kernel_task is due to high temperature on the left thunderbolt ports caused by charging. Plug your charger into the right side instead and it'll go away in about 15 seconds.
Are you sure it was that good with large files? I used to be a TextWrangler evangelist until I noticed it was mangling large files. You'd change one character but the saved file differed by far more than one character. Medium-sized nowadays, hundreds of MB TSV files. About 10 years ago.
Will L2 congestion become a problem once JWST and a few more missions join this craft? Will we have to manage that orbit like we do with geostationary?
I saw this post on the bus, installed the app. Checked what it says for my transfer. I had about 8 minutes.
Google said my bus comes at 7:29. Transit app says it comes at 7:37. The bus came at 7:29. The app didn't match the posted time table either.
This is in a major tech area next to a major university. Their About screen lists dozens of local transit authorities as data sources, with the one running the bus in question as a licensed source.
Because your $apt_repo_url is useless to a yum user.
The package management community needs to stop balkanizing.
How else do you propose releasing software to all Unix-like OSes without learning a half dozen package manager formats and quirks? Only other choice is a container.
American rail is built for cargo, so there are no sections as fast as a freeway.
Also there are several sections that are particularly slow. The track through the Rockies is the original 1869 route, with original curvy single track and original tunnels. Speed there is around 20mph. Sometimes you come close to I-70, where cars are going 60-80mph. The Sierras also have slow sections. That said, those were my favorite parts of the trip.
Patent Research Foundation | Front and backend | Full time | Seattle | ONSITE | www.patrf.org
We are a patent analytics startup in Bellevue, WA with a great culture and a great team. We have veterans of legal, data science, and scientific computing backgrounds. Our goal is to bring rigor to patent anaylitics.
We're recently funded and looking for senior front and backend engineers. Remote work is ok, but ideally we'd see you in our Bellevue offices relatively frequently.
The front-end person will help us build our web frontend platform.
The back-end person will work on our analytics platform, both on building features and helping with research.
I got a matte screen for the first time in several years, and immediately had a reflection problem from window light. On a glossy screen the problem is a relatively small highlight, but on the matte screen it was a giant blob that took up a significant part of the screen. The glossy screen is usable, the matte is not.
Anti glare has improved significantly on glossy screens, too.
I propose marking bots as such. Keep following useful bots. But users will at least know that a troll is automated and are able to act accordingly without wasting time and energy.
I agree that it's important to educate people both to question what they see and also how to question it. I disagree that this is the answer. We've been teaching that for years.
It's impossible to take everything critically, and honestly few will. So the entity with the largest bot army still has the longest propaganda lever.
Twitter and the other social networks know who is a bot, or else they haven't bothered looking. Something needs to force them to act.
I do see one mechanism: the bots go too far, and users don't want to be on a platform where they just interact with bots, so they go to more curated places to get their fill. So the business health of the platform depends on having trust. FB has a leg up on this since your friend list probably has people you've actually met. There the problem is your gullible friend forwarding you crap. A deputized bot, if you will. No level of education helps there.
Microsoft was on the other side of such a scheme when OS/2 got the ability to run Windows apps. That's when you stopped getting native OS/2 apps, because why write one when the Windows one works on both?
A big problem is that the Linux world loves to fragment itself.
There is no "Linux" to bundle. You have to pick a distro.
On the surface all major distros look like a good choice, but for unclear reasons they're all incompatible with each other despite doing the same things in the same ways.
For techies it doesn't matter because they can get any distro to work if one works. But it matters a lot for non-techies because things like tribal knowledge and even packages are all incompatible.
This is true on other OSes as well. C extensions are assumed to be compiled with the same compiler as Python itself. If you use distutils it will also use the same flags.
This is less problematic on other OSes where GCC dominates and clang is compatible, but it can bite on HPC systems that use PGI compilers.