While I firmly believe that changing the platform won't change much. I gave it a try to self-host my own data server for bsky and you will be able to find me there https://t.co/1wLGgMG9Ya if you search for me
Really happy to see that the knowledge of proper input lag testing is spreading little by little. Great article!
Also, I'm sorry you experienced the wrath of frustrated gamers, I had my fair share of that.
Last week, I received a small care package from @noodalls with a kit for building a lag tester!
I whipped up some Python scripts to automate testing and have generated tons of new findings (including the one above).
Refer to the link for full details.
https://t.co/eknbkZOVt4
@jav1ts@thefleecer@Arlieth@HiFightTH@fubarduck@GelatinLab When I made my script to reliably send inputs, I had to make it as part of the gamepad API injection to make it consistent. Just sleeping 16ms had way too many flaky results.
I didnt look in details the quoted post but I feel we are in this territory.
@jav1ts@thefleecer@Arlieth@HiFightTH@fubarduck@GelatinLab Connecting back to this platform and getting 20+ notifs already 😆?
I havent dug into it lately, however what I can tell is that input parsing happens one per game-tick (60 times/s) and the exact poll timing cant be trusted at the ms. Nothing new afaics.
Welcome to the public domain, THE SKELETON DANCE (1929). 🎞️ 🎶👻 Walt Disney's first Silly Symphony cartoon is a masterpiece of synchronized animation & music featuring dancing skeletons in a graveyard.
Learn more. ➡️ https://t.co/jNLDikrT0y
#PublicDomainDay#CopyrightFree
Nouvelle vidéo sur le Bazar ! 1heure de vidéo (oui) pour parler de la fatigue des open World !
Le problème des Open Worlds https://t.co/Apiw5sMuM8 via @YouTube
@MDZ_jimmY I would like to do that eventually. I have people I want to follow there so I created something there. Didnt think much more ahead.
In any case it wont change much considering the amount of tweets I've sent over the last months/years.
While I firmly believe that changing the platform won't change much. I gave it a try to self-host my own data server for bsky and you will be able to find me there https://t.co/1wLGgMG9Ya if you search for me
Le Projet de loi de finances pour 2025 a été publié hier soir. Vous pouvez dès ce matin visualiser ce que cela va changer et même réaliser vos propres modifications : https://t.co/XVXVo8kJqs
Le nuit fut courte mais le résultat est là !
https://t.co/SJ29JRIUFu
Seems when I updated the site it made the URL wrong.
This will be a site that contains links to my results as well as a description of what' going on.
@Ragnarork Would have strongly disagreed. But then I saw the definition in french is "Fidélité à tenir parole, à faire honneur à ses engagements". While I see in english is "Faithfulness or devotion to some person, cause or nation" or similar. I firmly adopt the former and reject the latter
@knkmotoh I tried tapping directly to the CPU or via the chipset same thing: the negotiated speed is FullSpeed. However, weirdly enough, I had 2ms also by picking the 2.0 port on the motherboard (??).
I feel it's windows usb driver picking the "wrong" value for the negotiated speed 🤷
Question to usb engineers. I have a USB 2.0/FullSpeed gamepad with a bInterval = 5 endpoint.
On windows, if I plug it to a SuperSpeed port, it has a 2ms polling rate. If I plug it to an old HighSpeed port, it has a 4ms polling rate.
Is it a bug?
The spec says that if it operates at HighSpeed it uses 16 microframes = 2ms. So I think that's where it comes from. However, USB Tree View tells me that both work at FullSpeed.