Jeremy Hinegardner

@johnbellone I'd like to put something that does this into the next version. Do you have something worked on yet, or shall I start on my own implementation. If you...

@johnbellone A week or so is what I'm thinking. As for more than one upstream? I would hope that it wouldn't matter and that the same version of a gem...

I wanted to see if this was an acceptable solution approach that would work for the project before seeing if I can generate data that causes it to fail. That...

@ohler55 is there another commit that does a similar thing on main? I’ve been away for a while - and I’ll be doing a fair bit of performance testing this...

@ohler55 oh I see 7280870c544767ab6f646495458836e67cc6803c - I’ll test the latest version this week and confirm.

@ohler55 I kicked off some long running processes using version ec9f0d175f3cefa11b322d2146327e9251199e02 and after running for almost an hour, they all get OOMKilled by linux. The only change i made to...

I had my system running with oj 3.11.2 which had the the issue that sometimes, it would segfault. I updated only oj to ec9f0d1, as in that was the only...

@ohler55 I'll plan on doing a long run against my patch before the end of next week too.

Well - unfortunately - it looks like my patch also suffers from the same long-term memory leak/reference leak. its takes about an hour of continually running, but it also gets...

The trimmed down version of the code is what I have in my PR in https://github.com/copiousfreetime/oj/blob/bugfix/issue-744-patch/test/test_744.rb and the data in https://github.com/copiousfreetime/oj/tree/bugfix/issue-744-patch/test/test_744-data - although I haven't tried this test on the...