hugelgupf

6 ms·

1.
by hugelgupf 3y ago
You should be able to disable write protection on the firmware flash using a suzyq cable to enable closed case debugging as well, and from there disable write protection. The screw is not the only way
2.
by hugelgupf 3y ago
I don't really get their headline. Google extended the AUE to 8 years. They're complaining that most Chromebooks only have 4 years of software life left. Are they buying devices released 4 years ago? Or refurbished devices?
3.
by hugelgupf 4y ago
[twitter OP] This is true. Appears once a day or so, you hit OK. Sometimes during a show you're watching, sometimes while it starts up. Don't know if it becomes blocking some day...
4.
by hugelgupf 4y ago
[OP here] I filed a thing for them and this person will get it in front of the right people. I have more faith in that more than generic Apple support anyway.
5.
by hugelgupf 7y ago
Why on earth did he agree to those FBI interviews where he admitted all this?
6.
by hugelgupf 7y ago
It is a failure, but not of the HTTP service. It's a failure of the RPC call. 404 means "this API doesn't exist", not "the API exists but it returns an error"
7.
by hugelgupf 8y ago
Brain fog, intestinal inflammation. Sounds so very familiar. Do you have research notes, on methodology and things tried? I'm now very curious for myself.
8.
by hugelgupf 8y ago
here are some basic linux tools in Go: https://github.com/u-root/u-root
9.
by hugelgupf 8y ago
That sounds much more attainable than what the blog post suggested - sorry if I took that the wrong way!
10.
by hugelgupf 8y ago
Given what I just read about IncludeOS (I'm not really familiar), it feels like implementing as much of musl as possible is probably the best you can do compatibility wise. There are going to be a lot of Linux features simply incompati
11.
by hugelgupf 8y ago
The number of system calls says close to nothing of the actual API surface present here. It's a poor measure. What actually matters is: what can I run? That said, 42 is not enough to run a lot of applications. It'll be interesting
12.
by hugelgupf 8y ago
I think the board that is pictured by Bloomberg is just an illustration. For example, Elemental never used that board. See my tweets here for the actual specs: https://twitter.com/hugelgupf/status/10481607945658613
13.
by hugelgupf 8y ago
Do you have a source for the Norwegians?
14.
by hugelgupf 8y ago
See the "Serverless Containers" section on https://www.google.com/amp/s/gweb-cloudblog-publish.appspot.... It's still in alpha, so Google is not advertising it wide. It was announced about a week be
15.
by hugelgupf 8y ago
Almost complete, but stalled, and based on Go 1.8: https://github.com/riscv/riscv-go
16.
by hugelgupf 8y ago
Haskell is 28 years old.
17.
by hugelgupf 8y ago
You can indeed safely run containers as root inside the sandbox. The UID/GID you run as is configurable.
18.
by hugelgupf 8y ago
You can implement a custom file system as a separate process (called Gofers) and have the sandbox connect to that. The custom file system has to adhere to the p9.File API (see pkg/p9/file.go) - the protocol is an extended version
19.
by hugelgupf 9y ago
No, what sigjuice said is right. The name was derived off u-boot, though, and u-root can be used (and is in use) with LinuxBoot.
20.
by hugelgupf 9y ago
Shameless self-promotion: also take a look at https://github.com/u-root/u-root :) (The gokrazy and u-root people all work at Google, I just contacted them to see if we can help each other. Not a competing project.)
21.
by hugelgupf 9y ago
They are effectively measuring how fast AWS infrastructure can reach whatever it needs (VM?) for each language. But these are not even cold start times. I'm assuming after the first request this is using the same VM (/backend job,
22.
by hugelgupf 9y ago
Binaries can also use dlopen to open libraries that aren't listed by ldd.
23.
by hugelgupf 9y ago
Seems to be down right now, so not sure what services are implemented, but... someone should port this stuff to upspin. :)
24.
by hugelgupf 9y ago
The front page of LinuxBoot has links to talks' videos and slides.
25.
by hugelgupf 9y ago
They didn't create this page. You can follow the development of the page at https://github.com/linuxboot/linuxboot.org
26.
by hugelgupf 9y ago
Booting via kexec is the intended idea here. Theoretically, it should be possible to boot any other OS with kexec. We may have to test that assumption for this project...
27.
by hugelgupf 9y ago
You could have a tiny initramfs that mounts the disk you want with your actual init and then switch_root to that. You'd have the kernel from flash, and init from disk.
28.
by hugelgupf 9y ago
We also have a mode that rewrites all commands' sources given so it can be compiled into one binary. It ahem needs some improvements, though :)
29.
by hugelgupf 9y ago
Yes, there should be firmware updates for the kernel in flash. (And we haven't really figured out how to handle that yet.) But such firmware updates likely have different constraints and a different schedule than your kernel-in-product
30.
by hugelgupf 9y ago
We should clarify that. Our example is an OCP Winterfell node, where boot time went from 8 minutes to 20 seconds.

More ›