The "What Are You Working On?" Thread

well we do have
https://www.racedepartment.com/threads/ac-modding-questions-thread.129295/
and

They are more direct but once again I'm thinking of taking the debate out of these threads where it can range further without causing the issues we've seen.
 
And here they are: The three blast furnaces of Schalker Verein, Gelsenkirchen.

1.jpg
 
Found a bug with his mipmap dithering/rotation not being turned off by DISALLOW_MESH_OPTIMIZATIONS =, because it was like a random hardcoded line outside the normal pipeline.
He really loves hardcoding random things for reasons mere mortals cannot understand.
 
on the other hand
we are not that many people

separate WIP thread CSP, and WIP thread modeling, and WIP etc.
or however the name, would maybe divide things a bit too much ?
Yes there is a danger in that I agree, my argument is that the more esoteric discussion about methods and best tools can often distract from the more centered modding threads.
Of course nobody is forced to join in that debate. But it's really up to the modders and I'll stand by their wishes.
 
inb4 this post gets deleted for "derailing" or "off-topic" but I'm absolutely sick and tired of people who never had to deal with this themselves telling me how there's nothing wrong with CSP and Ilja is the god coder while the entire thing is held together by hopes and dreams.
I deal with Ilja quite often on our project, and things like this are completely normal, if you implement something, something else gets broken along the way, and some wonky code/design choises usually have blood & sweat behind the reason of implementation.
But still ranting this much about something that's implemented by one guy, who does so much code at the same time - IMO either help him improve things with direct contact, and more constructive attitude (asking why's that made that way, for example, instead of personal attacks on him), or don't use it.
 
Last edited:
help him improve things with direct contact, and more constructive attitude
I have been dumping bugs and forwarding modders' feature requests into his DMs since early 2019 up until he closed shop and disappeared from everywhere. If he couldn't be bothered to reach out and provide some kind of contact info at least for the people who have been testing development builds for the past year or so then I don't think I have much of a reason to be bothered either.

if you implement something, something else gets broken along the way
I can understand that, what I cannot understand is why it takes him multiple months to fix bugs that have been confirmed and have reproduction steps provided yet he adds 10 new features with their own bugs in the meantime and then rewrites the code because it's beyond fixable instead of fixing the base first before adding new things to a solid foundation. Or when he fixes bugs in something, then decides to rewrite the feature entirely and the once fixed issues come back like a boomerang. How?

I know the last thing Patreon cares about is fixing issues but it seems counterproductive no?
 
Last edited:
I can understand that, what I cannot understand is why it takes him multiple months to fix bugs that have been confirmed and have reproduction steps provided yet he adds 10 new features with their own bugs in the meantime and then rewrites the code because it's beyond fixable instead of fixing the base first before adding new things to a solid foundation.
Well you kinda answered your own question here. From what i understand, he's not the know-it-all guru, but one who's constantly learning new things, and implements something temporarily before there's better way to handle it, which sometimes requires these new features first, to check if it's working or not.
It's frustrating, but code modding didnt exist in racing sims that much before AC, so it's something new to handle. Usually in studios, there is art/general direction, and those people think of user experience, or workflow of other team members, so they decide if one should continue working on new features or better halt it or stop it for good. (which is also mega-frustrating, but for the guy who's working on those features, you invest so much time in that, and then someone tells you it's not viable, just stop). Not quite the case with CSP. You could argue then, he'd better then implement it all first how he wants, then publish it, but well, then we wouldnt have CSP at all, for some 5-10 years more, and by then ac modding would decline anyway.
 
Last edited:
implements something temporarily before there's better way to handle it, which sometimes requires these new features first, to check if it's working or not.
That's fair. The question is, why is "temporarily" often 6-12 months? Why is /proper/ implementation taking so long while he's adding new features with their own share of issues because they're either
A) experimental
B) relying on a broken/patched code
C) implemented incorrectly from the start due to insufficient knowledge/experience at the time
D) all of the above?

Usually in studios, there is art/general direction, and those people think of user experience, or workflow of other team members, so they decide if one should continue working on new features or better halt it or stop it for good.
And that's something I noticed long ago. There's nobody that can say stop to his endless feature creep so it just keeps getting more and more unstable. Last recommended build (v0.1.60) has been released in MAY 2020. That is 17 months ago.

You could argue then, he'd better then implement it all first how he wants, then publish it, but well, then we wouldnt have CSP at all, for some 5-10 years more, and by then ac modding would decline anyway.
No, I'm not saying that. Feedback and testing is integral part of development. What I'm saying is, maybe it's time to take a step back and start prioritizing fixing the whole thing before the code becomes an unfixable mess full of uncommented hacks and fixes that cannot be changed without breaking 3 other things depending on aforementioned hacks in the process.

Focus on one thing at a time, make sure it's fully working, move on to the next. Especially if new features rely on existing code.

Edit: On top of that, I don't even think he has estabilished any "proper" way to keep track of or report issues, at least not that I know of. So yeah, given that he's no longer reachable via PM means there's no way to report anything now.
 
Last edited:
Would a group of coders help with the situation ?
Instead of ilja/x4fab only, why not creating a small and competent group of people able to work together and release some of the pressure of working alone with a gigantic spaghetti code ?
I guess it's a matter of either ilja wanting to do that / people wanting to work on CSP ?
 
needs to have the glass doors on that corner added still, but looks alright I think. Also fixed a lot of buildings' heights after notiving I actually had not deleted the "bigger" lidar mesh from my main work file. Not fine enough for buildings, but should be good for roof heights.
1.jpg

I need to figure out that fresnel shader to "fill in" the satellite image I use for much of the tracks surroundings. First trials weren't good.
 
Would a group of coders help with the situation ?
Instead of ilja/x4fab only, why not creating a small and competent group of people able to work together and release some of the pressure of working alone with a gigantic spaghetti code ?
I guess it's a matter of either ilja wanting to do that / people wanting to work on CSP ?
this would be not so much defense of Ilja this time :p but i guess it's how capitalism works. He's full-time on his shaders most of the time, and he makes the most of living from that. And you can't sell much to general interested public by publishing "fixed mipmap code" or other backend sort of things. Same with most of the studios these days, can't just fix existing content without releasing new one (and thus have enough funds for fixing old, but more limited time for that sadly). And general public in this case would rather throw money at beautiful rain, reflections and things like that.
 
Last edited:
this would be not so much defense of Ilja this time :p but i guess it's how capitalism works. He's full-time on his shaders most of the time, and he makes the most of living from that. And you can't sell much to general interested public by publishing "fixed mipmap code" or other backend sort of things. Same with most of the studios these days, can't just fix existing content without releasing new one (and thus have enough funds for fixing old, but more limited time for that sadly). And general public in this case would rather throw money at beautiful rain, reflections and things like that.
You don't know that Ilja has a 9-5 and CSP is a side-thing?
 
Seems to me Ilja is man of fantasy these days.

Before csp was side gig but now full time gig??

He has been writing shaders for ams2 recently and did even some things for rf2.

If csp is now full time thing how is it on Discord he doesn't even reply to pms.

Recent patron post had video soft collision tyre wall with no documentation nothing its like its there good luck.

There is so many possibilities with csp but nobody can understand/use most of it and if you do next update could brake the whole thing.

csp code is already huge.

 
I really like following the wide range of new and interesting stuff in this thread, but it slightly blows my mind that people have very strong opinions (and apparently even expectations) about how another modder chooses to go about developing their mod (i.e. effectively their hobby afaik)- particularly when the mod for the most part seems to make a whole lot of AC players very happy for free (or a small monthly donation).

Sure, they might not follow industry best practice, or answer DMs, or listen to well considered advice, or whatever - and you might find it genuinely frustrating - but just because a lot of people choose to support a thing does not necessarily make the thing a democracy, or imply a requirement to provide customer support and documented source code for the thing. Other mod(der)s are available. :O_o:
 

Latest News

How often do you meet up (IRL) with your simracing friends?

  • Weekly

    Votes: 47 9.3%
  • Monthly

    Votes: 21 4.2%
  • Yearly

    Votes: 31 6.1%
  • Weekly at lan events

    Votes: 3 0.6%
  • Monthly at lan events

    Votes: 2 0.4%
  • Yearly at lan events

    Votes: 12 2.4%
  • Never have

    Votes: 398 78.7%
Back
Top