hi great video series just missing some different software tests to see the capabilities of each board and at the end of the video would be great to see what score thoose sbc get and have some sort of chart
15:55 I actually ran ADB on the Milk-V Duo S, you just need to copy adbd binary to it, and then run it and specify some port, then you can use adb connect ip:port and it will work great, adb shell works, adb reboot works, etc.
Oh that is excellent, but did that require an ARM system image, not RISC-V then? The main issue is the default with all Milk-V is of course RISC-V, so the adb binaries likely wouldn't work unless you had RISC-V builds of them?
@@catto88 OH awesome - want to stick it on Github or share a link? Email it to me if UA-cam eats the comment. I can include it in my video description for others.
@@PlatimaTinkers Done, I have sent you a link in the previous comment, I hope UA-cam didn't eat the comment. I tried to setup ADB over USB, but no luck, I'll try to make it work later. The command to run was something like this: ./adbd --tcp-port=5555 On the host: adb connect 192.168.42.1:5555
@@PlatimaTinkers Looks like UA-cam ate it, I'll email it too. Also additional info: Why did I even try running ADB? I was looking at V2 sdk and saw a closed pull request that should "fix ADB for Duo 256M".
fishwaldo's port and the milkv duo..probably the most fun I've had so far with these riscv projects. I've had problems with luckfox boards though. And licheerv nano is another one that just plain works with fishwaldo. Good shit.
After playing with the luckfox, the wiki is really good but i wish they had a brain dead simple (for dummies like me) demo without a camera to play with the NPU. I got lost in the weeds pretty quick :(
In my opinion the MilkV Duo wins since it can run mainline linux. Working with outdated linux is a pain and will become more painful as time goes on. (Due to gcc or program incompatibilities, Yes i have ran into that before) From what i can tell MilkV is really pushing to mainline their board and they've been making amazing progress. Meanwhile rockchip has not done anything to bring mainline support to their SoC. (As per usual)
That is a REALLY valid point, and I didn't actually consider that in the comparison. I will in future! That being said, the newer Rockchip SOCs coming out are all 6.1+ SDKs thankfully, and you may be interested in checking out github.com/radxa/kernel/tree/linux-6.1-stan-rkr1 and github.com/Joshua-Riek/ubuntu-rockchip From what I can find, Rockchip is bringing 5.10 BSP code up to 6.1.25, but just not public yet.
I’m starting to distrust rockchip as of late. sbcs are not really their selling point anymore so i don’t expect them to really care about mainlining anything new. Also 6.1 is also not that new and it’s an lts kernel meaning they will keep it until eol. What milkv is doing on the other hand is comparable with what the rpi foundation is doing, Mainlining as soon as possible while the boards are still in use. I’m personally interested in the upcoming radxa orin o6, would love to see you cover it! :D
@@System64-Fumo Hey yeah I do completely get that, but they are still contributing much more than many other companies I see building ARM hardware, eg Imagination Technologies (not ARM specific, but a key pain in the ass for these issues), Texas Instruments with their AM series, etc. Rockchip do actually commit a LOT too, but just don't make a big deal about it. Eg if you have a look at the Linux kernel commit history, in the last three weeks there's been over 200 commits. As best I can see, Rockchip has made 1014 commits in the 12 months to date. There are also contributing for a wide range of SoCs, where as Milk-V are contributing for just 4 SoCs, so that does skew things a bit too IMHO. Milk-V and Raspberry Pi are also SBC vendors, so they're contributing more specifically for their boards, where as Rockchip makes the SoCs, not end user products, so it's also a bit different in that regard. The O6 does look awesome though, cannot wait for it to arrive!
Nope, it's basically the standard for magnetic card readers (you know, the magnetic track on credit/security cards). That has been largely replaced with smart cards and NFC, but magnetic cards are still relatively common as room keys in hotels, and that's one of the markets CVitek (the company behind the CV1800B, the SoC on the original milkV Duo) was targetting. The other big market being security cameras (which explains the camera interface and the TPU).
2025 will be Year of the Linux Desktop, of course :P
Absolutely, bring it on! 🤘
hi great video series just missing some different software tests to see the capabilities of each board and at the end of the video would be great to see what score thoose sbc get and have some sort of chart
15:55 I actually ran ADB on the Milk-V Duo S, you just need to copy adbd binary to it, and then run it and specify some port, then you can use adb connect ip:port and it will work great, adb shell works, adb reboot works, etc.
Oh that is excellent, but did that require an ARM system image, not RISC-V then? The main issue is the default with all Milk-V is of course RISC-V, so the adb binaries likely wouldn't work unless you had RISC-V builds of them?
@@PlatimaTinkers Nope, I compiled it for RISC-V
@@catto88 OH awesome - want to stick it on Github or share a link? Email it to me if UA-cam eats the comment. I can include it in my video description for others.
@@PlatimaTinkers Done, I have sent you a link in the previous comment, I hope UA-cam didn't eat the comment.
I tried to setup ADB over USB, but no luck, I'll try to make it work later.
The command to run was something like this:
./adbd --tcp-port=5555
On the host:
adb connect 192.168.42.1:5555
@@PlatimaTinkers Looks like UA-cam ate it, I'll email it too. Also additional info:
Why did I even try running ADB?
I was looking at V2 sdk and saw a closed pull request that should "fix ADB for Duo 256M".
Can you make a step step guide to set up luckfox pico mini with the camera and yolo s5 I am a beginners can you plsssss
Easy man... Hack around a bit with chatGPT to help you.
A few people asked for this, so I think I may have to! Cheers
fishwaldo's port and the milkv duo..probably the most fun I've had so far with these riscv projects. I've had problems with luckfox boards though. And licheerv nano is another one that just plain works with fishwaldo. Good shit.
When are you going to a vid on luckfox lyra series?
Likely this month - I already have stock on the way, and they also sent me sample units for a video 😁
The humane animal trap project sounds interesting.
Yeah I need to get back to that one! Just had stuff all video views, so it got down- prioritised a bit!
After playing with the luckfox, the wiki is really good but i wish they had a brain dead simple (for dummies like me) demo without a camera to play with the NPU. I got lost in the weeds pretty quick :(
I've had a few requests for this, so I think I may have to do a video on it! Thanks for the idea.
Meow? :3
@@catto88 Miao Miao :3
Meow!
In my opinion the MilkV Duo wins since it can run mainline linux.
Working with outdated linux is a pain and will become more painful as time goes on. (Due to gcc or program incompatibilities, Yes i have ran into that before)
From what i can tell MilkV is really pushing to mainline their board and they've been making amazing progress.
Meanwhile rockchip has not done anything to bring mainline support to their SoC. (As per usual)
That is a REALLY valid point, and I didn't actually consider that in the comparison. I will in future!
That being said, the newer Rockchip SOCs coming out are all 6.1+ SDKs thankfully, and you may be interested in checking out github.com/radxa/kernel/tree/linux-6.1-stan-rkr1 and github.com/Joshua-Riek/ubuntu-rockchip
From what I can find, Rockchip is bringing 5.10 BSP code up to 6.1.25, but just not public yet.
I’m starting to distrust rockchip as of late.
sbcs are not really their selling point anymore so i don’t expect them to really care about mainlining anything new.
Also 6.1 is also not that new and it’s an lts kernel meaning they will keep it until eol.
What milkv is doing on the other hand is comparable with what the rpi foundation is doing,
Mainlining as soon as possible while the boards are still in use.
I’m personally interested in the upcoming radxa orin o6, would love to see you cover it! :D
@@System64-Fumo Hey yeah I do completely get that, but they are still contributing much more than many other companies I see building ARM hardware, eg Imagination Technologies (not ARM specific, but a key pain in the ass for these issues), Texas Instruments with their AM series, etc.
Rockchip do actually commit a LOT too, but just don't make a big deal about it. Eg if you have a look at the Linux kernel commit history, in the last three weeks there's been over 200 commits. As best I can see, Rockchip has made 1014 commits in the 12 months to date. There are also contributing for a wide range of SoCs, where as Milk-V are contributing for just 4 SoCs, so that does skew things a bit too IMHO. Milk-V and Raspberry Pi are also SBC vendors, so they're contributing more specifically for their boards, where as Rockchip makes the SoCs, not end user products, so it's also a bit different in that regard.
The O6 does look awesome though, cannot wait for it to arrive!
I think wiegand is some rfid standard.
It's for keypads and card readers (door entry systems etc)
Nope, it's basically the standard for magnetic card readers (you know, the magnetic track on credit/security cards). That has been largely replaced with smart cards and NFC, but magnetic cards are still relatively common as room keys in hotels, and that's one of the markets CVitek (the company behind the CV1800B, the SoC on the original milkV Duo) was targetting. The other big market being security cameras (which explains the camera interface and the TPU).
Hah right you are! I just never heard of it before, or even saw it in the block diagram previously.
@joseoncrack Makes a hell of a lot of sense! Thanks for the input mate