Interesting, my only 2 encounters with Packt have been positive. One was a talk from a guy who wrote a book about a Python library who said Packt were good to work with, and the other was one of those pearls, "Python Machine Learning" (https://www.packtpub.com/big-data-and-business-intelligence/...)
The content of some of their books is written by very capable technical folks. Sadly, being a techy does not make you a good writer. Especially when you're not a native English speaker and are writing in English.
What's worse is that their review process seems to be worse than useless. Recently tried to read one of their books, which credited maybe 10 reviewers, and every other line had me cringing so hard it was unreadable.
Yep, I've recently read a Packt book and I found I think at least a hundred (!) of typos, grammatical errors etc. These kind of things give very bad evidence about the publisher, and (not always justly) indirectly also about the author, even if he is very competent technically. Better avoid this kind of publisher.
I've read an HN thread a few months ago and there was a lot of folks, both readers and writers, complaining about Packt's quality.
I wrote a book for Packt (and later for O'Reilly -- much better!) and I still cringe when I think about people reading it. They gave me three months to write it, and it's been out for three years now, so code samples are slowly breaking as technology moves forward, and some of the content is badly out of date. The principles are good, and there are definitely some useful things in it, but, my god, what I wouldn't have given for a copy editor and someone to actually review the code. It was all written in MS Word, as well, which made formatting and editing a nightmare. Their editors actually put mistakes into the book. I think I took most of them out, but, like I said, their deadlines are nuts.
