4 ms·

Again. Packt publication books are rarely good. Nothing against the author. But I have been time and again disappointed with their content quality. They go

Again. Packt publication books are rarely good. Nothing against the author. But I have been time and again disappointed with their content quality. They go for quantity of books over even the basic quality.

blowski 3y ago
Indeed, they do very little work in selecting authors and editing the books, but they still take a big revenue cut. So they tend to attract lower-quality authors. There are a few exceptions but I agree with the general opinion of Packt. They approached me a few years ago to write something, and Leanpub seemed a much more compelling offer.
rich_sasha 3y ago
I have one or two Packt books that are very good. But overall agreed. This seems like a strange mix of "hard" algorithms like sorting or binary search, and ML things like k-means clustering. Not sure who the intended audience is.
- wyclif 3y ago
  The intended audience is the legions of people grinding Leetcode and going through our industry's ridiculous interviewing process.
  - usrbinbash 3y ago
    They would be better served with more focused books however.
    - mlrtime 3y ago
      Better served pre or post interview? I'd argue they're better served doing leetcode (given finite time) to land the job, THEN reading this book to become better at their job.
      - rtaylorgarlock 3y ago
        I came here to specifically look for this comment, and... well, the rest says a lot about HN ;)
wyclif 3y ago
That's because Packt, unlike O'Reilly titles, are the shovelware of the programming textbook market. They have a pretty sorry business model. Lots of copypasta-style content slapped up into a book, then sending free copies of the book in exchange for reviews.
- bradley13 3y ago
  I wondered about that. I just looked at the table of contents, and there is far, far too much stuff in this book. Either it is 2000 pages long (it's not), or it is incredibly superficial.
- zerkten 3y ago
  They will take on literally anyone as an author. Authorship gets you kudos in the increasingly large segment of programmer influencers.
0RQ2kaTBBlQp 3y ago
Any opinion on this specific book?
- W0lf 3y ago
  Judging from the table of contents it seems that the first half of this book covers basic CS stuff (what is an algorithm, Big O notation, sorting and searching, data structures etc) and then discusses some machine learning basics before looking at the more recent deep learning architectures. I would not buy this book as there are plenty of (better) online resources and even better books IMHO such as the books from Bishop or Marsland.
- lxgr 3y ago
  I just randomly picked a chapter discussing a topic I'm somewhat familiar with (the one on TLS handshakes), and I'm not impressed with what I'm seeing. I don't think anybody that doesn't already know the basics of TLS would be able to make sense of it as presented there: There's an implicit assumption that an RSA public key is used for key establishment that's never mentioned, and the diagram (which is sloppily drawn too) doesn't help things either. In the end, "a" secret key is used to encrypt something – but the core of TLS key establishment (again, only the RSA variants) is how that key is agreed upon using only the server's public key. That's arguably the entire algorithm here (the rest is protocol negotiation)!
cengizIO 3y ago
Ditto. I was asked to write a Rust for Linux book back in 2020. Rust was (and still is) not actually driving any part of the Linux kernel so I kindly refused. They very clearly stated that they are only interested in being the first publisher with a book on this supposedly green field.
- matsemann 3y ago
  I was asked to be a reviewer / editor for a book back in 2014, as a uni student having a somewhat viral github library on top of a framework, but really no deeper understanding of the framework the book was supposed to be about. And what I would get for that work was a short bio in the book, a copy of the finished book, and another book of my choice. While I at the time was flattered and almost said yes, it also felt a bit weird knowing they didn't vet people more. What did that say about other books from them I had read?
flor1s 3y ago
It's weird how mixed quality their books are, and how many overlapping books they have. From Packt I like Sebastian Raschka's books on Python Machine Learning. Other than that from time to time they do have some good "recipes", but hardly any of their books seems worth reading cover to cover.
- zerkten 3y ago
  I think that's a case where a good author has chosen the publisher for some reason. They are good enough to be published by someone else but went with Packt possibly because it was an easy onboarding.
_the_inflator 3y ago
Did you read the book? Or is this just some general judgment you are passing? I own the book, and I enjoy it. Bought it for Xmas.
- sidcool 3y ago
  I read it superficially. And it's okayish. Again, nothing against the author.
jmaker 3y ago
Opposite to Manning, O’Reilly proper, and No Starch Press books, Packt’s aren’t even properly edited. Apress books are the worst in my experience, really zero editorial work noticeable, lots of grammatical and logical mistakes, many obvious code errors. There are a couple better Packt books, but I don’t even bother looking inside anymore.
- wyclif 3y ago
  Just a thought: if I were a decision-maker at O'Reilly, there's no way I'd allow my brand to get diluted by Packt with these oreilly.com joint publications. Reputation is everything (or almost everything) in the publishing world, and it's why when you see O'Reilly's name on the spine, or Wiley, Addison-Wesley, Manning, or No Starch, you can expect a certain quality.
3abiton 3y ago
Any better recommendations?
- harshreality 3y ago
  Algorithms? - Algorithm Design Manual (3rd ed.) (by Skiena) - Art of Computer Programming vols 1, 2, 3, 4A, 4B (editions 3,3,2,1,1) (collectively, "TAOCP") (by Knuth) - Algorithms (by Erickson) [CC-BY-4.0] - Algorithms (by Sedgewick & Wayne) - Introduction to Algorithms (4th ed.) ("CLRS") (by Cormen et al.) - How to Design Programs (by Felleisen et al.) - crossover between basic programming advice, functional programming, and some limited/basic algorithms coverage [CC-BY-NC-ND-4.0] - Pearls of Functional Algorithm Design (by Bird) - Purely Functional Data Structures (by Okasaki) Publishers? Pragmatic Bookshelf, or O'Reilly (not oreilly.com-selling-Packt). Find what everyone is using as a learning/reference resource for a specific topic and use that, regardless of publisher; it's rarely if ever Packt.
  - tmtvl 3y ago
    I would also add James Aspnes' Notes on Data Structures and Programming Techniques (<https://cs.yale.edu/homes/aspnes/classes/223/notes.html https://cs.yale.edu/homes/aspnes/classes/223/notes.html>, CC BY-SA 4.0).
  - emmanueloga_ 3y ago
    From all these I feel like the best intro is Sedgewick & Wayne. CLRS is less solo-learner friendly, in my opinion. I skimmed Erickson and I liked it too. In the "friendly+actually-useful" category, you missed "Open Data Structures" [1] ... I think it has one of the best descriptions of trees, including B-Trees which seem to be only superficially described in other books. The "Think" series is also pretty good and covers more than just algorithms [2]. TAOCP, Okasaki, Bird: these feel really academic and/or less useful for industry work. Reading HTDP wasn't as useful to me, at least: I feel like Felleisen's books describe what he "imagines" would be a good way to write software on a "real" job :-p (non-academic setting). Skiena: I have mixed feelings about it... It seems to have gained popularity since Steve Yegge's blog post about Google interviews, but imho many concepts in that book are better explained elsewhere. I suppose the intended purpose of the book was to serve as a reference and an index to other resources. However, it often refers to sources that are 10 years old or older, academic implementations that could be challenging to apply in a real job setting, or things that are no longer available online afaict (e.g., LEDA), or hard to read papers, most of which lack a working implementation of the described concepts, etc. I wonder if anybody can really vouch for it or have a personal anecdote of the book being useful or enjoyable. -- 1: https://github.com/patmorin/ods https://github.com/patmorin/ods 2: https://greenteapress.com/wp/ https://greenteapress.com/wp/
raincole 3y ago
Packt feels like Udemy to me. I know there are some good stuff, but...
- wyclif 3y ago
  It's a comparable situation, I tend to agree. I wouldn't make a 1:1 comparison but Packt is not known for publishing well-edited and written books.
- aerhardt 3y ago
  Personally had an order of magnitude better experience with Udemy than Packt. Hard to find advanced content on Udemy but there’s a lot of beginner content that I’d consider rigorous for its level - can’t say the same about Packt.
ruiycombinator 3y ago
second this, it seems that anyone can publish with them and the books most of the times are terrible, like reading clickbait articles on medium. Terrible.
bogomipz 3y ago
I completely agree. Packt as a publisher is hot garbage. That's not to say every author is bad but but even when you get a quality author with good content, you will have horrific formatting and nonsense like pages with one sentence on them, illegible screenshots and there's often a complete absence of proof-reading or editing. The Packt game seems to be to simply flood search results hence quantity over quality.

blowski 3y ago

Indeed, they do very little work in selecting authors and editing the books, but they still take a big revenue cut. So they tend to attract lower-quality authors. There are a few exceptions but I agree with the general opinion of Packt. They approached me a few years ago to write something, and Leanpub seemed a much more compelling offer.

rich_sasha 3y ago

I have one or two Packt books that are very good. But overall agreed. This seems like a strange mix of "hard" algorithms like sorting or binary search, and ML things like k-means clustering. Not sure who the intended audience is.

wyclif 3y ago
The intended audience is the legions of people grinding Leetcode and going through our industry's ridiculous interviewing process.
- usrbinbash 3y ago
  They would be better served with more focused books however.
  - mlrtime 3y ago
    Better served pre or post interview? I'd argue they're better served doing leetcode (given finite time) to land the job, THEN reading this book to become better at their job.
    - rtaylorgarlock 3y ago
      I came here to specifically look for this comment, and... well, the rest says a lot about HN ;)

wyclif 3y ago

That's because Packt, unlike O'Reilly titles, are the shovelware of the programming textbook market. They have a pretty sorry business model. Lots of copypasta-style content slapped up into a book, then sending free copies of the book in exchange for reviews.

bradley13 3y ago
I wondered about that. I just looked at the table of contents, and there is far, far too much stuff in this book. Either it is 2000 pages long (it's not), or it is incredibly superficial.
zerkten 3y ago
They will take on literally anyone as an author. Authorship gets you kudos in the increasingly large segment of programmer influencers.

0RQ2kaTBBlQp 3y ago

Any opinion on this specific book?

W0lf 3y ago
Judging from the table of contents it seems that the first half of this book covers basic CS stuff (what is an algorithm, Big O notation, sorting and searching, data structures etc) and then discusses some machine learning basics before looking at the more recent deep learning architectures. I would not buy this book as there are plenty of (better) online resources and even better books IMHO such as the books from Bishop or Marsland.
lxgr 3y ago
I just randomly picked a chapter discussing a topic I'm somewhat familiar with (the one on TLS handshakes), and I'm not impressed with what I'm seeing. I don't think anybody that doesn't already know the basics of TLS would be able to make sense of it as presented there: There's an implicit assumption that an RSA public key is used for key establishment that's never mentioned, and the diagram (which is sloppily drawn too) doesn't help things either. In the end, "a" secret key is used to encrypt something – but the core of TLS key establishment (again, only the RSA variants) is how that key is agreed upon using only the server's public key. That's arguably the entire algorithm here (the rest is protocol negotiation)!

cengizIO 3y ago

Ditto. I was asked to write a Rust for Linux book back in 2020. Rust was (and still is) not actually driving any part of the Linux kernel so I kindly refused. They very clearly stated that they are only interested in being the first publisher with a book on this supposedly green field.

matsemann 3y ago
I was asked to be a reviewer / editor for a book back in 2014, as a uni student having a somewhat viral github library on top of a framework, but really no deeper understanding of the framework the book was supposed to be about. And what I would get for that work was a short bio in the book, a copy of the finished book, and another book of my choice. While I at the time was flattered and almost said yes, it also felt a bit weird knowing they didn't vet people more. What did that say about other books from them I had read?

flor1s 3y ago

It's weird how mixed quality their books are, and how many overlapping books they have. From Packt I like Sebastian Raschka's books on Python Machine Learning. Other than that from time to time they do have some good "recipes", but hardly any of their books seems worth reading cover to cover.

zerkten 3y ago
I think that's a case where a good author has chosen the publisher for some reason. They are good enough to be published by someone else but went with Packt possibly because it was an easy onboarding.

_the_inflator 3y ago

Did you read the book? Or is this just some general judgment you are passing? I own the book, and I enjoy it. Bought it for Xmas.

sidcool 3y ago
I read it superficially. And it's okayish. Again, nothing against the author.

jmaker 3y ago

Opposite to Manning, O’Reilly proper, and No Starch Press books, Packt’s aren’t even properly edited. Apress books are the worst in my experience, really zero editorial work noticeable, lots of grammatical and logical mistakes, many obvious code errors. There are a couple better Packt books, but I don’t even bother looking inside anymore.

wyclif 3y ago
Just a thought: if I were a decision-maker at O'Reilly, there's no way I'd allow my brand to get diluted by Packt with these oreilly.com joint publications. Reputation is everything (or almost everything) in the publishing world, and it's why when you see O'Reilly's name on the spine, or Wiley, Addison-Wesley, Manning, or No Starch, you can expect a certain quality.

3abiton 3y ago

Any better recommendations?

harshreality 3y ago
Algorithms? - Algorithm Design Manual (3rd ed.) (by Skiena) - Art of Computer Programming vols 1, 2, 3, 4A, 4B (editions 3,3,2,1,1) (collectively, "TAOCP") (by Knuth) - Algorithms (by Erickson) [CC-BY-4.0] - Algorithms (by Sedgewick & Wayne) - Introduction to Algorithms (4th ed.) ("CLRS") (by Cormen et al.) - How to Design Programs (by Felleisen et al.) - crossover between basic programming advice, functional programming, and some limited/basic algorithms coverage [CC-BY-NC-ND-4.0] - Pearls of Functional Algorithm Design (by Bird) - Purely Functional Data Structures (by Okasaki) Publishers? Pragmatic Bookshelf, or O'Reilly (not oreilly.com-selling-Packt). Find what everyone is using as a learning/reference resource for a specific topic and use that, regardless of publisher; it's rarely if ever Packt.
- tmtvl 3y ago
  I would also add James Aspnes' Notes on Data Structures and Programming Techniques (<https://cs.yale.edu/homes/aspnes/classes/223/notes.html https://cs.yale.edu/homes/aspnes/classes/223/notes.html>, CC BY-SA 4.0).
- emmanueloga_ 3y ago
  From all these I feel like the best intro is Sedgewick & Wayne. CLRS is less solo-learner friendly, in my opinion. I skimmed Erickson and I liked it too. In the "friendly+actually-useful" category, you missed "Open Data Structures" [1] ... I think it has one of the best descriptions of trees, including B-Trees which seem to be only superficially described in other books. The "Think" series is also pretty good and covers more than just algorithms [2]. TAOCP, Okasaki, Bird: these feel really academic and/or less useful for industry work. Reading HTDP wasn't as useful to me, at least: I feel like Felleisen's books describe what he "imagines" would be a good way to write software on a "real" job :-p (non-academic setting). Skiena: I have mixed feelings about it... It seems to have gained popularity since Steve Yegge's blog post about Google interviews, but imho many concepts in that book are better explained elsewhere. I suppose the intended purpose of the book was to serve as a reference and an index to other resources. However, it often refers to sources that are 10 years old or older, academic implementations that could be challenging to apply in a real job setting, or things that are no longer available online afaict (e.g., LEDA), or hard to read papers, most of which lack a working implementation of the described concepts, etc. I wonder if anybody can really vouch for it or have a personal anecdote of the book being useful or enjoyable. -- 1: https://github.com/patmorin/ods https://github.com/patmorin/ods 2: https://greenteapress.com/wp/ https://greenteapress.com/wp/

raincole 3y ago

Packt feels like Udemy to me. I know there are some good stuff, but...

wyclif 3y ago
It's a comparable situation, I tend to agree. I wouldn't make a 1:1 comparison but Packt is not known for publishing well-edited and written books.
aerhardt 3y ago
Personally had an order of magnitude better experience with Udemy than Packt. Hard to find advanced content on Udemy but there’s a lot of beginner content that I’d consider rigorous for its level - can’t say the same about Packt.

ruiycombinator 3y ago

second this, it seems that anyone can publish with them and the books most of the times are terrible, like reading clickbait articles on medium. Terrible.

bogomipz 3y ago

I completely agree. Packt as a publisher is hot garbage. That's not to say every author is bad but but even when you get a quality author with good content, you will have horrific formatting and nonsense like pages with one sentence on them, illegible screenshots and there's often a complete absence of proof-reading or editing. The Packt game seems to be to simply flood search results hence quantity over quality.