Facebook Apologizes For Bug That Unblocked 800,000 People (gizmodo.com)

An anonymous reader quotes a report from Gizmodo: Facebook disclosed a new "bug" on Monday that temporarily let some users who'd been blocked on the service send messages to the people who had blocked them. The bug also let some previously-blocked users view posts that were shared "to a wider audience," such as publicly or with friends of friends, Facebook said. Facebook's privacy boss Erin Egan apologized for the error, writing in a blog that the company is reaching out to "over 800,000" users about the screw-up. The "blocking bug" was active between May 29 and June 5, for eight days, though the company now says Messenger should be acting normally. According to Egan's post: "[the bug] did not reinstate any friend connections that had been severed; 83% of people affected by the bug had only one person they had blocked temporarily unblocked; and Someone who was unblocked might have been able to contact people on Messenger who had blocked them."

Comments

Re:Would Rust have prevented this bug?

[Anonymous Coward]

No. RUST is not a modern language, it's RUSTY, like ASSembly or C/C++.

Java would have prevented the bug. Java programs, running on a java VM, written in java, running on Java hardware are 100% bug free, and execute faster than C/C++ or RUSTy languages.

Of course, for the ultimate in speed and security, we need an OS written in Java, running in a VM written in Java, running programs written in java, on a cpu running a java vm.

but nobody is perfect, except java programmers.

Blocking is so low

[loufoque]

Why would you even block someone, that's so lame.

Re:Blocking is so low

[cre1mer]

You block people when you want to have the last word and not read any follow up replies thereafter.

Re:Blocking is so low

[cre1mer]

Not surprisingly, I got modded down as a troll. But not because of my comment.

Re:Blocking is so low

[alvinrod]

In general I'd agree, but there are people out there who have nothing better to do with their lives and some tickle in their brain or bug up their ass that leads them to devote as much of their time as possible to wasting your time or making you miserable. Here's but one example of countless many.

If you're blocking someone just because they disagree with you or aren't validating your belief structure, I'd say that kind of makes you a bit of a dink. If someone's just endlessly spamming you, why devote your attention to any of it?

Re:Blocking is so low

[antdude]

Time to block you! :P

Re: Blocking is so low

[houghi]

How do you know that?

Re:Blocking is so low

[tokengeekgrrl]

Clearly, you don't know what it's like to have someone spam/harass you online with insults and threats because you had the audacity to determine you didn't want to be in a relationship with them anymore. I witnessed it happen firsthand to a close friend of mine, They had to block their ex online, change their cell phone number and file a restraining order.

Re: Blocking is so low

[cre1mer]

making videos about dolls and Comic Cons

If you're going to grow a YouTube channel, you need to tap into a niche with a built-in audience. Funko POP! and comic cons have complementary audiences.

Funko has been doubling each year for the last five years. I now own more shares of Funko stock than actual Funkos. I was so impressed with the investor presentation that I adopted it for my own YouTube content strategy.

Silicon Valley Comic Con 2018 has the same number of visitors (~70,000) and the same economic impact as Super Bowl 2018. SVCC is not even one of the larger comic cons in the US.

Re: Blocking is so low

[cre1mer]

No one asked that. And if we did, we wouldn't ask you.

Nope. You would look at the biggest YouTuber, see that every video has 1M+ views and tell everyone not to bother with YouTube.if you can't get 1M+ views on a video from DAY 1. Never mind that it took many big YouTubers five years to get established. I'm only seven months into my five year journey.

Congratulations.

That's why out of all my dividend-paying stocks that I've own, Funko is one of the riskiest stock that I own. Most anaylsts peg Funko as a speculative buyout target. It's more liklely to be bought out before the Funko bubble bursts.

Re: Blocking is so low

[cre1mer]

Crammar's off to a good start, I guess I hit a nerve. Again.

I'm not the one arguing with random ACs.

Re: Blocking is so low

[cre1mer]

Funko has never paid a dividend.

Correct. Which makes it a risky bet in my non-retirment, dividend-paying stock portfolio.

Also earlier you claimed it's doubled every 5 years.

Funko is a 19 years old company. As a private company, it had doubled every year for the last five years.

They've only been listed for about half a year.

A failed IPO from last year. Starting off at $12, sinking down to $7. I picked up shares for $11 last week. It's currently $13.

I meant actually invest in it and not lie about it on slashdot.

You need reading glasses.

Re: Blocking is so low

[cre1mer]

The only time I purchase a dozen shares of stock is when the dividend gets reinvested in the DRIP for free.

Re: Blocking is so low

[cre1mer]

Your numbers are shit. You're missing quite a few zeros for quantity and purchase price. Also, I'm a buy and hold investor. I don't see shares off unless the market raises the share price to unrealistic levels (thanks, Trump!) and I can get favorable tax treatment.

No they are not

[DarkRookie]

They are not sorry for doing this
They are sorry that they got caught.

Re: No they are not

[houghi]

Or they are sorry they need to inform people because of the GDPR.

Re:Would Rust have prevented this bug?

[ShanghaiBill]

Would using a modern programming language like Rust have prevented this bug from happening?

Possibly, but unlikely. Facebook uses a hodgepodge of mostly PHP and JavaScript, but other languages as well. But this was most likely a design or algorithm flaw rather than an error at the language level.

Their real problem is a lack of testing. A company of Facebook's scale should have a large suite of unit tests, regression tests, functionality tests, and usability tests, that are automatically run both pre- and post-deployment. They should have a rack of VPSes continuously testing and probing. There is no excuse for something like this going 8 days before detection.

LOL!

[MerlTurkin]

Facebook! ROTFLMFAO!

Re:Interesting

[MerlTurkin]

A lot of people just never learn. Or they're masochists.

What Ever

[dohzer]

Facebook say 'sorry', whatever-Cambridge-Analytica-are-called-now say 'thank you'.

Not surprising

[thePsychologist]

I've never had a Facebook, but the people I know who have one have complained repeatedly over the years about random unblockings.

I wouldn't be surprised if the team responsible for managing the permissions system in Facebook were told "hey guys, don't pay too much attention to the quality of your code here. Wink, wink."

That's okay, Facebook...

[Hallux-F-Sinister]

...since I blocked all of Facebook by deleting my Facebook account.

Comment removed

[account_deleted]

Comment removed based on user account deletion

Re:Interesting

[DigiShaman]

They're masochists

Comment removed

[account_deleted]

Comment removed based on user account deletion

Comment removed

[account_deleted]

Comment removed based on user account deletion

Comment removed

[account_deleted]

Comment removed based on user account deletion

Re:Free advice for CRE!MER!!!!

[cre1mer]

Trump is going to wuss out on the trade war. Just as he wussed out on the North Koreans last month and will wussed out on the Russians this month.