an article that might be useful, for reference

[EvoLandEco]

@richelbilderbeek I re-examined the code and I suppose there might still be some misunderstandings, I was planning to map the two key bindings to action_type::none, and this action type has long been existed in the code, I interpreted it as an idle status, thus I believed that re-assigning the key bindings from shoot and shoot stun rocket to none/idle was legit? Perhaps we can change none to idle here?

[richelbilderbeek]

Hi @EvoLandEco ,

Could we just use std:optional without using? Some consider using namespace std a bad practice, I suppose using optional = std::optional would be somewhat similar?

Yes, using namespace std in a header file is a bad idea (due to Koening lookup), using optional = std::optional can be useful while transitioning to C++17.

I was planning to map the two key bindings to action_type::none, and this action type has long been existed in the code, I interpreted it as an idle status, thus I believed that re-assigning the key bindings from shoot and shoot stun rocket to none/idle was legit? Perhaps we can change none to idle here?

I think it is a bad idea to map a key to doing nothing, whether it is called idle or none, as, well, you program how to do nothing...?

Still, I am happy to have my mind changed: show me some code that really needs this :-) . I predict, however, there is none.

[EvoLandEco]

closed by #648