Lol. My new flagship top of the line Galaxy phone won't let me take pictures if a face is detected now without a 3-5 second delay from when I press the shutter.
If subject turns around before then, the picture is instantly taken, and I'm instantly getting photos if no faces are detected as well. I've reproduced this across a number of conditions.
This is great for the family photos..
They're really going strong with their software now.
I noticed some long delays taking photos of people (though not quite as long as yours) and found that disabling the "beauty" filter, which smoothens skin among other effects, reduced that delay.
Wow, somebody at Samsung is sure having an incredibly bad day... Flashing firmware to devices in the field is genuinely a really tough problem. Due to space, memory and processing limitations you can end up backed into a corner as an engineer. It may be the case where there is genuinely no way to recover from this - turning a $10k remote software patch job into a several million dollar product recall.
We used to have a simple rule where you always had two firmwares in storage, a primary image and secondary image. If you see tonnes of unplanned reboots recently you fallback to the secondary image. It can also be manually rolled back by holding some button during boot. (Other reasons to falling back could also be a failed hash check or bad signing for example.) To be clear, this method also had potential flaws and two bad software updates in a row could be bad (it tries to mark "good" firmwares, but it isn't foolproof).
I've seen other methods also make use of a minimalist boot system that is just about enough to flash a new firmware. Stick a USB device in and hold a button during boot for example - if no button press you pay very little in startup time and the main application boots instead. You then * really * test this fallback firmware to make sure it will be very reliable in the field from cold boot.
This gets even more troubling when OTA (over the air) methods are more widely employed too, where there may not be any exposed programming port on the device. I remember seeing somebody brick a few hundred devices via bluetooth before somebody said "have you tested that firmware works?" - turns out it was a bad unchecked-in local build...
We had to get Samsung to update their firmware to stop the 6,000 TVs we installed on a cruise ship from saturating the physically isolated 10GbE network with attempts to check for firmware updates.
> We used to have a simple rule where you always had two firmwares in storage, a primary image and secondary image
If you've got a 64MB image, that could be 50 cents per unit. That works out to a non-trivial drop in margins, and Blu-ray players have gotten commoditized enough that you might not be able to afford that. And I guarantee you it's not just this corner. Lots of corners were cut. I assume, anyway. I don't actually know for this model.
The funny thing is I'd happily pay $10-20 more for a Blu-ray player that had less cost-centric engineering, the problem is consumers have no way to easily see this.
the problem is consumers have no way to easily see this.
I remember seeing several motherboards advertising "dual BIOS" quite prominently, although the customer base there would understand the advantages of the feature far more than if something similar was advertised for a Blu-ray player.
> If you've got a 64MB image, that could be 50 cents per
> unit.
I remember at times, even 1 cent savings was "rewarded". It really can get cut throat on cost savings. Luckily some of the things I worked on were automotive, so the potential cost of recall really was a consideration that we could hold over the heads of management, otherwise you would really struggle to make the case.
> And I guarantee you it's not just this corner. Lots of
> corners were cut.
I think very large books could be filled with engineering stories about corner cutting. This is the sort of thing that wakes me up in the middle of the night in a cold sweat.
> problem is consumers have no way to easily see this.
This is why I end up buying slightly older hardware/software that has had some real user testing, but of course this doesn't always work.
Do not connect Samsung home media devices to the internet.[0][1][2][3] If you really want the internet-based multimedia features, get a separate, more secure device such as a Chromecast or Apple TV—and even those aren’t perfect.[4]
How many times are we going to keep making the same mistakes? There is no benefit to connecting a Samsung TV or Blu-ray player to the internet when a $35 Chromecast works far better, has all the same features (plus more), doesn’t break a year or two after you buy it, and is more secure. There’s just no excuse.
Yup, if you’re willing to tinker, you can make a pretty nifty home media setup for practically nothing, customized to your liking.
But for everyone else, Chromecast et al are plug-’n-play. If you’re an Apple person and are willing to spend a bit extra, Apple TV’s UI/UX is hard to beat without making your own system.
I find Apple TV's UX incredibly annoying and inaccessible. Gesture based navigation with ambiguous rules, lack of clear contrast on selected items and modal UI make me wanna ditch Apple TV, but I don't know of good alternatives.
I prefer it primarily because it "just works". I don't use my TV often these days, so I stopped maintaining all my custom hardware (e.g., Retro Pie). I've found that my Chromecast can be finicky; sometimes it doesn't like to stay connected to Wi-Fi--and it has no excuse, seeing how I'm using Google Wifi.
That being said, I do like the UI. I don't mess with it much, though; I mostly just talk to it. It does a fairly good job of figuring out what show I'm telling it to play. It works, it updates on its own, it doesn't brick itself, it doesn't disconnect from the network (because it's wired in), it's reasonably secure, it plays what I tell it to play. Good enough for me.
Depending on your use case, you might want to give Kodi a try if you're not satisfied with the mainstream alternatives. I haven't used it in ages, but I hear it works quite well these days.
The plex and steamlink apps on my Samsung TV get daily use. So does the Chromecast connected to it for other things.
In the case of plex, having a native app on the TV is far superior than fiddling with my phone or tablet or starting something on my computer and running into the next room to watch it.
For steamlink, I've suddenly started playing some of my old steam backlog, and now that epic is throwing games at me some of those as well. And I've also got a serious retroarch habit that works great over the steamlink app.
I used to use apps on my Samsung TV, but it became clear that Samsung wasn’t interested in ensuring that the device was secure, so I had to look for alternatives. They stopped releasing updates for my TV about a year after I bought it, but security vulnerabilities were still being found.
Maybe my hardware is just on the older side—it’s a whole 5 years old now!—but Samsung TV apps are pretty slow compared to apps on third-party hardware and always have been. For example, my Samsung TV natively supports Miracast, but the latency is terrible compared to an external Microsoft Wireless Display dongle.
Here’s what I use for everything you’ve mentioned, no phone required, although it might be overkill for casual use:
Plex: Apple TV
Steam Link: Steam Link (although I should probably stop using that deprecated device and just use the app on Apple TV). I tend to favor Parsec over Steam Link, though; it’s less convenient but has lower latency and isn’t limited to games. I also have a Microsoft Wireless Display dongle, which is generally the most convenient of the three and still has low enough latency for casual use, although I wouldn’t use it for gaming.
Retro: RetroPie
I know it’s inconvenient to have additional devices for all of these when you could just use the apps on your TV. Strong passwords are also inconvenient. But is it worth getting hacked or dealing with Samsung’s problematic updates? Maybe someday the situation will improve, but until then, either choose a different brand or use external devices.
Not quite related, but I picked up a used LG Blu-Ray player that claimed to have broken Wifi. Apparently, whoever programmed the firmware set it up to use LG's custom DNS servers, now shut down, and it had no public DNS fallback for this situation. It's amazing how little planning there is for when an IoT server goes down.
It's amazing how little planning there is for when an IoT server goes down.
It's out of warranty, isn't it? Sounds like flawless planning to me.
I mean, seriously. What incentive could a company possibly have to keep those servers running five minutes after the last product shipped goes out of support? Sure, you'll get mad and swear to buy your next IoT widget from some other company. But that company will do the same thing, because they have exactly the same incentives.
This will have to be fixed legislatively... and as a liberty-minded kind of guy, I'm truly depressed by how often I find myself typing those words these days.
Ha! This happened to mine yesterday, spent an hour doing all the resets, googling, did a quick worthless chat with samsung support. I ended up just going to Target and buying a Sony Blu Ray Player. I told my wife maybe somebody might post a fix on the internet in a month or so.
The guy asked me no questions and just said make a service request. I figured they already knew there was nothing to do and just trying to close the chat as quickly as possible.
I was working once in a product company with big customer support section. There was almost zero communication between product and support. We never received any info about what problems customers were facing, and never were asked to provide technical support for the app we built.
> How can they not know this is happening minutes after the update is pushed?
- Maybe it only happens from a cold start, or maybe only when waking from sleep more.
- The push likely takes hours
- It might be hard to capture the right metric for if the push was good in time to stop the push. I wouldn't be shocked if support realized the issue first, but didn't have a way to quickly escalate to the firmware team.
- When the push fails, by the time that team figured it out, at a big company, it could take at least a day to make it into the customer service script.
Just musing aloud, but do we hold electronics companies to different standards than other companies?
For example, imagine that the latest Slack update made all desktop clients crash, and that Slack support simply said to try restarting the client. There would be mass outrage, and the expectation of a prompt fix.
On the other hand, there doesn't seem to be much expectation of an immediate fix from Samsung in this thread.
Or if a comparison more closer to same industry is desired, imagine if the issue was with Netflix instead.
This makes me wonder aloud, are electronics behemoths held to lower standards?
I think they're held to higher standards than software companies, or at least were, since in the latter, "try restarting it" has been a commonly repeated mantra for decades.
Or if a comparison more closer to same industry is desired, imagine if the issue was with Netflix instead.
That's slightly different since it's a service with ongoing payment, and customers who will stop payment if the service fails to serve, whereas Samsung has already been paid by those who bought their hardware.
First thought coming to mind is if this is indeed a bad firmware update as some suggest. Or maybe an x509 certificate used internally to verify something like firmware signatures has expired? I guess a bit of digging through these threads might shed some light on if people accepted updates recently.
I scrolled through way too many posts, and not a single person mentioning whether they tried to reset the firmware via USB, and saw no support person suggesting it either.
I've actually used their firmware update procedure to flash the update to my Samsung Blu-ray player when I got a movie that had unsupported codecs or something. The procedure (download a gzipped directory onto a USB drive, insert into USB port, boot) was well documented and straightforward (easy for anybody who's done related things like installing an O/S), worked the first try, and I've never missed a beat since).
I miss the days when companies cared about the longevity of their products. Companies like IBM and Compaq were infamous for well-placed updates to drivers, firmware, BIOS, and documentation.
I thought about that, but when I looked it seemed like you needed to use the menu to do it, and since it resets about every 5 seconds I don't think it's feasible to actually use the menu. I could get the one I had to open/close the tray reliably.
I would beg to venture expired certificate, firmware goes to validate itself, comes back expired, crashes because this scenario wasn’t considered, repeat.
The Players without Internet access are not affected. The system goes into a bootloop after that, even after disconnecting from the net. So something is apparently changed on the device. Maybe a temp/log file is filled with exceptions and the disk is now full.
Does not looks like that it is an update that broke it, look at that comment:
<<
Re: Blu-ray player BD-JM57C, keeps cycling on/off whenever plugged in
yesterday • last updated yesterday
I have a BD J5700, it started doing the same thing on 06/18/2020. It has never been connected to an Ethernet and as far as I know it has never been updated. I believe it is a preprogrammed failure.
>>
Assuming it is even possible to remotely recover? From the description it doesn't sound like it ever completes startup, and there's no USB port, so I'm not quite sure how local recovery would work?
Keep in mind this is a 2015 BluRay player, so outside a class action I doubt Samsung will do anything and just act like they all just coincidentally failed.
These players do have a "Hold Eject for 5 seconds" factory reset mode, but it doesn't appear to re-flash the factory firmware, only clear the settings.
Also, Samsung doesn’t even make Blu-ray players anymore. They got out of that game in feb 2019. So it’s not like they could even offer someone a discount on a new one.
Class action suits should require at the minimum replacement or repair of the product in question, including shipping to the user. Otherwise, lawyers get paid and I get $0.21 cent check in the mail.
So if that’s the route they go then by 2033 an impacted owner can register in a 85 step process to obtain a $5 check to use toward a new blu ray player that expires 90 days after it’s issued. Meanwhile the lawyers get to keep all the additional proceeds.
Are you sure it doesn't have a USB port? Most blu-ray players do (some of the discs let you download something to usb storage for some reason), and the front panel has a USB logo according to the Amazon listing:
In the legal department, maybe. From one of the replies:
"I also got assistance from chat with Samsung Pro Care reps. I am trying to troubleshoot one of the players. Samsung kindly gave me the reset proceedure for my model which is to hold down eject button for 15 seconds. did not work for me.
I requested if I could try firmware flash by USB. They kindly gave me a firmware zipped file which I unzipped and wrote to a USB stick. However since the players won't stay on the firmware flash proceedure instructions provided could not be initiated."
Oopth. This is going to be expensive. It looks like well over a dozen models were affected.
Could it be an issue with surpassed time relative to when it was initially built? A bunch of people mentioned their hardware is roughly 2 years old: 86400(seconds) * 365(days) * 2(years) = 63,072,000
The only things like this to look out for with sane implementations on modern hardware are at 24 and 49 days after 2^31 and 2^32 milliseconds have passed. At least until 2038.
When is it finally going to trickle into common sense that needlessly hooking an embedded device up to the Internet is a bad idea? An ethernet jack is an attractive nuisance, not a feature.
Why is it stupid if it works? Is my 6 year old smartphone stupid to use on the internet? Is my 20 year old keyboard which works excellent obsolete? Is my 25 year old microwave a hazard?
Money doesn't grow on trees, why would I stop using a device only because it is 5 years old?
I just wonder when they're going to make arrangements with isps to automatically connect to xfinitiwifi or attwifi for telemetry.
Telsa kind of does this with their cars. If you drive near a tesla service center your car will automatically connect to the the tesla wifi. You cannot disable this (although you can turn off wifi one-time until the next time the car is turned on, and maybe that works)
All Bluray video players are required to support BD-J to pass. [0] And BD-J is Java (Supposedly to make better bonus content). Support for BD-J in hardware has been required since '07. Things like internet access are part of the spec.
That's one of the big complications with ripping Blu-Ray discs. Even once you've broken the copy protection it's a guessing game what's played when. Also why VLC still doesn't have menu support.
It's better to separate devices. Use a Bluray player just to play Blurays, and use a TV just to display video. To play network content use a RPi+Kodi or at least a dedicated mass-market streaming device that will be highly incentivized to maintain first-class support rather than its firmware being an afterthought.
That's kind of my point though, word needs to get around that most "smart" features are actually anti-features only there for marketing. I'm even getting some downvotes here, as if this outcome was inevitable. It's not inevitable with a bit of forethought. People as a group need to develop the wisdom to know best practices when buying A/V gear, especially expensive A/V gear, rather than getting suckered into nonsensical anti-features that manufacturers are pushing.
I've got a decade+ old Blu-ray player with an ethernet jack. A network switch is a standard part of my entertainment center. Every time I've moved things around, I've resisted the call to add an extra cable and find out what network services are available. With my network setup I can easily block it from having Internet access, and it's still a hard no. There is just no upside, and a huge downside of it being more likely to break, and then having to buy a new player for the rare times I want to play a legacy disc.
My receiver is hooked to the network, but with no Internet access. It has an embedded webserver that is a remote control. But now that it's out of the warranty period I would never let it update itself, as that's a good way for a manufacturer to break a device with no recourse. Firmware used to be shipped with no updates, and that worked well. Weighing the chance of it being bricked from a manufacturer's update versus an attacker with vulnerability also getting past my router, the latter seems much less likely.
I know I'm bailing out a leaking boat here (cheap embedded cell modems are around the corner, which will necessitate doing surgery). But the market has to catch on some time...
Then make a phone app for it that can push software updates if approved in the app and not automatically. Login with the app, use it as a remote, etc. I dont know, just brain storming.
Not validate discs, but one of the issues is Blu-ray discs have Java apps on them, and some of the Java apps on the discs are, uh... buggy. So pushing out updates to compensate for said discs is often necessary.
Don't these devices have one of those hardware reset holes like routers do? I find them very useful because worst case scenario, just insert a pin and wait for 10s which is great if your menu/software side is broken.
I've played with OpenWRT a lot. This might have changed, but for a long time, routers stored the firmware on a flash chip and settings on an nvram chip. They might be the same chip, but you can think of settings and the OS being logically separated. The reset button reset nvram. If the vendor did an OTA firmware update (which no one used to to) and sent a bad update, the reset switch won't help. Some routers have recovery procedures that flash using the bootloader (also on the flash chip, but it's obviously best not to mess with it), downloading the firmware from a PC. If that fails, that's the end of the officially supported options that a 80th percentile user can do.
Anyone in the 29 pages of comments figure out the problem? It seems to be full of users saying "me too!" and moaning about "I'm not paying to get this fixed!".
I guess today's customer forums is a version of Borges' Library of Babel, but one with very dull contents.
And if the 2nd, 3rd, 4th, etc person said to themselves “Oh look someone has already posted about this. No need to kick up a fuss about it so I’ll be on my way. The community managers or who ever run the forum may not flag the issue as quickly with the next level of tech support.
But if you get an inrush of customers all saying “yo, shits broke...” The fire alarms get trigger sooner, if it’s a bad update it can get pulled sooner to lessen the number of affected customers and start the whole recovery process sooner.
Sure a single high quality post from someone in the tech space who’s curiosity made them break out wireshark and give you a detailed bug report can do wonders. But such reports take time, you got to have someone actually decide it’s worth there time to investigate and report back.
Or you can hear the high numbers of low detail reports and pull the fire alarm cause shits on fire somewhere.
If subject turns around before then, the picture is instantly taken, and I'm instantly getting photos if no faces are detected as well. I've reproduced this across a number of conditions.
This is great for the family photos..
They're really going strong with their software now.