Other battery types, like laptop batteries (uses I2C) are not supported.
This info should really be in the top-level README
9,701 karma · joined May 4, 2016
Other battery types, like laptop batteries (uses I2C) are not supported.
This info should really be in the top-level README
It has a list of regex'es which behave differently in different regex implementations, but does not name those implementations. Pretty disappointing.
Most of the Chromium-based browsers remove Google's telemetry and integrations, so if someone has a problem with Chrome "looking over my shoulder", then they have many more alternatives.
And sure, Manifest V2 will disappear in all those forks, but many browsers have built-in adblocker instead, for example both Opera and Brave claim to. And other than adblockers, what else is lost in Manifest V2?
Mind you, I am still using Firefox, but saying "the only alternatives are Chrome and Firefox" is completely incorrect.
It seems that any time I find a website with a horrible design and start looking for that "reader" button, that button is disabled in Firefox. Sometimes I tolerate the design, sometimes I switch to Chrome - because reader mode works on all sites there.
Firefox on my laptop has a propensity for eating lots of RAM, and constant CPU usage, even if nothing is going on. Task manager is useless. Profiler just shows "GC collecting".
Mobile firefox is OK, it simply feels like it's designed for a different person. For example even Chrome has "force reader mode" button.. Firefox does not. And don't even get me started on the "new tab" [age - it gives me 6 choices and all of them are "AI" ones which never show what I want. The only way to show a manually curated link list is "collections" and those do not get synced, and occasionally disappear for no reason.
If not, I recommend renaming it to something else, to avoid confusion. LehnABC perhaps?
[0] https://en.wikipedia.org/wiki/ABC_(programming_language)
Since it was a fixed vuln in closed-source cloud-based service, we will likely never have any more information. There are no patches to analyze or logs to check.
But I guess it's fine if author wants to make a political statement? Plus it does not matter that much anyway, if one needs this under regular license, they can always ask agent to rewrite all code...
That's the most frustrating part, really.
If you don't have co-routines, you need to have something else. That "something else" is most commonly threads, and each thread has non-trivial overhead, in scheduling and memory. A coroutine is much more memory efficient.
Removing GIL has nothing to do with that, coroutines are even worse re GIL problems. And of course, there are plenty of thread-based frameworks for Python, already - Flask and Django are most well known examples, but there are tons of others.
So my first question for author would be: "how is this different from Flask"?
Granted, there is a healthy dose of luck and talent involved, but saying that the only way to get rich is to have rich parents is a lazy and wrong take.
Every technology like this makes testing faster at the expense of introducing false positives / false negatives. AI is really not going to change that much.
(Unless AI will be solely responsible for grading, and there will be no human fallback / appeals process.. in which case this will be hell and I don't envy students stuck with this)
Instead of a single config dir where you just deploy the configs and you are done, IIS settings were all over the registry. And don't forget component registration - a single, global registry per machine, used by IIS. Want to run staging and dev server on same machine? Tough luck. Random software installs can upgrade components and cause failures in completely unrelated failures...
- The take-home projects and papers are useless for grading purposes because of AI cheating. Time for more oral exams.
Yep, the only real change from AI. This used to be $50/page from diploma mill and this was expensive enough to dissuade most people. Now it's almost free, and cheating is rampant.
I think author's hope for oral exams is too optimistic, that is too expensive. Good old written exams, either in blue books or on special locked-down devices, proctored to prevent cheating, will be a way to go.
- It makes no sense for professors to stand in front of large classes and deliver lectures. They should do small, interactive seminars.
I don't see what this has to do with AI? Ever since VCR was invented, professors doing non-interactive lectures could be replaced with recordings. Either there is a "human element" in those, in which case it's still there, AI or not; or it's just pure information transfer, in which case AI is just an inferior version of recorded lecture with worse presentation style and hallucinations.
- Professors should teach about AI, include how to prompt efficiently and how to critically scrutinize the output
100% agree, as with any important technology. But this is just a small curriculum change, just one more topic to cover in the lecture, and maybe a class.
---
Overall, I believe that modern university is pretty bad and should be changed, but I hardly think that AI made this problem significantly worse, and this essay really did not convince me of that.
My first job I was in ~10 people robotics/hardware startup, and we had CI cluster of about 4 workers, later extended to ~10 machines (managed by buildbot, those were the times...). And the parallel jobs happen all the time in commercial developments, no one likes to wait for a long time when they could wait for a short time. This is not done for various OS'es, instead one machine might be testing service A, while other tests service B, while the third one tests integration.
That said, there is nothing wrong with "Dead Simple" niche - I am sure that there are some users who could use it. Although those would probably want something with less lock-in, your current approach is pretty weird: the jobs which _look_ like they are written in regular bash/python/etc.. but actually they have extra functions ("run_task", "task_var" etc..) magically injected, so you cannot run the file directly nor can you run linters on those. I'd much prefer pretty much anything else which can run unmodified code.
That DSCI is not a forge - it's a task runner attached to gitweb frontend. There are _none_ of the forge-like features like user management, pull requests, etc... Even in CI area, most of the features are missing: from the quick perusal of the doc, even something as basic as "have runner on remote machine" or "run two jobs at once" is not implemented.
Which is a pity - as I think most projects just need a git host + web UI + easy pull requests + a way to clone other's repos, and tangled seems to do this pretty well.
- You want to test multiple build configurations in parallel, on multiple machines. How do you do locally?
- You need some test data, how do you handle auth to your file storage?
- You want it to be easy to jump directly to test results, and you want to see what takes all that time, how do you deal with that?
- Some of your tests require specific hardware (like GPU), how do you make sure it's provided?
- You have secured publish pipelines with OIDC and temporary token, so developer's machine compromise cannot upload rogue package. How does this work without CI?
Docker containers / nix is a great start, but it's nowhere close to the full picture. And yes, you can solve all of those problems, and get backup method for the cases when github is down, but it often means re-implementing things twice.. and a lot of teams simply do not bother.
That said, if you are so dependent on CI that you cannot even test the code locally, then there is something very wrong with your project. Every developer should be able to at least build the code and run basic/quick tests without CI's involvement.
> None of this is to say that RISC-V is doomed. As I said, I fully expect it to take over the space currently occupied by [...] Much like the linux kernel -- the price is right.
Also tons of AI-generated slop. Fr example the "shields do not protect from capacitive coupling" part is absolutely nonsense. Ground reference jumping, within reason, is totally OK for RS485. The disturbances do not accumulate. While DC/DC converters _do_ have fast switching transients, those are blocked by input/output capacitors and do not appear on the cables, unless those converts are defective.
Also, the title seems to lie. Author tried original cable and one more "OEM cable", and that's it.
Worst kind of AI slop.
I strongly suspect a measurement artifact.
They also had 2 small kids, so lots of more expensive resort activities could not be done.
No wonder they left $33k out of $40k unspent, this is as non-representative as it gets.
"Reading another user’s files" won't work, "Writing as another user" won't work, "Local impersonation via SUID binary" won't work either. You need root to bind to low port, and if you have it, there is no need to mess with NFS.
It's not to say that NFS is a secure filesystem by modem means - if there are clients where attacker has root access, they can pretend to be any user, and try to spoof IP to bypass access lists. But if used in the circumstances it was designed for - a secure network with only trusted computers connected - it's actually not that bad.
breaking news! If you completely control a computer, you can completely control a computer! This means doing anything the computer can do, such as logging in with passkeys, and using "passkey sync" functionality to sync passkeys.
> The helper gets on with the job. Moving quietly through your home, doing the work. Removing mounting plates, screwing in replacements with a cordless drill, testing each -- beep -- one, moving the ladder, sweeping cobwebs and dust away. All silently, with nary a peep out of them.
... until they eventually come to the alarm that was mounted near the light. The electrician did not know this, but the light was installed incorrectly, and its wires ere running just under the place where the new alarm was going to be installed.
When the helper installs the alarm, the screw goes in - and there is a loud snap and buzz, light blinks a few times and starts glowing really dimply. The helper ignores it - they were only told to install fire alarms. They put in the second screw and move on. Installing the smoke detector is a bit tricky, with the light out, strong smell of smoke and mounting plate deforming from heat, but the helper manages to do it - as they are pretty smart! Helper feels really good now that all the assigned tasks are done. They go to find the owner, but looks like he stepped out. No matter, helper will send the bill electronically for the job well done.
Good, this should have been done long time ago. And good for Apple for not cooperating.
Once this goes in, I'd expect to see 89%, which is zero nines. (I'd like to say, "a new low!", but sadly we've had this before)
Instead of trying to buy into that, I found out that individual tools are much nicer with. Minicom for fancy conversations. Custom scripts/picocom/tio for basic ones. Perl (later python) to script. zmodem to transfer files (or xmodem if the memory size is extremely limited).
But the biggest replacement is SSH of course. For the last 20 years, most of the devices which have serial ports also have Ethernet access. Use serial port to bootstrap, and then switch to SSH. "dropbear" is tiny and has very few dependencies, and yet gives a vastly superior experience compared to old serial terminal: multiple connections, file transfer, scriptability, etc...