|
I recently downloaded some large code bases and have had light reads through them, but what can I gain from this? How can I be sure what the author is doing is the right way to go about things? (One of the the code bases was the Zeus Trojan source code ;D) |
|||||||
|
|
(Answer assumes that you've picked code bases from a couple of projects that are relatively successful and well used). If you mean the best technical solution then you can't but there are lots of definitions of right. If this is working code from a popular application then that in itself is a reasonable definition of right, and that's one big part of what you can learn - you can see how people have solved problems in the real world and trying to work out the compromises and trade-offs they've made (and perhaps why they've made them). The other thing is that you can see how people have solved particular problems you might be interested in so pick a piece of functionality you're curious about, or one that's relevant to the sort of programming you do or want to do, and start digging through it. But I'd definitely try and focus what you're doing. Skim reading code will teach you very little. Pick an area you want to understand, or find a bug or change you'd like to make and focus on that. Just reading code is a passive activity and for most people that's a fairly ineffective way of learning. |
|||||
|
|
It depends what aspect of the program you are looking at e.g. code, architecture, tests? It will be better to read a good book on good practices for the area you are looking at beforehand (Code Complete 2nd Edition is fantastic all-round if you're just starting) so you'll know why code that looks good is good. Otherwise, as you say, you won't know whether the code as an example is "correct", and may not understand the intentions and reasons of particular aspects of the code without context of higher level concepts. |
||||
|
|
|
doing it the right way - A programmer is doing it the right way if his code fulfills all requirements. Which will be proven through tests. So, on the other hand, if you don't know, what the code should do and which programming guidelines (styleguide) had to be fulfilled and if you don't have test cases and test data at hand, then you won't be able to determine, if it was done the right way. Even if you find some nice looking algorithm, you should know, what it was supposed to do and you should know if it passes all tests before you add it to your own repository. |
|||||||
|
Well to start off i would say code is written once read many. The first thing you would ever realize is can i understand what is being done here or what the process is. Not to often do you find code that is readable and have rational comments which indeed are unto date with the code. To gain on you will need to benchmark the code and if needed try your own version of some part of the code, see how you fair of in terms of readability and later on for performance. At times you may put up an better algorithm also and also likely if understood fully you could use the patterns used in the code. In short it will always expand your knowledge base and fuel your passion to program/code and above all Read code which is a difficult skill |
|||
|
|
