4 ms·

this is cool! but I'm curious to see a git blame implementation.

by billconan 6y ago

phist_mcgee 6y ago
You mean the most conflict driven and passive aggressive feature in any programming tooling?
- asxd 6y ago
  I mean, it just tells you the commit that last changed a line...
  - billconan 6y ago
    it's not that simple. git is a snapshot, i.e. it saves an entire file. How to recover changed lines from it efficiently? And How to do it for the entire history? Basically, I want to understand this code https://github.com/git/git/blob/master/blame.c https://github.com/git/git/blob/master/blame.c
    - adkadskhj 6y ago
      I'm so confused on what you're saying. Are you saying it _isn't_ what commit last occurred on the given line? How would you explain this UI? https://github.com/git/git/blame/master/blame.c https://github.com/git/git/blame/master/blame.c Ie, what does blame mean to you?
    - asxd 6y ago
      I was replying to another comment that has since been deleted. It said something roughly along the lines of git blame being a contentious and passive aggressive command. I totally agree the technical implementation would be cool to study :)
      - Stratoscope 6y ago
        If you ever want to see "deleted" (flagged/dead) comments like that one, turn on "showdead" in your profile.
        asxd 6y ago
        Ah I see it now, thanks for the heads up!
- SamBam 6y ago
  Something tells me I wouldn't want to work in whatever team you work with. git blame is extremely useful for knowing when something was added, and what the context was. It is not about "blaming" people for mistakes.
  - adkadskhj 6y ago
    Yea, it strikes me as a poor name. Cheeky perhaps. I once had someone take me aside and ask me if it was rude to use "blame" in github to figure out who touched a specific line. They thought the name implied some sort of finger pointing and fault.
    - masklinn 6y ago
      > Yea, it strikes me as a poor name. Other VCS have aliases like « annotate » for the operation.
- NobodyNada 6y ago
  `blame` is one of Git’s most useful features for me, because if I’m confused about a piece of code I can instantly check to see when that code was added and why. Does it fix a bug; if so what’s the bug tracker ID and how was it resolved? Why did the programmer chose to do things this way instead of some other way that seems more obvious? How did this change fit into the rest of the codebase, and were any other files or functions changed to match? (Of course, the utility of this depends on how well people write commit messages. There’s nothing more annoying than looking at a blame and seeing “update somefile.java.”) If your organization is using blame to stir up conflict and point fingers, that sounds like a human problem rather than a technical problem, and a problem that you would still have even if `git blame` never existed.
- sdesol 6y ago
  My tool is designed to obtain insights from Git history and I've never understood how people can take such offense to git blame. I do agree calling it "blame" was probably not the best thing to do, but the fact is, it serves a purpose, which is, it's designed to help you understand how a piece of code came about. For you to call this feature "passive aggressive" must have been borne out of some traumatizing experience and for that, you have my sympathy.
- sodapopcan 6y ago
  That's Linus for you, I guess? I dunno, I always found `git blame` really funny yet at the same time, I empathize with those who are a little taken aback by it (including... what's the company's name who makes IDEs who calls it "annotate"... ?). I largely love it because most of the time when I `blame`, that's exactly what I want to do. I don't need software to give me a nicer verb so I can pretend that's not what I'm actually doing. And most of the time, the author is exactly who I thought it was going to be (which can result in jovial convos) or, more commonly, ITS ME! Oh, the humility. I guess I can see it being a problem if you work on a toxic team (shit, sorry) and yes, I do understand desire for a sheen of positivity in the language of the software we use, but pretty much every of the time I go to call `git blame`, I mean `git BLAME`. While it certainly does happen, I rarely find myself thinking, "Who wrote this amazing code!?? I MUST KNOW!!!" and in those circumstances, it actually soothes my ego to "blame" them for it :D edited for grammar/spelling
  - cjohansson 6y ago
    Git can also be an offensive word just like blame. It just depends on what attitude people have towards the context. I am not offended by git, blame, master and slave and I think it would be best if people who are offended just change their attitude towards the context
- 29athrowaway 6y ago
  git blame is mostly used to answer the question "when did this change and how?" Being able to answer that question is key for maintainers.
- quickthrower2 6y ago
  Nah, that award is for team time logging tools. Blame is normally used as a "who to ask" tool. Often the answer it gives you is "who last refactored this file" or "who originally moved this file from another location and lost the link to the prehistoric era when the interesting stuff happened"
- lostcolony 6y ago
  Another word for blame is 'responsibility'. In a political organization, it's good to take responsibility for something, but bad to blame. A generous interpretation of that is that the person/group who made the decision should raise their own hand, and that it's a faux pas for others to point them out. In a healthy one, you can recognize it's the same thing, and the important thing is how you respond to it (i.e., it's important to understand the responsible party so you can understand the reasons/motivations for the action, and accept them or address them).
- hu3 6y ago
  I'm sorry if you had bad experiences with git blame usage. People can be assholes sometimes. It's not meant to be a finger pointing tool but can be used for that.
- dang 6y ago
  Please don't take HN threads on flamewar tangents. That's what we're trying to avoid here. https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html

this is cool! but I'm curious to see a git blame implementation.

phist_mcgee 6y ago

You mean the most conflict driven and passive aggressive feature in any programming tooling?

asxd 6y ago
I mean, it just tells you the commit that last changed a line...
- billconan 6y ago
  it's not that simple. git is a snapshot, i.e. it saves an entire file. How to recover changed lines from it efficiently? And How to do it for the entire history? Basically, I want to understand this code https://github.com/git/git/blob/master/blame.c https://github.com/git/git/blob/master/blame.c
  - adkadskhj 6y ago
    I'm so confused on what you're saying. Are you saying it _isn't_ what commit last occurred on the given line? How would you explain this UI? https://github.com/git/git/blame/master/blame.c https://github.com/git/git/blame/master/blame.c Ie, what does blame mean to you?
  - asxd 6y ago
    I was replying to another comment that has since been deleted. It said something roughly along the lines of git blame being a contentious and passive aggressive command. I totally agree the technical implementation would be cool to study :)
    - Stratoscope 6y ago
      If you ever want to see "deleted" (flagged/dead) comments like that one, turn on "showdead" in your profile.
      - asxd 6y ago
        Ah I see it now, thanks for the heads up!
SamBam 6y ago
Something tells me I wouldn't want to work in whatever team you work with. git blame is extremely useful for knowing when something was added, and what the context was. It is not about "blaming" people for mistakes.
- adkadskhj 6y ago
  Yea, it strikes me as a poor name. Cheeky perhaps. I once had someone take me aside and ask me if it was rude to use "blame" in github to figure out who touched a specific line. They thought the name implied some sort of finger pointing and fault.
  - masklinn 6y ago
    > Yea, it strikes me as a poor name. Other VCS have aliases like « annotate » for the operation.
NobodyNada 6y ago
`blame` is one of Git’s most useful features for me, because if I’m confused about a piece of code I can instantly check to see when that code was added and why. Does it fix a bug; if so what’s the bug tracker ID and how was it resolved? Why did the programmer chose to do things this way instead of some other way that seems more obvious? How did this change fit into the rest of the codebase, and were any other files or functions changed to match? (Of course, the utility of this depends on how well people write commit messages. There’s nothing more annoying than looking at a blame and seeing “update somefile.java.”) If your organization is using blame to stir up conflict and point fingers, that sounds like a human problem rather than a technical problem, and a problem that you would still have even if `git blame` never existed.
sdesol 6y ago
My tool is designed to obtain insights from Git history and I've never understood how people can take such offense to git blame. I do agree calling it "blame" was probably not the best thing to do, but the fact is, it serves a purpose, which is, it's designed to help you understand how a piece of code came about. For you to call this feature "passive aggressive" must have been borne out of some traumatizing experience and for that, you have my sympathy.
sodapopcan 6y ago
That's Linus for you, I guess? I dunno, I always found `git blame` really funny yet at the same time, I empathize with those who are a little taken aback by it (including... what's the company's name who makes IDEs who calls it "annotate"... ?). I largely love it because most of the time when I `blame`, that's exactly what I want to do. I don't need software to give me a nicer verb so I can pretend that's not what I'm actually doing. And most of the time, the author is exactly who I thought it was going to be (which can result in jovial convos) or, more commonly, ITS ME! Oh, the humility. I guess I can see it being a problem if you work on a toxic team (shit, sorry) and yes, I do understand desire for a sheen of positivity in the language of the software we use, but pretty much every of the time I go to call `git blame`, I mean `git BLAME`. While it certainly does happen, I rarely find myself thinking, "Who wrote this amazing code!?? I MUST KNOW!!!" and in those circumstances, it actually soothes my ego to "blame" them for it :D edited for grammar/spelling
- cjohansson 6y ago
  Git can also be an offensive word just like blame. It just depends on what attitude people have towards the context. I am not offended by git, blame, master and slave and I think it would be best if people who are offended just change their attitude towards the context
29athrowaway 6y ago
git blame is mostly used to answer the question "when did this change and how?" Being able to answer that question is key for maintainers.
quickthrower2 6y ago
Nah, that award is for team time logging tools. Blame is normally used as a "who to ask" tool. Often the answer it gives you is "who last refactored this file" or "who originally moved this file from another location and lost the link to the prehistoric era when the interesting stuff happened"
lostcolony 6y ago
Another word for blame is 'responsibility'. In a political organization, it's good to take responsibility for something, but bad to blame. A generous interpretation of that is that the person/group who made the decision should raise their own hand, and that it's a faux pas for others to point them out. In a healthy one, you can recognize it's the same thing, and the important thing is how you respond to it (i.e., it's important to understand the responsible party so you can understand the reasons/motivations for the action, and accept them or address them).
hu3 6y ago
I'm sorry if you had bad experiences with git blame usage. People can be assholes sometimes. It's not meant to be a finger pointing tool but can be used for that.
dang 6y ago
Please don't take HN threads on flamewar tangents. That's what we're trying to avoid here. https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html