Hacker News Reader

Comment by m463 | Hacker News Reader

I was thinking about buying an air fryer. My search came up with cookbooks specific to that air fryer, and I was intrigued. I found a good 5-star book, but then I found that ALL the 5-star reviews were submitted the same day.

I complained, but Amazon defended the book as legitimate, and since I hadn't purchased it, they would not take any action. (to be honest, I assume frontline customer service reps don't have much experience or power)

So I purchased it, complained, got a refund and then they were able to accept my complaint (after passing the complaint higher in the food chain).

Seriously, how hard was it amazon? I guess they're starting to notice.

Take a look at air fryer cookbooks - there are books specific to most makes and models. But everything is ML copypasta all the way up and down - the title, the recipes and the reviews all seem to be generated garbage.

View on HN

I'm the author of Python Crash Course, the best selling introductory Python book for some time now. Years ago, someone put out a book listing two authors: Mark Matthes and Eric Lutz. That's just a simple juxtaposition of my name and Mark Lutz, the author of O'Reilly's Learning Python. The subtitle is obviously taken from my book's subtitle as well. I assume the text is an ML-generated mess, but I haven't bought a copy to verify that.

I used to comment on reviews for books like these explaining what was happening, but Amazon turned off the ability to comment on reviews a long time ago.

I've spoken with other tech authors, and almost all of us get emails from people new to programming who have bought these kinds of books. If you're an experienced programmer, you probably know how to recognize a legitimate technical book. But people who are just starting to learn their first language don't always know what to look for. This is squarely on Amazon; they have blocked most or all of the channels for people to directly call out bad products, and they have allowed fake reviews to flourish and drown out authentic reviews.

i stopped frequenting the dev.to community because the average quality of articles just got so low it stopped being worth my time

Conversely if you post something sophisticated there it will likely bomb. A bunch of emojis and explaining JS closures for the hundredth time. Does well!

dev.to is blocked on HN for this reason (try submitting a dev.to link; it won't appear under New.)

There's an old thread where dang explains that it's blacklisted (along with many many other sites) due to the consistently poor article quality.

> try submitting a dev.to link; it won't appear under New.

I think you'll see it if you're logged in and have showdead turned on.

I think the best way to recognize a legitimate tech book is... visit a Barnes and Noble. If it's a publisher or series you can find printed on the shelf, books are legit.

Unfortunately online market "platforms" are pretty much widely untrustworthy for any sort of informational purposes.

also, just doing your research on any platform other than Amazon helps.

> If it's a publisher or series you can find printed on the shelf, books are legit.

Not even that is a guarantee, there have been cases of rip-offs making it through a bunch of book-on-demand services.

All "marketplaces" allowing third parties unlimited, unmonitored access to product listings suffer from that issue.

Why don't beginners start at Python.org, though? It's such a great resource to learn the language.

- it's free, unlike books

- always up-to-date, unlike even the best book after a few months

- easy to choose: heck, there's only one official documentation! No chance of making a mistake here!

I guess book authors don't like my perspective...

Are you suggesting people just go read the documentation like an encyclopedia? I don’t know a single person who got their start programming by doing that - just about everyone wants some sort of guide to help lead them in good directions.

I did. On Windows, Python had (still have?) a good offline help. And it included a nice getting started tutorial. The only book I had was “The C Programming Language”. But they ignited my interest enough to start researching, and I landed on the "Site du Zero" (now OpenClassrooms) platform. The web was sparser, but better, in these days (2010).

That's more or less exactly how I learned to program. From books, with a few friends. Only after it got to a certain level and I started frequenting more places where we met other people working with computers some of which were professional programmers.

I still have some of them. They've aged surprisingly well.

> That's more or less exactly how I learned to program. From books

What kind of books? The person you're replying to is arguing in favor of books, but saying that the documentation in particular is not a good one to start with.

I think you are assuming disagreement where there is none.

I wasn't assuming it, but I felt like directly asking wouldn't make my comment any clearer.

The problem is "That's more or less exactly how I learned to program." sounds like it's a response to the question in the comment you replied to. Which would mean there's either disagreement or you didn't understand the question. And the way you mentioned other people sounded like you might be interpreting "guide" as a human guide.

I guess instead you were skipping the question and talking about the second sentence?

In that case I recommend quoting which sentence you're replying to in a situation like this.

Do the official docs even have tutorials? I'd send beginners to Khan Academy instead.

Yeah, https://docs.python.org/3/tutorial/index.html. But I would say it is good for those who already know another programming language and not for complete beginners.

Many beginners do start at python.org. However, if you don't know anything about programming, and you don't know someone who can answer all the little questions that come up, it's really hard to learn from documentation alone. Even the official Python tutorial is fairly inaccessible to many people who are trying learn a language for the first time.

Almost every Python author I've spoken with recognizes that no one resource works best for everyone. We each write to offer our particular take on a subject, and hope to find an audience that our perspective resonates with. I've never steered people away from documentation; in fact one of my goals is to steer people to the sections of documentation that they're ready to make sense of. One of my end goals is that people no longer need me as a teacher. That was my goal as a classroom teacher, and it's one of my goals as an author.

The idea that there are no mistakes in official documentation is pretty unrealistic. Technical documentation has certainly improved over the last decade or so, but it will never be perfect. Most of us recognize that some areas of programming are better handled by third party libraries. In a similar way, there will always be room for learning resources that are maintained outside of official documentation sources.

I didn't claim the official docs have no mistakes.

Since there's only one documentation, beginners can't get wrong with which docs to use.

Ad opposed to books, which have tons of bad choices available (hence the current discussion).

Ugh. I hate the, "You're not a customer yet so our CRM system won't let me talk to you."

And what happens when my problem is that your system won't let me place an order?

CRM should never mean Sales Prevention as a Service.

The great thing about filtering is that you don't have to hear the screams.

These accidents play out in slow motion until someone corners you at a family reunion and asks why their friends can't create accounts and when you ask them how long they say "months".

You'd think... but in a growing b2b company, the CRM is where sales get prevented under a certain threshold. heh.

False Negatives and False Positives are always connected. On the other side of the equation, there are plenty of bad actors who will casually flag their competitors to score a quick win. Crime doesn't like to go uphill - raising the stakes for feedback lowers the prevalence of bad actors.

I think that's a different issue. Amazon has thorny problems with takedowns. Company A trying to get rival company B's listing taken down probably happens 100's of times a day. I believe Amazon uses "proof of purchase" kinda like a CAPTCHA or proof of work - an extra hoop to jump through to reduce the volume of these things they have to adjudicate.

It should be a term of service that you’re not allowed to interfere with other customer’s listings.

If I found out one of the tenants on my multi tenant system was trying to mess with another’s, I would be livid.

> Seriously, how hard was it amazon? I guess they're starting to notice.

It's not hard. It's a cost center, and they're in the business of making money - not providing the best service.

They're biggest risk has always been the perception they peddle fraudulent simulacrums of worthy products.

It’s the same across all big tech. The size/volume for complaint handling doesn’t scale. It’s either filtered out by some machine learning algorithm or some poor person in a 3rd world country getting paid next to nothing who reviews the complaints so quality isn’t of importance.

There been a recent influx of scammers on Facebook local groups. Air con cleaning, car valeting, everyone’s calling out the scammers in the comments yet when you click report to FB the response is we have reviewed the post and it has not breached our guidelines, would you like to block the user.

If I don't get where I want to be with the front door customer service within a decent amount of time, I have always had good success contacting jeff@amazon.com. Their executive support team gets back quickly via email or phone and they really seem to care.

Reply on news.ycombinator.com