Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

> I think this is a really thorough mea culpa which is quite impressive

It took almost 6 months of badgering for that to finally happen though, despite an extensive and widely signed (by experts) op-ed pointing out how wrong and dangerous it was barely a week after the original article was published, and calls for corrections/retractions started within days if not hours.

> given the frequent failure of other newspapers to publish a prominent apology when they have got things far more wrong than this.

Getting things "far more wrong" matters less when it doesn't put lives at stakes, and you don't need to believe me on that, this correction article makes that point:

> During the review I independently confirmed that a Turkish government official had used the article when, in effect, attempting to deter users from WhatsApp.



> It took almost 6 months of badgering for that to finally happen though, despite an extensive and widely signed (by experts) op-ed pointing out how wrong and dangerous it was barely a week after the original article was published, and calls for corrections/retractions started within days if not hours.

Completely agreed, but they do fully concede this in their retraction:

>This made a relatively small, expert, vocal and persistent audience very angry. Guardian editors did not react to an open letter co-signed by 72 experts in a way commensurate with the combined stature of the critics and the huge number of people potentially affected by the story.

It might have taken a while for their internal review to complete but, honestly, what more can we ask for but an honest admission like this? If more media outlets could meet this standard we'd be in a much better place.


It's not a retraction.


Agreed!

Could they have come back to us faster, yes, but an honest, well thought through apology with all these details included is a really strong response and I'm not going to complain.

We can always complain about why not this or that, but it seems we are never satisfied even when we get what we want.


>what more can we ask for but an honest admission like this?

We can ask people to simply not trust traditional newspapers for reporting on stories concerning highly technical and scientific fields.


Who should we trust? Business Insider? TechCrunch?

They all get it wrong sometimes. We should be equally sceptical of all sources of news.


Yes, we should be skeptical of all news sources, but from my experience, traditional newspapers tend to consistently misreport information security news.

I personally only trust security blogs and Twitters operated by certain security experts when it comes to news about security. I have a list of about 30 or so experts I trust.

I know that's not really practical advice for a typical person, though.

I think the Guardian handled the retraction and apology as best they could, and they deserve props for that, but it seems the hit-miss ratio for infosec stories is very poor for most "mainstream" sources out there (as much as I despise the "MSM" term).


As an ordinary guy for whom IT security is just one of many topics in which I'm interested, what do you suggest? I can't follow 1000s of blogs that cover all my interests.

And how do I choose reputable blogs in the first place? Do I trust reputation on HN and Reddit?

In the end, sure some articles will get some things wrong. But I would like to see evidence that they get it wrong more often than any other general source of news.


I don't have time to list everyone, but a few off-hand:

For crypto, Matt Green: https://twitter.com/matthew_d_green https://blog.cryptographyengineering.com/

For cybercrime, Brian Krebs (though I would call him an investigative journalist rather than an expert): https://krebsonsecurity.com/

For intelligence/counterintelligence/OPSEC, thegrugq: https://twitter.com/thegrugq https://medium.com/@thegrugq

A few of these can occasionally be biased when there's a political edge to something, but some others I trust: Moxie Marlinspike, Daniel Bernstein, Dan Kaminsky, Rob Graham, Thomas Ptacek, Michał Zalewski, @SwiftOnSecurity (semi-parody account, but trustworthy info), Tavis Ormandy


Which other infosec stories did the Guardian screw up?


I don't read it, but just from a quick glance, this headline is misleading: https://www.theguardian.com/world/2017/jun/07/russian-hacker...

>An investigation by the FBI has concluded that Russian hackers were responsible for sending out fake messages from the Qatari government, sparking the Gulf’s biggest diplomatic crisis in decades.

>It is believed that the Russian government was not involved in the hacks; instead, freelance hackers were paid to undertake the work on behalf of some other state or individual.

They could've easily made the headline "FBI: Qatar hackers of Russian nationality". By making the first 2 words of the headline "Russian hackers", they're obviously trying to take advantage of the recent surge in reports over Russian state-sponsored hacking. Most readers who see that headline are going to assume they meant "Russian state hackers", until they read the second paragraph.

That said, I don't see any factual errors in the article itself.




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

Search: