The marketing move of offering an unlimited plan reveals that storage and traffic are not that expensive and someone made a choice that light users will subsidize heavy users. With that, hiding your data from you and subsequently deleting it, at least without first encouraging you to download it within some post-downgrade grace period, would be a choice, not necessity, and is user-hostile.
If it is an actual necessity—a service chose to market an unlimited plan to attract more users, and then realized they are losing money on storage and traffic so much that they would unapologetically burn bridges with existing users who showed themselves as willing to pay (who maybe needed to downgrade temporarily for whatever reason) with the above move—and yet their strategy is apparently to keep offering that plan (in hopes to turn things around with more light users joining?), I would question whether that service has serious issues with even medium term planning.
No matter their actual costs to provide the service, I'm struggling to see why they should not immediately delete all of your stored files upon cancellation of the storage service.
They are a European company, so you are the customer, not the product and recipient of subsidies. They use less manipulation and dark patterns than an equivalent American company.
You pay, you get service. You don't pay, you don't get service. If they can't bill you, they should try to communicate with you for a few months before treating it as a cancellation. If you cancel, then your choice is clear and you should expect your service to be immediately terminated at the end of the current billing period. If their service is storing files for you, termination of the service means deletion of the files.
There is no need for a grace period when you knowingly and voluntarily make the decision to terminate a file storage service.
> you are the customer, not the product and recipient of subsidies
They also do advertisement (promoted tracks and audio ads) but this is irrelevant to my point, what I described applies regardless, including the fact that heavy users of the unlimited plan and free users definitely receive subsidies, both from light users and from ad revenue of the platform.
> You pay, you get service. You don't pay, you don't get service
The definition of the service you receive and how good it is includes what happens when you decide to off-ramp from receiving it. Changing your service plan is your indication that you want to change service, what happens after that is how they handle it. There is no stipulation whatsoever that things stop being available to you immediately.
In fact, in case of SoundCloud, they themselves prove this, because they did not delete data but instead continued to keep data for free, which means providing you a service that you presumably stopped paying for. The silly move of them was to do that and not allow you to download it, and then emailing the victim urging them to pay to access this data, which makes it 100% a dark pattern and means they are effectively blackmailing customers with proven ability and willingness to pay.
If I remember right, Apple (an American company) handles it better and gives you a month to download excess data if you downgrade, but sure, “dark patterns”.
> There is no need for a grace period when you knowingly and voluntarily make the decision to terminate a file storage service.
If you terminate your use of a file storage service, you would expect your personal data to be deleted. However, no one terminated their use of a service, somebody apparently downgraded their payment plan (temporarily or not).
Sounds like they will warn you about your storage limit for a while, so you can choose which data to delete to be under the limit, before deleting your data at random to force you under the limit. Quite reasonable.
You mean Apple? I don’t think they actually delete any minor excess data that may occur incidentally due to race condition or eventual consistency. Just if you actually downgrade, they do… After a month or so, during which you can still download.
Recently I decided to evaluate it for serious use and start posting there again, only until their new uploader told me I need to switch to a paid plan, even though I triple-checked I was well within free limits and under my old now unused username I uploaded a lot more (mostly of experimental things I am not that proud of anymore).
It looks like their microservices architecture is in chaos and some system overrides the limits outlined in the docs with stricter ones. How can I be sure they respect the new limits once I do pay, instead of upselling me the next plan in line?
Adding to that things like the general jankiness or the never-ending spam from “get more fake listeners for $$$” accounts (which seem to be in an obvious symbiosis with the platform, boosting the numbers for optics), the last year’s ambiguous change in ToS allowing them to train ML systems on your work, it was enough for me to drop it. Thankfully, it was a trial run and I did not publish any pending releases.
If you still publish on SoundCloud, and you do original music (as opposed to publishing, say, DJ sets, where dealing with IP is problematic), ask yourself whether it is timr to grow up and do proper publishing!
This sounds like a classic consistency vs latency trade-off. Enforcing strict quotas across distributed services usually requires coordination that kills performance. They likely rely on asynchronous counters that drift, meaning the frontend check passes but the backend reconciliation fails later. It is surprisingly hard to solve this without making the uploader feel sluggish.
That would explain why the front-end would allow you to attempt something that goes over your limits, but not why the back-end would reject something that doesn't go over your limits.
My bet at the time was that they have a bunch of hidden extra limits based on account age, IP/user agent information, etc. If that is true, their problem is that they advertise the larger limits instead of the smaller limits (to get more users signed up), and that they do not communicate when their extra limits apply and instead straight up upsell you, which are both dark patterns.
That sounds plausible. I've had to implement similar reputation-based limits on my own backend just to keep inference costs from exploding, so I sympathize with the fraud prevention angle. Masking that as a generic quota issue to push an upsell is pretty hostile though.
The feeling of being gaslit, when I calculated and recalculated the length of my tracks and compared it with limits on their pricing page, was quite unpleasant.
Another possibility is maybe they reduced their limits from 3 to 2 hours of audio around the same time. I don’t know if it happened before or after my experience, did not read their blogs or press releases, only made sure I was well under whatever limits were currently listed on their pricing & plans page (I was probably under 2 hours as well, but as this point can’t be bothered to check). Perhaps that transition was chaotic and for some time their left hand did not know what the right hand is doing.
Fair point. I suspect it comes down to ghost reservations or stale caches. If a previous upload failed mid-flight but didn't roll back the quota reservation immediately, the backend thinks you're over the limit until a TTL expires. Or you delete something to free up space, but the decrement hasn't propagated to the replica checking your quota yet.
Fair point. I suspect it comes down to how they handle retries. If an upload times out but the counter already incremented, the system sees the space as used until an async cleanup job runs. It is really common to have ghost usage in eventually consistent systems.
Yes, TCAS II warns all the way down to 100m AGL (around 320ft above the ground), and they were already between 1000ft and 1500ft (~400m).
It may or may not have advised what to do (to climb/descent/etc.) because that is turned off below 1000ft, and they were approximately at that altitude at the time.
Edit: They were handed off to departures before tower’s traffic warning. The near-collision occurred in the middle of tower-departures handoff. Tower was warning them of traffic in hopes they were still on the frequency but they probably weren’t, and they noticed traffic just before they contacted departures.
On ATC side, maybe departures could have been more proactive and warn AA of traffic together with tower. On AA side, maybe they could have been listening to tower for a while as they are tuning in to departures (there were 10–20 seconds where AA was not listening to tower anymore and did not come in on departures yet). Seems hard to blame either of them in particular.
Original comment as is:
If the video is to be believed, the tower did tell American right away (at 1:36 in the video, way before any visible corrections by either plane were made) that there is traffic and to stop the climb. It’s unclear whether American paid attention to tower, because seconds later they came in on another frequency saying they have traffic in sight. When asked afterwards whether tower gave them a heads-up they denied it.
Of course, ITA paid even less attention, considering how they were the original cause of this all and how for 30 seconds they ignored ATC’s request to turn right immediately (issued at about the same time that AA was warned about traffic).
This doesn’t contradict that what AA did was proactive and possibly life-saving, but I have a suspicion that the initial deviation by ITA could have been benign if both crews paid their full attention to comms: what if ITA started to turn 270 immediately as they are told to (while continuing to climb up from 1500), and American simply stopped their climb at 1500? I am not 100% confident.
That said, I would also agree ATC could have been more proactive, harder on ITA (instead of just telling them to turn again 30 seconds later). Presumably they are strapped for resources right now.
(There could be errors in the above in case the chart and different radio communication tracks in the video are out of sync with each other, which is possible.)
If “in response” means replying back to tower on tower’s frequency, then no. After the lady on tower frequency told them about traffic (twice), they came in on departures frequency (it was a fresh contact, they started with “good afternoon, American 4 with you”) and said they have traffic in sight.
Edited after I rewatched the video:
1. Tower handed them off to departures.
2. They said bye and stopped listening to tower.
3. ITA veered left.
4. Tower noticed it and warned them, hoping they are still listening.
5. They were evidently not listening to tower anymore, and did not contact departures yet, when they noticed traffic themselves.
6. They greeted departures saying they see traffic, and veered left.
Later at 2:45 American said tower didn’t give them a heads-up. The fact that departures asked them about it could mean that departures thought they were still listening to tower.
Pretty sure the pilots have a second radio and could be listening to both departures and tower during handoff, but it’s unclear whether that’s routine. If they did it, they would have heard tower’s original warning.
> I believe their avoidance maneuver was a climb change.
According to the chart in the video, AA veered to the left. This maneuver started around 1:51 in the video, which is at least 10 seconds after tower warned them of traffic and instructed to stop the climb for the first time around 1:38.
I don’t know if they stopped the climb around 1:38. If we know for sure that they stopped the climb around 1:38 when tower told them so, then there is a good chance they were indeed still listening to tower and heard the traffic warning. If that’s the case, maybe they thought that stopping the climb 10 seconds earlier was insufficient (and tower was wrong about it).
Note that VASAviation's visualisations are not always 100% synched with ATC radio recordings, and the radio usually has gaps removed. It's a useful overview to see the tracks, but take the video's timing with a grain of salt.
Unless it is out of sync by tens of seconds, however, it is clear that they were handed off to departures and were neither responding nor even listening to tower.
One theory in the comments was that ITA loaded the wrong departure in their computer and just flew it without noticing that they were on the wrong side of the airport and/or ATC's prior instructions contradicted the electronic plan.
They were taking off facing such direction that made runway 24 right be visually on the left, which may or may not have been a factor in them apparently loading and flying the left one even as they confirmed on the radio that they are flying the right one. Possibly tired or distracted pilot.
It seems to be nearly impossible for me to advocate for myself at a place like a hospital.
It might be easier to do this for someone else, but it seems narcissistic to assume I of all the patients is so special. If there’s nobody to advocate for me, clearly I’m not!
Let’s say I try it anyway. I tend to be a slow rational thinker in real-time situations, especially under pressure. If I try to advocate for myself and ask questions, I would need to have time to consider the responses (did I even get the information I requested, what are the implications) and maybe do some research in order to make an informed choice as to whether to proceed or not, or whether to ask further questions. However, if I actually request time and have people wait for me, I enter a high-pressure mode in which I can’t think well. The clock is ticking, the stakes are high.
Even if it’s a simple routine case, I am entrusting myself to people who have the power to kill me. If it’s anything beyond routine, killing or harming me may not even be consequential to them (mistakes happen). It is a very particular type of situation.
The natural thing for me to believe is that all of these people are professionals. If I have reasons to supervise them, it automatically implies I believe they are either unprofessional or malicious, in which case I really should not be there in the first place. The arrangement is that I am not supposed to know better than them. If I try to supervise them, that implies I think I do. At worst it would be disrespectful or offensive and would make them hostile on a personal level (which is always at play between humans, regardless of the protocol), at best it would make me look like a crackpot not to be taken seriously anyway. Besides, if I already assume they make mistakes or are unprofessional, their answers can be false anyway.
On the other hand, I am aware that many, many mistakes are made in hospitals daily, so I know they are not such infallible professionals.
As a result, this makes me very reluctant to go to a hospital or a clinic for any reason. It’s probably bad.
Anyone has advice for overcoming this? Maybe training to think quickly and finding ways out in high-stakes situations like this? Tricking yourself into a mode where you feel natural advocating for yourself and act in a way that makes people treat you seriously without being offensive to them (considering the power they have over you)? Learning to not care what people think in a healthy way? (Please don’t suggest LLMs.)
For anything more than basic software-processed output and utility snaps or selfies, this high-end phone loses pretty terribly to an average hybrid consumer camera.
> The best camera is always the one you have on you .
And if all you have is a phone, then you will only ever have phone camera quality photos. For many, that is good enough, but it’s not really an argument to not buy a dedicated camera, so that you may carry it, and even use it to shoot better photos than your phone could.
Whether you post it on social media or not, if you want to do photography more or less for its own sake, a phone (particularly mid-range) is unlikely to have a satisfactory camera. If you need an ability to make utility snaps, then absolutely.
At this point, whenever there is an incident with neurology-related crew incapacitation or degradation (cf. the Jeju air disaster and how many professional pilots fail to make sense of crew’s actions) and the aircraft uses bleed air, I am donning tinfoil and suspecting a fume event.
For a mind-boggling near miss account (actually, not even a miss, since one of the crew has died in the aftermath), see https://avherald.com/h?article=4b6eb830. It quotes research estimating that in the US alone the number of air fume events is around 2000 per year, while the number of reported fume events is less than 10. Mental degradation is insidious, because you may not even be aware it is happening in the first place, so if you have arrived safely you have nothing to report. Yet the industry is silent (including FAA, which had no record of the Spirit incident).
It should be straightforward to mandate a subsystem that presents mental challenges to pilots to test their awareness, and that simplifies access (e.g., disables the emergency override timeout that apparently slowed down the entry in the Lufthansa incident) if one pilot is out to pee and the one remaining is not responsive.
I am curious if pilots in Spirit and Lufthansa cases exhibit similar neuro-degenerative damage.
Wait until physical camera makers not only license you the unit, but also make everything you shoot belong to them, like software camera apps (e.g., Filmic Pro) do now.
DJI can just add some mandatory firmware upgrade process that offloads your footage to the mothership, and 99.9999% will agree to everything without reading.
Canon's high-end DSLRs used to have a module to sign the RAW files as they came off the sensor, for use in law enforcement and other sectors. This was back as far as 2011.
Couldn't you replace the CCD with an adapter, connect the adapter to the video out of a computer, and then use the camera to "take a picture" of your already edited picture?
It seems to me that any "paper trail" scheme of the sort you describe would have to solve the problems of DRM to work: making the elements that report on the real world (in this case, the CCD) tamper-proof, making the encryption key impossible to extract, designing robust watermarks to avoid analog holes, etc.
So the photo print market is really weird right now.
Remember how, in the 1970s and 80s, they used to have little booths surrounded by parking-lot, and you could drive up to the Fotomat booth and drop off your 110 or 35mm film, and they would go develop it and bring back your negatives and prints, and you could drive your Dodge Charger or your Ford Fairlane to come pick them up?
And then, the pharmacies got in on this, because pharmacies are where the chemicals are at anyway. And at a pharmacy, you could have film developed, and you could also get prints, and reprints, and larger-sized prints, and framed photos and albums and greeting cards and all sorts of things.
And this pharmaceutical extension tradition carries on into the present-day. Now you can waltz into CVS or Walgreens or Wal-Mart, you can bring your USB or your microSD card, or just your phone with a cable, and you can plug in your USB or thunk down a disc, and load it into their kiosk computer, and some even have scanners. And then you can order instant photo prints! And they still can sell you albums, and framed photos, and large-format prints, and posters and whatnot.
Here's the trouble, though: phone cameras don't generate the right-sized images.
I was at a Walgreens and they were selling, like, 8x10 and 5x7 and other standard photo-sized frames and prints. And I upload a photo, and the kiosk complains. Kiosk says it's low-resolution. Kiosk shows me a sample preview, and the edges are cut off.
So I chat with the clerk there, and she tells me to just take a screenshot of the image and it'll work. LOL a screenshot, when the resolution is too low already?
And so eventually I figured out that, even if I took a 50 megapixel photograph with the phone's sophisticated camera, it would not print correctly. I told the clerk: this phone takes photos like a TV set. It's in 16:9 or 4:3 aspect ratios. Those are not the same as 8x10 photos!
So the pharmacies have all this tooling for conventional cameras. I suppose a DSLR could still turn out 8x10 photos. I suppose I could "crop" a photo down in my smartphone on Android. But what I really wanted was to download a PD photo from Commons.wikimedia.org and print that out in an 11x17 or larger. And that was not working out so well.
Phone cameras today are producing really impeccable photos of really impossible aspect ratios. There's a ton of tooling that is specifically made for photographs that were based on the size of negatives and the size of photo paper in the last 70 decades or so. Kodak and Fujifilm and their ilk are still haunting us.
Thankfully there are more online services. Everything I put now into Google Photos. Google Photos will happily generate a photobook and they'll even drop-ship them to my family. I have sent them cool photobooks in the past. I never got to peek at them. No complaints. Google Photos doesn't mind when your photos are a weird aspect-ratio. Google Photos will adapt. Resistance is futile. Prepare to be shown your memories.
> Google Photos doesn't mind when your photos are a weird aspect-ratio.
> ...I have sent them cool photobooks [printed and shipped by Google Photos] in the past. I never got to peek at them.
So you have no idea if the photos are stretched or cut off. (Given how many folks fail to complain about [0] godawfully misconfigured televisions that stretch, squash, or otherwise mangle what they're displaying, I wouldn't take the absence of complaints as evidence of correctly printed images.)
Each episode is chewed up slowly to the lowest common denominator, meaning half of an episode is condescending explanation of trivial matter. The misplaced air of faux mystery, true crime podcast style, does not help. The narrator being occasionally a free speech absolutist and explicitly in favour of money laundering services does not help.
Risky Biz covered Anom in https://risky.biz/RB751/ via a dense if a bit short interview with Joseph Cox, who wrote a book about it, starting around minute 35. Listen to that and respective Darknet Diaries episode to compare. In general Risky Biz offers more balanced, less boring, and up-to-date takes on cybercrime (sans interviews with actual criminals).
Could you reference some of this free speech absolutist background?
I'm willing to reexamine my appreciation for the darknet diaries,but I have never gotten this political take from the content itself. Rather he seems to simply be willing to engage with folks he disagrees with morally, as a matter similar to any journalist might. He does not engage with active crime or anything like that either.
I really like darknet diaries. I don't think it is just about cyber crime though. More like a human social journalism with a heavy cyber crime angle. So that's how you have such a range of content in my opinion.
Also, maybe listen at 1.3X? That makes the explanations and pacing imo easier
Yes. Sorry, I edited my comment for brevity. The episode about Tornado Cash was the one I pulled the plug on Darknet Diaries. His suddenly very opinionated take just did not jive with reality, and considering it is targeted at laypeople I wasn’t a fan of that.
If it is an actual necessity—a service chose to market an unlimited plan to attract more users, and then realized they are losing money on storage and traffic so much that they would unapologetically burn bridges with existing users who showed themselves as willing to pay (who maybe needed to downgrade temporarily for whatever reason) with the above move—and yet their strategy is apparently to keep offering that plan (in hopes to turn things around with more light users joining?), I would question whether that service has serious issues with even medium term planning.
reply