Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
raggi
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
42 ms
·
631.
▲
by
raggi
5y ago
> Usually (as per ASF rules) the team should wait 72 hours after creating a release candidate before publishing the release to give the community enough time to review and cast their votes. We are building consensus to shorten that windo
632.
▲
by
raggi
5y ago
As with earlier comments this seems to oversimplify the problem of throwing people at a problem. Adding people to a project puts more pressure on the current maintainers, to authenticate, validate, train and support newcomers. Apache has pr
633.
▲
by
raggi
5y ago
This commentary is exactly the problematic commentary the authors were referring to in the quote that David included near the top of his article. You may not be being so directly brash about their state, but you still imply with added drama
634.
▲
by
raggi
5y ago
It's a decent use case for planetlab.
635.
▲
by
raggi
5y ago
Oh so it does ( https://youtu.be/Zju8YaRcSk4?t=217 ), I missed it :)
636.
▲
by
raggi
5y ago
Preface: the product is probably fine, and certainly provides a nice simple UI which is perfect for it's intended use cases. There is something fishy about the video, though. You see in the video that Zoom has been keeping audio &
637.
▲
by
raggi
5y ago
I wrote some OpenCL code recently, wrapped in a rust program using opencl3. My CL code was pretty slow early on, it's a math heavy problem implemented initially with a bunch of for loops. I largely "optimized" it by smatterin
638.
▲
by
raggi
5y ago
Maybe, but to tell a story, these projects ended up making flights to a non-existent place. I'm from Bermuda, a small group of islands in the Atlantic ocean. What's probably now approaching a decade ago, we were alarmed to see air
639.
▲
by
raggi
5y ago
It works well in terms of programs working. It has similar challenges to C++ that various commonly used abstractions end up producing a lot of code on the program side regardless. It's nonetheless worthwhile, if you have enough program
640.
▲
by
raggi
5y ago
The problem they'll have is that they don't update the whole system atomically, so the generations of different packages may be different. A distribution like Clear or Nix could probably wrangle things appropriately.
641.
▲
by
raggi
5y ago
This is definitely an area it would be great to see more work in. On Fuchsia we started linking the rust stdlib dynamically, but there are some significant ABI constraints to doing this in a traditional Unix shared library ecosystem that we
642.
▲
by
raggi
5y ago
% gpg -c hello gpg: problem with the agent: No pinentry gpg: error creating passphrase: Operation cancelled gpg: symmetric encryption of 'hello' failed: Operation cancelled
643.
▲
by
raggi
5y ago
A lot of the tools now in fuchsias fx started life as personal standalone scripts later exported to the team. Nowadays contrib contains tons more, some of which also started similarly. These all now have a long history from many contributor
644.
▲
by
raggi
5y ago
I can absolutely confirm that this is what it's like to diagnose weird bugs on nascent operating systems, particularly on or nearby "portability boundaries", which are not always what you think they are. Pull that thread, but
645.
▲
by
raggi
5y ago
Oh, self-correction, Gandi is a Sectico delegate.
646.
▲
by
raggi
5y ago
How do you determine a users privacy requirements by site alone? The content does not imply relevance to privacy abuse; abusers come in all manners and may take abusive objection to content like w3.org that would to most people seem innocuo
647.
▲
by
raggi
5y ago
Which it appears that it was not, at the time that cert was issued! I have no idea if CAB have made CAA a requirement, but I presume not.
648.
▲
by
raggi
5y ago
Interestingly, it also doesn't allow the CA that it's serving, at least per the CAA records: w3.org. 3599 IN SOA ns1.w3.org. hostmaster.w3.org. ( 2021060201 ; serial
649.
▲
by
raggi
5y ago
Sadly Pi has some significant IP blocks that don't have open manuals or permissive drivers to work from, so targeting it for a home brewed OS is extremely hard work.
650.
▲
by
raggi
5y ago
It seems to only target very commonly ported and generally understood hardware, and not many things. 32bit Intel, no arm. Bochs VBE and Intel graphics, e1000 and rtl8192 nics. They're very sensible targets as they'll let you run w
651.
▲
by
raggi
5y ago
https://fuchsia.googlesource.com/fuchsia/+/main/scripts/gce may work better. I will ask around if there's something up with some ACLs in code search, it seems to work for me on personal machines and
652.
▲
by
raggi
5y ago
FEMU is really AEMU, it forked from QEMU many years ago for a variety of reasons. Today this configuration offers Vulkan support, which will hopefully make it's way back to QEMU eventually. We use both QEMU and FEMU in our work, if you
653.
▲
by
raggi
5y ago
fwiw, I got us booting on GCE years ago, and the scripts may still work. https://cs.opensource.google/fuchsia/fuchsia/+/main:scripts/...
654.
▲
by
raggi
5y ago
There are many ways to cut metrics, but here's some line counts: C 365017 C Header 894250 CMake 396 C++ 2025857 C++ Header
655.
▲
by
raggi
5y ago
This is pleasing to hear :)
656.
▲
by
raggi
5y ago
Yes. Do not connect with agent forwarding, as doing so would allow the server operator to connect to other locations as you. Do not forward environment information, though the typical ssh default is not to. You will likely leak your usernam
657.
▲
by
raggi
5y ago
Do you get rejected if you connect with agent forwarding enabled?
658.
▲
by
raggi
5y ago
Test using pixel 4 xl built in speakers at 2/3 volume. You don't need high end gear, you can pick artifacts in half the test cases almost more easily on low end speakers. A few of the tracks only become noticeable once theyre busy
659.
▲
by
raggi
5y ago
It's called a Buffet Overrun
660.
▲
by
raggi
6y ago
Aren't VDA's cheaper?
More ›