And today I got a G+ event invitation to a suspended G+ account (they don't consider "seebs" a real name). Since it's suspended, I can't opt out of notifications or mailings, nor can I use the help forums to contact them. Or anything else, so far as I can tell.

I know a few people who encountered this at Google. They found that it was absolutely lethal to team morale, because by definition it was actively harmful to you to help other people who report to the same manager; people worked around this at least some by forming teams of people reporting to different managers. But basically, of the people I know who work at Google, roughly 0% think the HR and staffing policies are reasonable, and I know more than one person who is being massively underemployed because of an arbitrary checklist of things that they have to do before they can be moved into a role that would use their stellar skills.

This will be a bit garbled, because my memory is vague and I want to shuffle details to keep people from being identified, but basically, imagine that you have a usual progression of programming roles from entry-level to senior, say. And similarly for sysadmin, and so on. And you have someone who is currently working as a relatively entry-level sysadmin, who would be an excellent senior programmer. You can't move that person to the programmer job because they haven't met the checklist of items for mid-level sysadmin yet, therefore they can't be evaluated for a possible change in job responsibilities. So your options are (1) acquire some meaningless credentials to do with obsolete operating systems no one still cares about or (2) look for work elsewhere, but not (3) move to a job inside Google where you'd be incredibly valuable to the company.

Last time I tried, I can't have an account under the only name I really care about ("seebs"). In theory they allow nyms, but in practice, I've never heard of anyone being able to get them to do this unless they're famous. Heck, I haven't even heard of anyone being able to get a non-form-letter response, or an opportunity to so much as write a single sentence in defense of their desire to be allowed to use a particular name.

So I do have a G+ account with a name they accept. It's not my name, but they've said at least once that it doesn't have to be your name, just a reasonably people-looking name. The real problem is, some people are uncomfortable if they see "fake names", by which they mean "names that don't follow a fairly conventional first/last name style and sound reasonably normal". So they aren't trying to solve the anonymity problem, they're trying to solve the problem of a few people getting uncomfortable when they see "handles" instead of "names". But not the problem of people getting uncomfortable when they're told they can't use the name they've been using for everything they do for the last couple of decades. Or the problem of people getting uncomfortable when their stalker has an easier time finding them.

Google wandered away from "don't be evil" a long time ago.

I am not at all convinced about this "almost all modern C compilers", given how many will do fairly awesome things once they determine that the behavior is undefined.

Pretty sure there is no intention whatsoever of turning that into defined behavior.

Uh-huh.

I did actually get one of those warnings popping up once, and... it was right. Hole in blog software was being exploited. Fixed hole, cleaned up, warning went away.

This doesn't sound right to me. The intX_t types, if present, have to be more 2s-complimenty, but they aren't really required to be present, as I recall.

Pretty sure the embedded systems guys wouldn't be super supportive of this, and they're by far the largest market for C.

And I just don't think these are big sources of trouble most of the time. If people would just go read Spencer's 10 Commandments for C Programmers, this would be pretty much solved.

gcc's been doing this for ages. We had a new compiler "break" the ARM kernel once. Turns out that something had a test for whether a pointer was null or not after a dereference of that pointer, and gcc threw out the test because it couldn't possibly apply.

Learn to play, noob! was written partially to address some of the things that make people gratuitiously unhappy.

I don't know, but it would probably be less damaging. The world does not benefit from this guy getting a ton of high-visibility options for advertising his militant refusal to even consider trying to comprehend anything about email or spam.

Yeah, but a whole lot of it will go to "generic host process for windows services", with no way to tell what each instance of that is doing.

More people oppose Obamacare than oppose the Affordable Care Act. Go figure.

I know a number of people who are supporters because they have health care for the first time in their lives. This has made me dislike the law significantly less. (Not that I ever thought the kinds of idiocy going on to try to stop it were appropriate.)

See, this one makes no sense to me, because all publishing and book-selling has always been like that: Private companies deciding what they will or won't sell. Self-published paperbacks have existed for a long time, and nothing other than economics prevented bookstores from picking up and selling physical copies of them, too. And they could do that, and they could also have stopped if they got complaints.

The key point is the feedback so you can tell whether it's working, and the structure that provides a reward mechanism to overcome the brain problem. Sounds useful to me.