Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


When's Melty on Steam?
ahaha that's no--wait, what?

Messages - mauve

Pages: [1] 2 3 ... 10
1
The way my caster works, a desync shouldn't happen due to a few dropped packets. However it may happen if a series of packets get dropped or delayed. It does seem really weird that it only happens on wired.

I'd like to point out that, although it's extremely rare, corrupted packets do occasionally make valid checksums, improbable as it seems. Even if you do everything 'right', you can't completely avoid desyncs on the internet. (Throwing some extra checksums in doesn't hurt though. Caster itself doesn't bother, and for rollcaster, the rollbacks cover a lot of these up anyway.)

Blocking the game while waiting for data shouldn't cause any desync errors though, provided your wait loop is at the right place. If you get worried about any sort of FPU shenanigans, store the FPU data when you block and reload it when it's time to unblock. But those shouldn't happen without video calls.

2
answer to all desyncs

fldcw 0x007f

3
w.

4
Source code would still be appreciated even if it's not caster based. Because I'm curious.
I second that, no smaller size text tho, since I'd actually like to help with something.

When you've released it, you were prepared to be everyone's slave right? =D
my view is that all of the existing caster derivatives would not have existed without the original caster source having been released, godawful and unreadable crap that it is. same goes for mtsp lunaport. if you want to nod back to something like it, following in its spirit doesn't hurt.

that being said, if it's still actively under development, whatever, not much value in that unless you throw it on git/svn somewhere.

5
huh. I didn't even hear about this until today. I'm actually sort of glad nobody came and pestered me about it!

Haven't tested it, but good stuff. Porting the thing is not easy.

edit: btw if this is based on caster code in any way, i'd appreciate source release just to pass on the tradition thanks. and if it's not i have no idea why the caster name is used.

6
NetPlay Community / Re: MBCaster 08/04/24 out.
« on: November 18, 2012, 09:55:32 PM »
if you guys want to do it I really don't care.

most of it's still linked from the wiki and the files themselves aren't going anywhere (for now).

7
NetPlay Community / Re: MBCaster 08/04/24 out.
« on: November 18, 2012, 11:02:04 AM »
pomf

8
NetPlay Community / Re: MBCaster 08/04/24 out.
« on: November 05, 2012, 08:38:17 PM »
Closing site in a couple weeks.

No questions, it's just going. Feel free to grab stuff before it vanishes.

also why the hell are there three stickied threads for mbcaster

also forgot i set this avatar, it's awesome

9
Other Games / Re: Dark's Souls's's's
« on: June 03, 2012, 08:23:39 PM »
Here's the steam page, if you're wondering.

Not a word about GFWL, but that doesn't mean much.

10
Melty Blood Auditorium / Re: The MB:AA:CC:PC thread: v.1.2 out!
« on: January 26, 2012, 08:45:48 AM »
The problem is that it's kinda... counterproductive for some people. I get 599~999 ping when using the delay tool, lol. I don't understand why it lags us so much.  :mystery:

It doesn't do a goddamn thing to the ping calculation, just the way it calculates the delay. Like, you see on screen "Delay d(n)", d is the value is changes the calculation of. Nothing else.

At best that might be because of the single-CPU setting but I have no idea why that'd do that.

11
Melty Blood Auditorium / Re: The MB:AA:CC:PC thread: v.1.2 out!
« on: January 15, 2012, 07:36:25 PM »
For what it's worth I've been a massive slowpoke / distracted / etc when it comes to updating stuff, so if he wants to take over tool duties I am entirely fine with this.

12
I dunno man.

I think I need to move away from dll injection stuff though, it seems to just be causing more and more problems as time goes on, sadly.

13
watch it be on some sort of jp-only download site or something

still, it's a step in the right direction. now if only they could fix their netcode...

14
Wonder how hard a VWF hack would be.

15
mod needs to be, well, modded to prevent keybindings

if i do another build i'll put that on it

if it was using default bindings even if you changed the values it uh probably isn't working at all though. that's weird.

16
Also AVG sucks.

Yeah my translations are probably considered viruses and stuff by some overzealous AVs too, since they use the same method.

Problem is there's only so many ways to unofficially mod something, and this is how you get in without modifying the original executable. I'll probably end up having to make an exe patch or something that loads it and I really don't want to do that.

17
http://mauve.sandwich.net/misc/mbaacctool-v11-delaytest.zip

testing tool version that reduces delay, launches mbaa.exe etc

the delay algorithm is fucking retarded, it takes average_ping*13/20 and then checks by multiples of 14 to see if it should use that delay or not. which has too many overall functions to it really and doesn't produce good results.

18
that's better i guess but there still seems to be severe problems with packet loss

19
against my better judgment i got annoyed and hunted down the delay calculation code last night

as expected, it's calculating incorrectly and increasing delay at increments of 14ms... if that value is ms anyway... which it isn't, i'm not sure what it is, but it's similar. i put a test version of my tool that tweaks it out last night but it seems my values were incorrect so i'll take another look

worse was when i found the recv function... is this game really using tcp? holy shit, that is unfixable.

also i have a massive headache today and am not even going to look at this crap

20
Automating is nice, thanks. Although setting it to 1 core was one of the first things I did :V

Because I doubt that French Bread has the ability to program multicore processes.
drivers don't care, they will multithread and beat ass

21
i quietly updated my tool the other night, with an option to set cpu affinity in there, since that seems to be all the rage right now. just set cpuID to 1 or something to activate

http://mauve.sandwich.net/misc/mbaacctool-v11.zip

22
Hello, I'm just getting started with linux and wine, any tips on getting MBAACC working on linux?  Does it work?
It does work but there's a bug with the font rendering code that will cause it to blow up in certain menus.

Haven't written a fix for it atm, too much crap to do, but it wouldn't be too hard.

23
Framedisplay v2.3 out.

If you were using one of the prerelease versions, this fixes a bunch of bugs.

I remember seeing versions of the framedisplay showing hitstun etc.. What option enables that?
That was for a version of the MBAC framedisplay. The additional stuff was done by Sprint.

The thing is, between MBAC and MBAA the format drastically changed. So I would have had to re-derive everything from the new format and I simply did not care that much.

I'm also not sure the original was 100% correct or trustworthy enough to actually use. I think you're probably better off just testing it as before.

24
Framedisplay v2.3 out.

If you were using one of the prerelease versions, this fixes a bunch of bugs.

25
Uh, well, yeah. Just had some quick words about it with him on IRC. Seems to be okay now. I'll keep such things in mind for the next time.
Yeah, sorry. I've been doing way too much lately and haven't really been checking the forums here much.

Pages: [1] 2 3 ... 10