Hacker Newsnew | past | comments | ask | show | jobs | submit | dambi0's commentslogin

The article you shared mentions a report and a committee that analyzed the report. You don’t happen to have any further details on those do you? I couldn’t immediately find the details in the article

A new account would tend not to have any history.

It seems uncharitable to immediately assume bad faith.

What is it about the content of the comment you disagree with?

I think it provides a further example to the parent post that regardless of what one thinks about the Act, the discourse isn’t entirely neutral.


That isn't true at all. Some of the argument relies on the experience of an Austrian education, but we are also encouraged to refer to other provided sources if we choose to seek other opinions.


Which is just as true of the argument

> Probably because you've seen it repeated so much in your hyper-propaganda bubble of reddit that you've started to believe it


Except that argument admits a means to evaluate it. You take the thing repeated ad nauseam and subject it to an evidentiary requirement. Are people actually being held without habeas corpus? Are people actually being executed based on their ethnicity? If anything in this nature is happening, has the rate of it significantly increased recently or has it been going on for decades?

The last question is pretty important if your argument is "Trump is a fascist and all we have to do is get him out", because then that argument is erroneous and you have to actually change the status quo instead of returning to it.


It could serve as a convenient distraction from rising prices or, more conveniently, risks introduced as a move to less staffing because of self-checkout.

It’s a bit easier to explain away the inconvenience of locked individual items as a response to a rise in crime rather than the store’s choices.

I doubt it’s a direct motivation but that isn’t to say it’s not a useful outcome either.


Or alternatively the word play wasn’t very good.


y. good nerd play.


Ruling out or refining an approach on the grounds it’s unlikely to lead to a suitable outcome (fixing and removing slop) is not the same as saying this code or approach represents a good enough outcome given what we currently know about the constraints of the problem (code review)


What if you include statistics on vans rather than just passenger cars?


You tell me. I sincerely doubt it would make any meaningful difference.


Well seeing as you alluded to having references amongst many I figured you might have the statistics to hand. Additionally why did you focus on new registrations rather than all registered vehicles?


How much plot could one or two paragraphs from a wider body of work contain?

We do learn more than it’s just muddy. For example the Lord Chancellor is somewhat introduced. We know the time of year. We know it’s been muddy for a while. We know the time of year. We know some term has finished so it’s likely that less people are around or it is quieter than usual. Whether these are relevant to the whole plot we can’t tell from such a short passage but that is true of any extract.


Doesn't Dickens want us to read the text slowly and imagine a dinosaur stomping around after the flood, horses and dogs dealing with the weather, and all sorts of visuals?

It's not about the amount of words but what is expected by the reader parsing them. The reader is expected to spend a lot of time imagining the non- plot stuff.


The fact we have labels for communication problems caused by failure to understand non-verbal cues doesn’t tell us that non-verbal cues are necessary for understanding


I didn't say they are always necessary so you probably replied in the wrong place.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: