Gunrunner

Let's see, a typical 1024x1024 BMP is always 3MB, the JPG version with the lowest compression (100% quality) is around 600-700kB, the DDS version with MIP images is around 650-750kB. The more complex the skin is, the heavier the JPG will be (conversely, a 1024x1024 skin consisting only of pixels of a single colour would be around 20kB), BMP and DDS however always have the same file size for the same spatial dimensions. BMP is frankly the worst format, it takes more space on disk and it takes longer to load (because nowadays it's faster to decompress a JPG/DDS than read a BMP, this hasn't always been the case). JPG is a nice intermediary, it takes the least space and it's fast to load, however it requires a bit more processing to load than DDS and takes up more VRAM than DDS (NB: with Kerbal Space Program, I'm not sure it holds true for Strike Fighters 2, I'll have to check). DDS are the best, they take only marginally more space than JPG in typical situations, they load the fastest and use the least amount of VRAM (see the NB above). P.S. : "Lossless" JPG produced by some graphic software as well as the one included as stock by ThirdWire may actually be larger than DDS in many cases. P.P.S. : Take into account that DDS is an accepted format only for skins, your best bet for mass conversion is JPG (and even then, there are a few instances where only BMP are read, demonstrating how "sub-optimal" some parts of the code are).

From my experimentations it seems video RAM related, I've tried the following, all with exactly the same settings, all using a mission that gives me a VRAM usage around 1.9GB with the 353 drivers. 353 Drivers : Windows 7 x64, GeForce GTX 760 2GB : No bugs. Windows 7 x64, GeForce GTX 960 4GB : No bugs. Windows 7 x64, GeForce GTX 970 4GB : No bugs. 355 Drivers : Windows 7 x64, GeForce GTX 760 2GB : Bugs, VRAM swapping. Windows 7 x64, GeForce GTX 950 2GB : Bugs, VRAM swapping. Windows 7 x64, GeForce GTX 960 4GB : No bugs, VRAM usage around 2.3GB. Windows 7 x64, GeForce GTX 970 4GB : No bugs, VRAM usage around 2.3GB. That led to three things to test : 1) What happens with 353 drivers when you use more than 2GB on a 2GB card. 2) What happens with 355 drivers when you use less than 2GB on a 2GB card. 3) What happens with 355 drivers when you use more than 4GB on a 4GB card. So I tailored my install and missions to test these... I only have a small sample because swapping cards and drivers is time consuming but so far : 1) Degraded performances, stuttering and similar bugs. 2) No bugs. 3) This is where things became interesting, for a variety of reasons I didn't manage to consistently have more than 4GB VRAM usage (varied between 3.5GB and 4.2GB depending on the run). However it showed something interesting; Above 4GB, both the GTX960 and GTX970 displayed the same bugs as the GTX760 and GTX950, at 3.5GB however, the GTX960 had no bugs while the GTX970 displayed them intermittently (there for a few dozen frames then gone, then some more appears and so on). It would seem that for some reason (frame buffer increase I'll guess) : a) 355 drivers lead to an increased VRAM consumption with Strike Fighters 2 with settings left to default. b) 355 drivers implement a different way to deal with VRAM when the card reaches full usage, favoring the conservation of framerate over texture display (with 353 drivers you slow down to a slideshow before having texture problems, with 355 you lose textures first). Anyone can try reproducing, confirming or discrediting that hypothesis (if true it would mean that stock installs should not have display problems on a 2GB card) ?

Too many mods. In short, you installed a third party RAF bird with decals setup to use a variant of the RAF nation sharing the same NationID as North Korea in your NATIONS.INI, that's why many mods are intended to be used on separate installs rather than merged with others mods.

Oasis ? *snorts derisively* Blur, baby, Blur...

Do335, once more, sorry for the tone but your own really annoyed me and the idea was, from my point of view, so ill-thought out it was insulting (come on, even defending it you admitted it wouldn't work on at last two levels if you were the target audience). But hey, if you have other ideas, I'm game, I'm much, much better at finding problems than solutions.

@Do335, as a matter of fact, yes (just not often for SF), but you like to tell little stories to illustrate your point, let me do the same for mine... - Hey girl (modder), what are you doing this week end ? - Well, I haven't add time to see my boyfriend (work on my Yak-36 project) for a long time so... you know... - Come on, that's lame, dump it and come spend some time with us instead (work on a F7U for us) ! - Nah, I'm not into that. - Come on, we'll pay you, I know you can use the money ! See the problem ? Donate money to modders without conditions ? Yes Refund modders their source materials, info gathering trips or necessary tools ? Yes Properly commission modders who advertise themselves as available for commision work ? Yes Exploit the financial needs of modders to have them work on what you want rather than what they'd rather do ? Fuck no What we need is something to patronize (in the noble sense) modders, and organize commission work, not a Mechanical Turk for Mods. Look, I understand where you're coming from and that it's well meant, but quite frankly even in an ideal world it'd fraught with problems and we're far from living in an ideal world. On another note, if you ever negotiate something with TK regarding the code, make sure to include some measure of support, I'd wager that the source code is not extensively commented (or at least not necessarily in any useful way for a 3rd party) and that documentation if it exists might be a few revisions late (That's not a jab at TK, it's just that for small outfits these things are certainly not a priority, been there, both as the original coder and the one tasked to make sense of someone else's work) and it might save you (or any coder ending up dealing with it) weeks if not months of work to have access to the man himself (not that he wouldn't help otherwise, but if it's part of the contract it's easier to guarantee it).

Do335, it doesn't matter, if it is free but has been lifted from another game, stolen from another modeller, used against the original licence, we'd still have a disaster on our hands. And yes, you would have a lot of foul play, we've only had a few so far and they've usually been swiftly taken care of, but once there's money involved, they'll be everywhere... The most basic will be some guy will buy/license a 3rd party model to adapt it, never minding he did so under the conditions that it's strictly for personal use, therefore his adaptation could not be distributed under terms compatible with your idea. Hell, many would do it without even thinking they are doing something wrong. As for your suggestion that the idea would only be for recognized modders, it is preposterous... You're telling me that you intend on solving the problem of too little modders with too little free time by... not adding any modders but using money to influence modders into reallocating their precious free time in ways that benefit you ? That's even less likely to work and I find fairly insulting for the modders because it suggests 1) that you think they should rather spend their modding time according to your piorities rather than their own, 2) that they are currently not contributing enough, 3) that money can motivate them to contribute more. They are not cats, you can't distract them from their current occupation by dangling a piece of foil in front of them, have some respect. And what do you think will happen the day most core modders will feel modding is becoming a chore ? Patronize them (as in, be their patron), commission them, but don't treat them like animals. Sure, some modders are pros or semi-pros and would be more interested in paid work and have retired from SF modding, but in that case... commission them the proper way, not through some pseudo-kickstarter. And sorry for the tone but you have royally pissed me off. @Stratos, he's probably suggesting that you two contact TK to negotiate one of the following : 1) Commission TK/TW to update the game engine to your specifications 2) Obtain a licence to the engine, with source code, documentation and content creation tools and the rights to continue development for the community 3) Outright buy the game, engine etc... And that anything less isn't enough.

Do335, with all due respect, it's not only not a good idea, it's also a potentially community destroying idea. Basically you're diminishing the efforts of volunteers and incentivizing less dedicated and scrupulous ones to take action potentially leading to decreasing quality control (discouraging existing high quality volunteer modders) and creating a host of problems (reverse engineering, use of asset without proper permissions, downright theft of models). I have no problem with an unofficial fund to help fund source material, fact finding trips and cover other expenses for recognized and high quality modders, or even a sort of Patreon for modders (even though most here would shudder at something they'd consider a handout I think), but your idea as it is formulated is unfortunately a recipe for disaster (remember, the community is not just us nice and reasonable forum-active people). I think the DAT schism is detrimental enough to the community, we don't need another one because we tried to implement a bad solution to a misdiagnosed perceived problem. As mentioned by others, it also puts the community into a whole other dimension from the legal and licensing point of view, and what were nice, tacit, gentlemen's agreements (or ignorance) will turn into a legal nightmare very quickly.

@Snailman, but the pre-purchase came with the season ticket, which would include the laser weapons and opponent plane hacking DLCs. It might be a bargain (or extremely expensive for no good reason).

Slick Cowboy, is that a 2GB or 4GB GTX 960 ? I have issues with a 2GB GTX 950 (which unfortunately can't use anything earlier thant 355.xx drivers) but none with a 4GB GTX 970 (even with 355.xx drivers). I'd wager this has to do with a combination of the game and drivers memory management on cards with "low" memory quantity/bandwidth when playing on high/unlimited with highly modded installs (large textures, modded environment.ini etc).

If you intend to use the dhimar decals included with the game, Dhimar HAS to be [nation256] to correctly use level 0 (country/service) decals. Also make sure that in decals.ini decals supposed to be insignias - and other country level decorations using the country numbering scheme - are set to DecalLevel=0. Another way is to create a nation specific skin duplicate and add 256 at the end of the FilenameFormat of concerned level 0 decals (i.e. FilenameFormat=Insignia should become FilenameFormat=Insignia256 if it's a Dhimar skin). Yet another way is to extract the Dhimar decals and rename them with the nation number you used in your modified nations.ini, but that's the least graceful way to do it. P.S.: The same applies for Paran, except that Paran should be [nation257]. Also, use placeholders between your last nation and Dhimar/Paran, if you skip numbers in the nations.ini it won't work properly.

Sex-ed for war planes, brilliant.

Nice; She's usually F-111B Manatee on my installs to match the equally ridiculously named Aardvark, but that wouldn't make for a nice patch...

Paran will invade Dhimar... again ?!

@russo, yeah, so Wrench has BIG hands, got a problem with that ? :)

Any chance of a 4000-5F (same pylons as the N, same nose as the M, with SPIRALE dispensers at the wingroot (at the back)) ?

Straw-man is already at least 200 years old. It's been used in it's current form and meaning in English for at least a century and with another (somewhat related) meaning in French for twice as long. Furthermore, unlike most other informal fallacies there is no other name for it anyway.

Wait, the door pops UP ? Of all the solutions they chose one probably requiring the most energy to actuate and creating the most problems when it fails (as it won't open cleanly naturally, neither would it close cleanly, while staying in the airflow) ? Ok, to be honest, it's probably the one adding the least weight. I just love this program...

Myth 1 : Can't turn, can't climb, can't run Answer: Doesn't really deny it, only puts it in the context where the F-35 will most often be "unladen" even though they carry coconuts while previous fighters were "laden", and reiterates the idea that "pah with all its gizmos it won't need aerodynamic performance anyway". It completely ignores the fact that over the life of the aircraft : A) changes in needs and requirement will mean that the plane will probably over time not rely only on internal carriage anymore. B) potential opposing planes will not be at a technological disadvantage, making the probability of classical engagement higher over time. Myth 2 : Can't fly at night or in the rain Answer: That's been solved, and thunderstorms are nearly solved too. I'm willing to accept that one. Myth 3 : The F-35B melts through decks Answer: No, it only diminishes their lifespan, everything's fine if we change the deck material. BULLSHIT, that's a straw-man, most serious commenters never suggested it did instantaneous, critical damage, but that over time it would wear it out much faster... which it does, to the point that they're acknowledging that they'll change deck material to solve that issue. Now on concrete spalling... this is entirely dismissed, however, to my knowledge this has not been tested with a fully loaded plane (requiring more thrust thus more heat and/or over a longer period) and not on every type of concrete revetment, I would not wager that all air bases are safe from F-35B yet. Myth 4 : The F-35 can't do CAS Answer: If we redefine what CAS is, provided all our gizmos work as intended and use expensive ammunitions, yes it can. It's all good on paper, but I don't see it working in a real war... Then there's the bullshit about using the gun without exposing itself to AA or MANPADS, the bit about the payload being higher than an A-10, yeah, with external payload, meaning you lose stealth, you can't run, you can't climb, you can't turn... It's enough to fight insurgents, but not a high intensity war. Then survivability... yes, on paper, let's wait their first conflict to say it actually works as intended shall we ? Myth 5 : Stealth doesn't work Answer: Noooo, LF radars need large antennas, we'll destroy them with cruise missiles before using our precious, precious planes... entirely missing the development of ground based radar networks since the 90's (granted, none of these are operational anywhere officially). Then this beautiful bit about advances in radar, making stealth not moot, but still an advantage over non-stealth aircrafts... yes of course, "stealth" aircrafts will still be detected later, but as long as they're detected far enough it doesn't matter anymore. And the final, "No stealth isn't a dead end, we've got event better things... for the next aircraft"... that will do a lot of good to the F-35 for sure. Myth 6 : The B variant compromises the others Answer: Straw-man, straw-man... the real claims about the compromise are about weight increase over the program and the development delays due to the focus over the F-35B and it's engineering problems. Myth 7 : A single engine us unsafe Answer: Non-sequitor about reliability of older engines... Nope, which is true, except that at present the reliability of the F-35 engine is not exactly at the point where it's entirely "safe", but even the most alarmist reports points that it's getting there... whether it will end up being the most reliable engine of all time as intended is another matter entirely, but it certainly will be safe enough. Myth 8 : The F-35 is too expensive, concurrency doesn't work Answer: All true, but only valid for the US and other air forces buying the F-35 over at least a decade. Doesn't address at all the cost problems for early adopters in low numbers like most european customers.

Thanks, guess I'll have to find a way to get my install files or abandon the idea of playing SF2 during my vacations, damn, guess I'll have to content myself with Kerbal Space Program...

Trying to reinstall everything from scratch, far from home, on a test machine, and unfortunately TW's download page seems to be down, as are the servers for the auto-installer (they fail to respond). Am I the only one facing that ? How long has it been going ? Seems like the end is nigh for SF2 on PC...

Deep Blue, without a doubt, any giant great white that can nearly beat Kasparov ought to be a force to be reckoned with !

Typhoid, care to detail your problems ? SF2 seems to work fine for everyone else so far, but you never detailed what your problems were. Likewise when it comes to display, on one hand you say quality is bad, but pictures are great then blame quality and complain about performance... Could you care to elaborate and have you tried updating your GPU driver before doing the rollback ?

It's not dead, it's pining for the fjords.

Campaigns are ok now, thanks.