4 ms·
I'm not asking you to prove a negative though, I'm asking if you have any proof that what Google says is untrue since you say that it is. Using a passive voice
by malvosenior 7y ago
I'm not asking you to prove a negative though, I'm asking if you have any proof that what Google says is untrue since you say that it is. Using a passive voice doesn't mean that it's untrue, I actually find the Google wording to be more natural than your version.
Is there anything beyond the semantics of the Google statement that makes you feel that it's false?
- shadowgovt 7y agoSemantics matter, since we're now at the phase of the process where legal counsel would be involved drafting these statements (even if they're internally targeted, because Google knows its internal confidential broadcasts to employees will be leaked).
- joshuamorton 7y ago> I'm asking if you have any proof that what Google says is untrue since you say that it is To be clear, I believe that everything Google has actually stated is factual. You're just drawing wrong conclusions from partial information. So in one sense, no, because Google hasn't lied. I don't have any proof that I could share. But yes, I have strong reasons to believe the things I'm stating. As an aside, are you a native English speaker? There's some very interesting cultural aspects when assigning blame. In American English, its the norm to assign blame directly "John leaked the documents.", whereas in other cultures/languages, a passive structure is more common "John had the documents. The documents were leaked." This would be a common and correct way of blaming John for leaking something in some cultures, but in English it isn't. Presumably Google and its lawyers are primarily English speakers.
- malvosenior 7y agoI am a native English speaker, I just find the original statement to be more natural and less forced. Out of curiosity if you think that the statement is true, but constructed in a way that hides information, what do you think actually happened? If I understand your position it looks something like this: 1. The four people mentioned setup watches on people's calendars, went looking for information on people... the targets of their searches weren't part of their every day work life. 2. They shared that information with someone (no one knows who) but not the press or externally. 3. The people who had the information shared with them then leaked it externally. 4. Google found out and fired the original four people from step 1 and wrote a memo deliberately masking step 2. It's possible the above is accurate (please let me know if I got your position incorrect), but I don't think the wording of Google's announcement alone is enough to prove that. I'm apt to take it at its face value, especially because step 1 is by far the worst part in my mind and that doesn't seem to be in doubt. That's why I'm curious if you have some inside knowledge or something (your bio says you work at Google).
- username90 7y agoNote that you set up a watch of someones calendar just by looking it up, so it is very easy to accidentally watch a lot of calendars. I wouldn't read too much into that statement.
- malvosenior 7y agoMaybe so, but the Google statement does say that the employees in question did all of this in a systematic way that was outside the scope of their job responsibilities. I'd actually like to know a lot more about step 1, I suspect it holds the key to figuring out what's going on.
- username90 7y agoI feel like they got very little, since they focused so much effort demonizing this calendar watching. I can understand when they say that you shouldn't share photographs of others calendars or harass them over calendars, but I don't see why viewing others calendars is a bad thing itself. It is kinda like looking at their code or commits, you can see what they do and how they work, it is very useful if you intend to transfer or just wonder how other teams are doing things.