All Activity

My name is Emils, I am from Latvia and create content about DCS World and I want to start a new endeavor! DCS World Baltics is a new community for people from Latvia, Estonia and Lithuania. Our goal is to create a community where we could fly semi realistic, immersive missions, learn together and create new friendships along the way, maybe even joining some international DCS operations in the future. Join DCS World Baltics today! I think we could find something to do both for intermediate and beginner pilots. Discord: https://discord.gg/8kwQNuT5qA

No sir. They are generalized estimates based on decades of my own experience in the software and aerospace industry and ones I have used to prove to my management that gaining access to source code is vital and worth the money. They may be meaningless to you but from my experience, having the original code for any type of software development improves future development and saves money in the long run.

great video in this reddit post!

Yeah. As per the last great development in AI, or whatever. And the one before that. I'm old enough to have seen all this before. Along with how fusion power was going to make electricity too cheap to be worth metering, automation was going to result in the working week reducing to 15 hours, and genetic engineering was going to eliminate hunger. And I'm still waiting for my flying car and my robot butler...

Please, the tech-bros need you believe that it is 'fast evolving' and be useful to... well, anybody at all 'soon'.

There seems to be a fair bit of evidence that the rate of LLM development is slowing down, to the extent that we may already have almost reached the limits of what it can practically achieve. Ultimately, they are nothing but next-word-predictor algorithms, built to fake 'intelligence' by being fed humongous amounts of dubiously-sourced data.

Может стоит просто подождать пока всё прогрузится? побольше головой повращать. А если в таск менеджер не тыкать а просто альт табится?

Which mission is this occurring so we can look at it, please.

thats a good catch Miki, good to read you.

Hello, I'm sorry for not following up with this thread as I haven't been receiving notifications and we've been working on other projects. We're working on squashing this (these) bugs and hope to have everything complete for the next update deadline. Thanks again, John

Considering IRL they did tests on the CV-59 Forrestal, landing on the Supercarrier isn't even a challenge for the Hook. Btw, they landed and took off from CV-59 with no arresting gear or catapult gear. So imagine doing that on the much more powerful C-130J, it wouldn't even be a challenge

However there is another situation, as I know: OH-58D team for a moment(?) have lost most if not all of its programmers. Module is supported and that's great. But what's the future of this module if Polychop CEO, or how it should be named correctly, will lost the ability to update the Kiowa?

Guys, the Uh-1 was released eons ago, and still cannot load infantry, neither can fly with gunners without those unrealistic hardpoints. And you expect to see Hummers loading trough the ramp?

Yep! Enabling proper steep dives....

What I tried to say is: An existing module would only become incompatible if the API changes. If the API changes, ED is well aware of that as they have changed the API. So keeping something "workable" should not require that many resources. Adding all the missing features might be a ton more work. Also, if a new version 3.0 requires new API calls etc. - the only one who knows very well what needs to be done is ED.

True, but fast evolving.

With REVERSE thrusting! I'm betting we'll see short landing on carriers competitions! This module will be anything BUT boring!

Don't get me wrong, I still really enjoy the hook, especially as it had gotten more updates recently to its systems and flight model. Having both modules in my hangar gives a lot of flexibility. I'm looking forward especially to the triple hook slingloading when it comes out.

Why would anyone want to translate the code? Ad for analysing DCS module source, there is zero possibility of LLMs being able to do any analysis worth the trouble, since they won't have been trained on similar material. THey'll just resort to their usual default of churning out plausible-sounding BS.

Couldn't resist and I gave the 1st mission a try, but for some reason I get stuck on this step: May I suggest, that since this is a Training Campaign, have all missions start with a base score of 50? ... this way you won't fail the Campaign for small issues like this? Cheers, and thanks again for sharing this Campaign Eduardo

I'd assume that ED knows their own API quite well, that helps to more quickly understand what has been done by whatever 3rd party. But it anyhow doesn't help if no source code is available or if no resources has been planned in to maintain and support something like this. The point of having 3rd party developers is... to NOT have to deal with this yourself.

Habe es mittlerweile „ein bisschen“ geschafft. Mit dieser XR Funktion UND einem Area Track rechnet er sich das Bild einigermaßen hübsch. Ich bin wohl einfach Opfer meiner Erwartung gewesen. Bei Begriffen wie „Neu“ „Advanced“ „Sniper“ Blablub habe ich wohl ein klein wenig zuviel erwartet.

And don´t forgett that we have fast evolving deep learning models that are quite good at analysing and translating code.