Junzi Zhang

Got it! Thanks a lot for sharing! I will take a look at this when working on the stabilization code tests.

I think the cvxgrp conda channel is managed by @SteveDiamond manually, so it's easy to get out of sync sometimes. But you can always download it through the conda-forge channel,...

The memory is on the order of `O(dim*mem)`, as you suggested. The implementation actually stores `x, f, g, g_prev, y, s, d` (each with size `dim`), `Y, S, D, M`...

This memory consumption is for both Type-1 and Type-2 AA. It indeed works for relatively large problems (`10^6` ~ `10^7` at least, on a laptop) as well. You can check...

Thanks a lot Luke for sharing this! Sorry it has been extremely busy at Amazon in the past two months and so I haven't got time to resolve this issue....

Everything looks good here except for two very minor items: 1) For MFOMO, let's display the full `optimizer` dict including `config`, not just `name`; 2) let's just make `INFO_PANEL_WIDTH` and...

> I'd really prefer to not support configurable INFO_PANEL_WIDTH and MAX_TABLE_LENGTH values. Allowing them to be configurable would force us to handle formatting across a range of widths (line-wrapping vs....

Looks like we already have the latest version of the doc pointing to the newest master branch while "stable" version pointing to the released version, which is good! But apart...

> There might be a confusion about "stable" versus "latest". "stable" points to the most recently published package version, while "latest" points to the codebase as it currently exists on...

@matteosantama Thanks for the edits! Btw is there a way to see the updated website? Seems that https://mfglib--51.org.readthedocs.build/en/51/ is still the initial version when the PR was created, making it...