Homebrew RELEASE Development Thread - RetroArch libnx

  • Thread starter m4xw
  • Start date
  • Views 677,645
  • Replies 4,272
  • Likes 69
Status
Not open for further replies.

wicksand420

Well-Known Member
Member
Joined
Nov 13, 2016
Messages
2,787
Trophies
1
Age
39
XP
2,296
Country
United States
Three questions:
  1. 4DO crashes the Switch with the Yabause config, modified with the BIOS loaded, as well as VSync off and threading on. Any fix or specific configuration needed to work?
  2. Any hope for a compiled build of Mednafen-PSX without the fullscreen hack on by default?
  3. Shouldn't anyone be working on why the current build of Yabause doesn't work?
I'd say that the more labor intensive cores should not be worked on now, get the basic ones established then work on the labored ones
 

DollyWhipDoll

Royalty & Unbothered. LMFAO!!! ;)
Member
Joined
May 18, 2018
Messages
451
Trophies
0
XP
991
Country
United States
Lmfao haha I guess we all have are terminology we use to refer to our graphics card.

It's an MSI GP62MVR

That's the laptop check out the specs.

https://www.msi.com/Laptop/GP62MVR-6RF-Leopard-Pro.html

Has great specs but probably should have held off for the 7th gen processor, would have gotten far more battery life.

Is this yours? That setting doesn't look familiar. Sounds like you. I didn't watch the whole thing.
 

m4xw

Ancient Deity
OP
Developer
Joined
May 25, 2018
Messages
2,442
Trophies
1
Age
119
XP
6,960
Country
Germany
@m4xw I think I fixed the root cause. Patch provided if you want to apply. Testing looks good on my side.

(I knew I didn't lose my T&F skillz that quickly)

FYI: this patch should fix all cores; not sure if it's entirely related but even the retroarch menu seems to respond much better now

tldr; the current input mechanism only recognizes buttons pushed at the precise time they're polled resulting in potential lost input. This patch ensures buttons pressed leading up to the poll within a frame cycle are not dropped => much better responsiveness in games

EDIT: dpad was messed up on the first patch, 2nd patch fixes that!

Thats great! Is the actual fix the | with hidKeysHeld?
Because AFAIK inputs should only be polled once every frame. (I am at work currently, cant check out the func/macro)

Could you try without the fix and change input polling type to early or normal rather than late, anyway that looks good to me, just curious.
This is honestly the first 3rd party fix for my code... finally... :D
Anyway your's is way cleaner so I like it. Could I maybe ask you to look at multiplayer support /2nd player
Should be easy

Three questions:
  1. 4DO crashes the Switch with the Yabause config, modified with the BIOS loaded, as well as VSync off and threading on. Any fix or specific configuration needed to work?
  2. Any hope for a compiled build of Mednafen-PSX without the fullscreen hack on by default?
  3. Shouldn't anyone be working on why the current build of Yabause doesn't work?

1. No Idea

2. I have the nro, you gotta wait 10h at least

3. No, who gives a fuck? If you don't see the Issue here https://git.m4xw.net/Switch/RetroArch/RetroArch/boards you can be sure I am not gonna work on that anytime soon.

Not like I get paid for giving you free shit. I've been working non stop in my free time to make this possible.

I would lie if I tell you my private life is not suffering, so... maybe practice for u.

Sorry but this "shouldn't", I really dislike the tone

I'd say that the more labor intensive cores should not be worked on now, get the basic ones established then work on the labored ones
+1

I think M4xw already has a private repo of Mednafen-PSX somewhere on his hard drive, lol
What we should begin with is (no surprise) the 4DO issue, I guess.

That doesn't make any sense, so no.
 
Last edited by m4xw,

stinkoman

Magnificent Bastard
Member
Joined
Dec 23, 2008
Messages
1,227
Trophies
1
Location
Hell.
XP
555
Country
United States
I think M4xw already has a private repo of Mednafen-PSX somewhere on his hard drive, lol
What we should begin with is (no surprise) the 4DO issue, I guess.
Lol why? 3DO emulation is probably the last thing on anyone's mind. It makes more sense to get stable NES, SNES, Genesis, etc. done before moving onto some weird shit.
 

m4xw

Ancient Deity
OP
Developer
Joined
May 25, 2018
Messages
2,442
Trophies
1
Age
119
XP
6,960
Country
Germany
Lol why? 3DO emulation is probably the last thing on anyone's mind. It makes more sense to get stable NES, SNES, Genesis, etc. done before moving onto some weird shit.

Idk why people wanna focus on fixing cores when there isn't a stable build that runs with good performance yet (non threaded).

I bet that 90% of all issues are caused by retroarch code, not the cores (with a few exceptions)

Literally any config or performance issue is retroarch (leading up to this point), not the cores (except psx)
 
Last edited by m4xw,

m4xw

Ancient Deity
OP
Developer
Joined
May 25, 2018
Messages
2,442
Trophies
1
Age
119
XP
6,960
Country
Germany
That's exciting! Not sure why any1 wouldn't want it in fullscreen though..looks amazing on the first psx release in fullscreen imo
Oh its still fullscreen right now on this version (had some issues, don't really feel like fixing that)
spoiler: performance feels the same, I would prefer using the already released version myself
 

SeongGino

The Average One
Member
Joined
May 20, 2015
Messages
939
Trophies
0
Age
25
Location
Jersey, Jersey
Website
www.twitter.com
XP
1,932
Country
United States
We REALLY should keep the pace going in working on having a sound backbone for all the other cores to work on, rather than adding more superfluous cores when RA itself is barely working on a thread (haaa!).
ALSO THRILLED IF THE INPUT POLLING HAS BEEN FIXED AHHHH
 
  • Like
Reactions: Ty_

m4xw

Ancient Deity
OP
Developer
Joined
May 25, 2018
Messages
2,442
Trophies
1
Age
119
XP
6,960
Country
Germany
We REALLY should keep the pace going in working on having a sound backbone for all the other cores to work on, rather than adding more superfluous cores when RA itself is barely working on a thread (haaa!).
ALSO THRILLED IF THE INPUT POLLING HAS BEEN FIXED AHHHH

Not live yet tho. :P

Will be home in 4h
 
Last edited by m4xw,

lifajucejo

Member
Newcomer
Joined
Jun 7, 2018
Messages
15
Trophies
0
XP
288
Country
Canada
Thats great! Is the actual fix the | with hidKeysHeld?
Because AFAIK inputs should only be polled once every frame. (I am at work currently, cant check out the func/macro)

It's actually the call to hidKeysDown that fixed the buttons, the hidKeysHeld fixed the dpad so events are registered if it's held down.

I doubt changing the polling type in retroarch will help as I reproduced the issue outside of retroarch separately but I agree it should be confirmed:

I won't be home for another 10 hours but your suggestion can be tested by anyone here with the existing fceumm build: load up Track & Field for NES and see how high you can register the cm/sec. Try the options @mx4w mentioned and see if you can sustain 800+ (I got stuck around 350-450).

Sent from my Pixel XL using Tapatalk
 
  • Like
Reactions: m4xw

m4xw

Ancient Deity
OP
Developer
Joined
May 25, 2018
Messages
2,442
Trophies
1
Age
119
XP
6,960
Country
Germany
It's actually the call to hidKeysDown that fixed the buttons, the hidKeysHeld fixed the dpad so events are registered if it's held down.

I doubt changing the polling type in retroarch will help as I reproduced the issue outside of retroarch separately but I agree it should be confirmed:

I won't be home for another 10 hours but your suggestion can be tested by anyone here with the existing fceumm build: load up Track & Field for NES and see how high you can register the cm/sec. Try the options @mx4w mentioned and see if you can sustain 800+ (I got stuck around 350-450).

Sent from my Pixel XL using Tapatalk

Ah yea I missread that. I thought I read "keys that have been pressed since the last frame" not held down. :P
 
Status
Not open for further replies.

Site & Scene News

Popular threads in this forum

General chit-chat
Help Users
    mthrnite tempBOT: @ mthrnite smacks Sicklyboy around a bit with a large leech