It is extremely flaky on GrapheneOS, at least on my Pixel 8 Pro. Just typing Ctrl-D to exit will corrupt it, requiring a full reinstallation of the Debian VM
The built-in terminal app seems to be similarly flaky on my Pixel 8. Also, the kernel it boots into is really stripped down, and it lacks a ton of essential features. I was not able to install VirtualHere client to pass through USB devices, and there's no built-in functionality. There's also no way to open it full-screen on the Pixel 8's DP-over-USB-C desktop mode. Hopefully it continues to improve, but it seems like Google is more into extracting value than they are improving their products at this point.
Oh wow. I did a very basic test this morning `ping google.com` and then ctrl+c and it seemed to work okay. Not done any more extensive testing than this though.
Could it be that it's just very flaky on all pixel devices? Or maybe something graphene is doing to harden the OS doesn't play nicely with how it's been implemented?
I just tried it on my S25. I can enable the option an open the APK but can't download it because it fails to create the VM because the S25 does not support Non-protected VMs, so I may require a rooted device. I guess I will stick to Termux but interesting feature nonetheless
Termux is doing a container. The android terminal is doing a virtual machine. That's the difference.
Termux would definitely be the light weight option, but you will be pinned to whatever version of the kernel your device is shipped with (may be a bit old.)
No, termux isn't a container, it's running directly in userspace on the host. The only weird thing is that because it's running directly on the host, it has to be built to use unusual paths, eg. /data/data/com.termux/files/usr/bin/bash instead of /usr/bin/bash. If it used containers (which IIRC it can't because Android doesn't really support it) that would actually be easier because then it could use a chroot to make the paths look normal.
Ah, well that stinks a little. I guess it makes sense, if android doesn't mandate a few kernel settings then working with containers might not be an option.
Couldn’t it implement a “fake chroot” by e.g. creating its own libc which wraps the real one but with path remapping, and then linking all its executables against that?
That would only work for things that use libc (so eg. most Go programs are probably not going to work). The main way that you can do an unprivileged fake chroot is proot, which termux does offer - see https://wiki.termux.com/wiki/PRoot - but that has a significant performance hit.
Thanks! I agree with you to some extent. I _do_ need to get better with these tools but I remain suspicious for reasons like skill atrophy and the fact that I'm perhaps experienced enough to know how much I don't know. Using AI to plug gaps in my own skillset feels like setting a dangerous precedent on the one hand, but on the other hand, if it works then what's the problem? People just want things that work at the end of the day.
A message that I maybe didn't land in my post is that it is a little bit ridiculous to demand or only deliver "Artisanal Code". It's more labour intensive and the end product is virtually the same at the end of the day.
I appreciate your response, and I will confess to a certain bias as at some point I think I made the leap of acceptance of AI, in the sense of "this is how it's going to be from now on so I better get on board with it".
Spot on. These are the exact tools I was referring to. They seem a little un-magical but the real value is the boilerplate they provide for context management. Essentially allowing coding agents to perform at their beat. For what it's worth, Taches is my tool of choice.
I don't think this is saying that the ratings are unreliable, but rather that searching by rating isn't a guarantee that a high-rated restaurant will show up due to the other factors at play.
You don't get a sorted list from highest rated to lowest rated, but rather, momentum of reviews, number of reviews, changes in rating etc.
My suspicion is that there probably is also a noticeable difference between companies that advertise on Google vs. those that don't. Anecdotally, the gym closest to me has higher ratings than all the other gyms in my area, but when I moved to the area it never showed up on Google Maps for me. It was only by walking by it that I decided to look it up on Google Maps specifically by name that it showed up for me.
I have the same opinion, but my worry with this attitude is that it's going to hold me back in the long run.
A common thread in articles about developers using AI is that they're not impressed at first but then write more precise instructions and provide context in a more intuitive manner for the AI to read and that's the point at which they start to see results.
Would these principles not apply to regular developers as well? I suspect that most of my disappointment with these tools is that I haven't spend enough time learning how to use them correctly.
With Claude Code you can tell it what it did wrong. It's a bit hit-or-miss as to whether it will take your comments on board (or take them too literally) but I do think it's too powerful a tool to just ignore.
I don't want someone to just come and eat my cake because they've figured out how to make themselves productive with it.
I think of current state LLMs as precocious but green assistants that are sometimes useful but often screw up. It requires a significant amount of hand holding, still usually a net positive in my workflow but only (arbitrarily) a modest productivity bump (e.g. 10-15%). I feel like if I can get better at reigning in LLMs I can improve this productivity enhancement a bit more, but the idea that we can wholesale replace technical people is not realistic yet.
If I were a non-tech, non-specialist and/or had no business skills/experience and my job was mostly office admin I would be retraining however, because those jobs may be over except as vanity positions.
You don't even need to consider an ally being a bad actor. When Ukraine was invaded, wheat prices (to name one) in Europe shot up. England in particular has fairly significant food security challenges.
I think the key difference is that they perform a search of the foods in Macrofactor's database which means that you're more likely to get a good estimate.
From someone who weighed and scanned a lot of foods, it has really improved the workflow
This mentality is why rail is always destined to fail in the US. That mentality didn't stop China, which has a comparable land area to the continental United States and allows it to benefit from highly efficient cargo operations all over the country.
But if you want to keep the France comparison, let's compare a journey from Paris to Marseille with a journey from San Francisco to Los Angeles:
Paris to Marseille:
- by car: 7h30m (773km/480mi)
- by train: 3h11m direct for 94 EUR/103 USD with luggage and WiFi, amongst others
San Francisco to Los Angeles:
- by car: 6h (613km/381mi) - note that I'm checking the route times at 8:24 BST and so it's night time in the US - hence less traffic.
- by train (and bus): 8h30m with a change for 61 EUR/67 USD
The SF to LA route covers less distance and requires a change, so the passenger would travel 6h by train and then a further 2h on a bus - and when you compare that to driving it doesn't make any sense to ever pick public transport.
I do think there's more to it than just Americans not liking rail travel or preferring their cars. To build significant passenger rail infrastructure requires coordination at state government levels with all the stakeholders; funding; purcahsing land; technology to build modern and efficient railways etc. I don't believe the US has the capability to just build rail anymore - it would require significant investment in skills and manufacturing to increase capability to the point where it will be able to build the type of modern railway they would need to actually be a viable alternative.
It's agonizingly slow (I used to take it to college in SLO) but between SF and LA you could take the Coast Starlight (you'd BART to Oakland first). It's 12 hours though. https://www.amtrak.com/coast-starlight-train . I used to take the Chinatown busses and they were faster, or ride share on Craigslist (or, really aging myself here - remember Zimride?)
My interpretation was that adolescent refers to the maturity with respect to life experiences.
The set of life experiences that a bachelor can experience is smaller than the set of life experiences that a married man with children can experience.
I don't think this is a particularly offensive thing to say, but instead, given that she's from a group whose voice has been historically subdued, I think it's a witty reparatee to refer to that philosophy as "adolescent". I don't think she's calling bachelors adolescent, but for the word of a bachelor with most likely more limited life experience, I do think it's fair to call out that it perhaps isn't the most rounded point of view.
