rFactor 2 Build 930 Now Available

I do get invisible fonts with this build. Selecting track and car is a guessing game. How well that i don't have more than a handfull mods, in addition to the the stock ISI content.

View attachment 15845
View attachment 15846

Apart from that i do get the first time ever a kind of cart feeling in the FR3_5_2014 on palm beach, while i have no idea if that is anyway related to the current build, because i didn't used this car before for more than some test drive now and than, but the feeling with some little setup work and my ffb settings and this build is already incredibly good.

I do feel a kind of friction and rolling resistance, but can't say if it is just related to my personal settings or something else. Howsoever, the car is already very nice to ride.

 

I understand and am sympathetic with the problems of one size fits all UI. However, Windows has three size options that apply to all display sizes. (100%, 125% & 150%) It's not like a lot of options to provide for and I'll bet that a lot of people use 125%.

 

Actually you can set it to 100, 101, 102, 103, 104..... you get the idea. It isn't just 100, 125 and 150. Regardless, you have to code around it and even then it's not perfect because you never find an exact fit with font sizes to compensate for the scaling perfectly. And not all fonts are equal. So ISI could make code changes (in the core) to fix their own UI because they know what fonts they are using, but then break other UIs that have different fonts. Again, this is Microsoft's mess IMO. They should have given developers a way to totally disable font scaling for an app, especially one running DirectX.

 

As annoying as it can be, I think it's important to remember ISI is a small team and they aren't just doing rF2; it's not like the person mainly responsible for the UI is sitting there watching youtube vids while people struggle with text sizes They would have allocated time for working on this, got it done, then moved on - that person could be working on something else entirely that will generate some revenue, or may even only work for ISI on a contract basis. And while the changes I've made are simple, you still need someone to do it, and then work out how to release it - having official patches for core components would quickly get messy, while a complete release might take extra time (that takes away from dev time) to prepare. ISI takes a fair bit of criticism, some of it seemingly warranted, but at the end of the day what they do affects their bottom line so if they were in a position to quickly fix things I'm sure they would.

Not sure if I made clear there were only 2 tests at that point; no, I'm not sure what was actually done and this wasn't the only issue for this person, either, so who knows

Yeah, the fact text size can do this in a game has to be a pain. Obviously I did 125 and 150 because they're the main options, I had a quick go at trying to make those additional ones smaller and reference the first one but didn't want to work for me (potentially a talent shortage on my part). Noel, is it possible to pick up the current scaling in order to try and adjust for it? I know it's not that simple with different fonts and could impact third party UIs, but I suspect this problem isn't going away soon!

 

I did find a lot of talk about this topic (ISI isn't the only one that fights this problem) and there are ways to look up the current scale factor which you can then use to know how many font sizes to go down. I suspect that is how they are handling it already. But even then, it doesn't work flawless. Resizing one box one pixel may be just enough to cause the font to not appear anymore. And it's a pain in the butt to test because you don't simply test it at 100% and quickly flip to 125%. Your forced to log off and then back on. So then you make it look right in 125 but now it doesn't look right in 100%.

It would probably be easier to just have three OSC files and then select them dynamically depending on what scale setting you're on. You could probably round to the nearest OSC file without too much trouble. Then at least it doesn't have to be a one size fits all problem. The OSC file is really small too. The other assets probably don't need any changes. So it shouldn't cause much bloat to have three UIs.

I am sure ISI will work it out. As it is it did work in testing on all three options at one point.

 

Lazza did a fix. it works fine for me. try it
http://isiforums.net/f/showthread.php/23801-930-UI-with-HDR-Core_UI_patch

125ger install in packages and modmanager
ingame you select the UIpatch in Optionen/Benutzermenue

 

I've noticed some Stuttering going on, the Green Bar (presume that's CPU? ctrl C) is all the way to the right with the purple bar 1/4 down.
Clios @ Indy & Loch Drummond with 19 AI. Not bad enough to stop racing but certainly noticeable.

Tried various Mods at different tracks with 39 AI (at times) since 930 was released with no previous issues.

 

Info and a test to try here:

http://isiforums.net/f/showthread.php/23794-Build-930-CPU-bar-maxed-out?goto=newpost

Yeah, I'm sure if it can't be mastered purely through code this would be the final workaround - being able to select different .osc files either manually or based on something picked up.

As an aside, the UI components I did are 99.9% a copy of the Core\UIData\UImaps.mas file and depending on the compression selected in the MAS tool came out between 10 and 14 megs; the UImaps is 28 megs. Usually decompression speed isn't affected by the level of compression, so not sure why the proper file has been left that big (not even sure how to get that size... lowest compression gives 14 megs, uncompressed is 180 lol)

 

I want to report that I see an improvement in this build with regard to the realism of the sim compared to my track car. It is most noticeable with the Eve F2 at Monte Carlo (I have only run a few cars and tracks so far). What I see is more control at the limit of adhesion and beyond. Most notably, it is now much easier to regain control when starting to spin. There are many places where previously once I started to rotate during braking I couldn't recover, and now I almost always can by brake modulation and steering inputs (just like a real car). Again - another significant improvement of realism. Great work and thanks!

 

I don´t where to pst this but will give a shot since i´m not sure if is related only to this build, so...

I can see rooster trails on replay but when gaming cannot see no mirrors. It could be something related to mirror resolution but i never change that configuration and, as i remember, i was able to see it in past builds.

 

Thanks and done

 

Tire contact patch improvements. The inability of the tire to deal with near-adhesion-loss gracefully was a long-standing problem that caused the unnatural spins. I always felt like the rear tires were unable to mould to the surface and instead rigidly tipped on their side a bit (reduced contact patch). Now it feels like the tire is really made of rubber and air-filled resulting in better sidewall flex and other characteristics that keeps more of the contact patch on the road (as in real life). Sometimes, the old tires (especially on some user-created mods) felt almost like solid rubber.

 

No green is GPU pink is CPU. I have this problem too, and I had it at Loch Drummond in Skippy. Was the stutter in the last corner but one by any chance?

For me there has been no pattern to it at all, except one time at LD it did it 2 times on the same corner in online race, the last corner but one.

 

aahh got that wrong then thanks for the clarification, does'nt explain the stutters, I'll respond in the other thread in the tech section.

 

There's a really easy way to remember which bar is which...

Physics = Purple
Graphics = Green

 

I think it's because I've always associated the colour Green with CPU since forever (regardless of rFactor1/2), it's like when Walkers Crisps changed the Salt & Vinegar packet colour....all hell broke loose and it took some getting used too

 

Thank you for the tip. I think i'll wait for an update, anyway I don't do switch like crazy between the content. Me is happy with any combo for longer than some minutes when it's good, and in this case, the current combo is very good already. Often less is more.

 

Good build, installed lite after uninstalling old first and restarted, always works like a champ.

Lester at 7am with everything maxed, incredible mod from 32pixels. Updated skips at LimeRock could be a stand-alone sim. (Although I'd like to improve the resolution in the car mirrors.) FR3.5/2014 at Silverstone, another great example.

Minor issues:

1) Scripted Weather set before entering 'race' is correct in session however in the Standings view, it shows a completely different set of weather icons.

2) Even though I set practice time to e.g. 120mins, the countdown timer in the Standings view shows 29 hours and 59 mins as the start. Also like this in previous builds.

3) UI is not limiting FPS (already reported I think), GPU shows a spike to 90%+ usage, once I start a race it drops to high 60%. I have it locked at 105 max FPS in player json. I would imagine a locked 60 FPS while in UI would be the default. *Using a 970 Strix.

4) CPU core behavior is odd (maybe), I see around 90% usage in Core 5, 30% in Core 3, and 10% in Cores 1 and 8. (Monitoring with MSI Afterburner.) *Using a 4770K OC'd to 4.2.

Everything looks great and runs smooth. Looking forward to what 2015 brings. Thanks ISI.

 

In previous builds (haven't been able to try 930 yet )it was only like that when you only select practice, and no other sessions. In that case the 'practice' changes to 'test day', and the length displayed is also the actual length of the session As soon as you select other sessions besides 'practice' (so 'qualification' and/or 'race'*) then the practice time should be the length you set it to (at least it has always been for me in previous builds)

* haven't tested 'practice'-'qualification' combination with 'race' unchecked, so not 100% sure about that one

 

The session length change in this case to 30 hours, nice bug