HankLloydRight
No Roads
So am I.
Tesla isn't going to have different 'teams' doing a dashcam. I'd rather they spend their resources fixing the hundreds of existing bugs in the crappy software they've already produced.
Upcoming integrated dashcam feature... How?
- Thread starter insaneoctane
- Start date
So am I.
Tesla isn't going to have different 'teams' doing a dashcam. I'd rather they spend their resources fixing the hundreds of existing bugs in the crappy software they've already produced.
HankLloydRight
No Roads
HankLloydRight
No Roads
If that's your requirement set, yeah, that's pretty easy. But that's just 5% of what a modern dashcam can do.
Multiple (synchronized) feeds (2-channels)
GPS location encoding
G-force sensor (Impact Detection / Speed Limit / Manual)
built-in Wifi connection
Parking modes
Motion detection (video, not g-force)
Cloud uploading/downloading
Real time remote viewing with mobile app
SMS Event notifications
File rotation (loop recording)
Sound recording
Now building all of those features, that work and play together nicely, in one easy-to-use hardware and UI package? Tesla isn't going to deliver anything close to that. And people will still demand the features of a REAL dashcam and install those anyway. The Tesla cam will be better than having nothing, but not by much.
Knightshade
Well-Known Member
Well, let's see what these "advanced" features are!
The model 3 has eight camera feeds, which could be synced easily.
It has GPS already. And more, and better, g-force sensors than most dashcams would.
It has wifi built in already.
Yeah...so could the Tesla one. Already covered that in previous posts in fact.
Still not seeing these magic "advanced" features.
Except the Model 3 can already do all that
The cameras (and ultrasonic sensors and radar) can all detect motion.
It does up/down loads to Teslas servers
It already has a mobile app that lets you communicate with the car they'd just have to add the video feed to it, and let you set what text notifications you wanted, set recording modes, etc.
The car already has a microphone for hands-free calling.
So again... the car already can do, or does, virtually everything you list.
None of it is AT ALL difficult to program on the back end either FWIW.
Since you haven't listed anything the HW can't already do I'm not seeing your argument holding much water- other than the UI to use these features might be a bit rough.
The only other place I could see an issue is data. There I could maybe see Tesla offering an optional, paid, "video cloud" feature if you want it, to cover their costs of the additional bandwidth uploading video files all the time.
Which will be convenient for thiefs to unplug and take with them after a break-in.
Except that any security violation (sensors detect collision like car gets backed into, door opened without key, etc...) could automatically store on internal card, and/or upload to cloud via LTE.
HankLloydRight
No Roads
blah blah blah... Just because most of the hardware may already be on the car, doesn't mean the software to integrate them all in new/different ways is so trivial.
Just saying "it has X already" doesn't make it easy to write NEW software to utilize "X".
Yes, GPS and g-force sensors are there for critical systems.. is Tesla going to allow outside access to them for a dashcam with hackable wi-fi or cellular access?
There's something to be said about a stand-alone device NOT tied to all the internal, critical systems inside the car.
Like I said, this is a ridiculous argument to have. We'll just revisit it in a year (or longer) when Tesla eventually releases their dashcam software (if they ever do) and do a feature-for-feature comparison. If these features are oh-so-trivial to add, then Tesla will add them all, right?
Last edited:
HankLloydRight
No Roads
Actually, the app never communicates with the car, it communicates with the Tesla API servers, which communicates with the car.
So they'd have to add a real-time video feed from the car to the servers, and relay that feed back to the app, in real-time.
Blackview already does this. But you think that's trivial to build the mobile UI and the back-end servers to handle that... for hundreds of thousands of cars? Ostensibly every car that has AP2 installed.
You keep saying "they just have to..." well, how many more "they just have tos" do you have, before you admit there are "just" a lot of complex things that have to happen, in complete synchronization, to get a system like this to work?
Knightshade
Well-Known Member
uh... yes it does.
The fact it goes through another device on the way doesn't change that.
If someone calls your phone, you realize they're not directly connecting with your phone, right? But they're still communicating with you.
Likewise when you use your phone to look something up on the internet, you're still communicating with whatever website, even though the signal goes to a cell tower, then through probably 10-20 other servers and routers and devices on the way, right?
Same deal here.
I mean... so do webcams that have been like $20 for a decade or two.
You keep acting like sending video over the internet is some REVOLUTIONARY MAGIC TECHNOLOGY
And it's really, really, really not.
Also- it's Blackvue, not Blackview.
Depends what you mean by trivial.
Technology-wise? Yes. That's incredibly trivial.
Also- very few companies have their own "back end servers" to handle that. They rent server space in the cloud from folks like AWS.
In fact- guess what the company you cited, Blackvue, does?
Hint: NOT host their own servers.
They use AWS.
https://helpcenter.blackvue.com/hc/en-us/articles/115001072291-Where-is-the-cloud-server-located-
Tesla would literally just need to make a phone call and the entire back end infrastructure would be spun up on demand almost instantly. Adding more servers (which are just VMs) takes a few clicks in their AWS account.
So... yes... trivial indeed.
Now, spinning that cloud service up does cost money- that might not be trivial... which is why I said Tesla would probably charge you for it as an option if you wanted it...and for folks happy enough with local storage/viewing they wouldn't need to pay.
I mean... any?
Because so far virtually none of the ones I've mentioned are complex at all.
The car already has all the data they need. The car already has all the hardware they need.
The server back end is hosted in the cloud- something they already use for other applications.
Writing a good UI is the only part that's complex. At all.
Adding the functionality is quite easy since it already is collecting virtually all the data needed anyway.
We weren't talking trivial. You were acting like the car wasn't capable of those features. I asked you, specifically, what "advanced dash cams" can do that the Model 3 can't
You then listed a bunch of stuff- all of which the model 3 can do
But most of the features you listed for "advanced" dash cams really are pretty trivial to program, or simply extensions of code that already exists on the car (power management for example, or location services for another)
Way to move the goalposts!
Though as I said relatively few of the things you list actually are non trivial to program since it's already doing a lot of it to one degree or another..
...what?
They don't need to "allow access"...the car already has access to them. How do you think the map software works to direct you places for example? How do you think the app lets you see where the car is?
You seem increasingly out of your depth regarding even what tesla cars and the app already can do
Yes there is.
It's more expensive, an eyesore, and requires additional installation.
None of those are really good things to say about it though.
True fact.
As I pointed out- most of them aren't even "features" they are just things the tesla HW already does, they simply need to use the data with the dashcam mode.
But even so I expect they will improve the software as it goes along too- see the updates to the map software or AP2.x for examples... or adding new features like easy exit, or additional functionality to the Model 3 scroll wheels, and so on...
HankLloydRight
No Roads
No, it doesn't. You said:
And I was educating you that the app does not communicate with the car.
Yes, it does. Words actually matter. Going through even one intermediary is a vastly different thing than a direct connection, be it one hop or 100 hops.
No need to mansplain to me how networks work. If that's what you meant, then why didn't you say that to begin with?
Webcams are not the same as dashcams. Show me ONE dashcam made 10 or more years ago that sent any video to the cloud. Just ONE. Are you smoking something, because that's just absolutely not true and your arguments are quickly losing credibility over it.
No need to be a condescending twit. I didn't say video couldn't be sent over the internet, I said that all the SOFTWARE to do that has to be developed on BOTH ENDS to send and receive the video in the formats available and for the requirements specified.
You keep insisting that all the HARDWARE to do this exists, which I agree, it does. BUT NONE OF THE SOFTWARE TO DO IT EXISTS.
I manage a large AWS infrastructure, so again, no need to lecture me on what AWS is and how it works.
Sure, you can spin up as many virtual servers as you wish.. but let me ask you this, genius. Who is going to develop and deploy all the custom SOFTWARE to run on those servers? You think there's some pre-packaged AMI out there that does everything and exactly what a Tesla dashcam service needs to do? ALL of that will need to be custom developed. But oh, that's right... that's also trivial. I forgot.
You don't seem to be well versed in security, do you? Just because certain parts of the car have access to the data, doesn't mean it's available to other, less secure parts. And once again, just because the data already exists doesn't mean there is one line of code to integrated it all together into a dashcam product.
You seem totally out of your depth regarding what level of software development is required to integrate all these disparate data feeds, hardware components, video feeds, user interfaces, and back-end servers. You make it sound like you could just throw it together over a weekend and spin up a few AWS instances.
I'm saying that a Tesla dashcam will NEVER have all of these features working at parity with an advanced dashcam of today.
Yeah, the first update after SIX YEARS. Great example. Here's another example: The browser -- took them oh, SIX YEARS just to update the rendering engine (which is already obsolete), but leaving the HORRIBLE user interface intact. The Media Player which was actually pretty good in FW6.x was completely decimated in FW 7.x and 8.x.. to the point that it's not even at feature parity with media players from the early 2000s. Gapless playback? Forget about it. Playlists? Guess again. Those things are actually trivial but still haven't happened in six+ years. I don't expect Tesla to be able to spend the time and resources necessary to fully develop an advanced, workable, dashcam product to today's standards.
Knightshade
Well-Known Member
Except you were, again, wrong.
The app does communicate with the car.
But like your cell phone, computer, or virtually every other electronic communication in the world, it goes through other devices on the way.
I even explained how- which part was unclear to you?
Then literally nothing "communicates" other than in person talking.
Your magic dash cam doesn't communicate with your app either- it also goes through servers. I even told you who hosts them (it's not the people who make the app or the dash cam)
I think there's pretty clearly a need based on the above. And still seems to be.
you are the one who still seems to incorrectly think your Magic DashCam works any different than these other examples.
Of course they are.
They are literally both cameras that send digital video from one place to another.
You realize "the cloud" is just "a server that's located somewhere else" right?
I mean- maybe not given your other posts so far...
But the concept and technology existed long before the marketing term came along.
So... all of them?
BUt sure- here you go-
Yahoo launches live cam site, but can't handle traffic
That's a story. From ten years ago about Yahoos live video from webcam service crashing from traffic.
And it mentions 3 or 4 other services that already existed at the time doing the same thing.
So, once again, you are factually wrong.
Except where it is, and I just gave you yet another source directly disproving your misinformation.
I mean- you just said that didn't exist 10 years ago- and it clearly did... so....
Again- not really.
The car already does that for crashes.
You'd have to change the conditions under which it sends it, and possibly tweak the stream, but the basic functionality is, once again, something the car already does
Except it does.
The car already knows your location, in software.
The car already knows gforce info, in software.
The car already captures video from the cameras, in software.
The car already uploads that video to cloud servers in certain conditions, in software
The car already makes that video available to view (only to tesla employees right now) in software.
The car already communicates with the owner via the app (again, through servers, since you seem to need that spelled out). In software.
The only thing that needs significant development is a UI- because I'm sure the one they use to view uploaded video now is ugly.
Show of hands on anyone who believes that based on your other posts?
Well, not the same guy for one.
Development and deployment are usually different people.
But again since their servers already do most of these functions you really just need a little integration and a decent UI on top of it.
It continues to not be the rocket science you seem to think it is.
Except, again, it doesn't.
Most of those pieces already exist
The car already does them
I mean.... more than you so far...
Sure. But we're talking about the end user app.
Which already has location data
So you don't seem to be familiar with what the car already does
Again.
In the case of location data- yeah, that's pretty much what it is.
Especially when the car already uploads video with GPS data to Teslas cloud servers
It does it now
No "new code" needed at all to do that.
I agree one of us does. I just disagree on which one
A weekend? naah. I could absolutely do it in a month, minus the pretty UI.
I'm more back-end that pretty-UI.
Which is why I mentioned, with some authority, that the rest is pretty trivial.
And you continue to be unable to support what you are saying.
davedavedave
Member
Geez you guys... Just admit it, you were made for each other.
HankLloydRight
No Roads
There's a big difference you are clearly not capable of understanding. A webcam is just a camera with a USB connection. They all rely on vendor software on the host, drivers, as well as an OS in order to function similar to a dashcam. You seem to be overlooking that part... that there's still SOFTWARE needed to make these things do what they do. A dashcam OTOH is a complete, embedded system that does it all, without the need for an additional host, drivers, or software or OS.
Stop being a condescending asshat. Yes, we agree that the car already does those things individually. You will still need quite a bit of software to integrate all those things together. For a "software engineer" you sure don't seem to understand that software is needed to build a complex system from individual parts.
So, I go buy an arduino with a GPS shield, accelerometer, 3G and wifi card, memory card, and a camera, and what, I snap my fingers and there's magically dashcam software?
Yes, but you still need those people to make such a system function. You can't wave your hands and make them go away.
WTF? Again, that's hardware. Where does this magic server software come from?
I'm not saying it's rocket science. I'm saying it's a lot more complicated than you claim it is.
Again, yes, but not together in a coordinated fashion necessary to function as a dashcam. It's like you think all these different parts that exist are just going to get together by themselves and magically make a dashcam.
Since you can't stop yourself from being a condescending asshat and just insulting me, you're now on my ignore list -- a list now comprised of ONE. Congratulations.
Knightshade
Well-Known Member
On the contrary- I explicitly sourced you a link showing you webcams sending video real-time to servers 10 years ago.
Something you claimed didn't exist.
Just like you didn't understand your own ADVANCED DASH CAM company uses cloud servers to do the same thing until I linked you a source showing that (and then you kept insisting that still wasn't "really" communicating directly even though it was your example of doing just that)
Sure. And that software is already in the car
You can tell, because it already uploads video to Tesla
That's a large part of how autopilot gets improved in fact.
Again you don't appear to understand how the very things you are discussing work.
A dashcam is not a magic lamp.
It's a device with a small computer in it, and a camera in it- it is the host, it runs DRIVERS for that camera using SOFTWARE AND AN OS (usually called an embedded OS for such devices).
The only difference between that MAGIC DEVICE and a desktop PC running a webcam is... yours is smaller. And much more limited in what it's capable of doing.
But again- the Tesla already has all that because it already captures and uploads video to servers
So no need to reinvent the wheel here.
I mean... now that I explained it we do.
But for most of your previous posts you insisted they'd need to "develop" the software to do them- since you clearly didn't know it already does them
Glad you finally admit you were wrong though
Again... not really.
The only "quite a bit" is the effort to put a pretty UI on it.
Virtually everything you want it to do, the existing software already does
Integrated.
It already uploads GPS stamped video, with embedded sensor data, to cloud servers.
Today.
Without ANY new software needed.
So explain again, specifically what you think they need to "add" or "integrate" other than a pretty GUI to access the video?
on the contrary- I understand you don't rewrite a ton of code that already does what you want.
You leverage that existing code as much as possible.
And since the Tesla code already does nearly all of it there's very little back-end work needed.
Pretty much all just deciding on settings/parameters for capture, and power saving decisions for when the car is "off"... plus a nice GUI for end users.
No- since, unlike the Tesla those individual pieces don't already do most of that in software already running
So that's a pretty poor comparison
If I need them, why would I want them to go away?
You're not making much sense again.
No, it isn't.
When I tell you the car already does those things- and you agree how can you possibly agree it does without knowing the software is already there to do it?
The only answer I can come up with is you have no understanding of how any of this works.
you're the only one who thinks it's magic.
or doesn't understand it's already there for the most part.
And you're factually wrong, since it's already doing nearly all of it today.
Except yeah, it kind of is.
It captures pictures and video from cameras, it includes GPS and other sensor data, and it uploads that data to cloud servers where people located elsewhere can view that.
Today.
With existing software.
No, I'm pointing out except for exposing this data to the end user app, and doing "car is off" modes, it already is one
You've spent the entire thread ignoring facts- so nothing new here
This thread has made me realize it's actually been a while since I saw a good old-fashioned pissing contest on a message board.
"I'm an engineer, I know these things." "Nuh-uh, I'm an engineer, and I know these things."
Well, I'm an engineer, I work on backend (Azure based) video processing of webcam and other video streams, and I just think this thread is funny.
Knightshade
Well-Known Member
My wife put forward a standing offer years ago to buy that on a t-shirt for me if XKCD ever adds to their merch
Canuck
Well-Known Member
I don't get the point of this thread. I've had a dashcam running 24/7 in my S for a couple of years now. I changed the relay to the cigarette lighter to keep it on 24/7. That also allows for my satellite radio to stay on when the car shuts down, since I'll often pause programs like Stern if I'm out doing errands while driving and then I don't miss anything and I can fast forward commercials when back in the car. The vehicle sleeping has no effect on the 12volt battery. I do carry a small but powerful lithium jumper just in case (haven't had to use it yet) since the good thing about Tesla is that it gives you fair bit of warning when your 12volt can no longer hold a proper charge.
I believe they were referring to the elon tweet about adding a dashcam functionality into the car using the existing AP cameras without buying a dashcam. person mentioned car going into a lowerpower sleep mode which keeps vampire drain to a minimum, but with an always on native dashcam functionality the car might not be able to "sleep" and contribute to bad vampire drain.
