Hacker News Reader

Comment by friedman23 | Hacker News Reader

I love jet brains but I'll say this. Their software is part of the reason I paid for 32 GB of ram on my new macbook pro.

View on HN

I'm one of those recent converts. I picked up programming in my late 30's and the first tool I ended up paying for was Pycharm. I never went through the 'print statement' phase as their debugger is incredible.

I'm using webstorm on an 8G Surface Go in a Starbucks today. Surprisingly not too bad.

How great is the 32 GB of ram? I have a 2017 macbook pro. I don't ever develop on a desktop machine to be able to know what I'm missing out on by only having 16 GB.

I run on 16GB at work and 8 at home. At home I run into issues with large Factorio factories or if I have too many browser tabs open (tbh that's an issue with me, not with my system, but more RAM would be a workaround for it), but at work the amount feels quite luxurious. I never got close to running out even with multiple VMs running at the same time, so I don't really see the need for more than 16. Maybe specialized applications need it like machine learning or so? Though that would seem like something you'd do on a server instead of your local laptop.

Unless you're running close to full most of the time, I also don't think the extra disk caching more RAM allows for is generally worth it. I have a few GB free most of the time (both at home and at work) and that seems to suffice to cache whatever files I'm currently working with.

If you keep tabs open for days and don't open them, like me, you might want to try a tab suspending extension. There's noticable lag when I open a tab I haven't opened in hours, but my FF on Ubuntu is fine with ~100 open tabs.

I did already, but nevertheless thanks for the tip!

I have chrome, webstorm, datagrip, fusion 360, an android emulator and an ios emulator open at the same time and there is 0 lag ever.

Check how much swap you’re using. If it’s very little, the answer is “not very”.

Thank Java.

The new breed of garbage collectors may change this significantly for the better. They need to be explicitly enabled, but there is a GC (G1) that may be better suited for desktop applications (low latency pauses, relinquishes memory to the OS more frequently while idle, etc.) available in Java 11.

It's been the default GC since Java 9 https://en.wikipedia.org/wiki/Garbage-first_collector

Ah, I'm confusing its availability with Shenandoah and zgc.

Unfortunately, the last time I used IntelliiJ it came with its own JDK8. Is that still the case, I wonder?

I just ran an update, IDEA and PyCharm use OpenJDK 11, Android Studio still uses 1.8.

They are all concurrent GCs though. You pay for the lower pauses with significant throughput hits.

Thank Garbage Collected languages. I just checked the IDEA config and found its maximum heap is set to 750 MB by default. You typically wouldn't need 32GB to run it.

I changed the maximum heap for my IntelliJ IDEA instances to 2GB and haven't seen a slowdown since. 750M is too low if you're working on a large project imo. Adding a dependency to my pom and immediately debugging used to cause the IDE to chug while it scanned everything.

Reply on news.ycombinator.com