Release your code. It’s a somewhat surprising fact but you can get away with publishing papers and not releasing your code. You will also feel a lot of incentive to not release your code: it can be a lot of work (research code can look like spaghetti since you iterate very quickly, you have to clean up a lot), it can be intimidating to think that others might judge you on your at most decent coding abilities, it is painful to maintain code and answer questions from other people about it (forever), and you might also be concerned that people could spot bugs that invalidate your results. However, it is precisely for some of these reasons that you should commit to releasing your code: it will force you to adopt better coding habits due to fear of public shaming (which will end up saving you time!), it will force you to learn better engineering practices, it will force you to be more thorough with your code (e.g. writing unit tests to make bugs much less likely), it will make others much more likely to follow up on your work (and hence lead to more citations of your papers) and of course it will be much more useful to everyone as a record of exactly what was done for posterity. When you do release your code I recommend taking advantage of docker containers; this will reduce the amount of headaches people email you about when they can’t get all the dependencies (and their precise versions) installed.
Divya Talwarand
。业内人士推荐搜狗输入法作为进阶阅读
В Иране раскрыли главные просчеты США в конфликте14:48
{ "move_npc", game_move_npc },