A menu with "我的帳戶" in it, and often a generic icon or no icon at all, doesn't really have sufficient context to determine what the button means. If the website is already translated into your language then great, but many websites aren't (because it's a small site, or you don't speak one of the most common languages, or it's aimed at a different audience, etc.)
Bad UX is the result, from the combination of disabling text selection and being in a language you don't understand. Ideally both would be fixed - since unselectable text causes UX issues even when in a language I understand (when I want to select as I'm reading to keep place, or copy a partial link, or right click -> search/define a technical term, or copy-paste to tell someone what button to click, etc.)
Did you provide translations for all of those? If we expand to the immediate vicinity you can also throw in Thai and Vietnamese as well. Plenty of Japanese and Korean people live in Singapore too.
If you want to experience the frustration of text not being text, take a look at one of the main train ticket booking websites in China https://www.12306.cn/index/
Plain old text that can be selected is always going to be the most user friendly to non-native speaker users.
The question then is on the balance of trade offs which user group experience is the one you want to cater more to, non native speakers or keyboard-only users.
Edit: I love how one of the icons is 票 - perfectly self explanatory to Chinese speakers. Good luck if you don't speak Chinese which goes to show that icons are cultural to some degree
> - No ⌘. sending CTRL-C (muscle memory and being advertised as native to the Mac is what one would expect).
For what it's worth, that shortcut also doesn't work in the Jetbrains terminal, Alacritty, Kitty, and even iTerm2. The only terminal emulator I've tried so far where that works is the built-in Terminal.app.
I recently noticed the Spotify app does this as well. Ever since I've just been trying it, and have been coming across a surprising amount of apps that have the same behaviour.
You usually hit the pathological case when you have your own system prompt (i.e. over an API) forcing it to one-shot an action. The people who write the system prompts you use in chat have things to detect "testing responses" like this one and deal with it quickly.
Personally a big fan of 1Password. On the topic of autofill, the only website it sometimes won't fill is Reddit, which you know, whatever, I never go there anymore anyway.
As a developer I also love their ssh and gpg integrations, very handy.
I do get it for free from work, but if I had to choose one myself I'd have to pay for I'd probably still pick 1Passwrod.
> I do get it for free from work, but if I had to choose one myself I'd have to pay for I'd probably still pick 1Passwrod.
I wanted to highlight that "getting it for free from work" isn't a sweetheart deal offered just to OP, but a feature of 1Password for Teams, meaning all employees of a business that uses 1Password automatically have a Family license for use at home https://support.1password.com/link-family/
And, for clarity, it's merely a financial relationship: the business cannot manage your Family account, cannot see its contents, and if you have a separation event you can retain the Family account forever in a read only capacity or you can take over the payment (or, heh, I presume move to another employer that also uses 1Password) and nothing changes for your home passwords
I am continually puzzled that sometimes people can't put together a denial without including an affirmation as a crucial part of that denial. It's like they're doing the opposite of question-begging, they're saying that you're wrong because you're right.
All phones are insecure to some extent, most phones compared to GrapheneOS/Pixels are less secure and this has largely proven out whenever there's been leaks of the capabilities of law enforcement phone cracking tools.
QubesOS is an OS for PCs which have a standardized hardware interface. Support for older systems is basically "free". Smartphones aren't standardized in the same way and the amount of effort it takes to properly support other phones has a considerably higher cost on developer bandwidth.
Anyone can fork GrapheneOS and build it for other phones if they want, instead of doing this the developers instead focus their time and effort on the most suitable hardware for their needs. This isn't a part of some agenda or a swipe at Linux, open source or Stallman's cholesterol filled heart, it's just pragmatism.
GrapheneOS has to do substantial work on each supported device to integrate the hardening features and fix the issues those uncover. Supporting other devices is not easy and involves a lot of resources. Those devices also need to provide the hardware-based features heavily used by GrapheneOS including hardware memory tagging, pointer authentication, verified boot, etc. which those devices don't provide.
Instead there's a bunch of other arguments that are just as reasonable which underline why deploying their security focused OS on such a hardware platform would be a waste of their time. This is your refutation?
It really seems like you're more concerned about hurt feelings than objective fact here. Every link you've provided thus far was framed by you as evidence of poor decisions or behaviour on the part of the GrapheneOS team but you've done nothing to elaborate, and after reading the content of those links for myself there is nothing there that support the things you've been implying.
It doesn't make a whole lot of sense, at least not unless I put myself into the mindset of a child and read any negativity expressed towards FOSS projects as an attack, or taking their choice to not target phones I like personally.
I have no idea where you managed to find any feelings in my replies, and I will ignore the personal attacks.
The linked security-related arguments aren't reasonable at all. They talk about improving users' security but instead the actual result is less security for the majority of people, due to (1) the high price of the supported hardware, (2) reliance on Google hardware not trusted by many users (https://news.ycombinator.com/item?id=45101524).
> I have no idea where you managed to find any feelings in my replies, and I will ignore the personal attacks.
Your username is fsflover and your posts clearly have an ideological bias that favours purely open source solutions even if it goes against reason.
> The linked security-related arguments aren't reasonable at all. They talk about improving users' security but instead the actual result is less security for the majority of people, due to (1) the high price of the supported hardware, (2) reliance on Google hardware not trusted by many users
All SoCs are a black box and all of them are made by untrustable companies that are likely already working with the security services of whatever country they're R&D'd or manufactured in. There is no good solution to this, so they picked the best worst option.
Nonetheless, most of the evidence that is available shows that GrapheneOS on Pixels are the most secure phones currently available. So, clearly not security theatre, whereas if they also supported phones that didn't even let you lock the bootloader it absolutely would be.
GrapheneOS isn't to blame for every other phone manufacturer dropping the ball.
Thanks for the clarification. Free software ideology is not like a religion, where people believe in a god. Every Stallman's essay explains a very practical reason for following his ideas. FLOSS protects you from the enshittification, walled gardens, backdoors (to a degree) and similar things.
GrapheneOS have put themselves in Google's walled garden in terms of the supported devices and now Google can easily make them less secure or even kill them completely at will.
This is like saying "you clearly have an ideological bias that favors democracy/ or freedom even if it goes against reason". Sometimes a tyranny is more efficient at forcing people to do a particular thing, e.g., produce weapons. It doesn't mean that choosing it can be reasonable sometimes.
> All SoCs are a black box and all of them are made by untrustable companies
You clearly can't understand that different people have different threat models. This is a huge problem of GrapheneOS developers: they never accept this possibility and force the single threat model upon everyone. This reminds me of Apple by the way: They do the same. In reality, some people can trust Chinese devices more than Google's ones (imagine that), or trust a particular company that didn't perform a ton of evil action like Google did (that's me and many others).
> There is no good solution to this
The good solution to this is security through compartmentalization, which is the best security approach ever invented. The more varied hardware people use, the harder it is to make a targeted attack or to mass compromise every single device sold.
> most of the evidence that is available shows that GrapheneOS on Pixels are the most secure phones currently available
I don't dispute that, and you won't find me saying that GrapheneOS is insecure in itself. I am saying that they did a wrong bet long-term, and their approach leaves a lot of people without Google's hardware insecure.
> not security theatre, whereas if they also supported phones that didn't even let you lock the bootloader it absolutely would be.
Once again, this is implying one single threat model upon everyone. I never leave my phone unattended, so nobody can secretly reflash it. And whenever I suspect a compromise, I reflash it myself using a disposable VM on Qubes OS. Does it look somewhat secure to you?
GrapheneOS is for people who want highly private and secure mobile devices. It has a very reasonable set of security requirements for hardware listed at https://grapheneos.org/faq#future-devices. Other devices meeting these standards do not currently allow using another OS or do not allow it to use the security features on this list. It is not the fault of GrapheneOS that other OEMs do not allow using it and do not provide comparable security.
The purpose of GrapheneOS is not an OS which people can install on as many devices as possible where substantial security sacrifices need to be made even compared to the stock OS and a reasonable level of privacy and security cannot be provided due to lack of firmware/driver updates. Without the hardware-based features we use as part of our work, it would also hardly actually be GrapheneOS.
Support for installing another OS on devices has been removed or is in the process of being removed by several OEMs. Providing an OS for most mobile devices isn't an option in the first place.
GrapheneOS is actively working with a major OEM since June 2025 on a small subset of their next generation devices meeting all of our official requirements and providing official GrapheneOS support. The initial phase of support may still require people to install it themselves, but it will be another option than Pixels and the plan is to do more than that. The OEM is very interested in GrapheneOS and there may be devices sold with it as an official option. We'll be able to start doing lower level hardening work on firmware rather than our work not going below the level of the hypervisor, kernel and kernel drivers beyond reporting vulnerabilities or making suggestions. We already do a large amount of low-level work specific to devices and will be doing much more of it in the future including at a lower level. We have a lot of improvements we want to make at the level of the boot chain and secure element.
GrapheneOS in the long term will be a hardware, firmware and software project working closely with one or more OEMs to make highly private and secure devices. We'll support the existing Pixel devices until end-of-life and will add support for new generations of Pixels as long as they continue meeting our requirements, but our focus will shift to devices made in partnership with OEMs.
The purpose of GrapheneOS is not something people can download for their existing device to make it less bad. That's not even generally possible due to lack of support for using another OS and crippling of devices when another OS is used, especially the security features. You're talking about doing something which has never been the project's purpose. The purpose requires using the best available devices and ideally working with an OEM to make better devices for it as we're working towards (the first generation will likely not be more secure than Pixels, but it will meet our official requirements and improve from there).
Thank you for taking time to write this reply. I understand your reasoning better now, and your plans look very promising. I hope you and the OEM will not forget about the user freedom, too.
Googling would show that any number of users run into issues with OO/LO file corruption, often from power interruption during saves. The applications seem to handle that in a suboptimal way, and maintainers are unwilling to address it. My suspicion is that their unspoken contention is that the problem is with Windows, not OO/LO.
I recommend backing up to a general file type simply because it's less likely to open in the offending application by default, if the user ever needs to access it.
It does, but I also found forum posts discussing similar issues with LO back when I had the original issue. I won't risk it; I would rather use applications without a hint of these problems. I've also dropped Evernote after it ate a few of my notes. It's almost impossible to get that user trust back without excising every bit of the concern, and then some, and unfortunately, the LO development community (like many FOSS communities, as well as many proprietary developers) is too self-involved to do that sort of thing.
> That always made me chuckle, since the entire point of the system is that you're supposed to be incentivized to return the cart to get your money back
I always kinda doubted that part, or at least its effectiveness. Iirc a 50 eurocent coin will unlock most trolleys, which is pretty cheap for a whole ass trolley.
And sure enough, there's a lot of elderly people that just have a shopping trolley in their yard or something. This morning I found one randomly in our bike shed.
It's not an incentive to "not steal the trolley", it's an incentive to put it back in its place for people who were already not planning on stealing one.
This way the store and the customers don't have to deal with trolleys strewn around everywhere and blocking parking spaces, among other advantages.
I think when they removed the coins during Covid they just noticed that most people were already well-behaved enough to return the carts to their places, so the incentive is just not needed anymore. Actually in Belgium, Colruyt had never had coins for their carts and it just works.
In the United States, carts are free. There is a stereotype that homeless people have shopping carts in which they keep their things.
There's no particular need to change this, because one person can only use so many shopping carts. If you maintain the price at "free", demand saturates and people stop stealing carts.
It's common for people to return carts to a designated area, and it's also not rare for people to just leave the carts somewhere convenient for them. Store employees periodically go around and move the carts back to the place where you expect to pick them up.
Costco is an interesting hybrid case. They make it easy to return the carts "correctly" by providing little depots scattered throughout their enormous parking lot. Realistically, the parking lot is so large that very few people would be willing to return a cart to the front of the store, where you get the cart from if you're going shopping.
However, people also aren't going to pick up carts from those depots deep within the parking lot and wheel them over to the store. So Costco employees still have to make rounds of the parking lot and move carts that have been left there to their correct location at the front of the store. But for Costco, you're supposed to leave the cart in the parking lot, but only in certain locations.
You say this like it’s only Costco that has return bays. I’ve rarely encountered ANY store that has shopping carts but doesn’t have return bays throughout the parking lot.
You should actually read the comment you reply to...
> If you maintain the price at "free", demand saturates and people stop stealing carts.
If the price is $1, the same people who'll steal them will keep stealing them (with a screwdriver it's easy to pry your coin back out of the slot anyway).
> it's also not rare for people to just leave the carts somewhere convenient for them
With the coin, guess what... it will be rarer, because the people have incentive to get their coin back. At least in theory. And if someone doesn't care about their change, some enterprising kids might return the carts anyway to gain some money, and the end result for the supermarket is the same: carts at their designated return locations. The worker just has to go to 3 or 4 of these locations instead of running up and down the parking lot collecting all the stray carts.
We have an Aldi that has coin slots on carts, only store in the area that does this. I rarely have cash on me and never carry coins so the few I do have stay in the car for the rare coin-only parking meter. My wife likes to be nice and not get the quarter back after we return the cart. She'll get mad at me if I make an effort to get the quarter back or if someone hands me their quarter in exchange for the one already in the cart. Like, I'm not trying to be stingy. I could give a shit about paying an extra 25¢ to grocery shop. The issue is that I don't have many quarters and don't feel like getting cash back or digging through the junk drawer to find another quarter. That quarter is worth more to me than face value.
Yeah, I like Aldi’s products and prices a lot but this is always frustrating if I stop there unexpectedly and don’t have a quarter.
It’s compounded because they’re also the only store near me that doesn’t have handbaskets, so you need to either grab a cart or hope there’s a loose cardboard box in the store (which there often is because they expect customers to use them, but you may not find it quickly).
> it's an incentive to put it back in its place for people who were already not planning on stealing one.
It’s also an incentive for anyone else.
If I put the coin in and then leave the cart in the lot anyways, someone who wanders by is also incentivized to grab it and put it back, as they would get a free coin.
The system is actually somewhat elegant, if you return the cart you pay nothing and if you don’t you pay a small fine to whoever does.
This brought back a memory of living in Byron Bay Australia in 1999 - there was a person who’s full time job was driving around town with a trailer, collecting shopping trolleys and returning them to the Woolworths supermarket.
I’d never seen that in the uk - but maybe that town was the sweet spot in size where it was small enough that you could actually get home with a trolley (and it was nice and flat), and maybe the number of visitors passing through meant rules got broken more - though the trolleys were more in the suburban areas than just where the hostels were.
A lot of developers think code should be self-documenting, which I fully agree with.
Unfortunately though I don't think I've ever worked on a project that was actually self-documented, even though that is what the leads wanted.
When you don't know the language or what "My Account" means? Not everyone speaks English.
reply