Back in the day, you flicked your computer's on switch, and it turned on. (And then it took ten minutes or so to load software from the tape drive – much faster from floppy, but still far from instant.) Now, our processors and storage devices are millions of times faster, and computers can take minutes just to start. Software needs to load itself into memory, then initialise the initialisation routines, then actually initialise, then you have to wait for the data to populate, then it's still slow until the cache has warmed up.
It's 2022 and if I want to play a rhythm game, I have to calibrate my input device because, since the days of PS/2, keyboard latency has gone up so much that I can hear the delay between the button click and the computer's reaction.
It's 2022, and when I draw something on one device, it takes several seconds for the drawing to make its way to the other, even if they're in the same room, because the data's being sent through four layers of format conversion in JavaScript, and the traffic's being sent half-way 'round the world.
It's 2022, and "blazing fast" means "slower, and probably more buggy, than Windows 95".
Most software I use feels more responsive than anything from the Windows 95 era.
[0] https://github.com/GNOME/vte/blob/master/src/vte.cc#L10704
My biggest gripe lately is being forced to move more and more of my daily work off of a very fast and beefy local workstation to an AWS Workspace that feels like I'm time traveling backwards 20 years to a laggy remote desktop connection.
30 million line problem - https://www.youtube.com/watch?v=kZRE7HIO3vk
Twitter and Visual Studio Rant - https://www.youtube.com/watch?v=GC-0tCy4P1U
Seems more rose-tinted glasses than anything.
It took under three minutes to cold boot (~5 seconds from hibernation), took less than a second to load a static HTML page from the 2GB hard drive, and could play tanks.swf[0] for hours without spinning up the fan.
Compare that to modernity: your average Windows 10 machine will normally refuse to cold boot in favour of "fast startup" (basically hibernation) because it takes so long to start normally. If you try that anytime, good luck getting to "I can type with less than a second's lag between a keypress and the character appearing on the screen" in less than ten minutes. If you have an SSD, 16GB of RAM and an octacore processor, you might be able to get that down to 3 minutes – juuuust about edging out the "wait for it to boot, double-click on a document, and wait two seconds for Wordpad to open" of 1999 laptop hardware and software. (And if you give those resources² to Windows 98 it'll boot the whole OS and have your file open in less time than it takes Microsoft Teams to hide its splash screen.)
In a modern web browser, it usually takes less than a second to load a static HTML page. But sometimes it just takes upwards of 10 seconds, sitting there and spinning on a blank page with a blank address bar and (aside from the fan noise) no apparent sign of life.
And… well, good luck with tanks.swf these days; without active cooling, you can feel the temperature difference through the case. My laptop³ went up over 10°C with the fan. A simple Flash game where the majority of the gameplay is stationary graphics idles at half a core, averages one-and-a half cores and peaks at ~2.8 cores (70% CPU utilisation).
I think it's fair to say that computers have got worse.
[0]: https://archive.org/details/tanks_flashgame
¹: Yes, it would beep incessantly during the DOS phase of the boot sequence (BIOS beep, then a pause, then a faster, higher-pitched beep I can only assume came from WIN.COM), then proceed to try to delete all the files on the desktop once the desktop loaded. If you timed it right, though, you could tap the delete key before the first beep, and save yourself the hassle of waiting for the keyboard buffer to drain.
²: Minus the RAM, of course; that's enough to overflow its 32-bit address size and make its memory allocator think there's no RAM.
³: 4 (logical) core Intel i5-5200U @ 2.20GHz, with integrated GPU
Can't you boot directly into a UEFI shell very, very quickly? Is this not equivalent to flipping a switch and jumping right into DOS?
Don't get me wrong. I agree with the sentiment and general message of your post, but it does seem like the goal posts have been moved.
I was at a vintage computer fest a few months back and got to play around on an old ~96 Pentium box running Win95. I was ASTOUNDED by how fast it was.
Most solutions rely on sending a file to some server in probably a different country, and then downloading said file on the other computer. Or, using a USB stick to pass the file from one to the other.
https://kdeconnect.kde.org/download.html
(It's also great as a remote control for your laptop when watching a movie or series on a projector)
I feel your pain, I tried tons of other solutions before discovering KDE Connect. It works well for my laptop/phone combination, but I'm not surprised if it doesn't work for others, given how horrible each other solution was that I've tried.
edit: I wonder if perhaps some part of your phone hardware overheats after N megabytes because it's not built for constant file transfer, or constant file transfer using these protocols at least? What if you try throttling it on purpose?
I think it is the fault of the network stack of one those two devices, otherwise I can't explain that. Maybe that the devices receive/send data faster than they are able to process. (Is this even possible?)
But I could be awfully wrong
Regarding your edit: That's a good idea, but I never attempted that. I can't test that, because I'm in a Wifi where no device is allowed to talk to another :(
Would be great with local network discovery of p2p peers but seemingly browsers are not really interested in enabling that...
$ python3 -m http.server 1337
Then just connect to the IP and Port of the machine hosting the python http server and transfer away.
Aside from Bluetooth I'm not sure how you'd do this without using a LAN or WAN.
Pythons web server is single threaded I believe so any simultaneously connections break.
So a discord server with just me in it is ... also gives you handy direct links for files up to 8mb.
Of course all of these solutions have some serious pre-requisites...
Bluetooth doesn't consistently work for connecting I/O devices to my computer, though. Only peer-to-peer file transfer.
Yes it was slow, but it worked straightforwardly and reliably, every, single, time.
Good times.
The use of depth and colors is streets ahead of the contemporary flat GUIs. You never have to guess the type or state of a widget, or where it begins or ends. Buttons typically have both icons and text. Icons to make it quicker to navigate, text makes it accessible without being fluent in Linear B ideograms.
Just check it: https://www.youtube.com/watch?v=kPv1gQ5Rs8A
It's not an intractable problem, just tricky and annoying. If a progress bar is a tiny fraction of an app's lifecycle it's a lot of work to make it accurate for little payoff.
Implementing a progress bar that works for things like "bytes uploaded" or "number of tasks in the queue" on the frontend is fairly trivial. The hard bit is how to transmit a progress value from the backend to the frontend. Typically that requires setting up either a socket or a keepalive connection, or polling an endpoint from the frontend, which is enough of a pain that very few web apps bother.
Ultimately everything required for this sort of feature is right there in every browser. It's the backend that's usually lacking.
Biggest offender is a large OS upgrade on Mac and it sits on that black screen. It often doesn't even say how long is left, but when it does the 'remaining time' is laughably wrong. Not really excusable when Apple have such a small range of hardware configs.
I've seen some Windows installers 'x remaining' consistently count upwards instead of downwards...or the other classic, it sits at 10% for a while and then suddenly jumps to 100
The hate for power users is real.
Users have become an instrument, rather than ends in themselves. You're attempting corral them so they do something you want instead of empowering them and offering them tools they may need to do what they want to do.
The very notion of power users is deeply incompatible with this.
Now power users are merely an inconvenience, since they usually advocate against dark patterns and prefer purely functional software.
I'm not sure how to really combat this. It might be too late.
Basically the only thing I do is check the autobill payees to make sure none of them are doing anything stupid. I have some trouble with executive function but I have not missed a bill in years because of this system. That might not seem impressive but losing money to late fees is a recurring theme in adhd communities.
I was very happy when it first appeared and my own bank made a big noise about supporting it, so went around to dig for the documentation. Had to sign up for a developer account first, that's all fine and dandy. Write some words about my use case, that's also fine I guess, not optimal but whatever. Finally arrived at the "Success" screen that said something like "Congratulations, we'll review your case and give you access to the sandbox if accepted" which was... Fiiiine.
Some days later, get sandbox + documentation access, write my ~20 line script to move 5 ImagineDollars from account 1 to account 2 within the same holder and want to see it working live but no, here comes a form with ~20 fields describing again the use case, how many users you expect, what you do in case there is errors and so on. I fill it out from the perspective that I'm just writing it for myself.
Wait ~2-3 weeks and receive a rejected email, they don't want individuals to write their own scripts, the service is for products to be built on top of their APIs, nothing more and nothing less.
Great.
Presumably this is some kind of off-the-shelf software though it's white-label so I don't know how to figure out what it's called, or how to find a credit union that uses it. That said, I find such credit unions are easy to ask questions of and get satisfying answers: the people actually working there will pick up the phone, rather than farming it out to a call center states or oceans away. They will therefore probably be able tell you if they offer this functionality or not before you open an account.
It's accepted to wait 6 seconds for your TV to start. Washing machine to light up, dishwasher to be responsive, or even apps to take seconds to boot up....
It makes me go crazy.
- Why don't TVs still have physical buttons? Sometimes the remote batteries die or I can't find it. Just let me turn things off or adjust volume.
- Why isn't there a sliding toggle "checkbox" in HTML? Why do I have to abuse a checkbox to get one?
The price of embedded systems really.
What doesn't work is just sharing stuff between different people and devices. I want to send a web page to my son on his desktop. I want to play a video on the TV. I want to play some music on the main speakers. I want to collaborate on annotating an image with my wife. The whole family wants access to the e-book collection, movies, music, etc.
All these things are possible now to an extent, but each and every one is some degree of difficult to set up, and it's own little thing. And not even mentioning the fractured wasteland that is messaging.
All of this should be baked into the very fabric of computing by now IMHO, a seamless landscape of shared data.
Specifically I volunteer for a group that uses SMS with embedded links for the volunteers to tap to check in and out of a job. Frequently these links are truncated or divided in a way that makes their URL structure invalid. This happens all the time due to the difference in message length and how iOS handles longer messages.
I know this boils down to an SMS character limit kludge, but our Android users do not experience this. The message comes as one block or "bubble" but the iOS users see multiple bubbles that are frequently out of order.
Fahgettabouttit.
Word is also remarkably bad with long (100+ page) documents and lots of different content sources. If you're not pasting as pure text into Microsoft Word, you're playing with fire.
Also, I suspect it's cruft inside the program that has made this problem actually worse over time, not just "still a problem". Thank god Markdown is now more widely supported, including on Microsoft-owned Github.
If you forgo trying to run a branded home router out of the box. You have to find a brand you can run OpenWRT on, and then install and configure OpenWRT. AKA, you need a somewhat specialized skillset that most folks won't have (it's always worth remembering that only 35% of people are able to complete tasks using only a computer and the internet).
In fact, just today a friend made a comment about his Internet that clearly spelled out buffer bloat. I didn't have a good recommendation to make him, and his current router doesn't even seem to have the most basic QoS.
This is not an easy problem to solve though, especially since connections fluctuate which means a static SQM configuration will not always be the ideal.
There are efforts to adjust the threshold dynamically but they're not quite there yet, and are even more specialised.
Get your account flagged by a black box and you could loose decades of emails, contacts, bookmarks and favorites,... Then it's a Don Quixote fight to get anything back.
Could not for the life of me get it working. Had to dig up a usb thumbdrive.
I used to have this working fine in 2002.
Ooh, just learnt that USB 3 fixes this problem! Never mind.
rtings.com measures the keyboard I'm using at 0.5ms latency[0]. I'm confident this is faster than any PS/2 keyboard.
[0] https://www.rtings.com/keyboard/reviews/razer/huntsman-v2
but I still can not scroll though a table of 5,000 items.
Well, something is wrong with your computer then probably. Here is a HTML table with more than 10,000 elements and it works fine for me even on my shitty computers: https://gist.github.com/CapableWeb/5714c672dc433c1963a45625e...
You can even confirm they are all being rendered with `document.querySelectorAll('tr').length` which confirms all rows are in the DOM and it renders smoothly at 60 fps for me.
Lots of other things mentioned piss me off too, but I can usually see (to my dismay) where the perverse incentives lie, or the lack of a market drive, or whatever.
This seems like a problem people would like to see solved, where nobody has reason to get in the way of solving it, and that should be solvable by now.
It’s not perfect, but it’s pretty good.
Google's keyboard is great for exactly this (I type in 4 different languages regularly).
Budget cuts every year, ministers more preoccupied by putting their name on a law or reform than to actually improve things.
It's basically printers all over again. It sucks!
[0]: Assuming you don't need color or images. Black and white images look awful.
Even 30+ years ago, it was simple. The phone has a docking station, and you simply place your phone into it.
Why do I have to fiddle with a cable to charge my smartphone? Some phones support wireless charging, but they're rare, and some manufacturers have dropped support for it.
I don't remember it being so simple. Every mobile phone I had in the 90s/early 2000 had a different charger, even within the same brand. They all had little holes with one pin in them, but all had different sizes, or completely different "couplers". I still have a assorted box with probably 10+ mobile chargers from the same amount of phones.
At least now all Android phones are mostly using the same kind of charger, and Apple being forced to support it as well. Hopefully we'll all be using USB-C for quite some time now.
I'm referring to cordless phones, not cell phones:
https://www.google.com/search?q=cordless+phone
No fiddling with any wires needed. Why can't we get docking stations like these for cell phones?
Not that I have seen them myself anywhere in real-life, but seems there is bunch of charging stations for various mobile phones (at least the ~three different models I've had last years (both Android and iPhones)). Maybe try one of those :)
Nowadays qi wireless charging is a good alternative.
Manually turn off bluetooth once disconnect from paired device, such as a car or headphones. Save battery and reduced in-store tracking.
Interactions with smart phones still requires multiple attempts to properly format text, fix corrects on unknown words or industry jargon, and copy and paste text.
Android 12+, Google camera no longer follows the standard of allowing to select the photo gallery application and requires open it manually. Older versions allowed for it. Now Google forces you to use their application and break standard. Another Do as I say, not as I do.
The linux touchpad experience... After shifting around responsibilities for a decade between libinput, compositor, desktop environment, applications and back, not many people seem to really see the importance of the touchpad as input device. Especially precisition, interial/kinetic scrolling + rubberbanding and palm detection do not feel optimal.
After digging around with libinput and libinput-config on fedora 36, I have achieved at least an acceptable level of touchpad usability and I'm looking forward to 2023 :-)
Open source devs have reverse-engineered the pointer acceleration curves for countless devices and should be thanked that one can use an off the shelf laptop with a working touchpad instead of thinking they're somehow to blame.
I have a term I know I used in some set of documents. I can't remember where I put those documents in some file structure. I use the search function. Friggin' 10 minutes later it finally stops searching, returning a gazillion files which have nothing to do with the term I was searching for.
Yeah, yeah, search is hard on local systems. Turn indexing on. Blah blah blah.
Like, I have endless Halo and ads in the home menu, but I don't have the ability to actually use the computer in any productive way.
My theory is they deliberately sabotage it so you have to use Google instead to try to re-find what you were looking at.
Special bonus: suspend/resume on even the most vanilla of laptops. The one I'm on now keeps hanging on wakeup because (I'm 90% sure) the fingerprint-reader driver is total trash. I've been hearing lots of reports of suspend/resume problems with 12th gen Intel processors, on multiple platforms. Windows or Linux, same. Again: how is this something that vendors can't even test properly, after all this time?
1. Your new language will die, but not because of the language, but because it's so hard to build out a standard library for it.
2. We still don't agree on the basics and fundamentals.
3. We are arguing from elitism (and its sibling, anti-elitism) more often than "make it easier".
HomeKit accessories in general have been extremely hit or miss, I had two Logitech Circle View cameras that constantly lost their network connection and had to be power cycled to get back online. I tried many things over the almost 2 years I had them set up in my house and nothing seemed to improve their connectivity. I replaced the indoor one with an Ecobee camera, and the outdoor one with an Eve camera, and haven't had any network issues since. I've seen similar behavior with "cheaper" HomeKit enabled light bulbs too. Certain brands seem to have major issues with connectivity, while others are 100% reliable. Maybe Thread will address this?
"Amazon Customer Service"
as a message's sender when the full name and email that come in the message are something like
"Amazon Customer Service" <contact@amazonfakedomain.com>
So many technically-sophisticated verification processes are currently run on incoming email to throw out spam and scam messages, yet this particular, rather clumsy form of attack remains unaddressed. Email clients will warn you about everything else but this! I can think of a few simple ways in which clients could detect suspicious sender emails and/or names and help users know that there's something seemingly fishy about a particular sender. Yet I'm surprised that so little attention is paid to this attack vector.
But - in many cases people need to transfer files asynchonously - either because they can't guarantee both machines will be online for the duration of the transfer, or the other person wants to receive the file at some other time other than "now".
So - email really isn't a solution to my mind.
You mean "works if you don't actively try to break it".
Most stories about Linux breakage involve esoteric distros (especially bleeding edge "rolling" distros or one-man Ubuntu-derived distros) and users trying to "customize" shit they shouldn't.
Install Ubuntu and it's fine.
No I mean "if you don't want to share your screen"
Apart from that the whole display servers world is just not working well. Some things still don't support Wayland, you can't share screen easily, X has it's own issues etc. HiDPI is still problematic with both; you get weird scaling issues with apps and and the occasional humongous mouse cursor. Using multiple displays with different DPIs you again get scaling issues. Setting up multiple displays still requires fiddling with xrandr, xinerama etc.
Just let me know if you need more examples. I tried to fully switch to Linux many times spending lots of time + money (bought one of the "recommended" devices).
HDR is just a marketing term for software that increases the colour and black saturation of content beyond the actual input. For photography that means more saturation than what you see in real life, for video that means tweaking it after you receive the signal.
An OS can't "support" or not support HDR; if a video has HDR and your display has the dynamic range then you'll see it in "HDR". Unless you actually want your OS to post-process everything on the screen to have marginally higher contrast than normal?
https://support.microsoft.com/en-us/windows/what-is-hdr-in-w... https://www.pcworld.com/article/546269/auto-hdr-deep-dive-ho...
> Setting up multiple displays still requires fiddling with xrandr, xinerama etc.
Dunno what you're doing but you can literally just plug in another display and it'll work... And per-monitor scaling works on Ubuntu/Gnome.
Not even remotely correct. It is definitely not a marketing term. HDR is set of specifications and formats for capturing, storing and displaying visual data with a higher than "Standard" dynamic range. It basically carries more data and it requires proper color management. It has to be supported by relevant drivers and the display server. Right now there is no way to view HDR content (video, photos or games) under Linux.
If you are interested; there is some effort going on to enable this but it will probably take a couple more years.
https://gitlab.freedesktop.org/wayland/wayland-protocols/-/m...
"HDR" is just metadata to pass along to the device to tell it how to optimally adjust the colours (or not adjust them). It's main use seems to be so that HDR displays can oversaturate "SDR" content in HDR mode.
> carries more data
Not really. The metadata isn't more colours or something, it just passes info to the device about certain aspects of the video.
HDR content will be HDR if the display has enough range regardless of whether or not the OS has HDR "mode". It's more about adjusting all the other content.
Dunno if you ever had a plasma TV back in the day but they could display way more contrast than LCDs of the day so you could adjust the settings to change the way all content looks. HDR metadata is basically that, but on a per-video basis. Useful enough but not some game changer... Which is why there's tons of articles about what HDR does and doesn't do.
Edit - where HDR really is great is photography... You take multiple photos of the same scene with different ISO and shutter settings then stitch them together for more colours than would otherwise be captured. But a screen can either display colours or it can't.
HDR content "will not" be HDR if you're playing it under Linux. It will however be HDR if you play it under Macos or Windows (keeping everything else like the screen and content same). There is no way around this right now. So in basic terms you will not see the dark blacks and bright whites under Linux.
> Edit - where HDR really is great is photography... You take multiple photos of the same scene with different ISO and shutter settings then stitch them together for more colours than would otherwise be captured. But a screen can either display colours or it can't.
It's the same for video or games really. With HDR you can record and display more details and contrast.
And that HDR photo you just described will not be shown as HDR even if your display supports HDR if viewed under Linux
But anyhow, Ubuntu supports 10 bit colour and has for awhile:
edit: and 10 bit color support on its own is not enough for HDR. I mean I don't know why we are even discussing this honestly. HDR under Linux is not supported at all. It's not an opinion; it's an objective fact which you can confirm in a couple of different ways. You can not view HDR content if you're using Linux no matter what content or display you have. There is an active effort in the community to have HDR support in Linux but it will take a couple more years probably.
I've "tried" Windows too and simply can't deal with how terrible it is.
Ever checked in on the youth lately? It's all smartphones, game consoles, iPads.
Students will use MacBooks/Laptops/Chromebooks for school if their course requires one but "devices" are the main computing device nowadays. Heck, even many non-technical business people have switched to iPads. Kiosks? Tablets.
The PC is no longer "the" personal computing device...
2. I keep giving different variants of the same command to my Google Home, which often gets answered with "I have no idea how to do that", or sometimes minute grammatical details change the command for the computer, or sometimes it refuses to remember a task that it did the day before with that exact phrase, or sometimes it randomly does something completely different. Why does "Turn living room off" do something different from "Turn off living room"?
I am slowly moving away from Google's ecosystem because nothing ever works seamlessly.
Sometimes double tap will select a word. Sometimes double tap will show the clipboard menu. Sometimes double tap will show the clipboard menu, then quickly dismiss it again. Sometimes to show the menu I have to single tap, but sometimes I have to long press.
If it's a single-line text box, I can't tap anywhere within it to get the clipboard menu, I have to tap (or hold, whatever it is today) on the tiny part at the beginning where the cursor is.
Then there's the whole nightmare of actually selecting the text you want correctly.
Pasted the wrong thing? There's no undo. Just suffer.
Horrible.
I have to type 25 character random passwords into AppleTV with no mistakes 5-10 times from a remote.
On that note, it's 2022 and I cannot set up an AppleTV in the Home app on a Mac. It's apparently impossible.
1. Devs incorporate an OAuth code based flow. Many apps do this. Passwords should rarely need to exist on a TV device.
2. Paste using the remote app if you have an iPhone.
The first is caused by dev laziness (and possibly better for rare scenarios where there isn't a second device) and the second is perhaps a cost thing (do you have an iphone?)
The typing experience on AppleTV is actively user hostile - the password is masked (you cannot reveal it) so if you even think you've made a mistake in one of your 25 characters, you need to start over.
The second doesn't work for me: we only have Android devices. It's unclear why our 3 MacBooks aren't sufficient for this purpose; I didn't have any suspicion that this would be the case.
I've seen a a database-centric version of this (I don't recall the name of the site though), but not a domain model version.
There's lots of other cloud cost stories like CloudFlare R2 egress saving people millions on large file storage.
I'm surprised decentralization hasn't hit cloud providers as hard yet either.
With that hotspot example my Telco in NZ (spark) blocks me from using a Wi-Fi hotspot on one of my android mobiles, which is ridiculous that they can turn off an Android feature on my phone for profit.
I still have to fight with it all the time, and cannot get it to work as expected. And that's with the most standard use cases:
- System always getting muted after unlock/reboot (for some reason)
- Playing sounds on the speaker despite jack headphones being connected (I have to manually switch the output)
- If I unmute the sound (that wrongly got automatically muted) with my earphones connected, it starts loudly playing my music on speakers at maximum volume.
- UI not reflecting the actual situation (it shows muted speakers despite playing stuff on it for example)
It's a credit card. It doesn't have a damn PIN in the US, and I have to go inside to pay for gas every time. I've quit buying gas at the area Mobil stations because of the hassle. Most of the other chains seem to get it right.
How is it that Europe has been doing chip cards (with PIN) for decades, and we can't get this right?
Oh, and Paypal's payments not working 50% of the time. Probably something with using firefox or being on linux, or having an adblocker. I have a secure password. I use 2fa. It still fails so often, I now avoid it.
Every blog post would have every reader be able to correct small mistakes, fix broken links etc. etc. Basically a distributed maintenance system for the web where stuff that is actually used by people is keep up-to-date.
It's no easier to send a confidential message using the most widespread communications protocol than in 1980.
Searching for anything serious/interesting I pause and think: Do I have it locally? Else, which specialist databases or sources I might try. Else use a bash script that hits a bunch of "alternative" search engines. Finally, use Google/Bing as last resorts. What happened to search?
What frustrates me about updates is that they nag at log-in when I've already been waiting for the computer to boot and just want to get on with the task at hand.
While I'm there, the path from media rental to "ownership". If I watch a movie for $4.99 and it's good, why not take the $4.99 off the price to own it rather than making me pay full price?
Why isn't there a single model for sharing video and alerts to a local PVR? Why do so many crappy cameras require internet access and 1st party apps?
I want cameras, I want alarms, I want heating and lighting automation, and I don't want to share access with some kid in China, or have nine different apps.
HomeKit? Sure, but on my terms, not Apple's.
I am in Germany.
or maybe im the only one who wants this.
It might sound like a joke, but AirPrint on my iPhone has been the most reliable way to connect to my printer by far.
At this point, I'm absolutely convinced I'm doing something wrong. Maybe this is painless in Chrome?
You are just describing PC fragmentation.
It is so incredibly slow and inept at finding anything, whereas the app Everything by voidtools has no problem and works instantaneously.
How have we not fixed this yet?
Ebikes are incredible. They are the most efficient form of transportation there is, they don't pollute, they are speedy, and they are compact!
- No weather protection
- Don't work (well, if at all) in snow, ice, heavy rains.
- Limited (or zero) safety features, despite going fast enough to cause injuries and using the same infrastructure as cars
- Limited speed and range
- 2d congestion limits throughput
And those "millions" of reasons all boil down to R&D costs. Reliability, inspections, takeoff-to-landing autopilots (they exist - have existed for over a decade - for military helicopter drones), etc. are all cost issues which are slowly being worked on.
The progress could be faster if it didn't rely on angel and startup-style investments and the associated NDAs and trade secrets, but I don't expect that to happen in today's political environment.
Most drivers (in the US, at least) have trouble enough navigating two dimensions. Adding a third would undoubtedly make matters worse. Hell, just look at the safety record of private jets.
RCS as standard/API available to use in any app.
No special drivers required, everything I've connected it to has worked flawlessly including macs, windows, my Samsung phone.
USB2 only seems bad on the surface, but the combined bandwidth of the webcam/mouse/keyboard/mic doesn't come near saturating it.
It goes "0 Days, 1 Day, 2 Days".
Not "0 Days, 1 Days, 2 Days". Or, worse, "0 Day(s), 1 Day(s), 2 Day(s)".
• are there 3 or 4? (Plural isn't just zero/one/many in every language.)
• is the quantity small-ish, or large-ish, compared to normal?
• what's the grammatical position of the noun you want to pluralise?
You'd have to have something a lot more sophisticated than string substitution. And there isn't really a pressing need for it. Day(s) works fine.