3 ms·

You ignored the part about "They couldn't even find the malicious code removals by 'the guy who she wouldn't let fuck her' that was one of her central claims."

• pyre 13y ago
  Depends on what kind of logs they have. Laying down a couple of assumptions: 1. They only have the git repo history to go on. 2. It's been long enough that errant objects have been garbage collected. It's possible that the code she committed removed via a push that removed it from history entirely (e.g. `git push +old-master:master`). If this push was never reverted, then the commit object would become an orphan that was up for removal at the next garbage collection interval. After that, the incident would absolutely just be 'he said, she said,' and a lack of evidence would not be proof that she was lying.
  • encoderer 13y ago
    Yeah, if they have no backups. Also, they said they looked at both the push log and the commit history.
  • kemayo 13y ago
    Unless the guy was hyper-vigilant about reverting her code (as in, within a few seconds of her pushing her own commit), wouldn't that be a ridiculously noticeable way to do it? Other people would have pulled, and would find themselves inconsistent with the origin repo when they next went to push... Not saying it's impossible, but it seems like it'd have generated some documentation if it happened, probably in the form of people sending out "WTF? Who's pushing history-changes???" emails. Edited to add: Also, the article says they reviewed git push logs, which I imagine would catch this.
    • pyre 13y ago
      > as in, within a few seconds of her pushing her own commit Is it really only possible for a revert to happen 'within a few seconds' to not cause issues? That's a strong statement to make based on little information. It might be true, but the repository in question could also have little traffic, or this could have happened over a weekend. > Also, the article says they reviewed git push logs, which I imagine would catch this. I see. I guess overlooked that (probably just read it as 'commit logs'). In general, I don't understand why they didn't just ask her about the commits in question. If it did happen, she would likely be able to point them in the right direction and at least narrow their search (e.g. 'the revert happened in repository X around these dates'). If she couldn't, then they are no better off, but no worse off either.
      • kemayo 13y ago
        > That's a strong statement to make based on little information. That's a fair point. It'd depend a great deal on how github has their repos set up, and I can't remember seeing anyone talk about that.
  • yawboakye 13y ago
    The investigation considered all possible commits surrounding the accusation of passive-aggressive code removal. One instance was found where the engineer updated and broke some CSS in an internal application, which was fixed in a later commit. The investigator determined this change did not appear malicious. The nature of the commit/revert.
    • pyre 13y ago
      Your reply does not address what I wrote. In the scenario that I mentioned the commits would no longer exist, and therefore they could review 'all possible commits' without reviewing the commits in question.
  • judk 13y ago
    The company's only product is an enterprise-targeted version control system. If they lose track of commits so easily , that would be rather damning for their product.
    • pyre 13y ago
      I listed the assumptions that I made to demonstrate the possibility of a scenario. That said, you've grossly simplified the situation. What if it is an unreferenced commit from a year ago. Would you really find it damning that you can't go to Github and say: Hey guys, I had a branch `foo` in my repository `bar` a year or so ago. I deleted the branch (which had no references) and don't have any backups of it. Do you think that you guys could restore it for me?