Jump to content

kazereal

Members
  • Posts

    671
  • Joined

  • Last visited

Everything posted by kazereal

  1. I don't know how many have tried with upgrade from 1.5 vs. clean install of 2.0. In my case the updater did not copy over files from 1.5 so I decided to continue and do clean install. I have same problem with both Mig-21 and Hawk now. Edit: Hawk works when using mission editor. So looks like instant action and mission generator have some trouble only.
  2. I think it is activation problem. There is second version of Mig-21 DLLs in bin-T4 - maybe that is what is supposed to be used with the Nevada? I don't know, trying it out next..
  3. Unfortunately torrent protocol is much more complex and predicting speeds depends a lot on what kinds of peers there are in the "swarm". If maximum reseeding is priority, then you might get situation where host A has piece 1 and host B has piece 2 which they should transmit to each other to reduce load on "master" seed. This would be fine if they have enough upload bandwidth, but many home users have only fast download and very slow upload (ADSL). This means that "master" is seeding as fast as it can as many as possible, but many peers may be useless in distributing the load due to slow upload. And this is hard to predict in "laboratory" conditions what kinds of peers there will be and how the load balancing will work out in practice.
  4. No need for freebies now. There is already one SSM module, another one would not make a difference in getting new people on board.
  5. Steam had to do those imports manually (since they were not Steam generated keys) and they were not getting any portion of those sales that happened outside of Steam. I can understand that they did not want to continue with it.
  6. Not to mention people started hammering servers with checking for updates in continuous loop. That is effectively creating situation similar to denial of service attacks which is not cool..
  7. At least ED will be seeding, otherwise nobody would get the original in the first place.. This likely is alternative to buying service from CDN or similar. I don't know what kind of cost that has but using torrent would be a lot cheaper for ED.
  8. That is systems modeling (ASM), not flight model.
  9. Either way, it seems to be leaking memory at some point and causing performance slowdown due to that.. Hard to figure out more without seeing actual code..
  10. I think the value is not calculated/shown correctly or there is another problem. Do this: * start instant action/fast mission with aircraft (let's say Hawk) * get in cockpit, note VRAM usage * return to main menu * start again instant action/fast mission but with another aircraft (let's say C-101) * get in cockpit, note VRAM usage * return to main menu.. repeat At first, I have something like 3.9GB usage (graphics card has only 3GB). On second loading, I have around 8GB usage. Third time, 9-10GB.. So either the value is not correctly shown/calculated or it leaks memory somewhere or loads same resources multiple times. When there is a performance "hiccup" (stuttering) I notice VRAM usage change a bit so that might be about moving things from system RAM into VRAM? Would that be avoided if memory usage was not so high? Edit: it is not about aircraft/module either, you can do same with TF-51 and P-51 as well with similar results. Edit2: loading into A-10C I get 3068MB usage and things work fine. Returning to main menu and loading into A-10A I get 3735MB usage and I get stuttering and bad framerate..Back to main menu and into A-10C it is now 4278MB..
  11. According to sysinfo you have Nvidia Geforce GTS 250, which is DirectX 10 GPU, not DirectX 11. Is that correct?
  12. Thanks for reporting. Take a rest, we don't want more problems to get in at this stage.. ;)
  13. Best way to support is to buy a module: correct people will get their share then. I donated extra P-51 module a while back to a friend who has bought two modules since then. Good way to get people interested in DCS I think ;)
  14. Calculations for drag from airbrake and gears is one thing.. Triggering the animation/state change is mostly in place if you can find out a way to get the input event into AFM code. That stopped me previously as I could not figure out why the keypress was not passed to FM code. I was guessing it would need better systems code in place to work properly (and binding the control to DCS in a different way) but might be something else too.
  15. Congratulations to everyone involved :)
  16. There's no talk about how long beta period is going to be. Beta is still a phase where new fixes and updates can be integrated to fix problems found during it (or before).
  17. Unfortunately it is not that simple and can't be "split" that way. Don't get me wrong here, I really like having good simulations and hope this one succeeds as well. Problem seems to be that people underestimate the complexity and amount of work involved in flight model. In personal experience software projects are often underestimated in amount of work so they end up very late, over the cost and/or cancelled. So I do hope whoever they have working on this realizes in detail what the work involves. And as said before, in a perfect world we would just input data and sim engine would magically give us perfect flight model. In real world it does not work that way. In a simulation aircraft performance depends on things like aerodynamic lift and drag at specific altitude, engine thrust at different settings and atmospheric conditions and so on. You can calculate various coefficients and factors separately at different parts of simulation and integrate at physics engine but there are limits and often things "feed back" into other parts of calculations. So until there is a video of working AFM on this I'm somewhat worried about the statements made before.
  18. Hmm.. I don't know how that sounded to you but I don't think it was insulting. Tone really does not carry well over text perhaps.. Anyway, there can be claimed that SFM flies well when you are flying level, constant speed, no crosswind and so on and so on. The thing is that does not matter since it is very rare to fly like that: these are combat aircraft and not passenger airliners. They are normally used in parts of flight envelope that civilian aircraft are not used at. And that means the "corner cases" will matter that much more. I think I've expressed my scepticism here.
  19. That is a really arrogant claim. And I would really like to see someone to manage pull that off.. Really, do go ahead. I can admit I'm not expert on aerodynamics but I'm curious to see what you can do. Would not be the first time someone making claims like that..
  20. I'll skip the part about what the difference is in generic case (theoretically speaking), but in the DCS simulation there are very clear differences with using SFM and AFM. Effect of damaged wings to flight model needs AFM. Proper ground contact handling needs AFM. Stalling, flatspin and so on are easy to notice. I would suggest your developer takes a step out from FSX and start comparing what the different flight models really are like.
×
×
  • Create New...