Christopher Jeffrey (JJ)

@arexkun, I'm definitely doing them in the main file. I just noticed the case insensitivity attribute selector too (I must have missed it the first time, but it's interesting that...

A lot more has been implemented. Although, I'm not sure if the `:past/:current/:future` pseudoclasses are possible to implement. The `:column` pseudoclasses seem like they would be slightly complicated to implement;...

I tried writing up a rough implementation of `:column`, and it was roughly ~100 lines. That is for a robust implementation that takes into account everything in the table model....

The only thing really missing now is :column. I have an implementation written up, but it needs to be tested.

Not right now, but [pty.js has the ability to set the uid and gid](https://github.com/chjj/pty.js/blob/master/src/unix/pty.cc#L99) of the forked process. So, it is technically possible to map each browser user to a...

I've personally never run into this. I just tested it in my terminal (as a blessed widget, which could conceivably have even worse performance issues especially if the terminal is...

You should probably pass `WNOHANG` to `waitpid` in case that exit code isn't available to read or you might hang the process. I wrote something similar to this a while...

To expand upon my concern. This is the line in question that I believe could cause interference. ``` bash $ grep waitpid ~/node/ -r ~/node/deps/uv/src/unix/process.c: pid = waitpid(-1, &status, WNOHANG);...

> But surely that race already exists with PTY.js, whether or not you attempt to harvest status at exit in the SIGCHLD handler? I suppose it could cause problems right...

> Well if that were to happen, waitpid would return -1 since the process no longer exists and the PID is therefore invalid. Alright, that's what I hoped. I'll review...