robscallsign

3 年前議論

My daily boot into Linux I'm prompted by no end of updates. Hundreds of megabytes of core libraries, fonts, office suites I never asked for but are installed by default, and of course, the weekly or so kernel update and getting to play the lottery as to whether your computer will boot again.

robscallsign

4 年前議論

I suspect it's the same reason why Microsoft just doesn't sell a version of Windows without telemetry for a flat rate: there is just too much money on the table for tracking and selling your private data over the lifetime of the vehicle (or computer).

robscallsign

4 年前議論

Interesting! One of my goto's for programming is the soundtrack from the game LA Noire, with rain/storm sounds in the background:

https://www.youtube.com/watch?v=xiGKxCAg_0o&list=PL0B80EA34D...

https://rainymood.com/

robscallsign

4 年前議論

Thanks for the reply. First, while I probably have a few thousand hours in the F18 in DCS, and a few thousand takeoffs and significantly fewer landings, I'm just a flight simmer, with no real experience in military or civilian aviation. Everything I say is just my speculations and guesses.

> Do you mean in, say, DCS F-18 the ECW systems and things like AIM-120 performance are fake, either because the devs don't have access to the real performance data or because they are under obligation not to make it too realistic? (That is, I would not be able to use the AIM-120 in real life in the same way it's used in DCS, because its performance is not the same?).

A bit of everything. First, everything in DCS is of course "fake". It's just a video game, and all physics, CFD, and electromagnetic waves are just simulated, and there is no way that DCS is doing high-resolution CFD and simulation of real radar returns in a real time game engine. Everything in DCS is approximated and faked. Everything. Some aspects better than others. The F18's radar performance uses a simple degredation factor to reduce the detection range when an object is below you, so many miles for so many degrees for example. Both simplistic, and incorrect.

> either because the devs don't have access to the real performance data or because they are under obligation not to make it too realistic?

I can only speculate of course, not being privy to the private deals Eagle Dynamics and subsidiary/parent companies may or may not have made with militaries. But both: real missile performance, especially for the modern AIM-120D is classified. There is absolutely zero chance that Raytheon/Lockheed/Hughes is going to give classified missile performance to Eagle Dynamics for a publicly consumed video game on the promise that they won't make it "too realistic". Second, even if there is a military grade classified version of the simulation somewhere, at least from the US perspective, they wouldn't be too worried about modelling the exact performance of the latest AIM-120D, for the simple reason that the AIM-120D is not a missile that the US navy or US airforce expects that they will have to defend against in a shooting war. It's their missile!

They would be more worried about modelling their latest estimates of the Russian and Chinese air-to-air missile performance. And, when it comes down to actually developing tactics to counter these threats, you don't necessarily need to do it in a real-time game engine. You can look at energy performance, range, etc, and determine ranges and work out a BVR timeline to defend against that threat.

Then, when you say, "AIM-120", do you mean the AIM-120 A,B,C, or D, or any subvariant, or export variants? For AMRAAMs, the F-18 in DCS has the AIM-120-B and the AIM-120-C-5 (along with the AIM-7M (and AIM-7MH) Sparrow, and AIM-9L/M/X). Whos to say if they're modelling all of these missiles correctly, as you'll be hard pressed to find an unclassified source of all the missiles characteristics and seeker/sensor performance.

As for the BVR timeline tactics, there is a lot of info out there. If you do a google search for various combinations of the terms like: BVR, Commit, Meld, Skate, Bonsai, Crank, Notch, F-pole, Short Skate, you will turn up various resources developed by both the flight sim community, and retired fighter pilots. These tactics were real BVR tactics at one point, but whos to say if what's trickled out to the flight sim community is accurate, and keep in mind a lot of these come from the F-14, or F-15 communities which were more focused on pure air-to-air.

Additionally, a lot of the material out there now is developed by flight simmers who are tailoring the tactics/timelines to do well in DCS! Not against real threats!

What the exact tactics are today for F-22s, F-35s and F-18E's going against Su-27s and Su-57s in a modern high threat environment? I have no idea.

> sent by data link from a nearby AWACS,

A big thing that DCS does not simulate at all is working in a modern high-threat ECM environment, particular detection and jamming. A radar pulse transmitted can be detected at much further ranges than it can get useful information from the return. Turning your radar on, or your datalink on to both tx and rx, is a huge liability in the modern battlefield - you tell your enemy exactly where you are. Jamming is another aspect - how well does datalink work in a modern environment against a peer threat when they can jam you? What if the AWACS aircraft gets taken out, what if the AWACS has to go EMCOM to protect itself? None of these are simulated at all in DCS (aside from the AWACS being shot down of course).

Anyway, I could go on, obviously I have a lot to say.

If you made it this far, here is a youtube video of a BVR engagement against AI aircraft in the F-18 with our virtual squadron.

https://www.youtube.com/watch?v=YKqypIQ8qiw

We had the support of a human Airborne Intercept Controller (AIC) using the LotATC software, and were doing a BARCAP (barrier combat air patrol), to protect an airspace so that other members of our group could attack ground targets. We had just had two previous BVR engagements, and were low on gas and trying to get to a tanker, when the AIC notified us of a new threat.

In the video we are following a BVR timeline to try to meld/sort at ~40NM, shoot at 30-35NM, and skate at 23NM, against AI Mig-29s with the russian equivalent to the AIM7M.

Anyway, just my ramblings on DCS, etc. Cheers.

robscallsign

4 年前議論

... and trained enough to handle any possible combination of inflight emergency, equipment failure, change of weather, safetly handle any last minute directive or change by ATC, etc.

robscallsign

4 年前議論

I've been a fairly serious flight simmer for about 20 years now, including 13 years of DCS, and have flown the DCS F-18 since it's initial release 4 years ago. This topic gets discussed a fair bit within the flight sim community, and we mostly conclude that we'd likely get the F-18 into the air, but would most likely kill ourselves: either passing out from lack of tolerance and training to handle the G forces, lack of familiarity with the sensation of flight which can wreak havoc on your inner ear and result in vertigo, dizzyness, nausea, or paying attention to any of the hundreds of small details and checks that real pilots do that you don't do in DCS (is the OBOGS working properly? cabin pressurization working properly, icing, etc). Simulator pilots would also not likely be able to handle any inflight emergency or problem in the air. Then, assuming we didn't already kill ourselves during the flight, we'd at best damage the aircraft during the landing, or kill ourselves and destroy the aircraft at worst.

Still, DCS offers a tremendous value as a low cost training platform. The DCS A-10C module was built for the US Air National Guard to use as a training simulator platform to train A-10C pilots, and other countries and airforces are increasingly using DCS to train their pilots. A Spanish company built the Aviojet C-101 module for DCS because it is used in the Spanish airforce and they wanted to use DCS as a training platform. A Chinese company built the JF-17 module for DCS. An Italian company is currently building an MB-339 module for DCS.

You can search online and find images of Chinese fighter pilots using DCS for training. There are a ton of things you can train to in DCS very cost effectively - practicing communications, tactical formations, administrative tasks and procedures, weapon switchology, etc. It doesn't completely replace real flight training of course, but it sure can help countries and militiaries with limited budgets stretch their training budgets.

> I also wonder how come most of this isn't classified

All of the "good stuff" is very, very classified. Particularly electronic warfare, radar performance, modern beyond visible range tactics, modern weapons performance porifiles, nuclear weapons delivery profiles.

A lot of the "nuts and bolts stuff" and basic training materials is unclassified and readily available. If you read through and study all of these documents you'll be well on your way to being a fairly competent virtual fighter pilot: https://www.cnatra.navy.mil/pubs-pat-pubs.asp

Most of the topics discussed there are fairly "traditional" fighter pilot stuff that have been discussed by airforces for over a hundred years now, so aren't really secret, even though they're being flown in a modern jet trainer like the T-45.

robscallsign

4 年前議論

It can be hard for Linux fans to admit when anything on Windows is better, but Voidtools's Everything is one. I've used (and still use) most file indexing and search tools in Linux, and Everything is just categorically better. Aside from the search being just straight up faster, the UI being faster, and the setup and configuration being absolutely painless, the killer app of Everything is that because of the way NTFS has a service announcing file changes any new/deleted/modify file is instantly re-indexed, across any disk, whereas inotify in Linux has many, many limitations in that regard.

While I've got the conch another thing that Windows does much better than Linux is Remote Desktop access.

robscallsign

4 年前議論

I'm curious if anyone has a similar image of the shoreline of the west coast of Alaska, British Columbia, Washington state, etc.

robscallsign

4 年前議論

I've been wondering how the EV adoption will play out in Canada. For example, the city I live in with a population of 162k had a record all time low of -48.3 C, and average lows in January of −17.9C, and 4 of months of the year have average lows below -10C.

robscallsign

5 年前議論

> This happened maybe 10-15 times

The fact that you tried this many times kind of makes you lose all credibility...

> Deliveroo are so much better it's not even funny.

It's like reading a reply from a bot account. Fool me once, fool me twice... but fool me 10-15 times, and deliveroo it is!

robscallsign

5 年前議論

If they're not connected, then they're just two antennas. What happens when you get far away from your wifi hotspot or a cellphone tower?

The voltage/energy transmitted becomes vanishingly small, and smaller than the thermal noise of the electrons in the other antenna, and undetectable. https://en.wikipedia.org/wiki/Johnson%E2%80%93Nyquist_noise

robscallsign

5 年前議論

Thats interesting, thanks for pointing it out

robscallsign

5 年前議論

I had read the LOTR several times, but it wasn't until I listened to it as an audiobook did it click for me just how much of the spoken dialog in the movies is pretty much word-for-word from the books. I get what you and others are saying about Tolkiens narrative style, but a lot of the dialog transitioned well to the big screen.

robscallsign

5 年前議論

It's also the ongoing cost of the voice and data plans, replacement cost of the phone if/when it breaks, replacement cost of the phone when the $3 Trillion dollar tech company forces it to be obsolete, replacement cost for the battery they don't let you service.

robscallsign

5 年前議論

Running some "last mile" type infrastructure would help hundreds of thousands, if not millions of people in Canada who live close to cities.

> Limited funds

Using the term "Limited funds" here you're trivializing the costs and challenges with starting up a Starlink competitor.

robscallsign

5 年前議論

Yes, I won't disagree that running cable or fiber is expensive, but I'd like to see a push to run cable or fiber to places like ours which are on a major highway, and only a few kilometers out of town. Somehow in these discussions running a few kilometers of infrastructure to outlying areas around major cities gets painted with the same brush as running thousands of kilometers of cable to hit every outpost in Nunavut.

The next would be more cellular towers with realistic data plans to service rural data. As an example these Telus plans are fairly representative of what's available. $90 for 20GB, $135 for 50GB. https://www.telus.com/en/mobility/mobile-internet?linktype=g...

robscallsign

5 年前議論

Is this really reasible to scale for a company without their own launch platform though?

As a Canadian living in a "rural" area, 15km from a municipality of 160,000 people with no access to wired broadband internet, I'd rather see whatever funding gets thrown at this startup into actually developing Canadian terrestrial infrastructure.

As a rant, in the 8 years since living here, I've seen cellular data prices double. Yes, double.

robscallsign

5 年前議論

I don't know enough orbital mechanics to understand the feasibility, but it's fun to consider that putting a shade over some of the hotter cities of the world - Phoenix for example, could have a big effect by reducing the need for air conditioning.

robscallsign

5 年前議論

It would be easier to attack/sabotage/JAM/DOS the ground stations.

robscallsign

5 年前議論

I second this. If your goal is to release it publically, chances are most of your dirty laundry and uglier hacks are either pain points that are systemically hard (installation, different OS versions/platform, dependencies, build environment) that everyone suffers with in which case the community can help, or it's in a part of the code that migth not be a core competency of yours, and so the community can also help.

コメント