0

I define "proper" as one that "correctly" and ideally uniquely represents a user for a given scope of work. Refer to how to configure a git identity.

Motivation: I have a colleague that seems to think its a "trivial", "non-issue" to use multiple identities; often one or more per machine from where he is working (i.e. noobie <noob@mylaptop.com>, nb <NooB@mypc.com>, boob <not.even.close> etc.). I haven't been able to convince him to clean up is act.

I'm looking for convincing 3rd party explanations of the pros / cons of using a correctly representative identity. Links to articles, manuals, docs, etc. welcome. Stick to factual pros / cons, no opinions please.

Community
  • 1
  • 1
i3ensays
  • 2,804
  • 3
  • 19
  • 23
  • 1
    Because it's the way to identify its work in a corporate process. Ask your manager to setup a development process that include being properly identified in the commits. It's not your work to convince him if it's a difficult personnality, it's your manager's. – CharlesB Aug 02 '13 at 17:32
  • https://help.github.com/articles/using-git-blame-to-trace-changes-in-a-file – dusual Aug 02 '13 at 17:33
  • Because you want an accurate record of who contributed what? If he wants to be an ass about it, point out that it's just going to make things very hard for him to demonstrate that he's actually contributing anything to the project... – user229044 Aug 02 '13 at 17:35
  • @CharlesB true. I'm working on that, however, said noobie was my manager's first hire 3 years ago and they are tight. Also the "Senior" in his title gives him (imaginary) powers. – i3ensays Aug 02 '13 at 17:47
  • @meagar that is not really a problem. It's pretty obvious who the author is as he's the only non-conformist. – i3ensays Aug 02 '13 at 17:48
  • I probably need to follow a technical reasoning train of thought to gain any ground (e.g. each identity uses up a Stash license so stop being wasteful!) although we're not actually using any hosted service. – i3ensays Aug 02 '13 at 17:56
  • 2
    This is a workplace politics issue. It *may* be welcome on http://workplace.stackexchange.com/ but check their [on-topic page](http://workplace.stackexchange.com/help/on-topic) first. – Todd A. Jacobs Aug 02 '13 at 18:09
  • 1
    @i3ensays Well, that's pretty much the only reason to enforce accurate author strings so... if you don't care about it, why are you posting this question? – user229044 Aug 02 '13 at 18:17
  • @meagar I didn't want to get into answering my own question, but there are other reasons. (e.g. post commit scripts that make use of email) – i3ensays Aug 02 '13 at 18:29
  • 1
    I can think of a variety of reasons that have to do with general reasons for attribution: budgeting, licensing, asking questions about some change after the fact and people just trying to make sense of the nature of how the codebase changed over time, but honestly it really depends on the type of project and work environment. – James Snyder Aug 02 '13 at 19:33

2 Answers2

1

You can use mailmap to allocate differing identities to one [current latest] canonical identity. Details in 'git shortlog'

Usually it is the 'signed-off-by:' line that actually identifies the author. This needs to be done so that you can allocate the code (your copyright) to the licence terms e.g. if you don't have a proper identity then there is disputed ownership and the code can't be made FOSS, or whatever your company/project needs.

In short, good code is supported by a proper author name.

Philip Oakley
  • 13,333
  • 9
  • 48
  • 71
0

Well I would say so you can use git blame

https://help.github.com/articles/using-git-blame-to-trace-changes-in-a-file

Well its important because it helps build a good culture of communication .You can always ask some one who wrote a particular piece before you go and delete it. Ownership is an important part of building good culture among team members

dusual
  • 2,097
  • 3
  • 19
  • 26