grumdan

Even if you don't inspect it personally, there's a greater community of people who don't get their paycheck from Apple who may be looking at the code.

Regarding the Intel comparison, you have no choice but to trust them, but by using Apple products, you are trusting Intel and Apple, which is worse than just trusting Intel.

I'm not saying that this is what happened and I don't know the background of this story, but it would have been easy (technically speaking) to just push an update to Chinese users that will extract their private key and send it back to Apple without any significant changes.

> In general, I agree with you that army should not do that. However, enemy rarely tries to help to achieve that, i.e. they are not putting markers "we are soldiers, shoot us" at their positions. So goal of typical army is to minimize loses, not to avoid them completely, because risk is non-zero for every shot.

I agree with that. This is why I was only asking for "reasonable effort" to try to avoid this. I realize this is vague, but there are some cases that are a clear violation.

> It looks like error (one bomb hit territory near to hospital), not a precise strike.

In the incident I was referring to (but can't find a link to for some reason right now), it was a sustained attack that didn't even stop for 30 minutes after the US forces were informed that it was in fact a hospital. I'm willing to attribute that to negligence and organizational dysfunction rather than malice, but even then this indicates a lack of concern for civilians and medical infrastructure in the country.

Another case was the bombing of that infant formula factory in the first gulf war that was, as far as I remember, also claimed to be a chemical weapons site, even though the administration knew this not to be the case at the time. I don't remember many details on this though.

EDIT: Okay, seems I was mistaken about this, they didn't know this to be false, but seemingly had no strong evidence to support their claims either. Here are some other examples of US strikes on civilian targets based on very flimsy evidence though, along with details on that original hospital case I brought up:

http://www.informationclearinghouse.info/article43079.htm

> Intelligence was not falsified in case of Iraq: chemical weapons are just moved to Syria few months prior to invasion.

This is the first time I've heard this claim. Most Western media reported that there was no intelligence to support the WMD claim at the time. e.g:

https://www.theguardian.com/world/2004/jul/09/usa.iraq2

> Not sure about Vietnam, but I know a bit about war in Vietnam from USSR side, and it not good. Not interested much in the topic.

In this case the administration lied about the reasons for the war to the general public (this was revealed by the Pentagon Papers).

I know very little about the war in Ukraine, so I'll take your word for it. My main argument was that the US military in particular (since this is the one most relevant to Silicon Valley) has a history of this type of behavior that significantly complicates the point of military research to keep us safe.

I didn't have a concrete moral framework in mind, but I was thinking of a few baseline rules that should hold in most sensible systems, such as not bombing hospitals (like what happened with that hospital in Syria a while ago), not killing civilians without due process in countries where there is formally no state of war (drone strikes in Pakistan and other countries), not starting wars of aggression based on falsified intelligence (Iraq, Vietnam), investing reasonable amounts effort to avoid harm to non-combatants (that "Collateral Murder" video, US support for the Saudis' actions in Yemen, drone strikes based on just cellphone signals); all of these points are arguably features of many different moral systems and should be much less controversial than the dilemmas you bring up. I'm not a lawyer, but as far as I know to they are all standard components of rules of engagement (like the Geneva Convention).

Also, all of these have been repeatedly violated by the US military or its contractors, making me more hesitant to want to help them, even under the assumption that they are necessary to protect our liberal societies from other regimes.

Helping a military that is behaving immorally and being unarmed are also not the only options. If more and more people refuse to work for the military on these grounds, this may in the long run put some pressure on them to change their behavior. I do admit that this is hopelessly idealistic though and I don't expect this to happen anytime soon.

> Is it really so terrible to research arms for the people who risk their lives keep us safe?

If this was the only thing the weapons were used for, then maybe people wouldn't be upset about this kind of stuff.

The same arms will however also be used to extrajudicially assassinate people based on nothing else than their phone signal, in countries we are not at war with; they might be sold to regimes like Saudi Arabia that kills countless civilians in places like Yemen; they might be used to fight wars based on reasons the administration knows to be false; they might be used to gun down civilians by private contractors; the list goes on.

I'd not be opposed to participating in military research if the weapons were actually used "responsibly" (I know that's a somewhat subjective terms, but most of my examples are quite clearly considered immoral by most people).

Cryptography deals with much more than encryption though, and also in real life often you want properties that go beyond what encryption provides (message integrity, forgery resistance, perfect forward secrecy, anonymity, ..).

I was assuming that Atom extensions were similarly powerful as in Emacs, if that's not the case, then my point is moot.

I agree that restricting what plugins can do can lead to better stability and speed at the expense of extensibility.

> I had the impression VSCode had the technical superior plugin system

From my very limited exposure to VSCode's plugins, it seems they are a lot more limited in how they can change the editor's behavior compared to elisp code in Emacs[1]; if Atom plugins are similarly flexible, then I'm not sure you can say it's (strictly) technically superior if you can only implement a small subset of what you can do in competitor's system. At most it's better at some tasks and worse at others if this is the case.

[1] I didn't say "Emacs plugins" since there is no distinction between user-written elisp code and core editor elisp code and you don't need to create any plugin project or such, which I think makes for a much more organic and pleasant customization experience.

I'm German and had the same immediate reaction as you, but as the other comments show, it's hard to tell how widespread that is.

On the other hand, life might be harder for people who don't fit in well and then don't have anywhere else to socialize; the lack of privacy and constant gossiping probably doesn't help either. I grew up in a 1000-people village (not on an island though) and experienced this to some extent, but would imagine this to be much worse in such tiny communities.

> No, they don't. Do you really think if, before she made this purchase, someone explained to her all the things that could theoratically go wrong with this device, she would have chosen not buy it?

I think this would only indicate that she might care, but thought theoretical worst case scenarios are not sufficient evidence for an actual, non-negligible risk to her privacy. If they had told her that exactly this would definitely happen, I think she would not have bought the device.

I think one problem of explaining privacy issues to non-techies is justifying why theoretical problems often pose a real threat without sounding (to them) like a conspiracy theory nutcase.

How harmful a security breach is depends on how much private data about users was stored though. So even though security breaches will still happen, they will cause less harm if less sensitive data is leaked as a result.

Steppenwolf by Herman Hesse is also a great read, dealing with feelings of meaninglessness and how they relate to one's identity.

This is somewhat orthogonal to what window manager you are running though. Most tiling WMs (e.g. xmonad) combine nicely with xfce[1], giving you all the convenience features you mention while still not having to mess around with moving and resizing windows. I've been using that combination for years and wouldn't want to switch back to either plain xmonad or plain xfce.

[1] https://wiki.haskell.org/Xmonad/Using_xmonad_in_XFCE

This is pretty subjective, but I spend a lot more time typing than using the touchpad and while the MBPs touchpad may be nicer, I think the MacBook keyboards can't hold a candle to ThinkPad keyboards. I'm happy to trade a worse touchpad for that any day.

In addition, their server-side code is open-source and they claimed that federation support is something they are considering to add: https://github.com/wireapp/wire/issues/160

Overall, I'm also not a big fan of Signal's hostility to federated protocols (https://signal.org/blog/the-ecosystem-is-moving/), so I'm more optimistic about Wire in the long run.

Your comment doesn't actually add anything to this conversation, so regardless of one's position on Trump vs. Hillary, the downvotes are justified. All you said was a vague commonplace sentiment that has very little to do with the topic at hand.

Even if we assume that we wanted to address this for real, I fear that it will be next to impossible to actually assess whether whatever mistake caused this has actually been addressed when all the technology behind it is proprietary. I can easily see people being swayed by a well-written PR speech about how "human safety" is their "top priority" without anything substantial actually being done behind the scenes.

I think any attempts to address such issues have to come with far-ranging transparency regulations on companies, possibly including open-sourcing (most of) their code. I don't think regulatory agencies alone would have the right incentives to actually check up on this properly.

> Let's say that halving the death rate is what we can reasonably expect from the first generation of self driving cars.

What's your evidence for why this is a reasonable expectation? The fatalities compared to the amount of miles driven by autonomous vehicles so far shows that this is not possible at the moment. What evidence is there that this will radically improve soon?

I would say that variability in ability to concentrate might be reasonable to assume as a starting point, given how much people vary in other features as well, even if there are similar connections to how humans evolved. For example, food preferences are also in part a result of evolutionary processes, but vary a lot between people.

Good public transport systems provide even better benefits for points 1 (accidents involving subways are probably still going to be rarer than for self-driving cars), 2 (you don't even need fossil fuels for subways / trams), and 5. Arguably you don't lose much on point 3 either, except that you may have to walk a few minutes to/from the nearest stop (which is probably good from a public health standpoint anyway).

I agree that this only applies in cities though, and public transport may not help much in rural areas.

> What's the argument for night owls being better learners at night?

I would say that the argument is that many people report this to be the case and there are some studies pointing to the existence of such an effect (sometimes called chronotype). Even in the absence of rigorous studies, a reasonable default view is to say that such an effect may or may not exist and we just don't know for sure if it does.

I would argue that the burden of proof is on the person making the claim that this effect doesn't exist, as that's a stronger statement about the world. It also seems to contradict a lot of anecdotes without proper evidence that casts doubt on their accuracy; the simplified model that you mentioned, about sleep flushing toxins from your body may not capture all the relevant aspects here (for example genetic components to concentration ability).

Wikipedia references a few studies, but I'm not an expert and can't tell how conclusive or rigorous they are: https://en.wikipedia.org/wiki/Chronotype

Just flipping the about:config bit alone doesn't help much though. Any malicious extension installed after changing the setting would still have to have some payload with similar malicious behavior.

Moreover, even if Firefox was compiled without support for that setting, malware could patch the Firefox binary (or download and install a malicious version) to disable the check and then it'd still be able to hide itself as a Firefox extension instead of a binary somewhere else on the system. Once there's malicious code running on the system, it's game over and a flag disabling installing more malware using one out of many possible methods is not going to help much.

However, not having the setting may help for users that get tricked into toggling it through some web page telling them to and then installing a malicious extension, but that's a different scenario than an already compromised system.

> Because any extension or sufficiently capable malware on your system can change about:config values, meaning that if Mozilla ever makes a mistake in the extension approval process and lets a malicious extension slip through, or you get such malware on your PC, then that extension/malware would be able to flip that config bit and open the flood gates for all malicious extensions.

If such a malicious extension or other malware got installed, it could already do anything another malicious unsigned extension could, so I don't see how having this setting weakens security under the assumption that there's already malicious code running.

If we assume a system is compromised, it may as well install other malware as normal binaries instead of a Firefox extension.