Maxime Vincent

@brendonovich if you’re running Node JS in a Tauri app, why did you use Tauri in the first place? You want to use Tauri to get *rid* of nodejs and write the part in Rust…

we've moved opencode desktop to electron. it's faster, more reliable, and will replace our tauri build soon. try it out in beta via the link below.

@ghidraninja Love that!

Sometimes it’s nice to go back to basics: Built firmware that’s just 246 bytes, uses no RAM at all, and runs parasitically from a 1-wire bus at just 0.2 mA. Yet it emulates a full jetski key! You can see voltage rising until the chip starts running based on the BOD threshold.

James just made my day (again)

Super helpful description, thanks ⁦@OTTOCAST11⁩

@SipeedIO Since #NanoKVM now supports switching to #PiKVM, is the source code of your PiKVM fork available somewhere?

@virkokka @tobi Yeah, lot of the main brands are that big. "RazerAppEngine" is 360 MB on disk here. The OSX download of Synapse 4 is ~170 MB. (The Windows download hides behind an "Installer" download that, which downloads the rest of the bloat after you run it). Logi GHUB is similar...

@maximevince @tobi >hundreds of megabytes Have you actually seen this?

Really neat development from the unlikely place of gaming peripherals that I appreciate: Both my mouse (finalmouse) and my keyboard (wooting) don't ship with software but have their entire control panels and firmware upgrade system implemented as websites that use WebUSB. So much better than shipping random always-on software that needs to be kept up to date. The web always wins.

@tobi “Don’t see” is what I wanted to say :)

@tobi 100% In the case of Finalmouse, we are technically using WebHID, but your point stands! We hate bloatware and see why a mouse driver should be a resident background app that’s hundreds of megabytes and requires updates every new moon…

@utfunderscore @tobi That’s why we also provide a downloadable fully offline version @finalmouse :)

@tobi This sounds great until the company decides they no longer want to host the website and your no longer able to configure your mouse. At least software you download can be archived

@douglascamata @tobi We made a downloadable, offline version as well @finalmouse :)

@tobi Except one day the website is down and then you can’t reconfigure your peripherals.

Introducing ULX SAKURA 桜 10,000 Limited Edition Units Drops September 20th · 11AM PT $175 · Ships next day With a color specially formulated from the blossoms of a wild weeping cherry tree. Available exclusively at finalmouse.com

@xTheWhale_ The Nordic chip can definitely go faster than 1000 times a second :)

I made it EASY so please say Goodbye Old Mouse Parts Some mice have special parts inside to make them work without wires. Two of these parts are the PAW3950 sensor and the Nordic nRF52840 chip. They worked very well for a long time, but now we know they can't go faster than a certain speed. Why Can't They Go Faster? The PAW3950 is a part inside the mouse that helps it see how you move it around. It is good and strong but it can't go faster than a certain speed, which we call 1,000 times each second. Another part, called the Nordic chip, helps the mouse talk to the computer without wires. This chip also can't handle going faster than 1,000 times each second. Even when some companies tried to make it look faster, it still wasn't really faster. Pretending to be Faster Some companies made mice and said they could go much faster, like 20,000 times each second! But that's not really true. They used a trick called "guessing" to pretend it was faster. This trick makes things look good, but it doesn't always work right. It's like guessing where a missing puzzle piece should go instead of putting the right one in. Learning from Older Good Mice Some older mice, like the Logitech G Pro, worked very well because they didn't pretend. They just stayed steady at the right speed. This made them feel better to use because they were stable and always acted the same. Saying Goodbye Now we say goodbye to these old mouse parts. They were good and helped mice work well for a long time, but now we need new parts that are truly faster and better without pretending. So, thank you old parts, and hello to the new and better mouse parts coming next!

@GothamChess why is this marked as not correct? I suppose it’s an alternative solution ? (Enjoying your book, btw!)

@dnak0v @radareorg How does it work in r2?

point and click reversers just discovering AI can do RE 😂 bro we've had this shit in radare2 for 2 years now

Zephyr RTOS + Segger System View is such an amazing combo. Real insights of exactly what's going on in your firmware; works out of the box! @ZephyrIoT @SEGGERMicro

@maartenweyn China Export again! Ongelofelijk dat zoiets zelfs mag/kan…

Dochter komt thuis met iets dat ze gekregen heeft uit de Action. Rara wat is er mis. tip zie laatste figuur…

@iamtechless Great use of XLAT with an external solenoid and external switch! I know I'm not exactly impartial here, but the Finalmouse ULX would have gotten the lowest weight and the best wireless tech (best polling stability, lowest latency numbers) in your review. So unfortunate it's missing from the list.

@FODFinancien jullie file upload module voor "antwoorden op een brief" is defect en faalt omwille van de CORS policy van elke moderne browser. Best even rechtzetten? (Ik slaagde er alsnog in de file digitaal te uploaden door de CORS policy per command line uit te schakelen)

@QVHenkel That is one sexy PCB layout!

To reduce component count and decrease PDN inductance at higher frequencies I´ve swapped the 0201 decoupling capacitors on the core supply rails for the SoC to 0402 feedthrough capacitors. Even though these caps are larger the mounting inductance is lower than on the two terminal 0201 caps

@ya_man_102 @xTheWhale_ @8lack5n1per What PCB is that?

@xTheWhale_ @8lack5n1per Don't worry. No problem. If you have already completed your PCBs, SMT takes only 30 seconds per one.

wootingのすごいところ。 「competitor keyboard」とか言うてフツーに他社品を比較展示するところ。