Exactly my reaction. Try to do sorting, aggregate / set functions and performance tune up database operations, you will meet with a wall named math.

C# at 8? Not bad!

I'm the one who must say so. Thank you! :)

I've just created /v/askdev. Here you can ask anything related to IT world, even networking, career or robotics software. Currently it have about 12 subscribers there.

Easy, it's not like this question is being responded anyway.

What's the different here with [HELP] flair and /v/LearnProgramming? I won't mind to x-post there

Oh crap. Seems like a small issue but big in programming world. In the simplest situation, someone need to think three times to say that some other's code is wrong, in any manner, because it can offend them? Err...

Looks like I misunderstand your writing. Do you write the article to discuss about the internal design of your game engine or API that will be exposed outside? I read it as the former and looks like you are referring to the latter one.

If the case is to develop API which will be used, then implicit flexibility (the condition is unknown to the user) is indeed, more harmful than good. Explicit flexibility (different parameter type for example) is arguable though. Having API with close to one operation and close to none configuration is better than the opposite (exceptions still apply though).

For the internal process, flexibility to certain degree is indeed required, otherwise it'll lead to major refactoring here and there. And no, DI with constructor injection is different than reflection-based flexibility. But it's another discussion about DI.

Moreover, documentation-driven-development will be more useful for API development.

Good post, but my golden rule is: extremely applying some guidance never bring any good. YAGNI is one. If you develop something without at least researching and developing some of useful-simple features then you will be faced with refactor-intensive code or WET (write everything twice) code out there.

That said, if you have enough manpower and resource to develop the engine, you can always develop with documentation-driven-development first then followed by test-driven-development. To be honest, I am always regret myself for developing a framework without documentation-driven-development due to lack of manpower.

If you don't have manpower and resource, focus to develop some simpler useful utilities features first. Example if you use stong-typed language as Java or C#, non-throwing data conversion tools will be useful (from string especially). For weak-typed language as javascript, data-type detection tools or explicit data-type conversion tools will be useful. Priority on tools that: 1. do primarily close to one job only (non-complex), 2. have close to none configuration, 3. certainly will be used.

And you get it wrong on flexibility. Flexibility can be achieved by many things, clean or ugly. Reflection flexibility is imho ugly, unless you know what you are doing with it (dynamic code compiler is one thing that can be helped with reflection). Reflection also slow for large data operation (AutoMapper suffer from this). Dependency injection is cleaner, but still the implementation is debatable whether it is clean or not.

After all, only experience can determine whether you can develop a good flexible usable features or resort to refactor-intensive code. However overall, good post.

Imagine that stackoverflow using this as captcha for posting question. Mind-blown.

true...

The other NSA already read this thread though. Maybe they will think: "we are popular that some newbie want to join us"?

Thanks. I've do some research and distributing the .net runtime is indeed bad practice and can cause license violation. The rest is fine.

Nice advice. Looks like I try to use this path though.

At first this seems like a bad, ignorant answer. But the sad truth is it is true :)

Woah the answers here are awesome and many. Tomorrow I'll try to respond each one and summarize them!

I am aware with .Net core and it's open source. But it's not yet completed (correct me please) and in progress. Currently I'm not sure that the version of .Net I'm currently using is open-sourced already or need to wait for .Net core to complete. Additionally, I am using Roslyn in this project, and Roslyn is also part of open-sourcing .Net. But still, it's in pre-release right now.

For me, knowing that .Net is now starting to open source is different with understanding that .Net is free to use for open-source project. But looks like I'll try to look for it whether it's really free to use for open-source project.

IMO it's as easy as put them under ODesk/Upwork if the capital is easy part here. I can't say that there is any incentive that I can give. However, given the fact that the final product itself can be beneficial to the programmers itself, is it enough incentive?

Honestly, yes. I am under impression that C# cannot be open-sourced. At least I can't ensure that distributing the binary files don't violate any license.

Roughly searching, I find a good thread in SO about open source C# project. here. At least I am not alone in considering license issue.

By the way, is there any documented article that stated the licensing activity with C#/.net?

Good point. I have a working viable product already in C# and as I stated before, I don't want license problems so I decided not to continue it. Looks like I need to re-develop it in another open-source language as well before continuing.

Thanks for feedback!

Haven't tried any new languages after expertise with C#. Looks like in next several months I want to begin learning new one. I have choices between Lisp, Python, Haskell, Ruby. Preferably Lisp since it's said to be most powerful. Or Python since it's seems popular.

Haven't really consider Go until now. Could you ELI5 me what's good about Go and how it stand between those open-source language?

Edit: add Python

Could someone ELI5 me what is code of conduct and why it is bad in github?

In C#, I just found this. It is using winapi or similar lower module to do direct access to harddisk.

I still believe that the majority of higher-level abstraction language such as C# and java don't have many harddisk operation. Most of them provide memory processing or they are using database as storage instead. The file saving operation usually lower than 10 mB. Don't know about C/C++, but I think it should be around the same.

So in non-intensive harddisk operation apps, the technique to direct access harddisk is overkill. A more general performance optimization such as sorting algorithm can be used almost anywhere that involve sorting.

I don't say that your tips isn't useful or significant. I just stated why there is lack of interest of direct-hardware operation, compared to general-applicable optimization such as algorithm. Moreover, hardware is different from one and another, such as NAS-hdd, make it harder to do programming through hardware.

You just updated the flair list. Sweet move.

Edit: The help/education/beginner IMO is redundant. Isn't help/education is sufficient here unless multiple flair can be used in same post?

In your post above it only shows 8 (and one of no flair). What is the other 2?

Or, where can I see the list of flair in one subverse?

Context is needed here.

For programming language closer with hardware such as C/C++ and assembly maybe your point is on the spot here.

For higher one such as Java and C#, the function that you are describing here is not applicable, since in those language the function has been abstracted and handled behind the scene.

And not every apps using hard disk to operate. I think the people will like to discuss more general-applicable case rather than specific like this.

Similar with stackoverflow eh? [Migrated/Closed/Duplicated] Nice one though, I prefer something like this. Hopefully we can search by flair later on.

Please add [Tips n Trick] or [Technical] flair for articles that is technical (ORM and design pattern for example). And I think something about [Discussion] will purpose good here.

And hopefully we can mark reposted article though.

Good luck :) And as suggestion, please differentiate the css style (color) between quotes and comment replies though. Sometimes it's swapped each other when I read them.

You have a good statement there. C# is not the best tools to develop small apps, and heavyweight as well. That is at least true for winform and wpf, but I don't think the same can be applied to console apps.

C#'s main strength comes from the abstraction, and more or less it is a verbose language (CMIIW).

This can be improved with subvoat synonym mechanism. Tagging is better, but require more effort and also require synonym, so subverse synonym should at least give some improvement.

The poster can tag the question as v/prog_design, v/prog_help, v/prog_architecture, v/prog_syntax, v/prog_language, v/prog_feature and all of them can be seen from v/programming. Meanwhile user can also filter into synonym.

That way, you can filter for whatever discussion you like without negating the rest invisible to everyone.

It does. Stackoverflow does not allow discussion that is mainly opinion-based or lack of research.

But many of programming question is opinion-based (architecture, design, language comparison). Some even goes into holy war, ex: surrogate vs natural keys, emacs vs vim, tab vs spaces for indentation.

And for the lack of research. Programming is all about terms, what if we want to do some research about something but don't know what we want to search?

Those discussion can leads here.