[pypy-dev] Re: removing transparent proxies

2022-05-13 Thread Nathaniel Smith
Does jinja2 still rely on them, or did that get fixed at some point? On Fri, May 13, 2022, 05:02 Carl Friedrich Bolz-Tereick wrote: > Hi all! > > If anybody is particularly against removing transparent proxies, please > speak up! The feature has been marked as deprecated since around 2014: > >

[pypy-dev] removing transparent proxies

2022-05-13 Thread Carl Friedrich Bolz-Tereick
Hi all! If anybody is particularly against removing transparent proxies, please speak up! The feature has been marked as deprecated since around 2014: https://doc.pypy.org/en/latest/objspace-proxies.html It contains a bit strange irregular code that is pushing against the objspace APIs in

[pypy-dev] Re: removing transparent proxies

2022-05-13 Thread Carl Friedrich Bolz-Tereick
Oh, fascinating, I didn't know that they were using tproxy at some point! Does not really look like it's still there though: https://github.com/pallets/jinja/search?q=tproxy= But will grep a bit more carefully later when I'm not on the phone. Carl Friedrich On May 13, 2022 2:55:09 PM

[pypy-dev] Re: removing transparent proxies

2022-05-13 Thread Alex Gaynor
it was removed here https://github.com/pallets/jinja/commit/5308c9588d50d49b18885a8864915d728477a433 Alex On Fri, May 13, 2022 at 10:23 AM Carl Friedrich Bolz-Tereick wrote: > > Oh, fascinating, I didn't know that they were using tproxy at some point! > Does not really look like it's still

[pypy-dev] Re: removing transparent proxies

2022-05-13 Thread Carl Friedrich Bolz-Tereick
Thanks Alex! it probably means we should not remove tproxy from pypy2 because the last compatible version of jinja2 *does* use tproxy. Carl Friedrich On May 13, 2022 4:25:42 PM GMT+02:00, Alex Gaynor wrote: >it was removed here