(Original post by alleycat393)
would you agree with the above?
This certainly isn't advice we would give out from the PS help team which I don't think you're a member of?
Experience isn't necessary though it won't hurt. Things like projects help yes but you can write a PS based just on your reading and be fine. And programming is just one aspect of compsci so we don't suggest that you focus on just that.
I'm speaking as an existing student. If the PS team are recommending that CompSci students read books and do nothing else then with all due respect, that's subpar advice.
I'm not saying that experience is necessary, nor that programming is the only aspect of CompSci that an applicant should focus on. There are plenty of students who go on to study CompSci with absolutely no prior experience, especially since it isn't offered at all colleges. In my experience, the strongest students (and therefore the strongest applicants) are the ones that get stuck in, not the ones who say they've read around the subject.
It doesn't matter what facet of CompSci an applicant is interested in, the subject as a whole is largely practical. It involves a lot of "doing". As a result, students who can demonstrate things they have done appear more favourably than students who can only say they've read. Because at the end of the day, the things you do are an application of the things you know. Reading it in a book is the first stage, applying it in a project is the second. Students who do are quite simply a stage higher than those who only read. And while programming isn't the only part of CompSci, it is one of the main focuses. An programmer worth their salt will tell you that you can only learn to program by doing, not by reading. As PQ says, books aren't bad, but they're not match for actually getting stuck in. It's also worth noting that the link provided by PQ provides a mix of books, practical activities and news sources. A good applicant will have a mix.
As you say, many students will be fine based purely on doing some reading. A lot of degrees expect no prior experience and will build experience from the ground up. But equally CompSci is highly focused on being able to do your own research and solve problems. Reading books will only get you so far. While not necessary, an applicant with practical experience is undeniably stronger than one who only has book knowledge. Many students will also find it easier to do something than to sit and read a book.
On a side note, there's a growing issue with books. They're becoming increasingly outdated. Some of my recommend reading is from 10-15 years ago. While many of the core concepts have not changed, the industry has continued to evolve. It is still evolving. Books are fine if you want to learn about FAT32 and NTFS, they're older file systems with plenty of documentation. But want a book on APFS, which is only 2 years old and in use across Apple devices? Good luck. Books on Computer Architecture? Easy to find. But will they mention Spectre and Meltdown, which are modern issues? Unlikely.
Books are fine for core underlying concepts that haven't changed much. But they fail massively when it comes to addressing modern technologies and in many cases, outright don't offer a modern approach. I have a book on Forensic Analysis of File Systems, it's part of my recommended reading. It's a really good book, but it's from 2005. It doesn't even address ext4, a filesystem from 2006 that I'm interacting with on a daily basis. As a result, books will only take a student so far and need to be supplemented with knowledge from elsewhere.
Reading books is fine. But it is not the most effective way to get experience with CompSci. There are other ways to make your application stand out that also help you grow as an applicant. Not to mention, if a student said they had read a book on the PS, I'd immediately have some questions about that. Can you demonstrate that you read the books? Anyone can say they've read something. How much of the book did you take in and understand? Reading a book is a waste of time if you don't get something out of it. And in all honesty, if someone tried to impress me by saying that they'd read a 500 page CompSci book from end to end, I'd think they're an idiot. I'd know for sure that they won't have taken it all in, assuming they read it at all, and that there would be better things they could have done with their time.
Quite simply, sitting down to read a reference book from end to end is a terrible approach to learning Computer Science.