Programmer coordination discussion

Not a programmer, but I'm usually on somewhere between 2130 and 0400 GMT.

I can be on from 14:00 to 24:00 GMT, and maybe a bit later if necessary

In order for me to become a part of this project, I need to explain what my capabilities are.

I am learning C++ programming and I have a considerable amount of experience in Java Eclipse.

all through my life I have admired they way games function, and wanted to create my own for a very long time, I always wait for oportunities like this so I can actually participate in a big project.

- Thank You

I am available at these times,

10:00am to 9:30pm US Eastern Time

I would also like to ask where to give most of my information to.

Like my phone number and ethnic background and stuff of that nature.

Thank You and I hope I can get an answer soon...

I am gonna try one of these things (:cyclops:)

Wait, what for?

This is an open-comunity project, meaning that anyone can just hop in and help at anything they want. You dont need to leave information about you Nor any contact info. You dont need to get accepted or anything either.

Thank you for clearing that up

I was a little confused on what I was actually supposed to do.

So, obviously this game is going to be programmed in C++ but, I myself have only learned like, functions in C++ :| . I was wondering what people who know Java are good for, maybe prototypes or examples. I don't know, I just really want to get involved. Anyone got any ideas?

Although not a programmer myself, I do know that languages other than C++ can be used for creating prototypes for the game. What's more, I'm pretty sure that knowledge of Lua is what is needed to mod the game, but first we need a game before we start modding it.

Alright, I can definitely work on some prototypes.

Update on progress? I was just wondering how we're doing?

Inca wrote:

Update on progress? I was just wondering how we're doing?

It looks to me like progress is currently isolated to the microbe stage development for which the status can be found here:
https://thrivegame.canadaboard.net/t866-microbe-stage-progress-report#17235

I think it would be benificial if there was added a small readme/help file to each release explaining the controls and features of the current release!

I moved your post here just to keep that thread empty, although a disclaimer should be added saying not to post there, anyways don't worry about it.

Good point, but I don't think we need one for every release, just for all upcoming releases.

I compiled a long list of mostly prorgamming related questions that i'd like to know the answers to before getting started!

I read through most threads here on the forum that wasn't too old and seemingly irrelevant, i also skimmed through the source code to answer as many of my own questions as i could. If i missed an important thread or part of the source code just throw it at me!

1. What exactly will the lua scripting be used for? I saw modding mentioned, will it only be that?
   
2. Is the prototype branch actually meant to be a prototype and be thrown away/refactored for trunk or is it just a funny name for the current "trunk"?
   
3. I see some remnants of a desire for unit-tests around, but don't see any actual tests in the code, is this just due to down-prioritizing with a desire to eventually have covering tests or am i missing something?
   
4. I've gathered the actual best way for programmers to coordinate what is to be done and what has been done is the issue [url=github.com/Revolutionary-Games/Thrive/issues?state=open]tracker[/url] or should this thread also be used?
   
5. I saw an early release of the microbe editor tool somewhere, what is the progress of it, which git branch is it on and who is working on it?
   
   
   Specifically regarding code standard:
   
   
   
6. Do we have any kind of transcribed code standard? or has it just been preference of the current lead programmer?
   
7. When do we use inlining? i've read a lot on the subject but never really reached a conclusion as to when it should be used if at all.
   
8. Do we prefer exceptions to error codes and output parameters? minor performance convern.
   
9. Do we use a specific convention for when to use struct and when to use class? Or just the C-like vs OOP-like distinction 'feel' based one.
   
10. Code like
    

    Code:

    void
    setOwner(
     EntityId owner
    ) {
     m_owner = owner;
    }

11. is a bit hard for me to read effeciently as opposed to a more traditional:
    

    Code:

    void setOwner(EntityId owner) {
     m_owner = owner;
    }

12. Is there a consensus on the preference of that form, or a requirement due to editor width? I don't want to start bossing around and i can get used to it i'm sure, just curious if i'm alone on that.
    
    
    Less programming relevant:
    
    
    
13. This component list for microbestage seems to not have been updated in a while, is it still relevant?
    
14. Is this post about signals still relevant? i couldn't find any related code in the prototype branch.
    
15. In the last entries of this thread, it was discussed a while ago how much to restrict the use of multithreading with no consensus reached. What would the answer be?
    

 

Ahead of time Thanks if you can answer one or more of my questions

EDIT: Removed question 1 as it was a dumb oversight of mine

Thanks for answering, very insightful!

Regarding:
3: My experience with unit testing is limited to theory. I'll try and write unit tests when appropriate following the existing convention, but they probably wont be pretty initially.
10: I would have written the second example as:

Code:

auto prepareData(const std::vector<AwesomeComponent::AwesomeData>& data,
                 std::function<bool(const AwesomeComponent::AwesomeData&)> filter)
{
  // ...
}

Letting the doxygen and return statement speak for the return type, long parameter lists are on several lines but with the first, on line with the function name.
But i can get used to it np!

Had nothing to add to the rest of the answers

EDIT: Looking at the existing unit tests, i don't see how i could mess that up, great syntax!

I'm not sure this is the right place for these kinds of questions; let me know.

I found this in engine.h with the purpose of making the engine object non-copyable:

Code:

public:
  ...
   Engine(const Engine& other) = delete;

How does that delete work? never seen it before. (hard to google) I assume it just deletes the new object instantly but how does the user know this happened?
Why not just make the copy constructor private?

EDIT: Also no assignment overload so it is still copyable, right?.

Well, the auto keyword for return types only works if the function body is readily available, which is not always the case, for example in header files. It's also only advisable if there is exactly one return statement in the function.

And yeah, you could look up the function signature in the documentation. But (with a good text editor and proper keyboard shortcuts) it's faster to just open the header in another window / tab and navigate to the function definition than to pick up the mouse, switch to the browser, and click through to the appropriate site. That's assuming that there even is documentation available for that particular function.

Quote :

I'm not sure this is the right place for these kinds of questions; let me know.

Probably not, but I don't think we have some thread for miscellaneous programming questions. Maybe we should start one.

Quote :

How does that delete work?

The C++ standards committee likes to reuse keywords. Officially, it's to ensure backwards compatibility as new keywords could already be in use as variable names. My theory, however, is that they want to cause maximum confusion among programmers to create an artificial scarcity of C++ experts that could end up in the standards committee and drink up their coffee. But I digress.

That weird delete syntax is a C++11 feature and does exactly the same as a private copy constructor, it's just slightly more explicit and causes the compiler to throw an error if someone tries to even implement the deleted function. There's another variant with default:

Code:

Engine(const Engine&) = default;

Which, instead of removing the overload, inserts the default implementation for that overload. Useful if you want to define some custom constructors (which implicitly removes the default copy constructor) and know that the default copy constructor should still work.

Quote :

Maybe we should start one.

Sounds like a good idea! Probably better if you are the author.

And interresting with the c++11 syntax! I'm not completely up to snuff with it yet, so pardon if i write some nasty 98 code from time to time

You didn't comment on the lack of assignment overload still allowing copy?

EDIT: And with the documentation i meant in the header files! The doxygen source should be right above the definition right? but yeah i can see the issue with auto. I won't hassle you with it more.

Quote :

You didn't comment on the lack of assignment overload still allowing copy?

You can "delete" assignment operators the same way and if the copy / move constructor is deleted, the copy / move assignment operator should almost definitely be deleted as well. And now that you mention it, that is not the case in the Engine class, which is an oversight on my part. Good catch!

Edit: A wild thread appears.

Regarding the git issue tracker, i can't figure how to assign myself to an issue (not that i am ready to yet). I assume this is because i don't have permission to, so in order to start working on an issue, do i

• ... inform someone with permission that i want to be assigned? (and how if so)
  
• ... get permission to assign myself to issues?
  
• ... just leave it open and work on it until i have a solution and then inform someone?
  

Yeah, seems like only "official" (in the sense of GitHub) project members can assign issues to people. If you want to start working on one, leave a comment underneath the issue (and maybe check beforehand if someone else was quicker) and I'll assign you when I get the chance.

Hi there,

First of all, I'm back in the team and ready to get some coding done

I just finish to build the project and everything is working. I got some warning along the way, but they seems to come from external libraries over which we have no control(Setup.ps1 log for the interested). The game also went on half my second screen instead of getting full on my main screen but, I remember this happening a while ago. I'll eventually look into this issue considering how bothering it is to me.

• Anyone else is using 2 monitors and not affected by that issue?
  

The most time consuming problem I had, was that I tried to run Thrive/build/Thrive.exe instead of Thrive/build/dist/bin/Thrive.exe. Running the former gave me a missing OgreMain_d.dll and I spend a lot of time trying to figure it out. The latter worked fine. This bring me to my next question :

• Is it normal behavior? If so, I'd be interested in modifying the Build's Readme to make this precision. This would be a good way for me to make a first pull request and get involved on the github project
  

This bring me to my second question. I haven't found many information about github on the forum. Is there somewhere I could find information about the github etiquette/usage for the project? Yesterday, I was considering to make some modifications and make a pull request but I was really confused by the fact that the 56_Compound_Dropping branch was really ahead of the Master branch.

I now see that the branchs have been merged and the Master is much closer to the actual date.

• What would've been the best practice to make a modification with yesterday branches system? Use the Master branch and hope I won't enter in conflict? Put a comment asking the branches to be merged? Look at what file/modification the other user is doing and try to not conflict?
  

I think it would really be helful to make a github guide for the project to help newcomer understand how are managed the fork in the project and how long should fork last and such... I used github a lot as a version control but not as a tool to work in team. The only way I did source control in team is with check out(take control of the file)/check in(update/free the file) that doesn't seems to be github's workflow.

I already spent the last 1-2 day looking at the github issue/pull request/discussion and forums discussions to get a good idea of where the project is at. I'll spend a good part of the day to look/modify the code to get a good feel of it.

Expect pull request from me soon enough

RodGame wrote:

Hi there,

First of all, I'm back in the team and ready to get some coding done

Hi Rod! Great to see you again!

RodGame wrote:

Is it normal behavior? If so, I'd be interested in modifying the Build's Readme to make this precision. This would be a good way for me to make a first pull request and get involved on the github project

Unfortunately yes, because Windows has no easy mechanism for telling it where to look for DLLs besides the directory where the executable is, so the DLLs (and our art assets, for that matter) first need to be put in a certain directory structure. If have an idea on how to adjust the build system to almost eliminate this step, but I haven't gotten around to trying it out yet.

RodGame wrote:

This bring me to my second question. I haven't found many information about github on the forum.

The last section of the style guide (found in the doc subdirectory of the code repository) contains some information on our git workflow. And yes, it's completely normal for some branches to be far ahead of master. In fact, master should usually be the "oldest" branch because all work is done in other branches. Only when a feature is done is it merged back into master.[/quote][/quote]

Nimbal if you are lurking still, could you take another look at the SVN server?
I'm ready to add some new models to it, but regardless of what commands I try on the server I get a: 405 'Method not allowed'

I have the user you gave me but from my understanding SVN is supposed to ask me for the authentication information, which it does not seem to be doing.

Thanks!


And since this thread is usually pretty silent I will add a reminder to people reading through, that most of the activity is happening at the github page: https://github.com/Revolutionary-Games/Thrive/issues?state=open