GermanyWhat does the XInput Plus program do? Does it pretend to be a controller somehow? How does it achieve that compared to Ginput?
I'm all for keeping the game as vanilla as possible, because the rule "use an official version without mods" is a pretty straightfoward way to put down the boundaries in which we can work to push our respective speedgames to the limit. This isn't mainly about how much time something exactly saves (althoug that surely does play a part), it's also a matter of principle to keep things simple and avoid possible unexpected side effects of a mod as well.
Of course in the real world there are always edge cases that have to be decided on a case-by-case basis. Clearly these mods like Dinput, Ginput or d3d9.dll have been deemed acceptable at some point to make it easier for people to run the game, like playing windowed when you only have one monitor (which in a way also levels the playing field).
Now I'm not entirely sure what changed that Ginput is suddenly not acceptable anymore. It seems like people became aware of some features that it adds to the game that you don't have otherwise, however from what I read in this thread there is some conflicting information:
I think with this decision it's also important to consider what options there are. If the method with Xinput works the same (or even better for the runner due to configurable control scheme) but is deemed "more legit" then it seems like banning Ginput could mean everyone is winning in the end. Now I'm not very familiar with Ginput, Xinput, Dinput or any of the other input stuff that may exist (simply because I never used it with GTA), so I don't really know how they work and what they really modify and if one is objectively a "worse mod" than the other. I think this should be explained by someone knowledgeable, especially before we decide to ban one thing and tell people to use something else. How is Xinput more legit than Ginput? Is it just that it doesn't add any "features" to the game. Or does it work completely differently in the first place? As I understand it both put a file into the game directory? There was also mention of different versions of Ginput, does that make any difference in what we're discussing?
In the end, maybe everyone will gladly switch to another method because it's simply better, or maybe not. Either way, I think more research and clarification and testing is needed before any real voting can take place.
Turbodog is correct. You may not pause the timer at all during an RTA run, whether you pause the game or not. One reason for this is that it's a simpler way of timing, you just measure the time between start and end of the run. As opposed to that, for example for the run you submitted before the rule clarification (it was always intended to be RTA, but not stated) I had to time every pause you did and calculate the end time accordingly to verify it. Not that I don't believe your time, but I need to verify it (I actually made the run a second shorter because you stopped the timer too late at the end :P). It just causes more potential issues to allow pausing of the timer at all.
As an addition to the guide I recorded how I do the searchlight part, which seems a bit faster:
I already knew the Farm Generator Skip although I never used it because I couldn't get onto the cart consistently enough. It seems like Alan just jumps different distances randomly, although maybe I'm just pressing the wrong buttons. I did seem to get it a bit more consistently if I sprint jump in the right way, which I may not be able to do everytime without more practice.
You call the skip stupid in the video description, is that because you also have trouble getting on the cart consistently or because of the jumping over the fence afterwards? If it's just the latter, that can actually go a lot faster and easier than you showed it in the video (still a bit risky though):
I didn't have time to verify all the runs last night. I was wondering about that Xbox360 thing. Also about the timing in the Special 1 run. You stopped the timer kind of during the fade at the end, but I'd prefer to have it end when the cutscene video starts, because it's a clearer cut. That would add about a second to your time. I hope that's ok with you.
I also added rules to the Specials and updated the rules of the main game to be hopefully more clear.
Thanks for the videos. As you guessed some of those aren't new, since my run is a pretty old no-reset run I did just to set a decent RTA time it unfortunately doesn't have the latest strats. Quite hilarious that you can just shoot the first fusebox. That's why it's nice to have more people run this and find strats, it's hard to think of everything.
Just wanted to throw this in here, comparison between skipping Grove4Life and doing it. I cut out the non-relevant parts (please tell me if you think I cut it wrong). I made some small mistakes in both videos (and of course the RNG in Grove4Life), so take it with a grain of salt.
It appears you can save roughly 2 minutes by skipping Grove4Life. So it depends on how much time you loose due to homies jumping out whether it's worth it. In my PB I lost about 1:30 to that (so still saved 30s), although I probably could have handled that better.
@netwizard It's not a matter of ruining anyones work (which it clearly doesn't) or that he invested less time (he may even have invested more). The issue isn't with using tricks or tools that aren't usually allowed, I'm sure everyone would welcome a TAS and be interested in how it works. The issue is just with trying to pass it off as a regular run without any explanation (whether due to ignorance of the rules or malice), even after he got called out for it.
Coollertmb was faster, but for what it's worth I also added this dupe to the wiki now.
Edit: I also uploaded a video with overlays and stuff.
I like how the video isn't available in Germany due to music, even though it has no sound..
I don't know about the onmission changer, but I made a CheatEngine script that also has that functionality, as well as an output of duping-related memory values.
I think any player that doesn't look at the sourcecode would see End of the Line as one mission with checkpoints. It's not really any different than unlocking a Trip Skip by failing a mission, just that it is differently implemented. The fact that it only sets "latest mission passed" to "End Of The Line" and not Part 1/2/3 supports it being just one mission.
The All Missions definition already seems rather player-perspective based, since it depends on what the player can clearly see in the game. So an implementation detail that isn't apparent to the player shouldn't be considered here. Imho. :)
Looks like you're using the address for 1.0. The most recent Steam version has an offset for global variables of 0x77970. The Onmission variable is at gta_sa.exe+649FC4 for 1.0, so it would be at gta-sa.exe+6C1934 for the Steam version.
Note that the "gta-sa.exe" module can be in different places for the Steam version (not just 0x400000 as with the other versions), so the actual address can change. You always have to add up the in-module address and the address of the module. For CheatEngine you can just add "gta-sa.exe"+6C1934 to the list.
I'm not an expert, so I can't really definitively say one way or the other, but this feels like a false-positive. On the page you linked, Avast says "Win32:Malware-gen". The generic name suggests it detected possible malicious behaviour through some heuristic (like maybe injecting code or changing files, which I assume the OHKO mod would do), rather than a specific known threat.
Another name on there is "Trojan.AHK.AD". The AHK hints to it being AutohotKey, which I think this Mod might be based on (which of course can or can not be used for malicious purposes depending what you do with it). For the specific "AHK.AD" variant I found something about it injecting code, so it may detect the same functions being used in this Mod and suspect it's the same thing.
Generally, I find these Malware names always pretty confusing. It's either some generic name that could literally be anything or they contain some hash values or whatever and you can barely find anything about it specifically. Either way, these kinds of things often seem like some heuristic finding similarities to something that could be dangerous.
As I said, I can't say if this file is dangerous, but if it's made or at least compiled by Ligthnat0r I don't think it would have any malicious intend. As a general rule I like to also publish the sourcecode alongside a compiled version. I couldn't immediately find this on Lightnat0r's Github, maybe he could shed some light on this. If anything, it's interesting how it exactly works and why it would trigger some alarms. :)
The LiveSplit version to use these new setting features isn't released yet.
Manually starting the timer when you load from a save would probably be best, but it might work with the "loading" memory value, which seems to be set when loading from a save. However I couldn't find the address for that for the Steam version.
Edit: Ok, I found the address for the Steam version as well. I added a setting for starting the timer when loading from a save, disabled by default, since I'm not quite sure if it will work well and if everyone even would want that to happen. You'll still have to reset manually though, because I'm not sure of a safe way to detect when to reset in that case. I really don't want to reset the timer accidentally during a run, that's why I'm using the ingame timer to check that it's really at the very beginning of the game (for resetting when starting a New Game).
After some more testing, I feel like I don't want to add too many branches. While it might be easier to find a specific mission, I think having to expand all those branches to get an overview of what is selected isn't that nice. So I'll just keep it organized by area alone and ordered roughly in the way you would do them. It's not like you have to edit the settings that often anyway, only if you edit your splits.
I already added some side missions under a "Side Missions" top-level branch and then separated by Zero, Wang Cars, Heist and Schools. In this case I think it makes more sense to organize them like this. Those are all enabled by default as well, since it doesn't really matter if you don't actually do them.
I'm not sure what other stuff to add, would be nice to get some feedback on what people would be interested in.
How do you think the missions should be presented in the settings? At the moment I have all the any% splits grouped by section (LS, Badlands, SF, ..) and ordered like the usual any% route. An alternative would be grouping them by section and mission chain (like LS/Sweet, LS/Ryder, SF/Woozie, ..).
Also is there any interest for splitting more than the usual missions? I already added collectibles (split on every single collectible/tag or when all of a type are completed), but I'm not sure if that's too useful except for silly splits of splitting every tag in an All Tags run. :P
Edit:
With the settings I mean either like this (grouped into mission groups):
Main Missions
- Los Santos -- Intro --- Big Smoke --- Ryder -- Sweet --- Tagging up Turf --- Cleaning the Hood --- Drive-Thru --- Nines and AKs --- ... -- Big Smoke --- OG Loc --- Running Dog --- ...
Or like this (only grouped by area):
Main Missions
- Los Santos -- Big Smoke -- Ryder -- Tagging up Turf -- Cleaning the Hood -- Drive-Thru -- Nines and AKs
