The Valentine’s Day Text Message Mystery(nytimes.com)
Oddly enough, I just saw this pop up on Reddit, as an awkward consequence of the vanishing/reappearing text messages:
tl;dr: someone thought her boyfriend was cheating because a message he'd likely sent months ago appeared in the middle of the night
There could be some very pretty big consequences to this screw up. It would seem the a prompt message of explanation might be a good idea. But it’s a teleco, so that would seem unlikely.
> to this screw up
While it's likely explained by other commenters as a possible replay of previously sent messages, AFAIK there is no SLA for the guaranteed windows of SMS delivery.
Also, "pretty big consequences" rarely apply when you enter into a "binding arbitration" clause with your cell phone provider.
Third party doctrine is a known loophole to the surveillance state for getting around pesky privacy laws, and my gut instinct tells me this third party sms gateway is a goldmine for whatever prism has become these days. TOS/EULAs probably say they can do whatever they want with your data including sharing with third parties so they have a legal out because it's not just the corporations influencing congress via K-street, but because the military industrial congressional complex has a vested interest in providers not protecting our privacy, and will railroad any provider that tries to stand up to them (see: Quest).
Congress on both sides is utterly corrupt and unrepresentative and until America wakes up to that fact and stops playing the two-party game nothing will change. (duvergers law has many weaknesses)
The carriers keep everything anyway.
This is true but slightly different regulations will apply to the data.
Regulations? Those rules the government makes? I bet they look the other way when they are the ones asking for data.
Worth noting that SMS does not guarantee delivery in any way and can take unspecified amount of time to deliver. I've often seen messages delivered days or weeks after being sent.
We need TCP over SMS!
But then you can't send texts if the other person's phone is off, out of range, or otherwise without service...
and I think you mean backwards, SMS over TCP.
Well, you typically layer TCP over an unreliable lower-level transport to get reliable transmission, so that works, too.
Also, SMS over TCP sounds somewhat reasonable, so I didn't wanna go with that.
SMS over TCP over SMS
Oh! Seems like I missed the joke. :)
TCP over BlockChain over SMS
Expect checks from VCs to start arriving on Monday
SMS delivery reports
SMS DRs are themselves SMSes, and subject to the same problems.
Remember how in the olden days, you sometimes got multiple copies of a single text? That was because the sender had marked "DR required", and some system along the way was dropping the receiver's DRs, so the SMSC kept retrying.
That's how all "reliable" protocols work when the message fits in a single transport unit. See the Two Generals Problem for why and how no protocol can tell you when to stop validating (though for practical purposes probability and the speed at which we can send messages allows us to proceed as if the problem were solvable)
Some networks will not deliver delivery reports. Worse still, some networks fake delivery reports.
Non-paywall link: http://archive.is/0v5YA
Connection refused ?
There was an amusing story on the most recent episode of the Verge's podcast, where they mentioned that when they assigned a reporter to cover this story, the bulk of the time he spent on it was just making sure it wasn't some kind of viral marketing scheme for an upcoming romcom film or something.
This is one of the reason why the Apple iOS messaging app is so popular. IIUC, when sending messages among users who have iPhones, SMS is not used.
iMessage has had some interesting bugs:
* 2014 -- if you switch from an iOS device to an Android device, iPhone users who try to send you text messages that are channeled via iMessage will have their messages silently vanish ( https://www.cnet.com/how-to/how-to-fix-apples-imessage-bug/ ; IIRC after a lawsuit, they added a remarkable little support page, https://selfsolve.apple.com/deregister-imessage/ )
* 2016 -- Apple iPhone devices were being used to send text messages which were going to the wrong recipient ( https://wp.josh.com/2016/10/26/divorced-by-apple-in-californ... , discussed at e.g. https://news.ycombinator.com/item?id=12797249 )
The flip-side to this is that if you have a regular number which you use without an Apple device and use the same number in your Apple settings, your SMS get hijacked, and only show up on the Apple devices, which you may not look at for weeks.
(Ask me how I know this.)
How do you know this? Is this old?
I have a few Google Fi data only sims and as soon as I put one in an iPhone, iMessages to the original sims number stop going through at all.
I remember there used to be issues like this when people switched to Android. Though this could still be an issue and maybe inserting a sim with a different number causes Apple to automatically unregister the original number, but just moving the sim does not.
I guess this is technically by design, but only if you have your iCloud account associate the number with iMessage (you toggle the send/receive from the number in the settings).
So even if your phone isn't using imessage, if your computer or ipad is, they'll receive the imessages and apple doesn't also send a text in addition to the imessage for the non-Apple device.
What prompted you to write an, essentially, Apple marketing piece when there's no good reason to do so?
Or WhatsApp, if you want to message people who aren't on ios.
Or signal, which has similar functionality but is always end to end encrypted and open source.
...or Matrix, or XMPP, which are decentralized and where attaching your phone number to account is optional.
And don’t mind Facebook having your data.
Keybase is where it is at. Even my parents are on it. Group messages that work and some attempt at security and privacy - all for free-99.
Don't forget about Signal. It works great as an sms replacement
It sounds like this company only exists to transfer text messages from one carrier's network to another. If that's the case, then any monoculture would have prevented this, it wouldn't have to be Apple's walled garden.
You can pick to use SMS, but by default they'll use iMessage.
It’s crushing when these phantom texts come back from people who had died or had long been missing.
some previous discussion: https://news.ycombinator.com/item?id=21474961
Apparently the NYT since then was able to coax a possible explanation from Verizon, and from the company that Verizon et al uses:
> On Thursday afternoon, after many unreciprocated emails were sent to carriers, a spokesman for Verizon suggested that the answer might lie with a third-party text message service provider called Syniverse Technologies, in Tampa, Fla.
> Was Syniverse the vendor responsible?
> “You really need to ask them that,” the Verizon spokesman said.
> ...Late in the afternoon on Thursday, a spokesman for the company, Kevin Petschow, said in an email: “During an internal maintenance cycle last night, 168,149 previously undelivered text messages were inadvertently sent to multiple mobile operators’ subscribers.”
Looks like that was speculated here https://news.ycombinator.com/item?id=21478218
That reads like someone with specific inside knowledge IMO.
Note that it's currently the only comment from that user, and the username looks like a random throwaway (unless vb6sp6 is a big fan of Visual Basic 6.0, Service Pack 6).
But maybe it's an obvious guess to a wide number of people with knowledge of that general ecosystem of companies.
I'm guessing it was someone within Verizon up-stream, given they were thrown under the bus. They'd know who was to blame, would potentially resent their employer being linked, and could pass off their comment as a guess if pressured.
That's plausible. But it also could be a good guess. Lots of messages from valentines day gives us a good hint that we have a time capsule from february. Machine with time capsule reboots and then sends everything is plausible for SMS systems (smtp too).
No strong carrier affiliation for senders or receivers suggests an intermediary. Not a lot of those in the US; Syniverse, maybe SAP (Sybase 365); I don't think Bandwidth does intercarrier work.
Maybe. But do you create throwaways to speculate for a single comment? I know I don't.
They even got the company name Syniverse right. That's cool.