Hacker News Reader

Comment by nxc18 | Hacker News Reader

I like this a lot. My uni just switched to Python from IDLE a few years ago. While its nicer than IDLE, the complexity of the IDE can be a bit much for novices. Heaven help you if you accidentally select the wrong python interpreter at project startup, for instance.

The other thing I'd like to see is more intelligent defaults for project location. On Windows, it defaults to the C drive, which is fine except for when students don't know where their documents folder is buried. Putting a git repo in a system folder breaks git because of permissions, but the UI will never tell you that.

View on HN

I disagree. Misspelling symbols and forgetting function names and signatures is a waste of time that distracts from what you're really learning.

I was dismayed to see the amount of time my first-time-programmer CS classmates had to spend navigating directories to locate functions and dealing with typo-based compile errors that would have been fixed immediately with an IDE. We should not be teaching new programmers that this is normal. Language and type aware autocomplete, inline syntax error and undefined symbol highlighing, and go-to-definition are things you should always expect from your programming environment.

Sure you can set up an editor to do these things, but that's a project in itself, and the results tend not to be nearly as good as "just use the JetBrains product."

IDLE has pretty decent auto-complete. I've found that teaching with IDLE first, then switching to PyCharm (or other more full-featured environments) after they've learned the basics produces better programmers.

Python also has the built-in ``help`` and ``dir`` for reminding yourself about features. No need to rely on the IDE when the language has you covered.

I agree. When I'm tutoring students and they have a problem I point out things like 'Find Usages' and the 'go to' menus. Those things are useful.

Unfortunately with Python, for to the nature of the language, there are often things pycharm can't call out. There are other times where pycharm says it won't work, but in fact it will (happens a lot with turtle, but imports in general are challenging). This isn't a criticism of pycharm so much as a challenge for any IDE with Python; VS Code gets it wrong more often than not as well.

>My uni just switched to Python from IDLE a few years ago

Another option is Thonny - a Python IDE for beginners.

http://thonny.org/

This is great. The step by step debugger should be very useful to beginners.

(I'm the PyCharm Developer advocate.) You're right, we need to do some thinking regarding a project's "environment". It's something we're talking about now.

Also jetbrains products are java-based and can bog down low powered machines to a point of frustration.

PyCharm works fine on my 2013 Macbook Air 11" (1,3GhZ i5 with 4GB RAM and a rather slow 128GB SSD).

I wouldn't call that a low powered machine. I started learning to program with a Gateway Netbook (sub 1ghz single core processor with 1gb ram) just a few years ago. It was so painfully slow to load anything other than a REPL and Vim (with no plugins - I tried compiling YouCompleteMe but all 3 times the machine ran out of memory). That's what I stuck with, and I'm glad :D

I'd say it's an average powered machine, at best.

I wouldn't have bought an i5 with 4 GB of RAM and 128 SSD for my sw development job in 2013. I had a Core Duo 2 with that much RAM from 2006 at the time (well past its end of life). That was still pretty ok CPU wise but the memory became really tight. I could make it faster with a SSD but it wasn't worth it.

I bought an i7, 16 GB (up to 32 GB) laptop at the beginning of 2014, with a 750 HDD that I swapped with a 1 TB SSD an year later.

The Air is my private machine and I bought for size/form factor (aka travel).

I had a MacBook Pro 15" (far better spec'd) at the time, but left it at home/work due to size and weight. The 11" Air fits "everywhere" and still has a full size keyboard and a good trackpad.

My Late 2013 MacBook Pro (bought early 2014) is a 2,3GhZ i7, 16GB memory and 500GB SSD. Compiling, VirtualBox/Parallels and Docker is less painful, but 99% of the time the old 11" Air is good enough.

Yeah, definitely not high powered, I just wanted to highlight that it's also not low powered compared to what others may be try to use PyCharm on, such as myself

It was considered slow 4 years ago, but I bought it for the form factor (and still use it daily).

PyCharm, Django and PostgreSQL runs ok, but it's too slow for docker.

Anything slower is useless for (modern) development and if you're a student on a budget there's loads of used or refurbished PCs on ebay costing $100 or even less.

A <$250 laptop purchased today is more powerful than the Air mentioned above. I think it's a reasonable baseline. It's still possible to get a 2GB laptop today, but it's becoming more rare, and the CPUs are very much more powerful than your old netbook.

Fair point. Most of my experience goes back to the "pre-macbook" stage of my life where I was running windows 7 on whatever hunk-of-junk laptop I could afford in college.

I have a lot of respect for the iterative improvements they make on all their products (I've use nearly all of them).

Reply on news.ycombinator.com