I think it exists because it's a concise way to communicate "this is ending, and will end in a predetermined amount of time, but it has not yet ended" in the same manner that sunset indicates that a day is ending but has not ended yet.
A Google+ private community allows social, collaborative photo sharing photos with friends with login required. This is not supported in Google Photos! With Photos shared albums you can select people with whom to share, but those people just get an email with a secret token embedded in a URL -- aside from trusting your non-technical friends not to accidentally leak this URL, there is no way to know who is viewing the photos or remove access from specific individuals. I really hope the Photos team upgrades their shared album permission model now that the alternative in Google+ has been sunsetted.
That strikes me as a pretty extreme reaction. If it's an issue you care about and you think it's a viable approach, you'd really withhold your contribution on the basis of a required email address?
Keep an eye on http://hypothes.is. I think they plan to support private annotations, and handle PDFs through pdf.js. It doesn't have an offline document management component, but it could be used as the annotation component in a larger system.
While it is clear that Broder exaggerated some of his claims, Musk also made several baseless and misleading claims in his responses and then threw a bunch of plots at people to "prove" his points. Most people cheering him on just didn't read carefully enough to notice.
I don't disagree with you, but as I understood it the author is partly dismayed by the fact that the delivery fee is so large. Even if every copy were delivered via 3G (which is unlikely), $2.58 for an 18 MB file seems steep. Sure, on one level it's not his place to complain -- he still probably made a boatload by leveraging Amazon's massive sales platform -- but I also understand that pricing schemes like that can feel like a bit of a rip-off (e.g. $20 printers with $40 ink cartridges).
Is it unique enough that you can be sure it's your password, and not someone else's? I ask because the cracked passwords seem to be the simple/obvious ones that are likely to be used by multiple people. If it is strong/unique though, it would effectively confine the hack time to the last 2 days.
I haven't done it yet, but I've been meaning to set up a very simple solution to this using a preprocessor (probably m4). I think wrapping a few blocks in if-statements based on the host name will go pretty far.
Google makes money by displaying ads, and they do so in a fairly unobtrusive way. Why are some people so determined not to provide anything in return to the company that offers them a valuable service?