From a shear managerial and business standpoint, it makes more sense to have a large group of people working on the software. More people means more bug fixes, tighter code, and more attention to the software. Having two or three people doing that ratchets up their value a ton, such that they can basically hold the company ransom. That is a bad idea. The way to do it is to basically have a code auditor or two, who goes over and verifies submitted code to the repository. That way you have some measure of quality control, and still allow a large group to work with/on the software.
Plus with a good repository set up you can basically have two people working on it simultaneously, as I understand it, and there is a check-in / check-out procedure that mixes changes
( ... )
We do have some repository software, though everyone has access and can easily break everything if they so choose. But it's not to big of a deal because if you can find the particular module that is broken you can roll back your version for it.
As for commenting, most people are really bad at it. I've been trying to give an over view of most of my code, but if I'm looking at someone else's things I usually have to ask a lot of questions.
Well, at the company I work at we have the only a few people write code model. Three people know the code inside and out and what every section does. One other guy knows the code inside and out but only works part-time and works on stuff that's sort of side-lined to the major stuff. Two other people work on new stuff that they put in themselves. Two other people work on tasks that the aforementioned three give them. Finally, four interns also do work on tasks that the aforementioned three give them. Not counting interns, we have nine full-time employees, and three part-time employees (four if you count Mark Egli, who I'm considering an intern above). Three of the four fulls and two of the three part-times do not code. Seven people code
( ... )
Well, sure our stuff is specialized, but not every part of it. That was my point when I talked about not worry about low level software components. These parts can be used for multiple projects so it's silly to have every software project redo the code. We keep these things in special folders you can include in any project. The guy's point was that if someone else is responsible for a part like this, then when you want to change something (or it breaks) then you have to wait for them to fix it before you go on. My point is that is that if you did all the coding yourself it would take more time than it would to just wait for the other person to fix once in every great while
( ... )
Comments 5
Reply
Reply
As for commenting, most people are really bad at it. I've been trying to give an over view of most of my code, but if I'm looking at someone else's things I usually have to ask a lot of questions.
Oh, and google is the shiznit.
Reply
Reply
Reply
Leave a comment