[BUG] Slows the the Rails server to a crawl after first use (30+ seconds page loads)

[nguyenquangminh0711]

[BUG] Slows the the Rails server to a crawl after first use (30+ seconds page loads)

[feliperaul]

An update: while the Chrome tab keeps on spinning, I got this error message:

[feliperaul]

Another update: I can confirm that with binding.pry everything is working as expected.

[nguyenquangminh0711]

Hi @feliperaul, let me confirm:

• The slowness happens when you inject jard, refresh the page on chrome, inspect, type exit, remove jard. Then, the next page is unacceptably slow right?
• Or, you inject jard, inspect, then continue and the page load is slow?

The message is interesting. That event should only be raised then the terminal emulator is resized.. Does it popup when you are in Overmind, or in a vanilla terminal emulator?

[feliperaul]

Hi @nguyenquangminh0711

First option: The slowness happens when I inject jard, refresh the page on chrome, inspect, type exit, remove jard. Then, the next page is unacceptably slow.

About the message, I'm not in front of the computer right now, but if I recall correctly, it was a vanilla Rails terminal.

What happens is, with iTerm2, I use CMD SHIFT ENTER all the time to resize the terminal windows to occupy fullscreen. I probably did that with jard even on the vanilla rails terminal, because when jard is activated only half the screen is too small for it.

Let me know if I can help debugging any further!

[nguyenquangminh0711]

Thank you. I think that's enough for me to debug. At the first glance, I think there are some conflicts between rails autoreload so that tons of scripts are running forever. During that time, when you resize, a race condition happens and pop that error out. I never go into this scenario before. Gonna test it out and let you know if I found some clues 😉

[nguyenquangminh0711]

Hi @feliperaul, I tested on my local, and I can confirm that it's truly a bug. The slowness is significant. I implemented a fix, and the situation is good (for me). Could you test it again by pointing jard to my branch?

gem 'ruby_jard', git: 'https://github.com/nguyenquangminh0711/ruby_jard', ref: 'bug/fix-slowness-after-exit'