In other news today...
-
Thinking: "Where did I go wrong? What was my first mistake?..."
-
@topspin said in In other news today...:
@loopback0 said in In other news today...:
Pushing out smaller batches of minor bug fixes more often isn't inherently a bad thing.
Moving the cheese every fucking day and breaking things is, though.
Yeah, blame the lure of the āNew And Improvedā¢ā. Professional apps tend not to do this, because serious business realizes the value of keeping the muscle memory valid by keeping things laid out still the same. But Basic Fred the User is more likely to pick an application that āis actively developedā so that's what Basic Fred the User getsāand we are unfortunately along for the ride.
-
-
@Bulb said in In other news today...:
To be honest I don't get why many mobile apps, many of the Google ones included, even exist.
-
@hungrier I don't think they can collect that much more than in browser with help of
s.
-
@Bulb Native apps can have more invasive permissions, and can run in the background and/or on phone startup.
-
@hungrier My experience is that getting to run on background on iOS is fairly close to impossible.
-
@Bulb That's a good point, I hadn't considered the iOS privacy/security features, which I don't know much about
-
@TimeBandit said in In other news today...:
That article mainly focuses on what users will see once Monarch rolls out. What they don't mention is that they will be changing how your data is retrieved. (tl;dr is at the end)
Current desktop versions of Outlook and Windows Mail/Calendar currently retrieve the data they need directly from the hosting server. For reasons that may be known to no one, the different versions are configured to use different sets of protocols to do this. The Mac version prefers EWS, but can also use POP and IMAP. The Windows version prefers to use MAPI, with a sprinkling of EWS, but is capable of using EAS, POP, and IMAP. Windows Mail/Calendar prefer EAS, but are also able to use IMAP and POP.
The current mobile versions of Outlook do not work the same way. Instead, they connect to a Microsoft server which then connects to the hosting server. To do this, the Microsoft server prefers EAS, but can fall back to IMAP and POP. In order to reduce calls to the hosting server - in case you use multiple mobile devices or are switching to a newer device - the Microsoft server stores all the retrieved data locally before relaying it to your mobile device(s).
Project Monarch is supposed to be switching to using the same methods as the current mobile versions. However, my company's microsoft contacts have also indicated that they will be dropping support for POP, leaving EAS and IMAP as the only supported protocols.
tl;dr: If you use a Microsoft email client after Monarch is released, or currently use Outlook on a mobile device, Microsoft will have a copy of all the data you retrive.
-
@abarker said in In other news today...:
tl;dr: If you use a Microsoft email client after Monarch is released, or currently use Outlook on a mobile device, Microsoft will have a copy of all the data you retrive.
I only use Outblind only with the work mail that is hosted on Azure, so they already do have those data.
ā¦ I am wondering whether the Google Mail client does anything similar though.
-
-
@dkf said in In other news today...:
@TimeBandit said in In other news today...:
New year, same Linus
ECC RAM isn't too big a deal for most consumer systems. Most users probably won't see any bit flips in the entire lifetime of their devices.
Didn't Torvalds specifically call that argument out as a self-serving lie by hardware manufacturers?
-
@Mason_Wheeler I only skimmed the article, but I think the implication was that ECC would have been useful in preventing stuff like rowhammer. Technically also bit flips, just from a different source, and not quite as random.
-
@dkf said in In other news today...:
@HardwareGeek said in In other news today...:
FPGA, so the configuration of the bit-flip-detection hardware could potentially be bit-flipped.
Unless you have a very large number of FPGAs, or are putting them into a nuclear reactor or somewhere else really horrible, you won't have much of a problem. My point wasn't that there was a high rate of problems, but rather that with enough area of silicon, you'll see them in production anyway.
IIRC (and I may not, because it was several years ago), it was a radiation-hardened system for space. Cosmic radiation may not be as bad as a nuclear reactor, but the system is probably a little harder to get to for repair/replacement.
-
-
@Mason_Wheeler said in In other news today...:
Didn't Torvalds specifically call that argument out as a self-serving lie by hardware manufacturers?
I'll put it a different way: you get a few football-fields' worth of silicon for RAM (with the feature size packed as densely as you can afford) and you'll definitely start to see random bit-flip events from background radiation. Error correction in the hardware is very helpful in such circumstances, since you don't really want your nice datacenter acting as a high energy physics detector.
The downside is that it increases the memory latency slightly and it means that you can only actually access a whole row at a time (but you can hide this if the ECC is in the memory controller hardware).
-
-
@dkf said in In other news today...:
@Mason_Wheeler said in In other news today...:
Didn't Torvalds specifically call that argument out as a self-serving lie by hardware manufacturers?
I'll put it a different way: you get a few football-fields' worth of silicon for RAM (with the feature size packed as densely as you can afford) and you'll definitely start to see random bit-flip events from background radiation. Error correction in the hardware is very helpful in such circumstances, since you don't really want your nice datacenter acting as a high energy physics detector.
The downside is that it increases the memory latency slightly and it means that you can only actually access a whole row at a time (but you can hide this if the ECC is in the memory controller hardware).
The other downside is that background radiation isn't the only thing that can cause bitflips. Just look at Rowhammer. ECC would stop that cold, so that's on the hardware manufacturers for refusing to put it in.
-
From here in Arizona today:
Who the hell is Wilson?
-
Shock shock horror horror shock shock horror!
-
The move, [A WhatsApp spokeswoman] said, is part of a previously disclosed move to allow businesses to store and manage WhatsApp chats using Facebook's infrastructure.
...
She said there will be no change in how WhatsApp shares provides data with Facebook for non-business chats and account data.
-
@loopback0 said in In other news today...:
The move, [A WhatsApp spokeswoman] said, is part of a previously disclosed move to allow businesses to store and manage WhatsApp chats using Facebook's infrastructure.
...
She said there will be no change in how WhatsApp shares provides data with Facebook for non-business chats and account data.And two days later there will be a change, because now they don't have to announce it.
-
-
-
-
@TimeBandit said in In other news today...:
There's something funny about an industry where Nintendo accidentally created two of the biggest players.
-
@DogsB said in In other news today...:
@TimeBandit said in In other news today...:
There's something funny about an industry where Nintendo accidentally created two of the biggest players.
"No biz like showbiz."
It applies to entertainment industry overall.
-
-
Put it in the cloud, they said. It will be better
-
@dkf said in In other news today...:
, since you don't really
Or maybe you do, and can sell the flip data to SETI or something! š
-
-
-
-
The probable aggressor may have caught fire himself during the act
EDIT: It seems the journalists couldn't see it end well:
There is no trace of the possible aggressor, who seems to have gone up in smoke.
-
-
@Rhywden Apparently, that information included the correct real-world identity information for a lot of users. Because they were using photos of driving licenses to verify identity, and a deleted flag is just the same as actually deleting stuff, right? Even though admins could access the whole databaseā¦
-
@dkf It's not entirely clear at this point whether the scraping process "only" includes the raw metadata for "publicly" available files or if they indeed got hold of everything, including the kitchen sink.
Considering that there are reports that, before their final shutdown, their sign-up process was broken to the point where you could just enter fake data, I'm not too sure that their security managed to seal that part away properly.
-
@Rhywden According to items on my Twitter feed, they weren't purging EXIF metadata at all, and were keeping all images (though with a ādeletedā flag set in the database for security, of course; posted messages apparently used the same approach). And the fake admin accounts in question could indeed actually read absolutely everything. I didn't notice whether that included kitchen sinks, but it might as well have.
This is a total privacy breach. If they have any EU users who exercised their right to be forgotten, Parler will be in a lot of trouble!
-
@dkf Yah, I did a bit more reading and it seems that the broken sign-up process accepting anything also extended to password recovery - which also accepted anything.
So, those driver licenses and SSNs might also just be out there now...
-
Although some of the content donk_enby pulled may include videos from deleted or private posts, reports of a Parler "hack" that have been circulating on Reddit and Twitter are false, she said.
"Only things that were available publicly via the web were archived," she clarified in a tweet today. "I don't have your e-mail address, phone or credit card number unless you posted it yourself on Parler."
Oh, well.
-
-
-
@Benjamin-Hall said in In other news today...:
I've said for years that the thing that's probably causing all those health problems for cigarette smokers is what's in the papers.
-
-
-
Meanwhile in B*****m:
-
Better swallow
the contrabandthe sandwiches:No, seriously:
The British Department for Environment, Food and Rural Affairs issued guidance suggesting drivers "use, consume, or dispose of" any personal food items in their vehicles before attempting to enter the European Union from Britain.
"From 1 January 2021 you will not be able to bring POAO (products of an animal origin) such as those containing meat or dairy (eg a ham and cheese sandwich) into the EU," the department said.
-
In related news: this eatery might be fined for littering and smuggling contraband across the border:
-
Just one more, because this one might not be such bad news:
-
@da-Doctah said in In other news today...:
@Benjamin-Hall said in In other news today...:
I've said for years that the thing that's probably causing all those health problems for cigarette smokers is what's in the papers.
What about pipe and cigar smokers who have similar health risks? What about people who chew tobacco that also have elevated risks for cancers in the mouth, throat, and stomach?
I don't think it's the just papers in cigarettes that are causing the health issues.