Quote from: VaNilla on December 18, 2022, 10:07:54 PM

I usually save 5-10 replays whenever I land a VC/SA stunt, by saving one replay, alt tabbing to make a copy, and repeating the process until I have multiple replay backups - this helps sidestep the risk of corruption.

I'm a bit curious about this one, have you ever done this and had one of the replays been corrupted but not the others?
From the source code of the game (at least in VC, not super knowledgable about SA) the saving functionality is just writing data directly from a buffer in memory. Any corruption is likely to have happened in this buffer rather than during the writing itself (although not impossible I suppose), so saving multiple of what would ultimately be the same corrupted buffer likely wouldn't fix anything, at least in my head.

I've had a couple of cases where corruption seemed to occur when F3ing quickly after F2. Or just fat fingering the keys. I figured it was probably something to do with the game attempting to access the file whilst it's still being written, but not too sure.

The problem for me with the saving multiple reps option is in VC I get crashes fairly regularly when alt-tabbing or maximizing the game again. Seems to be common no matter the executable I use. Having said that, it's no added risk I guess since unless you have the file state viewable on a second screen, you wouldn't know if the rep is corrupted or not until you tab out anyway. Which I've considered doing. But it's kinda a pain in the ass.

Maybe at some point we could have a tool that verifies rep integrity without erasing the replay buffer (i.e. via F3). Would let you resave via F2 if something went wrong.

Would also say my corruption fears stem mostly from my SA days where I lost a handful of reps to it. In VC it doesn't seem as frequent.

This was fun, nice

as I saw in the past people landing harder combos.

...you saw someone land something harder than a stalled rear wheel grind to csm to prec?

Cmon man. With maybe only one or two exceptions, these spots are possible but not doable.

https://www.youtube.com/watch?v=POVuKPNgatU

Awesome stuff both

Some people collect stamps, or build model railways, or write weird erotic fiction.

Seems like some of us will be doing tricks in ancient GTA games into our old age 

Yes Dannye's as 'truth' makes sense given it's ubiquitous. I know there's some that prefer to use older versions though, and for those I guess they use other speedos. Hence this thread.

Key thing to remember is 02E3 * 3.6 = KPH. (02E3 * 3.6)/1.609 = MPH. With that most other speedos should be comparable. For those that aren't, someone needs to test if they care.

Is there a way to check if your speedo is using that opcode? Outside of driving the same run-up twice to see if it matches some other tool.

But since Dannye implemented the speedo into the USCM I don't see a reason anymore to use any other speedometers, if only for the fact we can be sure it doesn't alter anything else in the game when installed. I never used one because it would just prove I'm slower than everyone else  But just having one standard so comparison is easy can only be a good thing. Thanks for the more detailed explanation.

Yeah I think using Dannye's is the easiest option for most. Although I'm not sure how easy it is to customize the position/colour etc of the stats display, when compared to a CLEO script. I was already using a CLEO speedo, and have customized it a bit in that respect. So I'll probably just convert it to mph using the above calculations. That way it has parity with the mph that USCM will give people.

I know Jason has customized the USCM display though, so it's definitiely doable.

You can check how the CLEO script (assuming it's a CLEO mod), by loading the .cs into a text or code editor. Then see how it's grabbing vehicle speed from the game. If it's not a CLEO script, I'm not sure what method it will be using.